You are on page 1of 510

Essentials in Ophthalmology

Series Editor: Arun D. Singh

Jorge L. Alió
Jorge L. Alió del Barrio, and
Francisco Arnalich-Montiel Editors

Corneal
Regeneration
Therapy and Surgery
Essentials in Ophthalmology

Series Editor
Arun D. Singh

More information about this series at http://www.springer.com/series/5332


Jorge L. Alió
Jorge L. Alió del Barrio
Francisco Arnalich-Montiel
Editors

Corneal Regeneration
Therapy and Surgery
Editors
Prof. Jorge L. Alió MD, PhD, FEBO
Professor and Chairman of Ophthalmology
University Miguel Hernandez
Vissum-Instituto Oftalmologico de Alicante
Alicante
Spain

Jorge L. Alió del Barrio Francisco Arnalich-Montiel


University Miguel Hernandez Vissum Corporation
Vissum-Instituto Oftalmologico Madrid
de Alicante Spain
Alicante
Spain

ISSN 1612-3212     ISSN 2196-890X (electronic)


Essentials in Ophthalmology
ISBN 978-3-030-01303-5    ISBN 978-3-030-01304-2 (eBook)
https://doi.org/10.1007/978-3-030-01304-2

Library of Congress Control Number: 2018963751

© Springer Nature Switzerland AG 2019


This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been made. The publisher remains
neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
This book is dedicated to our families and our patients, who
are those who enlighten our daily life with happiness and ideas
and who make our lives useful.
Acknowledgement

Prof. Jorge L. Alió would like to thank his supporting co-editors Jorge L. Alió
del Barrio, and Francisco Arnalich-Montiel for their contributions to this
work.

Alicante, Spain Prof. Jorge L. Alió

vii
Preface

 orneal Regeneration, the Present and the Future


C
of an Emerging Surgical Solution for Corneal Disease

Regeneration of human tissue is one of the most challenging and promising


topics in surgery and indeed in the whole of medicine. Eye care today, based
mainly on surgical options and medication, is finding the way to provide bet-
ter care for our patients affected by ocular disease using biological methods
of therapy, including tissue bioengineering which offers a new alternative that
may lead to a totally different perspective in the way in which we consider the
treatment of eye disease today.
Eye disease severely affects our quality of life. The possibility of success-
fully treating many ocular diseases which lead to visual impairment with
methods which are today the standards of care that do not involve surgical
options based on tissue substitution (transplantation), use of long-term ther-
apy or even mutilation caused by the elimination of diseased tissue and
replacement by new tissue (like in pseudophakia) is indeed extremely attrac-
tive. When accomplished, well-developed tissue regeneration therapy will be
considered as one of the major steps forward in human medicine.
This alternative is now being pursued by researchers and innovative clini-
cians and surgeons finding a way to obtain new methods that are more ana-
tomical and customized than those that we have today, even though these are
successful.
The cornea is a privileged tissue for studies on tissue regeneration. It is the
forefront of the eye, and due to its transparency, it provides the first step for
vision. Corneal blindness is one of the leading causes of blindness in the
world and can be cured only by tissue substitution, which is not always pos-
sible or even available. Corneal transplantation has been evolving over the
last 20 years into different modalities which include lamellar grafts of the
anterior and posterior surfaces that are, in fact, today, the most popular ways
to treat corneal disease. However, we are all aware of the limitations of cor-
neal grafting based on the availability of tissue and the biological hazard
always associated to these procedures.
The cornea, due to its accessibility, is where the most important advances
in corneal regeneration therapy and surgery are currently being developed. In
this book, the reader will find the most advanced and promising authors,
offering the alternative step that corneal regeneration may provide to clinical

ix
x Preface

practice today. From animal models to real applications based on data


obtained in clinical human studies, corneal regeneration is taking a great step
ahead towards an application that will change corneal surgery, creating a new
revolution. From ocular surface regeneration, corneal stroma regeneration
and corneal endothelial cell therapy, we may consider that there is an imme-
diate future in which not only shall we be able to regenerate these three parts
of the cornea (surface, stroma and endothelium), but, considered altogether,
we could regenerate the whole cornea based on autologous or cell bank tis-
sue. This is in part recreation and involves many challenges, from biological,
ethical, organizational and probably financial costs. Of all these, the funda-
mental use of autologous stem cells for these purposes harbours huge benefits
in terms of ethics and availability although we can already imagine the
emerging role that cell therapy banks can offer in the future.
We welcome the reader to this book which is the first of its type on this
topic and the first to offer in a systematic and improved way the developments
that we have obtained today in corneal regeneration. We hope that this will
open and stimulate the imagination of the reader and researchers in order to
go further and faster in this new and most promising way and stage of eye
surgery.
Just to finish this preface, I would like to thank all the participants, authors
and investigators of different origins who have participated directly or indi-
rectly in this book. Without their generous contribution, their ideas and their
hard work over many years in the background, something like this would not
have been possible. They have enlightened the future and this book aims to be
the platform to support all this light that will take our practice to a higher level
for the benefit of our corneal diseased patients.

Alicante, Spain Prof. Jorge L. Alió


Perspectives on Regenerative Medicine

It is now widely recognized that regenerative medicine is sure to open novel


treatment pathways deep into the twenty-first century, as it follows along the
line of organ and tissue transplantation, as well as stem cell-related transla-
tional research [1]. In regard to the former, it is set to replace organ replace-
ment therapies, such as heart, liver and kidney transplantations, via the use of
sophisticated cell therapies and/or tissues and organs created by tissue engi-
neering procedures [2, 3]. In regard to the latter, specific haematopoietic stem
cell transplantation, mesenchymal stem cell transplantation, cultured skin
sheets, corneal epithelial stem cell sheets, etc., can now be observed. In regard
to the cells used for regenerative medicine, there are two different strategies.
The first involves the direct use of stem cells in culture, while the second
involves the use of tissue-specific well-differentiated cells in culture. In
today’s regenerative medicine, two different concepts and strategies exist:
one is the direct transfer of cells and/or tissues that, expectedly, possess nor-
mal cell functions, while the other aims at releasing miracle substances,
including some cytokines, from the transferred cultured cells to the recipient
damaged/diseased organ, an example of which can now be seen in the heart-­
related field of regenerative medicine.
In the corneal field, upcoming medical advancements include superior
methods of donor corneal graft preparation prior to surgery. Previous cornea-­
related surgical procedures, such as keratophakia or epikeratophakia, as well
as more modern corneal endothelial grafting techniques, such as DSAEK and
DMEK, are prime examples [4]. From the aspect of pioneering corneal regen-
erative medicine techniques, autologous cultured corneal epithelial [5, 6] and
oral mucosal epithelial cell sheets [7, 8], as well as synthetic artificial corneal
grafts [9], have all been used in the clinical setting. Moreover, mesenchymal
stem cells have recently been applied for the reduction of corneal scarring
and restoration of corneal transparency, as have autologous adipose-derived
adult stem cells for the treatment of keratoconus [10, 11], a disorder that
results in a progressive thinning of the cornea. In addition, a novel cultured
corneal endothelial cell injection therapy for the treatment of bullous kera-
topathy [12] has recently been introduced. Some of these are truly ‘stem cell-­
specific’ therapies, while others are regenerative medicine-oriented treatments
that involve the use of cultured well-differentiated cells.
It is widely known that there are extremely important issues that must
firmly be addressed when considering the introduction of regenerative medi-
cine products to the clinical stage worldwide. This is clearly illustrated by the

xi
xii Perspectives on Regenerative Medicine

fact that although an impressive number of novel and significant academic


research findings have been published in the field of corneal regenerative
medicine, very few cell products have been officially approved by govern-
ment agencies. In many countries worldwide, the approved medical products
are categorized into three large groups, including pharmaceutical agents,
medical devices and regenerative medicine products. Of those, regenerative
medicine products fall into the newest category and are strictly regulated by
laws and/or guidelines set by each government. The primary point of govern-
ment laws and guidelines is to eliminate vague and/or false regenerative med-
icine products as well as to strictly control the products from the aspect of
safety and efficacy. While it is true that academic research scientists tend to
focus on the efficacy, regulatory authorities place the most importance on
safety, which is a vital perspective of regulatory science. Since most regen-
erative medicine products deal directly with in vitro human cells, the possi-
bility exists that recipients risk adverse events via dangerous unexpected
chemicals and/or viruses, thus illustrating why strict government regulation is
vitally important. With that aside, it remains undeniable that product efficacy
is a primary aspect [13, 14]. To reach that goal, the applied concept of ‘quality
by design’ [15] assures equal quality and flawless repeatability of the prod-
ucts, which is essential. To verify efficacy and safety in a clinical trial, one
must set the primary endpoint from not only the viewpoint of clinicians and
researchers but also via a general consensus of our society, which is the res-
toration of visual function. Moreover, the concept of a well-organized surro-
gate endpoint is important. In some ways, there are several important points
that slightly differ between the process of research and development and the
product development. Cutting-edge medical products originating from
embryonic stem cells, induced pluripotent stem cells and direct conversion
cells will assuredly emerge, as they may be key to unlocking the door to
future treatment pathways, thus once again illustrating the importance of
regulatory science, especially in regenerative medicine.

Shigeru Kinoshita
Department of Frontier Medical Science and Technology for
Ophthalmology
Kyoto Prefectural University of Medicine
Kyoto, Japan

References
1. Editorial. Advancing regenerative medicine. Nature Med. 2014;20(8):795.
2. Langer R, Vacanti JP. Tissue engineering. Science. 1993;260(5110):920-6.
3. Dimmeler S, Ding S, Rando TA, Trounson A. Translational strategies and challenges in
regenerative medicine. Nature Med. 2014;20(8):814-21.
4. Tan DT, Dart JK, Holland EJ, Kinoshita S.  Corneal transplantation. Lancet.
2012;379(9827):1749-61.
5. Pellegrini G1, Traverso CE, Franzi AT, Zingirian M, Cancedda R, De Luca M. Long-
term restoration of damaged corneal surfaces with autologous cultivated corneal epithe-
lium. Lancet. 1997;349(9057):990-3.
Perspectives on Regenerative Medicine xiii

6. Rama P, Matuska S, Paganoni G, Spinelli A, De Luca M, Pellegrini G. Limbal stem-


cell therapy and long-term corneal regeneration. N Engl J Med. 2010;363(2):147-55.
7. Nishida K, Yamato M, Hayashida Y, et  al. Corneal reconstruction with tissue-engi-
neered cell sheets composed of autologous oral mucosal epithelium. N Engl J Med.
2004;351(12):1187-96.
8. Sotozono C, Inatomi T, Nakamura T, et al. Visual improvement after cultivated oral
mucosal epithelial transplantation. Ophthalmology. 2013;120:193-200.
9. Fagerholm P, Lagali NS, Merrett K, et al. A biosynthetic alternative to human donor tis-
sue for inducing corneal regeneration: 24-month follow-­up of a phase 1 clinical study.
Sci Transl Med. 2010;2(46):46-61.
10. Alio Del Barrio JL, El Zarif M, de Miguel MP, et  al. Cellular therapy with human
autologous adipose-derived adult stem cells for advanced keratoconus. Cornea.
2017;36(8):952-60.
11. Alio Del Barrio JL, El Zarif M, Azzar A, et  al. Corneal stroma enhancement with
decellularized stromal laminas with or without stem cell recellularization for advanced
keratoconus. Am J Ophthalmol. 2018;186:47-58.
12. Kinoshita S, Koizumi N, Ueno M, et al. Injection of cultured cells with a ROCK inhibi-
tor for bullous keratopathy. N Engl J Med. 2018;378:995-1003.
13. Charo RA, Sipp D.  Rejuvenating regenerative medicine regulation. N Engl J Med.
2018;378(6):504-5.
14. Marks P, Gottlieb S. Balancing safety and innovation for cell-based regenerative medi-
cine. N Engl J Med. 2018;378(10):954-9.
15. Lipsitz YY, Timmins NE, Zandstra PW. Quality cell therapy manufacturing by design.
Nat Biotech. 2016;34:393-400.
Contents

Part I Corneal Regeneration: The Concept, the Facts, the Potential

1 Corneal Anatomy ����������������������������������������������������������������������������   3


Miguel Gonzalez-Andrades, Pablo Argüeso, and Ilene Gipson
2 Corneal Healing ������������������������������������������������������������������������������  13
Veronica Vargas, Francisco Arnalich-Montiel,
and Jorge L. Alió del Barrio
3 Corneal Tissue Engineering������������������������������������������������������������  23
Mohammad Mirazul Islam, Roholah Sharifi,
and Miguel Gonzalez-Andrades

Part II The Stem Cell

4 Stem Cells: Concept, Properties, and Characterization��������������  41


Natalia Escacena-Acosta, Javier Lopez-Beas,
Christian Claude Lachaud, Mehrdad Vakilian,
Juan Rigoberto Tejedo, Vivian Capilla-González,
Francisco Javier Bedoya, Franz Martin, Abdelkrim Hmadcha,
and Bernat Soria
5 Corneal Stem Cells: Identification and Methods of Ex Vivo
Expansion�����������������������������������������������������������������������������������������  57
Christian Claude Lachaud, Abdelkrim Hmadcha,
and Bernat Soria
6 Corneal Epithelial Stem Cells: Methods for Ex Vivo
Expansion�����������������������������������������������������������������������������������������  77
Gustavo S. Figueiredo, Hardeep Singh Mudhar,
Majlinda Lako, and Francisco C. Figueiredo
7 Corneal Stromal Stem Cell: Methods for Ex Vivo Expansion ����  99
Olena Al-Shymali, Jorge L. Alió del Barrio,
and James L. Funderburgh
8 Corneal Endothelial Cells: Methods for Ex Vivo Expansion ������ 109
Stephen Wahlig, Matthew Lovatt, Gary Swee-­Lim Peh,
and Jodhbir S. Mehta

xv
xvi Contents

9 Corneal Regeneration: Use of Extracorneal Stem Cells�������������� 123


Christian Claude Lachaud, Abdelkrim Hmadcha,
and Bernat Soria
10 One Cell, Two Phenotypes: Capturing Pluripotency for Corneal
Regeneration������������������������������������������������������������������������������������ 145
Trevor Sherwin, Carol Ann Greene, Colin R. Green,
and Kushant R. Kapadia
11 Corneal Stem Cell-Based Therapies���������������������������������������������� 155
Yuzuru Sasamoto, Yoshinori Oie, and Kohji Nishida

Part III Regenerative Surgery and Therapy of the Ocular Surface


Epithelium

12 Ocular Surface Epithelium: Applied Anatomy ���������������������������� 175


Harminder Singh Dua and Dalia G. Said
13 Classical Techniques for Limbal Transplantation������������������������ 191
Rafael I. Barraquer and Juan Alvarez de Toledo
14 Simple Limbal Epithelial Transplantation: An Update���������������� 213
Nandini Venkateswaran and Guillermo Amescua
15 Cell Therapy Using Ex Vivo Cultured Limbal Cells:
CLET and Equivalent���������������������������������������������������������������������� 221
Paolo Rama and Giulio Ferrari
16 Cell Therapy Using Cultivated Oral Mucosal Epithelial
Transplant (COMET)���������������������������������������������������������������������� 225
Roberto Fernández Buenaga and Sajjad Ahmad
17 Cell Therapy Using Extraocular Mesenchymal Stem Cells �������� 231
Teresa Nieto-Miguel, Sara Galindo, Marina López-­Paniagua,
Inmaculada Pérez, José M. Herreras, and Margarita Calonge
18 Cell-based Therapy Using Induced Plutipotent Stem Cell���������� 263
Ricardo Pedro Casaroli-Marano
19 Cultivated Limbal Stem Cell Transplantation: Indications
and Technique���������������������������������������������������������������������������������� 277
Joséphine Behaegel, Sorcha Ní Dhubhghaill,
and Marie-José Tassignon
20 Optimizing the Ocular Surface for Regenerative Surgery:
What Is Important and What Is Essential for the Outcome�������� 291
Kai B. Kang and Ali R. D’jalilian
21 Stem Cell Spheres for Corneal Regeneration�������������������������������� 299
Salim Ismail, Jennifer J. McGhee, Ye Li, Jeremy John
Mathan, Jinny Jung Yoon, Himanshu Wadhwa,
Stephanie U-Shane Huang, and Trevor Sherwin
22 Eye Platelet-Rich Plasma (E-PRP) for Corneal Regeneration���� 317
Alejandra E. Rodríguez and Jorge L. Alió
Contents xvii

Part IV Regenerative Surgery of the Corneal Stroma


23 Applied Anatomy of the Corneal Stroma�������������������������������������� 349
Harminder Singh Dua and Dalia G. Said
24 Confocal Microscopy of the Cornea in a Clinical Model
of Corneal Stromal Expansion Using Adipose Stem Cells
and Corneal Decellularized Laminas in Patients
with Keratoconus ���������������������������������������������������������������������������� 363
Mona El Zarif, Karim Abdul Jawad, and Jorge L. Alió
25 Limbal Stromal Stem Cells in Corneal Wound Healing:
Current Perspectives and Future Applications ���������������������������� 387
Noopur Mitragotri, Mukesh Damala, Vivek Singh,
and Sayan Basu
26 Cell Therapy of the Corneal Stroma Using Ex Vivo Cultured
Extraocular Cells ���������������������������������������������������������������������������� 403
Jorge L. Alió del Barrio

Part V Regenerative Surgery of the Corneal Endothelium

27 Corneal Endothelium: Applied Anatomy�������������������������������������� 419


Francisco Arnalich-Montiel
28 Corneal Endothelium: Isolation and Cultivation Methods���������� 425
David Mingo-Botín, Marie Joan Therese D. Balgos,
and Francisco Arnalich-Montiel
29 Corneal Endothelial Cell Transplantation: Animal Models�������� 437
Brad P. Barnett and Albert S. Jun
30 Cell Based Therapy for Corneal Endothelial Regeneration�������� 455
Noriko Koizumi and Naoki Okumura
31 Corneal Endothelium Regeneration: Future Prospects �������������� 463
Wei-Ting Ho, Hsin-Yu Liu, Fung-Rong Hu, and I-Jong Wang

Part VI Bioengineering Cornea Surgery

32 Umbilical Cord Stem Cells in the Treatment of Corneal


Diseases �������������������������������������������������������������������������������������������� 477
Mohammed Ziaei, Jie Zhang, Dipika V. Patel,
and Charles N. J. McGhee
33 Dysfunctional Corneal Endothelium:
Delivery of Cell Therapy ���������������������������������������������������������������� 485
Stephen Wahlig, Gary Swee-Lim Peh, Matthew Lovatt,
and Jodhbir S. Mehta
Index���������������������������������������������������������������������������������������������������������� 499
Contributors

Sajjad Ahmad, MB, BS, FRCOphth, PhD  Cornea and External Diseases


Department, Moorfields Eye Hospital, London, UK
Jorge L. Alió, MD, PhD, FEBO  Professor and Chairman of Ophthalmology,
University Miguel Hernandez, Vissum-Instituto Oftalmologico de Alicante,
Alicante, Spain
Jorge  L. Alió  del Barrio, MD, PhD, FEBOS-CR University Miguel
Hernandez, Vissum-Instituto Oftalmologico de Alicante, Alicante, Spain
Olena Al-Shymali, MD  Vissum Corporation, Alicante, Spain
Juan  Alvarez  de Toledo, MD, PhD, FEBO-CR  Centro de Oftalmología
Barraquer, Anterior Segment, Department, Barcelona, Spain
Institut Universitari Barraquer, Universitat Autónoma de Barcelona, Barcelona,
Spain
Guillermo  Amescua, MD Bascom Palmer Eye Institute, University of
Miami, Department of Ophthalmology, Miami, FL, USA
Pablo  Argüeso, PhD Massachusetts Eye and Ear and Schepens Eye
Research Institute, Department of Ophthalmology, Harvard Medical School,
Boston, MA, USA
Francisco  Arnalich-Montiel, MD, PhD, FEBOS-CR Vissum Corporation,
Madrid, Spain
Marie Joan Therese D. Balgos, MD  Research and Development Department,
Cornea, Cataract and Refractive Surgery Department, Vissum Alicante,
Alicante, Spain
Brad  P.  Barnett, MD, PhD Wilmer Eye Institute at Johns Hopkins,
Department of Ophthalmology, Baltimore, MD, USA
Rafael I. Barraquer, MD, PhD  Centro de Oftalmología Barraquer, Anterior
Segment Department, Barcelona, Spain
Sayan  Basu, MBBS, MS Center for Ocular Regeneration (CORE), LV
Prasad Eye Institute, Banajara Hills, Hyderabad, Telangana, India

xix
xx Contributors

Francisco Javier Bedoya, PhD, MD  Department of Cell Regeneration and


Advanced Therapies, Andalusian Center of Molecular Biology and
Regenerative Medicine-CABIMER, Junta de Andalucía-University of Pablo
Olavide-University of Seville-CSIC, Seville, Andalusia, Spain
Joséphine  Behaegel, MD Faculty of Medicine and Health Sciences,
Department of Ophthalmology, Visual Optics and Visual Rehabilitation,
University of Antwerp, Campus Drie Eiken, Antwerp, Belgium
Department of Ophthalmology, Antwerp University Hospital, Edegem, Belgium
Center for Cell Therapy and Regenerative Medicine, Antwerp University
Hospital, CCRG-Oogheelkunde, Edegem, Belgium
Department of Ophthalmology, Brussels University Hospital, Jette, Belgium
Roberto Fernández Buenaga, MD, PhD  Cornea, Cataract and Refractive
Surgery Department, Vissum Madrid, Madrid, Spain
Margarita  Calonge, MD, PhD CIBER-BBN (Biomedical Research
Networking Centre in Bioengineering, Biomaterials and Nanomedicine),
Carlos III National Institute of Health, Valladolid, Spain
IOBA (Institute of Applied Ophthalmobiology), University of Valladolid,
Valladolid, Spain
Vivian  Capilla-González, PhD Department of Cell Regeneration and
Advanced Therapies, Andalusian Center of Molecular Biology and
Regenerative Medicine-CABIMER, Junta de Andalucía-University of Pablo
Olavide-University of Seville-CSIC, Seville, Andalusia, Spain
Ricardo Pedro Casaroli-Marano, MD, MSc, PhD  Department of Surgery,
School of Medicine and Hospital Clinic de Barcelona, University of
Barcelona, Barcelona, Spain
Institute of Biomedical Research (IIB-Sant Pau) and Barcelona Tissue Bank,
Banc de Sang i Teixits, Barcelona, Spain
Mukesh  Damala, MSc Biochemistry Center to Ocular Regeneration
(CORE); and Brien Holden Eye Research Center, LV Prasad Eye Institute,
Hyderabad, Telangana, India
School of Life Sciences, University of Hyderabad, Hyderabad, Telangana, India
Harminder Singh Dua, MBBS, DO, MS, MD, PhD  Academic Section of
Ophthalmology, Division of Clinical Neuroscience, University of Nottingham,
Nottingham, UK
Department of Ophthalmology, Queens Medical Centre, University Hospitals
NHS Trust, Nottingham, UK
Mona El Zarif, Master in Optometry, OD  Optica General Sarl, Department
of Optometry and contactology, Saida, Lebanon
Vissum Instituto Oftalmologico de Alicante, Alicante, Spain
Natalia  Escacena-Acosta, PhD Department of Cell Regeneration and
Advanced Therapies, Andalusian Center of Molecular Biology and
Regenerative Medicine-CABIMER, Junta de Andalucía-University of Pablo
Olavide-University of Seville-CSIC, Seville, Andalusia, Spain
Contributors xxi

Giulio  Ferrari, MD, PhD San Raffaele Hospital, Department of


Ophthalmology—Cornea and Ocular Surface, Milan, Italy
Francisco  C.  Figueiredo Newcastle University, Institute of Genetic
Medicine, International Centre for Life, Newcastle upon Tyne, Tyne and
Wear, UK
Royal Victoria Infirmary, Newcastle upon Tyne, UK
Gustavo S. Figueiredo, MB, ChB, PhD  Newcastle University, Institute of
Genetic Medicine, International Centre for Life, Newcastle upon Tyne, Tyne
and Wear, UK
James  L.  Funderburgh, PhD University of Pittsburgh, Department of
Ophthalmology, Pittsburgh, PA, USA
Sara Galindo, PhD  CIBER-BBN (Biomedical Research Networking Centre
in Bioengineering, Biomaterials and Nanomedicine) Carlos III National
Institute of Health, Madrid, Spain
IOBA (Institute of Applied Ophthalmobiology), University of Valladolid,
Valladolid, Spain
Ilene Gipson, PhD  Massachusetts Eye and Ear and Schepens Eye Research
Institute, Department of Ophthalmology, Harvard Medical School, Boston,
MA, USA
Ali  R.  D’jalilian, MD University of Illinois at Chicago, Department of
Ophthalmology and Visual Sciences, Chicago, IL, USA
Miguel  Gonzalez-Andrades, MD, PhD Massachusetts Eye and Ear and
Schepens Eye Research Institute, Department of Ophthalmology, Harvard
Medical School, Boston, MA, USA
Carol  Ann  Greene, PhD Department of Ophthalmology, New Zealand
National Eye Centre, Faculty of Medical and Health Sciences, The University
of Auckland, Auckland, New Zealand
Colin  R.  Green, BSc, MSc, PhD, DSc Department of Ophthalmology,
New Zealand National Eye Centre, Faculty of Medical and Health Sciences,
The University of Auckland, Auckland, New Zealand
José  M.  Herreras, MD, PhD CIBER-BBN (Biomedical Research
Networking Centre in Bioengineering, Biomaterials and Nanomedicine)
Carlos III National Institute of Health, Valladolid, Spain
IOBA (Institute of Applied Ophthalmobiology), University of Valladolid,
Valladolid, Spain
Department of Ophthalmology, Clinic University Hospital, Valladolid, Spain
Abdelkrim Hmadcha, PhD  Department of Cell Regeneration and Advanced
Therapies, Andalusian Center of Molecular Biology and Regenerative
Medicien-CABIMER, Junta de Andalucía-University of Pablo Olavide-­
University of Seville-CSIC, Seville, Andalusia, Spain
xxii Contributors

Wei-Ting  Ho, MD  Department of Ophthalmology, Far Eastern Memorial


Hospital, New Taipei City, Taiwan
Fung-Rong  Hu, MD Department of Ophthalmology, National Taiwan
University Hospital, Taipei, Taiwan
College of Medicine, National Taiwan University, Taipei, Taiwan
Stephanie  U-Shane  Huang, Bachelor of Science (Hons)  Department of
Ophthalmology, New Zealand National Eye Centre, Faculty of Medical and
Health Sciences, The University of Auckland, Auckland, New Zealand
Mohammad  Mirazul  Islam, MSc, PhD  Massachusetts Eye and Ear and
Schepens Eye Research Institute, Department of Ophthalmology, Harvard
Medical School, Boston, MA, USA
Salim Ismail, BSC, MSc (1st Class Hons)  Department of Ophthalmology,
New Zealand National Eye Centre, Faculty of Medical and Health Sciences,
The University of Auckland, Auckland, New Zealand
Karim Abdul Jawad, Bachelor of Science  Optica General Sarl, Department
of Optometry and Contactology, Saida, Lebanon
University of Nicosia, Department of Life and Health Sciences, Nicosia, Cyprus
Albert S. Jun, MD, PhD  Wilmer Eye Institute at Johns Hopkins, Department
of Ophthalmology, Baltimore, MD, USA
Division of Cornea, Cataract and External Eye Diseases, Baltimore, MD, USA
Kushant R. Kapadia, BSc, PGDipSci, MSc  Department of Ophthalmology,
New Zealand National Eye Centre, Faculty of Medical and Health Sciences,
The University of Auckland, Auckland, New Zealand
Christian  Claude  Lachaud, PhD Department of Cell Regeneration and
Advanced Therapies, Andalusian Center of Molecular Biology and
Regenerative Medicine-CABIMER, Junta de Andalucía-University of Pablo
Olavide-University of Seville-CSIC, Seville, Andalusia, Spain
Majlinda  Lako, PhD, MSc, BSc Newcastle University, Institute of Genetic
Medicine, International Centre for Life, Newcastle upon Tyne, Tyne and Wear, UK
Javier Lopez-Beas, PhD  Department of Cell Regeneration and Advanced
Therapies, Andalusian Center of Molecular Biology and Regenerative
Medicine-CABIMER, Junta de Andalucía-University of Pablo Olavide-­
University of Seville-CSIC, Seville, Andalusia, Spain
Kai  B.  Kang, MD University of Illinois at Chicago, Department of
Ophthalmology and Visual Sciences, Chicago, IL, USA
Noriko  Koizumi, MD, PhD Department of Biomedical Engineering,
Faculty of Life and Medical Sciences, Doshisha University, Kyotanabe,
Kyoto, Japan
Ye  Li, MBBS (Student) Department of Ophthalmology, New Zealand
National Eye Centre, Faculty of Medical and Health Sciences, The University
of Auckland, Auckland, New Zealand
Contributors xxiii

Hsin-Yu  Liu, MD Department of Ophthalmology, National Taiwan


University Hospital, Taipei, Taiwan
Marina  López-Paniagua, PhD CIBER-BBN (Biomedical Research
Networking Centre in Bioengineering, Biomaterials and Nanomedicine)
Carlos III National Institute of Health, Valladolid, Spain
IOBA (Institute of Applied Ophthalmobiology), University of Valladolid,
Valladolid, Spain
Matthew  Lovatt, PhD Tissue Engineering and Stem Cell Group, The
Academia, Singapore Eye Research Institute, Singapore, Singapore
Franz Martin, PhD, MD  Department of Cell Regeneration and Advanced
Therapies, Andalusian Center of Molecular Biology and Regenerative
Medicine-CABIMER, Junta de Andalucía-University of Pablo Olavide-­
University of Seville-CSIC, Seville, Andalusia, Spain
Jeremy  John  Mathan, BMedSc (Hons), MBChB Department of
Ophthalmology, New Zealand National Eye Centre, Faculty of Medical and
Health Sciences, The University of Auckland, Auckland, New Zealand
Charles N. J. McGhee, DSc, FRCOphth  Department of Ophthalmology,
New Zealand National Eye Centre, Faculty of Medical and Health Sciences,
The University of Auckland, Auckland, New Zealand
Jennifer  J.  McGhee, BSc Department of Ophthalmology, New Zealand
National Eye Centre, Faculty of Medical and Health Sciences, The University
of Auckland, Auckland, New Zealand
Jodhbir  S.  Mehta, BSc, MBBS, FRCSCEd Singapore National Eye
Centre, Department of Cornea and External Disease, Singapore, Singapore
David Mingo-Botín, MD, PhD  Cornea Unit, Department of Ophthalmology,
Hospital Universitario Ramón y Cajal, Madrid, Spain
Noopur  Mitragotri, MSc Biochemistry Center to Ocular Regeneration
(CORE); and Brien Holden Eye Research Center, LV Prasad Eye Institute,
Hyderabad, Telangana, India
Hardeep Singh Mudhar, BSc, PhD, MBBCHir, FRCPath  Royal Hallamshire
Hospital, Department of Histopathology, Sheffield, South Yorkshire, UK
Sorcha Ní Dhubhghaill, MD, PhD  Faculty of Medicine and Health Sciences,
Department of Ophthalmology, Visual Optics and Visual Rehabilitation,
University of Antwerp, Campus Drie Eiken, Antwerp, Belgium
Department of Ophthalmology, Antwerp University Hospital, Edegem, Belgium
Center for Cell Therapy and Regenerative Medicine, Antwerp University
Hospital, CCRG-Oogheelkunde, Edegem, Belgium
Teresa Nieto-Miguel, PhD  CIBER-BBN (Biomedical Research Networking
Centre in Bioengineering, Biomaterials and Nanomedicine), Carlos III
National Institute of Health, Valladolid, Spain
IOBA (Institute of Applied Ophthalmobiology), University of Valladolid,
Valladolid, Spain
xxiv Contributors

Kohji Nishida, MD, PhD  Department of Ophthalmology, Osaka University


Graduate School of Medicine, Suita, Osaka, Japan
Yoshinori Oie, MD, PhD  Department of Ophthalmology, Osaka University
Graduate School of Medicine, Suita, Osaka, Japan
Naoki  Okumura, MD, PhD Department of Biomedical Engineering,
Faculty of Life and Medical Sciences, Doshisha University, Kyotanabe,
Kyoto, Japan
Dipika  V.  Patel, PhD, MRCOphth Department of Ophthalmology, New
Zealand National Eye Centre, Faculty of Medical and Health Sciences, The
University of Auckland, Auckland, New Zealand
Gary Swee-Lim Peh, PhD  Tissue Engineering and Stem Cell Group, The
Academia, Singapore Eye Research Institute, Singapore, Singapore
Inmaculada  Pérez, PhD  IOBA (Institute of Applied Ophthalmobiology),
University of Valladolid, Valladolid, Spain
Paolo Rama, MD  San Raffaele Hospital, Department of Ophthalmology—
Cornea and Ocular Surface, Milan, Italy
Alejandra  E.  Rodríguez, PhD  Laboratory of the Research, Development
and Innovation Department, Vissum Innovation, Alicante, Spain
Dalia  G.  Said, MB, BCh, MSc, MD, FRCS Academic Section of
Ophthalmology, Division of Clinical Neuroscience, University of Nottingham,
Nottingham, UK
Department of Ophthalmology, Queens Medical Centre, University Hospitals
NHS Trust, Nottingham, UK
Research Institute of Ophthalmology (RIO), Cairo, Egypt
Yuzuru  Sasamoto, MD, PhD Division of Genetics, Department of
Medicine, Brigham and Women’s Hospital, Boston, MA, USA
Roholah  Sharifi, PhD Massachusetts Eye and Ear and Schepens Eye
Research Institute, Department of Ophthalmology, Harvard Medical School,
Boston, MA, USA
Trevor  Sherwin, BSc (Hons), PhD Department of Ophthalmology, New
Zealand National Eye Centre, Faculty of Medical and Health Sciences, The
University of Auckland, Auckland, New Zealand
Vivek Singh, MSc, PhD  Center to Ocular Regeneration (CORE); and Brien
Holden Eye Research Center, LV Prasad Eye Institute, Hyderabad, Telangana,
India
Bernat  Soria, MD, PhD, FRCP Department of Cell Regeneration and
Advanced Therapies, Andalusian Center of Molecular Biology and
Regenerative Medicine-CABIMER, Junta de Andalucía-University of Pablo
Olavide-University of Seville-CSIC, Seville, Andalusia, Spain
Centro de Investigación Biomédica en Red de Diabetes y Enfermedades
Metabólicas Asociadas (CIBERDEM), Madrid, Spain
Contributors xxv

Marie-José Tassignon, MD, PhD  Faculty of Medicine and Health Sciences,


Department of Ophthalmology, Visual Optics and Visual Rehabilitation,
University of Antwerp, Campus Drie Eiken, Antwerp, Belgium
Department of Ophthalmology, Antwerp University Hospital, Edegem, Belgium
Juan  Rigoberto  Tejedo, PhD Department of Cell Regeneration and
Advanced Therapies, Andalusian Center of Molecular Biology and
Regenerative Medicine-CABIMER, Junta de Andalucía-University of Pablo
Olavide-University of Seville-CSIC, Seville, Seville, Andalusia, Spain
Centro de Investigación Biomédica en Red de Diabetes y Enfermedades
Metabólicas Asociadas (CIBERDEM), Madrid, Spain University Pablo de
Olavide, Seville, Spain
Mehrdad  Vakilian, PhD Student Department of Cell Regeneration and
Advanced Therapies, Andalusian Center of Molecular Biology and
Regenerative Medicine-CABIMER, Junta de Andalucía-University of Pablo
Olavide-University of Seville-CSIC, Seville, Andalusia, Spain
Veronica Vargas, MD  Vissum Instituto Oftalmologico de Alicante, Alicante,
Spain
Nandini Venkateswaran, MD  Bascom Palmer Eye Institute, University of
Miami, Department of Ophthalmology, Miami, FL, USA
Himanshu  Wadhwa, MBChB, BMedSc (Hons) Department of
Ophthalmology, New Zealand National Eye Centre, Faculty of Medical and
Health Sciences, The University of Auckland, Auckland, New Zealand
Stephen  Wahlig, BSc Tissue Engineering and Stem Cell Group, The
Academia, Singapore Eye Research Institute, Singapore, Singapore
I-Jong  Wang, MD, PhD  Department of Ophthalmology, National Taiwan
University Hospital, Taipei, Taiwan
College of Medicine, National Taiwan University, Taipei, Taiwan
Jinny  Jung  Yoon, PhD, MBChB Department of Ophthalmology, New
Zealand National Eye Centre, Faculty of Medical and Health Sciences, The
University of Auckland, Auckland, New Zealand
Jie Zhang, PhD  Department of Ophthalmology, New Zealand National Eye
Centre, Faculty of Medical and Health Sciences, The University of Auckland,
Auckland, New Zealand
Mohammed  Ziaei, FRCOphth  Department of Ophthalmology,
New Zealand National Eye Centre, Faculty of Medical and Health Sciences,
The University of Auckland, Auckland, New Zealand
Part I
Corneal Regeneration: The Concept, the
Facts, the Potential
Corneal Anatomy
1
Miguel Gonzalez-Andrades, Pablo Argüeso,
and Ilene Gipson

1.1 Introduction derive from neural crest cells that migrate


between the primitive corneal epithelium and the
The cornea is a uniquely translucent, avascular lens [5–7].
tissue located on the anterior segment of the eye The anatomical dimensions of the human cor-
(Fig. 1.1). It is surrounded and maintained by the nea are variable among individuals (Fig.  1.1).
adjacent corneoscleral limbus and the connective The average corneal horizontal diameter (white-­
tissue of the conjunctiva with its adnexa [1]. It to-­white distance or limbus to limbus) is greater
plays a vital role in visual function by providing than the vertical one (11.71 ± 0.42 mm and 10.63
(1) the major refractive component of the visual ± 0.63 mm, respectively) [8, 9]. The central cor-
system [2], (2) a translucent tissue that allows neal thickness ranges from 0.50 to 0.60 mm [10–
light passage to the lens and retina, and (3) a bar- 12], which increases progressively toward the
rier that protects the eye from fluid loss and the periphery, where thickness can be 26% greater
external environment. These crucial functions than central point values [12].
result from the structure of the cornea, which is
composed of three anatomical layers: epithelium,
stroma, and endothelium (Fig. 1.2). The limbus is 1.2 The Corneal Epithelium
the reservoir for the adult stem cell population
that replenishes the cornea and is the site of ter- The corneal epithelium is a nonkeratinized strati-
mination of the vasculature and entry of the fied squamous epithelium that constitutes the
nerves that provide an extraordinarily rich inner- outermost layer of the cornea. The epithelium is
vation environment (Fig. 1.2). comprised of 5–7 cell layers including flattened
The embryology of the cornea is based on apical squamous cells, subapical cells that have
inductive interactions that take place in the cra- winglike structure – hence they are called “wing
nial ectoderm [3]. The corneal epithelium origi- cells” – and a basal cell layer that is columnar in
nates from the cross talk of the neural ectoderm shape (Fig.  1.3a). The corneal epithelium is
with the optic vesicle [4]. The corneal stromal unique among other stratified squamous epithelia
cells or keratocytes and the corneal endothelium in that it is of a very uniform thickness and has a
unique smooth wet surface that, along with the
M. Gonzalez-Andrades (*) · P. Argüeso · I. Gipson tear film, provides its refractive power. Cells of
Massachusetts Eye and Ear and Schepens Eye each layer have a specialized structure and func-
Research Institute, Department of Ophthalmology,
Harvard Medical School, Boston, MA, USA
tion, and apical and basal cells show specific pro-
e-mail: Miguel_gonzalez@meei.harvard.edu tein expression patterns.

© Springer Nature Switzerland AG 2019 3


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_1
4 M. Gonzalez-Andrades et al.

Fig. 1.1  The cornea (1) is located on the anterior seg- zontal diameter (orange arrow) is greater than the vertical
ment of the eye. It is surrounded by different structures: one (blue arrow). The central corneal thickness ranges
the sclera (2), the iris (3), the ciliary body (4), the lens (5), from 0.50 to 0.60 mm in the central area (dotted red line),
and the zonule of Zinn (6). The anatomical dimensions of which increases progressively toward the periphery (dot-
the human cornea are variable. The average corneal hori- ted green line)

Fig. 1.2  The cornea is composed of three anatomical lay- involved in corneal epithelial regeneration: corneal stro-
ers: epithelium (1), stroma (3), and endothelium (5). Two mal stem cells (CSSCs), limbal epithelial stem cells
distinguishable areas flank the stroma: the Bowman layer (LESC), suprabasal corneal epithelial cells (SCEC), tran-
(2) and Descemet’s membrane (4). The limbus (green sient amplifying cells (TAC), terminally differentiated
circle) is the reservoir for the adult stem cell population cells (TDC) (Conj., conjunctiva)
that replenishes the cornea. There are several cells

The most apical cell layer of the corneal epi- vides a barrier to the penetrance of large mole-
thelium expresses membrane-associated mucins cules and pathogens [16, 17] with MUC16
MUC1 and MUC16 that form the surface glyco- functioning as the major component of the bar-
calyx [13, 14] (Fig. 1.3b). The apical cell mem- rier preventing, for example, rose bengal dye and
brane of this cell layer exhibits folds or ridges pathogen penetrance [16–18]. Apical cells exhibit
termed microplicae [15] (Fig.  1.3c). The tight junctions or zonula occludens that consti-
membrane-­associated mucins emanate from the tutes the paracellular barrier for fluid loss and
tips of these microplicae to form the glycocalyx external molecules and microorganisms.
(Fig. 1.3d). The glycocalyx not only contributes The basal cell layer is specialized and pro-
to formation of an optimal tear film, but it pro- vides tight anchorage of the epithelium to the
1  Corneal Anatomy 5

a b

c d

Fig. 1.3 (a) The corneal epithelium is comprised of 5–7 cell layer exhibits folds or ridges termed microplicae
cell layers. (b) The most apical cell layer of the corneal (MP). (d) The membrane-­associated mucins emanate from
epithelium expresses membrane-­associated mucins, such the tips of these microplicae to form the glycocalyx.
as MUC16 (red staining), that form the surface glycocalyx (Modified from Argueso et al. [18] and Gipson et al. [2])
(cell nuclei in blue). (c) The apical cell membrane of this

underlying stroma. To provide these anchoring neal epithelium its cohesion and resistance to abra-
functions, the basal cells secrete the basement sion. The corneal epithelium also has gap junctions
membrane, composed of type IV collagen, type for intercellular communication purposes, and
VII collagen, laminin, fibronectin, nidogen, and they are found throughout the epithelium but are
perlecan, among other proteins [19–22]. Further, of different character apical to basal [26].
they assemble specialized anchoring junctions, The cornea is the most densely innervated tis-
hemidesmosomes, which link the cells’ cytoskel- sue in the body [27, 28] (Fig. 1.4). Corneal inner-
eton by way of alpha6 beta4 integrin to the type vation is not only responsible for corneal
VII collagen anchoring fibrils that penetrate into, sensitivity but also provides trophic factors that
and intersperse between, collagen fibrils in the are essentials for wound healing and corneal
anterior stroma [23]. regeneration [29–31]. Intraepithelial nerve termi-
Corneal epithelial cells express the cytokeratins nals innervate all corneal epithelial layers [32].
3 and 12, and all are connected to their neighbors Nerve terminals are branches from a continuous
via intercellular desmosomes [24]. The cytokera- subbasal nerve plexus, which originate from the
tin intermediate filaments attach to the desmo- anastomosis of epithelial leashes in the central
somes’ placque [25] and, along with adherens and paracentral cornea. Epithelial leashes are
junctions and hemidesmosomes, confer to the cor- subbasal nerve fibers branched from stromal
6 M. Gonzalez-Andrades et al.

Long ciliary
nerves (Va)

Corneal Limbus
INT epithelium

SNP SNP
Epithelial
leashes
Stromal
bundle

Corneal
stroma

Fig. 1.4  The cornea is the most densely innervated tissue in the central and paracentral cornea. Epithelial leashes
in the body. Intraepithelial nerve terminals (INT) inner- are subbasal nerve fibers branched from stromal nerves
vate all corneal epithelial layers. Nerve terminals are that penetrate into the cornea from the corneoscleral
branches from a continuous subbasal nerve plexus (SNP), limbus
which originate from the anastomosis of epithelial leashes

nerves that penetrate into the cornea from the hypothesis has been put forward by Connon et al.
corneoscleral limbus [32]. which suggests that differences in substrate stiff-
The corneal epithelium is renewed within ness beneath the epithelial cells (a centripetal
periods lasting between 7 and 14 days (for review stiffness gradient) can drive both the inward
see [33]). The basal cells of the corneal epithe- migration and differentiation of the limbal stem
lium possess limited mitotic activity, presenting a cells via mechanotransductional forces [43].
density of 6000  cells/mm2 [34, 35]. These cells Corneal epithelial cells sensing these changes
derive from limbal epithelial stem cells (LESC) have been shown to undergo differentiation in
that reside in the corneoscleral limbus in an response to substrates of differing stiffness [44].
undifferentiated state [36]. The capacity of LESC The limbus is a specialized region that acts as a
for self-renewal is limitless and their mitotic 1.5 mm-wide transition zone between the cornea
activity is low. These cells may represent less and sclera [45]. It is densely vascularized, inner-
than 10% of the total limbal basal cell population vated, and protected from UV light damage by
[36, 37]. Different markers such as ABCG2, melanin pigmentation [39]. The limbus is com-
ΔNp63α, N-cadherin, cytokeratin 19, vimentin, posed of an epithelium (which lies between the
KGF-R, and ABCB5 have been used to identify corneal epithelium and the conjunctival epithe-
putative LESC [36, 38–41]. The LESC divide lium, basal cells of which house a population of
and give rise to transient amplifying cells capable limbal epithelial stem cells – LESC – in an undif-
of a finite number of cell divisions [22]. Finally, ferentiated state [36]) and a stroma that contains
the cells migrate centripetally to differentiate into vessels [46] and different types of cells such as
suprabasal corneal epithelial cells, which give melanocytes [47] and mesenchymal stem cells
rise to terminally differentiated cells of the cor- [48]. Two specialized structures in the basal epi-
neal epithelium surface [42]. The mechanism thelium of the corneoscleral limbus make
underpinning this process of corneal epithelial epithelial-­stromal interactions possible through a
homeostasis is poorly defined. Recently a new basement membrane along the limbus. The pali-
1  Corneal Anatomy 7

sades of Vogt are papilla-like structures located in membrane, provide an optimum microenviron-
the subepithelial connective tissue that provide a ment for limbal basal cells [22, 36, 39].
protective environment for LESC [49]. Among
these palisades are limbal epithelial crypts, finger-­
shaped projections where the LESC are located. 1.3 The Corneal Stroma
These structures protect the cells from shear stress
and supply them with nutrients from the neighbor- The corneal stroma is a specialized connective
ing blood vessels [22, 40]. At the palisades of Vogt, tissue layer, between the corneal epithelium and
LESC might interact with corneal stromal stem endothelium. It constitutes 90% of the human
cells, i.e., neural crest-derived mesenchymal stem corneal thickness, representing the structural axis
cells (MSC), that maintain the LESC niche and of the cornea. It is flanked by two distinguishable
can differentiate into functional keratocytes, which areas: the Bowman layer on the anterior side and
in turn might regenerate the normal corneal stroma Descemet’s membrane on the posterior side
[48, 50–52]. The limbal stroma is not transparent, (Fig.  1.2). The corneal stroma is considered an
and its collagen fibrils are not as orderly arranged immunologically privileged structure because of
as the adjacent corneal stroma, and it begins to the absence of blood and lymph vessels [54].
resemble the scleral stroma in both the arrange- This feature maintains optimal transparency and
ment and the size of its collagen fibers [53]. Cell- increases the chances of survival of the donor
matrix interactions, partially mediated by local cornea when it is transplanted into the host.
variations in the composition of the extracellular The corneal stroma is composed of extracellu-
matrix and the heterogeneity of the basement lar matrix and its embedded cells (Fig. 1.5a). The

Fig. 1.5 (a) The corneal


a
stroma is populated by
keratocytes embedded
between the collagen
lamellae. (b)
Keratocytes, which are
quiescent in adults, have
a flat dendritic
morphology which gives
them a large area. (c)
Collagen fibers are
grouped together to form
lamellae that are
oriented mainly parallel
to the tissue surface at
all depths
b c
8 M. Gonzalez-Andrades et al.

extracellular matrix synthesized by these cells is transparency and minimize fluctuations in the
mainly formed by collagen, primarily type I col- refractive index, reducing light scatter by express-
lagen, although type V and VI collagen are also ing a series of water-soluble proteins called crys-
present. Type I collagen accounts for 68% of the tallins [70]. These proteins control optical changes
dry weight of the cornea. Other types of colla- and reduce the dispersion of light that occurs in the
gen fibers present in much lower amounts in the cytoplasm of both cell types [71].
corneal stroma are types III, XII, and XIV [22, Other cell types resident in the corneal stroma that
55]. Collagen fibers are grouped together to form respond to injury and pathogens are bone marrow-
lamellae that are oriented mainly parallel to the derived cells. These cells are involved in the anti-
tissue surface at all depths [56] (Fig.  1.5c). This gen presentation and immune response and include
pattern of orientation is an important factor affect- dendritic cells and macrophages [72–76]. Dendritic
ing the transparency and strength of the cornea cells are also considered resident cells in the cor-
[57–62]. In addition, the stroma’s unique unifor- neal epithelium, where they received the name of
mity of collagen fibril diameter, controlled by the Langerhans cells (similar to the dendritic cells found
precisely regulated amounts of type I and type in the epidermis of the skin) or intraepithelial cor-
V collagen, contributes to transparency. A third neal dendritic cells [76, 77]. Immune cells can influ-
factor contributing to stromal transparency is the ence the avascularity status of the cornea, which is
amorphous ground substance or matrix surround- based on the balance of anti-angiogenic factors that
ing the collagen fibrils. This material is primarily ­counterbalance pro-angiogenic/lymphangiogenic
dermatan and keratan sulfate proteoglycans, which factors upregulated during wound healing [78, 79].
regulates interfibrillar distance between the colla- The two specialized acellular areas that border
gen fibrils thus contributing to the regular spacing the corneal stroma are Bowman’s layer and
of the fibrils that is required for transparency [52, Descemet’s membrane. Bowman’s layer is
63]. Water makes up 78% of the corneal volume. 8–14 μm in thickness and consists primarily of col-
The stroma is populated by keratocytes embed- lagen fibers type I, III, and V produced by epithelial
ded between the collagen lamellae, occupying cells and intertwined with each other apparently
approximately 3% of the stromal volume [52] without any architectural pattern [46, 80].
(Fig.  1.5b). Keratocytes, which are quiescent in Descemet’s membrane is a uniquely thick base-
adults, have a flat dendritic morphology which ment membrane produced by the corneal endothe-
gives them a large area; they make contact with each lium. It is composed of type IV, type VIII, and type
other through gap junctions [64]. The keratocyte XII collagen, perlecan, nidogen, netrin, fibronectin,
density changes according to the stromal region, and laminin, among other proteins [81, 82]. This
decreasing in number from anterior to posterior membrane is composed of an anterior banded layer
[65, 66]. The functional activities of keratocytes developed in the fetus and a posterior nonbanded
include the synthesis of extracellular matrix and layer that is continuously synthetized by the endo-
stromal remodeling; these cells produce collagen thelial cells during adult life [83, 84].
lamellae and proteoglycans including keratocan,
decorin, lumican, and mimecan [67]. Keratocytes
express mesenchymal and hematopoietic markers 1.4 The Corneal Endothelium
and can be identified by expression of the follow-
ing proteins: keratocan, aldehyde dehydrogenase, The corneal endothelium is a simple low cuboi-
crystallins, CD133, CD34, and cytokeratin 3 [67, dal epithelium [81, 85]. These cells show a hex-
68]. The expression of these markers is affected in agonal morphology on their apical surface that
response to injury, which induces keratocytes to abuts the aqueous humor (Fig. 1.6), whereas their
adopt a fibroblast or myofibroblast scar-forming basal surface, in contact with Descemet’s mem-
phenotype that is reversible depending on the envi- brane, is irregular [86–88]. The human corneal
ronment [68, 69]. Keratocytes, as well as epithe- endothelium shows an increased cell density in
lial cells, have developed a mechanism to increase the paracentral and peripheral areas compared
1  Corneal Anatomy 9

location, cell-cell contacts, and lack of growth


factor stimulation [98, 99]. Recent findings sug-
gest that endothelial progenitor cells exist in the
corneal periphery and can differentiate into func-
tional corneal endothelial cells [100–104].

Compliance with Ethical Requirements  Miguel


Gonzalez-Andrades, Pablo Argüeso, and Ilene
K.  Gipson declare that they have no conflict of
interest. No human or animal studies were car-
ried out by the authors for this article.

Fig. 1.6  The corneal endothelium is a simple low cuboi- References


dal epithelium, whose cells show a hexagonal morphol-
ogy on their apical surface that abuts the aqueous humor 1. Gipson IK, Joyce NC.  Anatomy and cell biology of
(in vivo confocal image of the corneal endothelium of a the cornea, superficial limbus and conjunctiva. In:
healthy person). (Courtesy of Dr. Caro-Magdaleno) Albert D, Miller J, Azar D, Blodi B, editors. Albert &
Jakobiec’s principles and practice of ophthalmology.
Philadelphia/Edinburgh: Saunders/Elsevier; 2008.
with the center [89]. Connections between the p. 423–40.
2. Gipson IK, Joyce NJ, Zieske JD.  The anatomy and
different cells are based on adherens junctions, cell biology of the human cornea, limbus, conjunc-
tight junctions, and desmosomes [46, 90]. tiva, and adnexa. In: Foster CS, Azar DT, Dohlman
The functional activity of corneal endothelial CH, editors. Smolin and Thoft’s the cornea : scien-
cells is to enable the exchange of ions and fluid tific foundations and clinical practice. Philadelphia:
Lippincott Williams & Wilkins; 2005. p. 1–35.
through active transport between the corneal 3. Lwigale PY.  Corneal development: different cells
stroma and the anterior chamber, where the aque- from a common progenitor. Prog Mol Biol Transl Sci.
ous humor is located [91]. This functional activity 2015;134:43–59.
ensures corneal transparency and avoids stromal 4. Collomb E, Yang Y, Foriel S, Cadau S, Pearton DJ,
Dhouailly D. The corneal epithelium and lens develop
edema [85, 92]. One of the main ionic pumps independently from a common pool of precursors.
expressed on the lateral membrane of the endo- Dev Dyn. 2013;242:401–13.
thelial cells is Na(+)/K(+)-ATPase, which is one 5. Creuzet S, Vincent C, Couly G. Neural crest deriva-
of the markers used to differentiate them from tives in ocular and periocular structures. Int J Dev
Biol. 2005;49:161–71.
other corneal cell types. Another marker com- 6. Lwigale PY, Cressy PA, Bronner-Fraser M.  Corneal
monly used for this purpose is ZO-1, even when keratocytes retain neural crest progenitor cell proper-
it is also present in the corneal epithelium [93]. ties. Dev Biol. 2005;288:284–93.
Recently, some researchers have proposed new 7. Greene CA, Green CR, Dickinson ME, Johnson V,
Sherwin T.  Keratocytes are induced to produce col-
markers that can be used to purify this cell popu- lagen type II: a new strategy for in vivo corneal matrix
lation, such as CD98, CD166, CD340, CLRN1, regeneration. Exp Cell Res. 2016;347:241–9.
MRGPRX3, HTR1D, GRIP1, and ZP4 [93, 94]. 8. Rufer F, Schroder A, Erb C.  White-to-white corneal
The mitotic activity of endothelial cells is min- diameter: normal values in healthy humans obtained
with the Orbscan II topography system. Cornea.
imal, with an annual loss of 0.6% approximately. 2005;24:259–61.
This cell loss is compensated by polymegathism 9. Khng C, Osher RH.  Evaluation of the relation-
and polymorphism, by which the cells increase in ship between corneal diameter and lens diameter. J
size and change shape, respectively [95–97]. The Cataract Refract Surg. 2008;34:475–9.
10. Doughty MJ, Zaman ML.  Human corneal thickness
regenerative capacity of the corneal endothelium and its impact on intraocular pressure measures: a
is very limited, and proliferation is reduced by review and meta-analysis approach. Surv Ophthalmol.
factors associated with age, central topographical 2000;44:367–408.
10 M. Gonzalez-Andrades et al.

11. Randleman JB, Lynn MJ, Perez-Straziota CE, Weissman nexins in the human corneal epithelium. Invest
HM, Kim SW.  Comparison of central and peripheral Ophthalmol Vis Sci. 2005;46:1957–65.
corneal thickness measurements with scanning-slit, 27. Ferrari G, Hajrasouliha AR, Sadrai Z, Ueno H,

Scheimpflug and Fourier-domain ocular coherence Chauhan SK, Dana R. Nerves and neovessels inhibit
tomography. Br J Ophthalmol. 2015;99:1176–81. each other in the cornea. Invest Ophthalmol Vis Sci.
12. Doughty MJ, Jonuscheit S. An assessment of regional 2013;54:813–20.
differences in corneal thickness in normal human 28. Bonini S, Rama P, Olzi D, Lambiase A. Neurotrophic
eyes, using the Orbscan II or ultrasound pachymetry. keratitis. Eye (Lond). 2003;17:989–95.
Optometry. 2007;78:181–90. 29. Muller LJ, Marfurt CF, Kruse F, Tervo TM. Corneal
13. Argueso P, Tisdale A, Mandel U, Letko E, Foster nerves: structure, contents and function. Exp Eye Res.
CS, Gipson IK. The cell-layer- and cell-type-specific 2003;76:521–42.
distribution of GalNAc-transferases in the ocular sur- 30. Shaheen BS, Bakir M, Jain S. Corneal nerves in health
face epithelia is altered during keratinization. Invest and disease. Surv Ophthalmol. 2014;59:263–85.
Ophthalmol Vis Sci. 2003;44:86–92. 31. Sacchetti M, Lambiase A.  Neurotrophic factors

14. Gipson IK. Distribution of mucins at the ocular sur- and corneal nerve regeneration. Neural Regen Res.
face. Exp Eye Res. 2004;78:379–88. 2017;12:1220–4.
15. Nichols B, Dawson CR, Togni B. Surface features of 32. Marfurt CF, Cox J, Deek S, Dvorscak L.  Anatomy
the conjunctiva and cornea. Invest Ophthalmol Vis of the human corneal innervation. Exp Eye Res.
Sci. 1983;24:570–6. 2010;90:478–92.
16. Gipson IK, Argueso P.  Role of mucins in the func- 33. West JD, Dora NJ, Collinson JM. Evaluating alterna-
tion of the corneal and conjunctival epithelia. Int Rev tive stem cell hypotheses for adult corneal epithelial
Cytol. 2003;231:1–49. maintenance. World J Stem Cells. 2015;7:281–99.
17. Gipson IK, Spurr-Michaud S, Tisdale A, Menon
34. Lavker RM, Dong G, Cheng SZ, Kudoh K, Cotsarelis
BB.  Comparison of the transmembrane mucins G, Sun TT. Relative proliferative rates of limbal and
MUC1 and MUC16  in epithelial barrier function. corneal epithelia. Implications of corneal epithelial
PLoS One. 2014;9:e100393. migration, circadian rhythm, and suprabasally located
18. Argueso P, Spurr-Michaud S, Russo CL, Tisdale
DNA-synthesizing keratinocytes. Invest Ophthalmol
A, Gipson IK.  MUC16 mucin is expressed by the Vis Sci. 1991;32:1864–75.
human ocular surface epithelia and carries the H185 35. Yanoff M, Fine BS.  Ocular pathology. 5th ed.

carbohydrate epitope. Invest Ophthalmol Vis Sci. Philadelphia: Mosby; 2002. xxi, 701 p.
2003;44:2487–95. 36. Schlotzer-Schrehardt U, Kruse FE. Identification and
19. Shafiq MA, Gemeinhart RA, Yue BY, Djalilian
characterization of limbal stem cells. Exp Eye Res.
AR. Decellularized human cornea for reconstructing 2005;81:247–64.
the corneal epithelium and anterior stroma. Tissue 37. Menzel-Severing J, Kruse FE, Schlotzer-Schrehardt
Eng Part C Methods. 2012;18:340–8. U.  Stem cell-based therapy for corneal epithelial
20. Tuori A, Uusitalo H, Burgeson RE, Terttunen J,
reconstruction: present and future. Can J Ophthalmol.
Virtanen I.  The immunohistochemical composition 2013;48:13–21.
of the human corneal basement membrane. Cornea. 38. Takacs L, Toth E, Berta A, Vereb G. Stem cells of the
1996;15:286–94. adult cornea: from cytometric markers to therapeutic
21.
Resch MD, Schlotzer-Schrehardt U, Hofmann-­ applications. Cytometry A. 2009;75:54–66.
Rummelt C, Kruse FE, Seitz B. Alterations of epithelial 39. Schlotzer-Schrehardt U, Dietrich T, Saito K, et  al.
adhesion molecules and basement membrane com- Characterization of extracellular matrix components
ponents in lattice corneal dystrophy (LCD). Graefes in the limbal epithelial stem cell compartment. Exp
Arch Clin Exp Ophthalmol. 2009;247:1081–8. Eye Res. 2007;85:845–60.
22. Castro-Munozledo F.  Review: corneal epithelial
40. Notara M, Alatza A, Gilfillan J, et al. In sickness and
stem cells, their niche and wound healing. Mol Vis. in health: corneal epithelial stem cell biology, pathol-
2013;19:1600–13. ogy and therapy. Exp Eye Res. 2010;90:188–95.
23. Gipson IK. Adhesive mechanisms of the corneal epi- 41. Ksander BR, Kolovou PE, Wilson BJ, et al. ABCB5 is
thelium. Acta Ophthalmol Suppl. 1992;202:13–7. a limbal stem cell gene required for corneal develop-
24. Gonzalez-Andrades M, Garzon I, Gascon MI, et  al. ment and repair. Nature. 2014;511:353–7.
Sequential development of intercellular junctions in 42. Ramos T, Scott D, Ahmad S.  An update on ocular
bioengineered human corneas. J Tissue Eng Regen surface epithelial stem cells: cornea and conjunctiva.
Med. 2009;3:442–9. Stem Cells Int. 2015;2015:601731.
25. Colabelli Gisoldi RA, Pocobelli A, Villani CM,
43. Foster JW, Jones RR, Bippes CA, Gouveia RM,

Amato D, Pellegrini G.  Evaluation of molecular Connon CJ. Differential nuclear expression of Yap in
markers in corneal regeneration by means of autolo- basal epithelial cells across the cornea and substrates
gous cultures of limbal cells and keratoplasty. Cornea. of differing stiffness. Exp Eye Res. 2014;127:37–41.
2010;29:715–22. 44. Jones RR, Hamley IW, Connon CJ. Ex vivo expansion
26. Shurman DL, Glazewski L, Gumpert A, Zieske JD, of limbal stem cells is affected by substrate properties.
Richard G.  In vivo and in  vitro expression of con- Stem Cell Res. 2012;8:403–9.
1  Corneal Anatomy 11

45. Li W, Hayashida Y, Chen YT, Tseng SC. Niche regu- dimensional electron tomography of bovine cornea.
lation of corneal epithelial stem cells at the limbus. Structure. 2010;18:239–45.
Cell Res. 2007;17:26–36. 64. Leibowitz HM, Waring GO. Corneal disorders : clini-
46. Hogan MJ, Alvarado JA, Weddell JE.  Histology of cal diagnosis and management. 2nd ed. Philadelphia:
the human eye. 1st ed. Philadelphia: W.C.  Saunders Saunders; 1998. xvi, 1172 p.
Company; 1971. 65. Petroll WM, Boettcher K, Barry P, Cavanagh HD,
47. Dziasko MA, Tuft SJ, Daniels JT.  Limbal melano- Jester JV.  Quantitative assessment of anteroposte-
cytes support limbal epithelial stem cells in 2D and rior keratocyte density in the normal rabbit cornea.
3D microenvironments. Exp Eye Res. 2015;138:70–9. Cornea. 1995;14:3–9.
48. Funderburgh JL, Funderburgh ML, Du Y. Stem cells 66. Patel S, McLaren J, Hodge D, Bourne W.  Normal
in the limbal stroma. Ocul Surf. 2016;14:113–20. human keratocyte density and corneal thick-
49. Notara M, Daniels JT.  Biological principals and
ness measurement by using confocal micros-
clinical potentials of limbal epithelial stem cells. Cell copy in  vivo. Invest Ophthalmol Vis Sci. 2001;42:
Tissue Res. 2008;331:135–43. 333–9.
50. Du Y, Funderburgh ML, Mann MM, SundarRaj N, 67.
Hashmani K, Branch MJ, Sidney LE, et  al.
Funderburgh JL.  Multipotent stem cells in human Characterization of corneal stromal stem cells with
corneal stroma. Stem Cells. 2005;23:1266–75. the potential for epithelial transdifferentiation. Stem
51. Funderburgh ML, Du Y, Mann MM, SundarRaj N, Cell Res Ther. 2013;4:75.
Funderburgh JL.  PAX6 expression identifies pro- 68. Branch MJ, Hashmani K, Dhillon P, Jones DR, Dua
genitor cells for corneal keratocytes. FASEB J: Off HS, Hopkinson A.  Mesenchymal stem cells in the
Publication Fed Am Soc Exp Biol. 2005;19:1371–3. human corneal limbal stroma. Invest Ophthalmol Vis
52. Pinnamaneni N, Funderburgh JL.  Concise review:
Sci. 2012;53:5109–16.
stem cells in the corneal stroma. Stem Cells. 69. Foster JW, Gouveia RM, Connon CJ.  Low-glucose
2012;30:1059–63. enhances keratocyte-characteristic phenotype from
53. Chen Z, de Paiva CS, Luo L, Kretzer FL, Pflugfelder corneal stromal cells in serum-free conditions. Sci
SC, Li DQ.  Characterization of putative stem cell Rep. 2015;5:10839.
phenotype in human limbal epithelia. Stem Cells. 70. Chen Y, Jester JV, Anderson DM, et al. Corneal haze
2004;22:355–66. phenotype in Aldh3a1-null mice: in  vivo confocal
54. Ma DH, Chen HC, Lai JY, et  al. Matrix revolution: microscopy and tissue imaging mass spectrometry.
molecular mechanism for inflammatory corneal neo- Chem Biol Interact. 2017;276:9–14.
vascularization and restoration of corneal avascular- 71. Jester JV.  Corneal crystallins and the develop-

ity by epithelial stem cell transplantation. Ocul Surf. ment of cellular transparency. Semin Cell Dev Biol.
2009;7:128–44. 2008;19:82–93.
55. Meek KM, Fullwood NJ.  Corneal and scleral
72. Yamagami S, Usui T, Amano S, Ebihara N.  Bone
collagens--a microscopist’s perspective. Micron. marrow-derived cells in mouse and human cornea.
2001;32:261–72. Cornea. 2005;24:S71–4.
56. Abass A, Hayes S, White N, Sorensen T, Meek
73. Nakamura T, Ishikawa F, Sonoda KH, et  al.

KM. Transverse depth-dependent changes in corneal Characterization and distribution of bone marrow-­
collagen lamellar orientation and distribution. J R Soc derived cells in mouse cornea. Invest Ophthalmol Vis
Interface. 2015;12:20140717. Sci. 2005;46:497–503.
57. Ruberti JW, Zieske JD. Prelude to corneal tissue engi- 74.
Takayama T, Kondo T, Kobayashi M, et  al.
neering  - gaining control of collagen organization. Characteristic morphology and distribution of
Prog Retin Eye Res. 2008;27:549–77. bone marrow derived cells in the cornea. Anat Rec
58. Quantock AJ, Young RD. Development of the corneal (Hoboken). 2009;292:756–63.
stroma, and the collagen-proteoglycan associations 75. Forrester JV, Xu H, Kuffova L, Dick AD, McMenamin
that help define its structure and function. Dev Dyn. PG. Dendritic cell physiology and function in the eye.
2008;237:2607–21. Immunol Rev. 2010;234:282–304.
59. Maurice DM.  The structure and transparency of the 76. Saban DR. The chemokine receptor CCR7 expressed
cornea. J Physiol. 1957;136:263–86. by dendritic cells: a key player in corneal and ocular
60. Goldman JN, Benedek GB. The relationship between surface inflammation. Ocul Surf. 2014;12:87–99.
morphology and transparency in the nonswelling 77. Hattori T, Takahashi H, Dana R.  Novel insights

corneal stroma of the shark. Investig Ophthalmol. into the immunoregulatory function and localiza-
1967;6:574–600. tion of dendritic cells. Cornea. 2016;35(Suppl 1):
61. Benedek GB. Theory of transparency of the eye. Appl S49–54.
Opt. 1971;10:459–73. 78. Ellenberg D, Azar DT, Hallak JA, et al. Novel aspects
62. Meek KM, Knupp C. Corneal structure and transpar- of corneal angiogenic and lymphangiogenic privilege.
ency. Prog Retin Eye Res. 2015;49:1–16. Prog Retin Eye Res. 2010;29:208–48.
63. Lewis PN, Pinali C, Young RD, Meek KM, Quantock 79. Zhong W, Montana M, Santosa SM, et al. Angiogenesis
AJ, Knupp C.  Structural interactions between col- and lymphangiogenesis in corneal transplantation-a
lagen and proteoglycans are elucidated by three-­ review. Surv Ophthalmol. 2018;63:453–79.
12 M. Gonzalez-Andrades et al.

80. Wilson SE, Hong JW. Bowman’s layer structure and 93. Okumura N, Hirano H, Numata R, et  al. Cell sur-
function: critical or dispensable to corneal function? face markers of functional phenotypic corneal endo-
A hypothesis. Cornea. 2000;19:417–20. thelial cells. Invest Ophthalmol Vis Sci. 2014;55:
81. Beuerman RW, Pedroza L.  Ultrastructure of the
7610–8.
human cornea. Microsc Res Tech. 1996;33:320–35. 94. Yoshihara M, Ohmiya H, Hara S, et  al. Discovery
82. Kabosova A, Azar DT, Bannikov GA, et  al.
of molecular markers to discriminate corneal
Compositional differences between infant and endothelial cells in the human body. PLoS One.
adult human corneal basement membranes. Invest 2015;10:e0117581.
Ophthalmol Vis Sci. 2007;48:4989–99. 95. Lass JH, Gal RL, Ruedy KJ, et  al. An evaluation
83. Johnson DH, Bourne WM, Campbell RJ.  The ultra- of image quality and accuracy of eye bank mea-
structure of Descemet’s membrane. I.  Changes surement of donor cornea endothelial cell den-
with age in normal corneas. Arch Ophthalmol. sity in the Specular Microscopy Ancillary Study.
1982;100:1942–7. Ophthalmology. 2005;112:431–40.
84. Weller JM, Schlotzer-Schrehardt U, Kruse FE,
96. Krachmer JH, Mannis MJ, Holland EJ. Cornea. 3rd
Tourtas T.  Splitting of the recipient’s descemet ed. Philadelphia: Elsevier/Mosby; 2011. 1–2 p.
membrane in descemet membrane endothelial 97. Schroeter J, Rieck P.  Endothelial evaluation in the
keratoplasty-­ultrastructure and clinical relevance. Am cornea bank. Dev Ophthalmol. 2009;43:47–62.
J Ophthalmol. 2016;172:1–6. 98. Joyce NC. Proliferative capacity of corneal endothe-
85. Srinivas SP.  Dynamic regulation of barrier integ-
lial cells. Exp Eye Res. 2012;95:16–23.
rity of the corneal endothelium. Optom Vis Sci. 99. Peh GS, Beuerman RW, Colman A, Tan DT, Mehta
2010;87:E239–54. JS.  Human corneal endothelial cell expansion for
86. He Z, Forest F, Gain P, et al. 3D map of the human corneal endothelium transplantation: an overview.
corneal endothelial cell. Sci Rep. 2016;6:29047. Transplantation. 2011;91:811–9.
87. Worner CH, Olguin A, Ruiz-Garcia JL, Garzon-­
100. Hara S, Hayashi R, Soma T, et al. Identification and
Jimenez N. Cell pattern in adult human corneal endo- potential application of human corneal endothelial
thelium. PLoS One. 2011;6:e19483. progenitor cells. Stem Cells Dev. 2014;23:2190–201.
88. Harrison TA, He Z, Boggs K, Thuret G, Liu HX, 101. Espana EM, Sun M, Birk DE. Existence of corneal
Defoe DM. Corneal endothelial cells possess an elab- endothelial slow-cycling cells. Invest Ophthalmol
orate multipolar shape to maximize the basolateral to Vis Sci. 2015;56:3827–37.
apical membrane area. Mol Vis. 2016;22:31–9. 102. Dirisamer M, Dapena I, Ham L, et  al. Patterns of
89.
Amann J, Holley GP, Lee SB, Edelhauser corneal endothelialization and corneal clearance
HF. Increased endothelial cell density in the paracen- after descemet membrane endothelial keratoplasty
tral and peripheral regions of the human cornea. Am J for fuchs endothelial dystrophy. Am J Ophthalmol.
Ophthalmol. 2003;135:584–90. 2011;152:543–555 e541.
90. Barry PA, Petroll WM, Andrews PM, Cavanagh HD, 103. Borkar DS, Veldman P, Colby KA.  Treatment of
Jester JV. The spatial organization of corneal endothe- Fuchs endothelial dystrophy by Descemet strip-
lial cytoskeletal proteins and their relationship to the ping without endothelial keratoplasty. Cornea.
apical junctional complex. Invest Ophthalmol Vis Sci. 2016;35:1267–73.
1995;36:1115–24. 104. Iovieno A, Neri A, Soldani AM, Adani C, Fontana
91.
Alaminos M, Gonzalez-Andrades M, Munoz-­ L.  Descemetorhexis without graft placement for
Avila JI, Garzon I, Sanchez-Quevedo MC, Campos the treatment of Fuchs endothelial dystrophy: pre-
A. Volumetric and ionic regulation during the in vitro liminary results and review of the literature. Cornea.
development of a corneal endothelial barrier. Exp Eye 2017;36:637–41.
Res. 2008;86:758–69.
92. Mergler S, Pleyer U. The human corneal endothelium:
new insights into electrophysiology and ion channels.
Prog Retin Eye Res. 2007;26:359–78.
Corneal Healing
2
Veronica Vargas, Francisco Arnalich-Montiel,
and Jorge L. Alió del Barrio

2.1 Introduction and basal epithelial cells. These cells have junc-
tional complexes between each other to prevent
The cornea covers the anterior 1/6 of the total the passage of foreign agents. They have an aver-
surface of the globe and, due to its refractive age lifespan of 7–10  days. Apoptosis of these
power and transparency, is the most important cells is induced by ultraviolet radiation, hypoxia,
optical element of the eye. and mechanical friction [2]. Because it is the
It has three major cell types, epithelial cells, outer layer of the cornea, it can be easily injured.
stromal keratocytes, and endothelial cells, and
consists of five layers, epithelium, Bowman’s
layer, stroma, Descemet membrane, and endothe- 2.1.2 Bowman’s Layer
lium [1], and a surgical layer called Dua’s layer.
We are going to describe briefly each one of them. It is an acellular layer composed by collagen
fibrils types I and III, with a mean thickness of
12 μm; it does not regenerate after injury [3].
2.1.1 Corneal Epithelium
2.1.3 Stroma
It is the first layer of the cornea and consists of a
nonkeratinized stratified squamous layer, and it It constitutes 90% of the corneal thickness, and it
is richly innervated. It has 4–6 cell layers, a is composed by collagen types I, III, V, VI, XII,
thickness of 50um approximately, and three dif- and XIV; the uniform arrangement of the collagen
ferent kinds of cells, squamous cells, wing cells, fibers is essential for the corneal transparency. Its
primary cells are the keratocytes which are usually
quiescent in the normal corneal; but in response to
stromal injury, they get activated and undergo
V. Vargas transdifferentiation into myofibroblasts [3].
Vissum Instituto Oftalmologico de Alicante,
Alicante, Spain
F. Arnalich-Montiel
2.1.4 Descemet Membrane
Vissum Corporation, Madrid, Spain
It is the basement membrane of the endothelium.
J. L. Alió del Barrio (*)
University Miguel Hernandez, Vissum-Instituto It has a thickness of 10 μm, and it is mainly com-
Oftalmologico de Alicante, Alicante, Spain posed of collagen types IV and VIII [3].

© Springer Nature Switzerland AG 2019 13


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_2
14 V. Vargas et al.

2.1.5 Endothelium 2.2.2 Integrins

It is a single layer of hexagonal cells. Young They are transmembrane receptors; their main
adults have a cell density of 3500  cells/mm2 function is to connect cells to the extracellular
approximately, and they do not regenerate. Its matrix, and they play an important role in the
main function is to regulate corneal hydration healing process. The α2β1 integrin, expressed in
through ion transport systems in order to main- the epithelial cells that migrate following a cor-
tain the corneal transparency [1, 3]. neal wound, interacts with type I collagen and
The corneal healing process is complex and upregulates the production and synthesis of
crucial because its transparency may be jeopar- matrix metalloproteinases [6].
dized during the process leading to scar forma-
tion and blindness.
2.2.3 Matrix Metalloproteinases
(MPPs)
2.2  ytokines and Growth
C
Factors Involved in Corneal These zinc-dependent endopeptidases influence
Healing cell migration by disengaging cell-cell and cell-­
matrix adhesion and extracellular matrix (ECM)
Corneal wound healing would not be possible degradation. They also modulate the activity of
without the interaction of cytokines, growth fac- growth factors and cytokine receptors [7].
tors, integrins, and proteases; so, first we are
going to describe the function of each one of
them: 2.2.4 Urokinase-Type Plasminogen
Activator (uPA)

2.2.1 Cytokines It is a serine protease synthesized by corneal epi-


thelial cells and corneal fibroblasts.
Interleukins 1α (IL-1α) and 6 initiate the pro- It mediates the release of inflammatory cyto-
cess of epithelial wound healing [2]; they are kines from fibroblasts, activates MMP-9 [8], and
proinflammatory cytokines, and its levels have a may stimulate cell migration [7].
correlation with the severity of the corneal
injury [4].
IL-6 stimulates epithelial migration and may 2.2.5 Growth Factors
influence the production of fibrotic material by
activated keratocytes [5]. Depending on the situation, growth factors can
IL-1α has several functions during corneal promote or inhibit the migration, proliferation,
healing, among them are [6]: and differentiation of cells [9]. They may act in
paracrine or autocrine mechanism [10].
• Mediation of keratocyte apoptosis
• Epithelial cell proliferation by inducing the
expression of keratinocyte growth factor 2.2.6 E
 pidermal Growth Factor
(KGF) and hepatocyte growth factor (HGF) (EGF)
on keratocytes
• Stimulation of epithelial wound healing in It is a polypeptide produced by macrophages and
conjunction with epidermal growth factor a variety of epithelial cells. It has a mitogen effect
(EGF), both accelerating epithelial wound on epithelial cells, and it is present in the tear film
closure [9]. After an epithelial wound, the EGF in tears
2  Corneal Healing 15

promotes cell proliferation [6]. It acts using both epithelial cell proliferation and inhibits the
paracrine and autocrine systems [10]. function and activation of inflammatory cells
in  vitro; the same study showed that topical
administration of HGF in an in  vivo model
2.2.7 T
 ransforming Growth Factor-β of corneal injury suppresses ocular inflamma-
(TGF-β) tion [12].

It is a polypeptide with three isoforms β1, β2,


and β3 and is produced by platelets, mononu- 2.2.9 F
 ibroblast Growth Factor
clear, and inflammatory cells. Its receptors are (FGF)
located at the stromal cells, limbus, and cen-
tral corneal epithelium. It acts through an It is produced by epithelial and corneal cells; has
autocrine mechanism [10] and has several a mitogenic effect in epithelial, stromal, and
functions [9]: endothelial cells; promotes cell migration; and
inhibits TGF-B1 expression in the stroma [9].
• Stimulates the production of proteoglycans,
fibronectin, and collagen.
• Limits inflammatory responses. 2.2.10 Keratinocyte Growth Factor 1
• Regulates (inhibits) epithelial cell (KGF-1)
proliferation.
• Induces proliferation and migration of stromal It is a member of the FGF family and functions in
fibroblasts. a paracrine fashion [7, 10]. It is mainly expressed
• Increases extracellular matrix synthesis. in the corneal stroma, and its receptor is expressed
• Antagonizes the mitogenic action of HGF, in limbal epithelial cells. It stimulates epithelial
EGF, and KGF in corneal epithelium [5]. cell proliferation [6].
• TGF-β1 and TGF-β2 trigger the transforma-
tion of fibroblasts to myofibroblasts.
• TGF-β3 mitigates scar formation after injury. 2.2.11 Platelet-Derived Growth
Factor (PDGF)

2.2.8 H
 epatocyte Growth Factor Produced by epithelial cells, it is made up of a
(HGF) combination of four polypeptide chains: A, B, C,
and D (PDFG-AA, PDFG-BB, PDFG-AB,
It is a glycoprotein produced by mesenchymal PDFG-CC, PDGF-DD). They regulate migration
cells (fibroblasts and macrophages) [11]; its and proliferation of keratocytes [7] through an
function is to regulate the normal proliferation of autocrine and paracrine mechanism [10].
central and peripheral epithelial cells [6] and epi-
thelial differentiation. It targets epithelial cells in
a paracrine manner [7, 10]; it is secreted by the 2.2.12 Nerve Growth Factor (NGF)
lacrimal gland, and most of its receptors (c-Met
[12]) are in the central epithelial cells [9, 11], It is a soluble protein that belongs to a family
although they can also be found in the endothe- of neurotrophic factors; it is normally released
lium and keratocytes [12]. in the aqueous humor and tear film [13]. During
It induces epithelial cell migration and prolif- corneal healing, it stimulates epithelial cell
eration and inhibits apoptosis [11]. migration [7], proliferation, innervation [9],
A study demonstrated that under an inflam- and keratocyte differentiation into myofibro-
matory environment, HGF promotes corneal blasts [7].
16 V. Vargas et al.

2.2.13 Insulin-Like Growth Factor the extracellular matrix (ECM) [2, 6], which
(IGF) is formed by fibronectin to allow cell migra-
tion [5].
This growth factor and its receptors are expressed
in the corneal epithelial cells and in keratocytes. 2. Epithelial Cell Migration
It regulates cell migration and differentiation; it The cells adjacent to the injury site flatten and
acts synergistically with substance P enhancing migrate toward the injured area to reestablish
cell proliferation and wound closure [7]. the integrity of the epithelium [6]. The ECM
facilitates this migration, and the limbal epi-
thelial stem cells are activated by growth fac-
2.3 Epithelial Wound Healing tors and cytokines.
The cell migration occurs at a constant rate
Of all the corneal layers, the epithelium is the of 0.05–0.06  mm/h and slows down as the
quickest to heal. The healing process consists of defect gets smaller [5]; it is initiated by EGF
four phases: (1) latent, (2) migration, (3) prolif- [7], and HGF and KGF facilitate it [9, 11].
eration, and (4) attachment (see Fig. 2.1). The expression of the latter growth factors is
One phase can begin before the previous one upregulated by PDGF [9].
is completed, but they always follow the sequence There is no mitotic activity in this phase [5].
of the phases [5].
3. Proliferation and Differentiation
1. Latent Phase During this phase, the epithelial cell density is
It is the time between the injury and the com- restored.
mencement of reepithelization [6]. Limbal epithelial stem cells proliferate dur-
Damaged cells undergo apoptosis and are ing wound healing and give rise to transient
shed into the tear film, intercellular junctions amplifying cells that differentiate and migrate
are dismantled, and the cell-substrate junc- to the center of the cornea [14]. There is mitotic
tions are replaced with weaker attachments activity in the cells outside the wound area [5],
[5]. Cytoskeletal proteins like talin, vinculin, and it is delayed within the wound area [2].
actin, and integrin are synthesized. The latter Daughter cells begin to differentiate into
dissociates from the hemidesmosomes and wing cells and then squamous cells, and the
desmosomes to distribute evenly on the cellu- intercellular junctions are reformed (the zonula
lar surface serving as an adherent molecule to occludens are the first ones in being reestab-

Migration Attachment
• Celular • Epithelial cell
reorganization • Cells at wound density is • Epithelial layer is
• Protein edge flattens restored adhered to the
synthesis and migrate underlying
substrate
Proliferation and
Lag phase
differentiation

Fig. 2.1  Epithelial healing phases


2  Corneal Healing 17

lished) restoring the epithelial barrier function. Upon proper healing, once that the EBM
A new basement membrane is formed [5]. This recovers its barrier function, the levels of TGF-β
stage is stimulated by HGF [11]. and PDGF fall, myofibroblasts undergo apoptosis
(promoted by HGF which induces degradation of
4. Attachment the ECM, specially fibronectin that is an anchor
In this phase, the epithelial layer strongly adheres for myofibroblasts [11]), keratocytes reoccupy
to the underlying substrate through hemidesmo- the anterior stroma, ECM is reabsorbed, and cor-
somes [5]. The permanent attachment of the neal transparency is restored [16].
hemidesmosomes depends if the basement
membrane remained intact at the time of wound-
ing [15]. If the anterior stroma is compromised, 2.4.1 Haze Formation
this phase can last a long period of time [5].
A prolonged persistence of myofibroblast after
healing and an excessive number of them corre-
2.4 Stromal Wound Healing late to corneal opacity [11]. Mature myofibro-
blasts express vimentin, desmin, and α-smooth
The first stromal healing response after injury is muscle actin (which directly correlates to corneal
the apoptosis of anterior keratocytes. This process wound contraction), which are opaque due to the
is triggered by cytokines like IL-1, Fas ligand, and decreased intracellular crystallin production [16,
necrosis tumor factor α [7, 16]. The keratocytes 19]. Myofibroblasts elaborate a disorganized
that are adjacent to the wounded area activate and ECM that interferes with the adequate distribu-
differentiate into fibroblasts due to an increased tion of collagen fibers [19]. TGF-β favors the
expression of actin [11]. Once they arrive at the deposition of excessive ECM [7]. This new ECM
wounded area, they repopulate that region that has accumulates aberrant proteins and forms scars
been depleted of keratocytes through apoptosis that may persist for a long time due to the slow
and differentiate into myofibroblasts [11]. These turnover of these proteins. These scars have com-
cells deposit a provisional ECM and generate con- ponents that are absent in a healthy stroma like
tractile forces to close the wound [7, 17]. collagen types III, VIII, XIV, and XVIII and
The keratocyte transformation to fibroblast embryonic fibronectin [7] (see Fig. 2.2).
and myofibroblast is triggered by TGF-β1, TGF-­ This type of abnormal healing of the cornea
β2, and PDGF [7, 11, 17, 18] can be seen after surface ablation surgeries like

Fig. 2.2  In stromal


healing process, after
injury, keratocytes
transform to fibroblasts, Keratocyte Fibroblast Myofibroblast
which transform to
myofibroblasts. If there
is a pathological
remodeling process
(large amount of Remodeling
myofibroblasts and a
disorganized ECM),
haze develops; if not, the
myofibroblasts undergo
apoptosis, and the Pathological Physiological
cornea keeps its
transparency

Haze Apoptosis
18 V. Vargas et al.

photorefractive keratectomy (PRK). During the persistent epithelial defect without corneal
first few weeks after PRK, a mild haze (early ulcers; and in stage 3, there is stromal edema,
haze) can be seen, which is secondary to corneal corneal ulcers, and perforation. The patients
fibroblasts that are opaque in the same manner included in this study had stages 2 and 3; they
as myofibroblast, but the ECM they produce is received one drop of NGF (50 μl) every 2 h for
more organized than the one produced by the 2  days and one drop six times a day until the
myofibroblasts. This type of haze usually disap- epithelial defect healed. Reported side effects
pears with time unlike the pathological haze like ocular pain, conjunctival hyperemia, and
(late haze) that is secondary to myofibroblasts, photophobia were secondary to the improve-
which appears 2–3 months [19] after the correc- ment of the corneal sensitivity. Just three
tion of (usually) high refractive errors and can patients that had a previous trigeminal nerve
last years [11]. resection and secondary corneal anesthesia
The epithelial basement membrane (EBM) required a retreatment [20].
plays an important role in this scenario. A defec- It has also been reported that one drop of NGF
tive EBM allows the penetration of TGF-β and (10 μL) helps restoring the corneal transparency
PDGF from the epithelium into the stroma induc- and thickness at the site of the surgical wound
ing myofibroblast generation that, as mentioned within 3 weeks following cataract (phacoemulsi-
before, produces a disorganized ECM that blocks fication) surgery [21].
the keratocytes that are in the posterior stroma Cases of autoimmune peripheral keratitis
from moving into the subepithelial area to con- have also been successfully treated with
tribute to the EBM regeneration [16]. NGF [22].
Mitomycin C inhibits the development of
myofibroblasts, and its use has reduced the inci-
dence of late haze after PRK [19]. 2.5.2 Insulin-Like Growth Factor
In LASIK, the creation of a flap preserves the
integrity of the EBM [19], so haze is not a risk Topical treatment with insulin-like growth factor-
because the stromal remodeling is not as signifi- 1 (IGF-1)-derived peptide (SSSR) combined
­
cant as it is in PRK [7], except at the edge of the with substance P-derived peptide (FGLM-amide)
flap where the EBM is damaged [19] and some triggered epithelial cell migration [23]. This
myofibroblasts may transiently appear [7]. combination has been used successfully to treat
epithelial defects secondary to neurotrophic kera-
titis in patients without limbal stem cell defi-
2.5 Clinical Applications ciency [24] and in patients with neurotrophic
of Growth Factors keratitis secondary to diabetes mellitus and neu-
rosurgical damage [23].
Some growth factors have been clinically used to
enhance corneal healing with good results, espe-
cially in cases of neurotrophic keratitis: 2.5.3 Epidermal Growth Factor

It has been used to treat epithelial defects of mul-


2.5.1 Nerve Growth Factor tiple etiologies; a study of nine eyes with epithe-
lial defects reported a complete reepithelization
A study showed that topically applied NGF in seven eyes after the use of a soft hydrogel con-
closed persistent epithelial defects in 45 eyes tact lens impregnated with EGF, while two eyes
with neurotrophic keratitis. There are three did not respond to the therapy and had a signifi-
clinical stages in neurotrophic keratitis: in stage cant ocular inflammation; no ocular toxicity was
1, there is a punctate corneal keratopathy with documented after the use of the EGF-impregnated
tear film abnormalities; in stage 2, there is a contact lens [25].
2  Corneal Healing 19

2.5.4 A
 utologous Eye Platelet-Rich 3. Penetrating trauma
Plasma (E-PRP) 4. Anterior segment infections or inflammations
including infectious or sterile endotheliitis
E-PRP is an autologous preparation of plasma 5. Pharmacologic toxicity leading to toxic

rich in platelets; the alpha granules in platelets anterior segment syndrome due to irrigating
contain several growth factors like EGF, solutions or viscoelastic devices incorrectly
PDGF-AB, IGF-1, and TGF-β. They modulate compounded or manufactured; ophthalmic
wound healing once they are released from acti- instrument contaminants; ocular medica-
vated platelets [26]. tions with incorrect pH, osmolality, or pre-
Alió et al. [27] used E-PRP for the treatment of servatives; contaminated water sources; and
dormant ulcers (neurotrophic keratopathy, her- intraocular lenses polishing, cleaning, and
petic keratopathy, and immunological ulcers) at a sterilizing compounds
dose of six times a day in addition to routine med-
ication. Clinical improvement was found in 92% The endothelium has a rather restricted
of the eyes; a complete resolution of the ulcer was response to injuries and is not thought to undergo
achieved in 50% of the cases. In the same clinical mitosis in vivo, so when cells are lost, the gap is
study, perforated eyes or high risk of perforation compensated by active sliding and enlargement
due to deep chronic corneal ulcers was treated of adjacent cells [30]. This results in the restora-
with amniotic membrane transplantation com- tion of the endothelial barrier at the expense of
bined with a clot of autologous E-PRP; 71% of reducing cell density. The process of resurfacing
the eyes had a complete resolution of the ulcer. the injured area is completed in three stages that
can take several weeks [31]. First, there is an ini-
tial coverage of the wound by migration of adja-
2.6 Physiological Mechanism cent endothelial cells, which forms a temporary
of Endothelial Regeneration incomplete barrier with reduced pump sites and
immature tight junctions. Following this, the bar-
Endothelial cell density slowly declines over rier (i.e., tight junctions) and subsequently the
time; however, corneal decompensation seldom number and quality of pump sites normalize, the
occurs. Besides normal aging or endothelial dis- endothelial cells adopt an irregular polygonal
ease, there may be many exogenous stresses that shape, and the corneal thickness typically returns
could potentially damage the corneal endothe- to normal at the same time that the transparency
lium over a person’s lifetime. These may reduce is restored. Finally, during the following months,
even further the endothelial cell density below a the endothelial cells reacquire their characteristic
level critical to the maintenance of corneal hexagonal form.
deturgescence, as the physiological pump-barrier There are a number of reasons that prevent
function fails and corneal edema establishes. endothelial cells from dividing and are currently
This critical density has been estimated to be still a matter of research and discussion. It is
10–15% of the normal cell count, which falls thought that the densely packed cells exhibit
between 300 and 500 cell/mm2 [28]. strong contact inhibition, mediated by p27kip1, a
The most common interventions that might cyclin-dependent kinase inhibitor that arrests
stress an individual’s corneal endothelium cells in the non-mitotic G1 phase [32, 33]. In
[29] are: addition to this, it has been described that in the
aqueous humor, there are low concentrations of
1. Contact lens wear, causing corneal hypoxia growth factors that promote mitosis, while inhib-
and leading to polymegathism and itory growth factor such as TGF-ß is found in
pleomorphism abundance [32]. Finally, the high rate of cellular
2. Intraocular surgeries such as cataract, glau- metabolism with chronic ultraviolet light expo-
coma, and refractive IOL surgery. sure results in nuclear oxidative DNA damage
20 V. Vargas et al.

that can promote a state of stress-induced senes- Pharmacological treatment may also have a
cence [34]. role on endothelial barrier reformation. Since
The current knowledge about the negligible Okumura and colleagues in 2009 identified a
regenerative capacity of the corneal endothelium specific ROCK inhibitor Y-27632 capable of
has been challenged after the popularization of promoting adhesion, survival, and proliferation
endothelial keratoplasty (EK), specifically with of primate-derived corneal endothelial cells
the Descemet membrane endothelial keratoplasty in vitro [37], other studies have shown that this
(DMEK). Reports of spontaneous regeneration drug is able to enhance corneal endothelial
of the corneal endothelium are increasingly wound healing in  vivo in animal models, both
appearing in the literature, but it remains unclear when injected intracamerally with a cell suspen-
whether this clinical improvement is due to sion and when applied topically as eye drops
enlargement and spreading of the cells or true [38, 39]. Following on from the animal data, pro-
cell division. spective studies with higher number of patients
Spontaneous endothelial barrier reformation are needed to demonstrate its efficacy, although
has been described in different surgical scenar- some preliminary studies have shown that rho-
ios. Corneal clearance after failed endothelial associated protein kinase (ROCK) inhibitors,
keratoplasty due to graft detachment has been either as Y-27632 or ripasudil, exert a certain
reported in Fuchs endothelial corneal dystrophy effect and may have a role as supplementation to
(FECD) patients [35] and also in normal subjects salvage procedures when there is no corneal
after the accidental removal of the Descemet edema resolution [40–42].
membrane in the course of cataract surgery [35]. There is not a single mechanism that explains
Bare areas of uncovered corneal stroma have the reformation of the endothelium in vivo, and
been repopulated with endothelial cells after the relevance of each mechanism is still specula-
free-floating endothelial roll injection, procedure tion. Cellular enlargement and migration to repair
known as Descemet membrane endothelial trans- a defect is probably the most important one. In
fer (DMET) in FECD but not in pseudophakic the cases of corneal endothelial transplantation,
bullous keratopathy [36]. Endothelial repopula- when a gap in the barrier is filled, it is often not
tion has been also seen after DMEK graft decen- possible to tell whether migrating donor or recip-
tration that leaves an arc or annulus of bare ient cells filled the gap. The centripetal pattern of
stroma or in recently described hemi-DMEK or corneal resolution in hemi-DMEK and quarter-­
quarter-DMEK, where the graft is bisected or DMEK implies that the cells on the donor graft
divided into four pieces, respectively, prior to are responsible for clearing [43]. Paracrine effect
implantation, which raises an interesting ques- of growth factors secreted by the transplanted
tion regarding the amount of endothelial graft donor graft may also contribute to endothelial
material needed for transplantation [35]. Even cell migration, and this mechanism has been con-
more, the outcomes of planned descemetorhexis sidered key in DMET success [44]. Finally, true
without grafting in FECD were surprising, as proliferation may also happen in the reformation
clear cornea was observed in more than 65% of of endothelium in  vivo, although there is not
patients (n = 31/47), with improvement detected enough evidence. Corneal endothelial mitosis
as early as 1 month postoperatively [35]. The rea- was described in patients in vivo as early as in the
son why recipient endothelium does not show 1980s when mitotic spindles were detected using
capacity to repair itself without surgery is proba- specular microscopy and thymidine incorpora-
bly because in the cases of FECD and PBK, the tion in transcorneal frozen eyes [44]. Recently a
damaged endothelial cells function very poorly study identified a rapidly proliferating subpopu-
but are not so dysfunctional as to detach from the lation of cells in the endothelium without any
Descemet membrane. They therefore occupy sign of cell aging, with the ability to form spheres
space and by contact inhibition prevent ingrowth and express stem cell and neural crest markers
of healthy cells in their place. [44, 45]. These results are suggestive, but
2  Corneal Healing 21

although it is tempting to think that there may be 3. Nishida T, Saika S, Morishige N. Cornea and sclera:
anatomy and physiology. In: Krachmer J, Mannis M,
an endothelial stem cell population, without Holland E, editors. Cornea. St. Louis: Mosby; 2011.
human studies using cellular labeling in vivo, it is p. 1–22.
difficult to conclude with certainty if the corneal 4. Sotozono C, He J, Matsimoto Y, Kita M, Imanishi J,
endothelial cells are proliferating. Kinoshita S. Cytokine expression in the alkali-burned
cornea. Curr Eye Res. 1997;16:670–6.
5. Ashby BD, Garrett Q, Willcox MDP. Corneal injuries
and wound healing – review of processes and thera-
Take-Home Messages pies. Austin J Clin Ophthalmol. 2014;1(4):1017.
• Corneal healing is a complex process in 6. Liu C-Y, Kao WW-Y. Corneal epithelial wound heal-
ing. Prog Mol Biol Transl Sci. 2015;5:1–11.
which numerous growth factors and 7. Ljubimov AV, Saghizadeh M.  Progress in corneal
cytokines are involved. wound healing. Prog Retin Eye Res. 2015;49:17.
• Of all corneal layers, the epithelium is https://doi.org/10.1016/j.preteyeres.2015.07.002.
the quickest to heal. 8. Sugioka K, Mishima H, Kodama A, Itahashi M,
Fukuda M, Shimomura Y.  Regulatory mechanism
• During the stromal healing process, of collagen degradation by keratocytes and corneal
haze may appear due to numerous myo- inflammation: the role of Urokinase-type plasmino-
fibroblasts and a disorganized ECM. gen activator. Cornea. 2016;35(Suppl):S59–64.
• After injury, endothelial cells adjacent 9. Klenkler B, Sheardown H, Jones L.  Growth fac-
tors in the tear film: role in tissue maintenance,
to the injury site slide and enlarge in wound healing, and ocular pathology. Ocul Surf.
order to fulfill the defect. 2007;5(3):228–39.
• There might be endothelial cell prolif- 10. Imanishi J, Kamiyama K, Iguchi I, Kita M, Sotozono
eration, but we need more evidence to C, Kinoshita S.  Growth factors: importance in
wound healing and maintenance of transparency
support it. of the cornea. Prog Retin Eye Res. 2000;19(1):
• Some growth factors have been success- 113–29.
fully used to enhance corneal healing. 11. Miyagi H, Thomasy SM, Russell P, Murphy CJ. The
role of hepatocyte growth factor in corneal wound
healing. Exp Eye Res. 2018;166:49–55.
12. Omoto M, Suri K, Amouzegar A, Li M, Katikireddy
KR, Mittal SK, et al. Hepatocyte growth factor sup-
Compliance with Ethical Requirements  Veronica presses inflammation and promotes epithelium repair
Vargas, Francisco Arnalich-Montiel, and Jorge L. in corneal injury. Mol Ther. 2017;25(8):1–8.
Alio del Barrio declare that they have no conflict 13. Lambiase A, Sacchetti M, Bonini S.  Nerve growth
of interest. No human studies were carried out by factor therapy for corneal disease. Curr Opin
Ophthalmol. 2012;23:296–302.
the authors of this article. No animal studies were 14.
Saghizadeh M, Kramerov AA, Svendsen CN,
carried out by the authors for this article. Ljubimov AV. Concise review: stem cells for corneal
wound healing. Stem Cells. 2017;35:2105–14.
15. Dua HS, Gomes JAP, Singh A.  Corneal epithelial
wound healing. Br J Ophthalmol. 1994;78:401–8.
16. Torricelli A, Santhanam A, Wu J, Singh V, Wilson
Financial Disclosure S. The corneal fibrosis response to epithelial-stromal
injury. Exp Eye Res. 2016;142:110–8.
17.
Bukowiecki A, Hos D, Cursiefen C, Eming
None of the authors have any financial disclosure SA.  Wound-healing studies in cornea and skin: par-
allels, differences and opportunities. Int J Mol Sci.
2017;18:1–24.
18. Wilson SE.  Corneal myofibroblast biology and

References pathobiology: generation, persistence and transpar-
ency. Exp Eye Res. 2012;99(1):78–88. https://doi.
1. Dawson D, Ubels JL, Edelhauser HF.  Cornea and org/10.1016/j.exer.2012.03.018.
sclera. In: Levin LA, Alm A, Nilsson SFE, Ver Hoeve 19.
Marino GK, Santhiago MR, Torricelli AAM,
J, Wu S, editors. Adler’s physiology of the eye. Santhanam A, Wilson SE. Corneal molecular and cel-
Edinburgh: Saunders; 2011. p. 71–130. lular biology for the refractive surgeon: the critical
2. Agrawal VB, Tsai RJF. Corneal epithelial wound heal- role of the epithelial basement membrane. J Refract
ing. Indian J Ophthalmol. 2003;51:5–1. Surg. 2016;32(2):118–25.
22 V. Vargas et al.

20. Bonini S, Lambiase A, Rama P, Caprioglio G,


the proliferation of the developing corneal endothe-
Aloe L.  Topical treatment with nerve growth fac- lium. Invest Ophthalmol Vis Sci. 2004;45(7):2163–7.
tor for neurotrophic keratitis. Ophthalmology. 34. Joyce NC, Zhu CC, Harris DL.  Relationship among
2000;107:1347–52. oxidative stress, DNA damage, and proliferative
21. Cellini M, Bendo E, Bravetti GO, Campos EC.  The capacity in human corneal endothelium. Invest
use of nerve growth factor in surgical wound healing Ophthalmol Vis Sci. 2009;50(5):2116–22.
of the cornea. Ophthalmic Res. 2006;38:177–81. 35. Van den Bogerd B, Dhubhghaill SN, Koppen C,

22. Lambiase A, Bonini S, Aloe L, et  al. Anti-­
Tassignon M-J, Zakaria N. A review of the evidence
inflammatory and healing properties of nerve growth for in  vivo corneal endothelial regeneration. Surv
factor in immune corneal ulcers with stromal melting. Ophthalmol. 2018;63:149–65.
Arch Ophthalmol. 2000;118:1446–9. 36. Dirisamer M, Ham L, Dapena I, van Dijk K, Melles
23. Nishida T, Chikama T, Morishige N, Yanai R, Yamada GRJ.  Descemet membrane endothelial transfer:
N, Saito J.  Persistent epithelial defects due to neu- “free-floating” donor descemet implantation as
rotrophic keratopathy treated with a substance P- a potential alternative to “keratoplasty”. Cornea.
derived peptide and insulin-like growth factor 1. Jpn J 2012;31(2):194–7.
Ophthalmol. 2007;51:442–7. 37. Okumura N, Ueno M, Koizumi N, Sakamoto Y, Hirata
24. Yamada N, Matsuda R, Morishige N, Yanai R,
K, Hamuro J, et al. Enhancement on primate corneal
Chikama T-i, Nishida T, Ishimitsu T, Kamiya A. Open endothelial cell survival in vitro by a ROCK inhibitor.
clinical study of eye-drops containing tetrapeptides Invest Ophthalmol Vis Sci. 2009;50(8):3680–7.
derived from substance P and insulin-like growth 38. Okumura N, Okazaki Y, Inoue R, Kakutani K, Nakano
factor-1 for treatment or persistent corneal epithelial S, Kinoshita S, et  al. Effect of the Rho-associated
defects associated with neurotrophic keratopathy. Br kinase inhibitor eye drop (Ripasudil) on corneal endo-
J Ophthalmol. 2008;92:896–900. thelial wound healing. Invest Ophthalmol Vis Sci.
25. Holland S, Morck D, Schultz C. Treatment of corneal 2016;57(3):1284–92.
defects with delayed re-epithelization with a medical 39. Okumura N, Sakamoto Y, Fujii K, Kitano J, Nakano
device/drug delivery system for epidermal growth S, Tsujimoto Y, et  al. Rho kinase inhibitor enables
factor. Clin Exp Ophthalmol. 2012;40:1–6. https:// cell-based therapy for corneal endothelial dysfunc-
doi.org/10.1111/j.1442-9071.2012.02795.x. tion. Sci Rep. 2016;6(1):26113.
26. Alió JL, Arnalich-Montiel F, Rodriguez AE.  The
40. Okumura N, Inoue R, Okazaki Y, Nakano S, Nakagawa
role of “eye platelet rich plasma” (E-Prp) for wound H, Kinoshita S, et al. Effect of the Rho kinase inhibi-
healing in ophthalmology. Curr Pharm Biotechnol. tor Y-27632 on corneal endothelial wound healing.
2012;13:1257–65. Investig Opthalmology Vis Sci. 2015;56(10):6067.
27. Alio JL, Abad M, Artola A, Rodriguez-Prats JL,
41. Moloney G, Petsoglou C, Ball M, Kerdraon Y,

Pastor S, Ruiz-Colecha J. Use of autologous platelet-­ Höllhumer R, Spiteri N, et  al. Descemetorhexis
rich plasma in the treatment of dormant corneal without grafting for fuchs endothelial dystrophy-­
ulcers. Ophthalmology. 2007;114(7):1286–93. supplementation with topical Ripasudil. Cornea.
28. Waring GO, Bourne WM, Edelhauser HF, Kenyon 2017;36(6):642–8.
KR.  The corneal endothelium. Normal and patho- 42. Koizumi N, Okumura N, Ueno M, Nakagawa H,

logic structure and function. Ophthalmology. Hamuro J, Kinoshita S.  Rho-associated kinase
1982;89(6):531–90. inhibitor eye drop treatment as a possible medi-
29. Edelhauser HF.  The resiliency of the corneal endo- cal treatment for Fuchs corneal dystrophy. Cornea.
thelium to refractive and intraocular surgery. Cornea. 2013;32(8):1167–70.
2000;19(3):263–73. 43. Lam FC, Baydoun L, Dirisamer M, Lie J, Dapena I,
30. Joyce NC.  Proliferative capacity of the corneal
Melles GRJ.  Hemi-Descemet membrane endothelial
endothelium. Prog Retin Eye Res. 2003;22(3): keratoplasty transplantation: a potential method for
359–89. increasing the pool of endothelial graft tissue. JAMA
31. Watsky MA, McDermott ML, Edelhauser HF. In vitro Ophthalmol. 2014;132(12):1469–73.
corneal endothelial permeability in rabbit and human: 44. Lam FC, Bruinsma M, Melles GRJ. Descemet mem-
the effects of age, cataract surgery and diabetes. Exp brane endothelial transfer. Curr Opin Ophthalmol.
Eye Res. 1989;49(5):751–67. 2014;25(4):353–7.
32. Senoo T, Joyce NC.  Cell cycle kinetics in corneal 45. Katikireddy KR, Schmedt T, Price MO, Price FW,
endothelium from old and young donors. Invest Jurkunas UV.  Existence of neural crest-derived
Ophthalmol Vis Sci. 2000;41(3):660–7. progenitor cells in Normal and Fuchs endothe-
33. Yoshida K, Kase S, Nakayama K, Nagahama H,
lial dystrophy corneal endothelium. Am J Pathol.
Harada T, Ikeda H, et al. Involvement of p27KIP1 in 2016;186(10):2736–50.
Corneal Tissue Engineering
3
Mohammad Mirazul Islam, Roholah Sharifi,
and Miguel Gonzalez-Andrades

3.1 Introduction tation [8], although the whole world suffers a


severe scarcity of donor corneas. This results in
Mankind has been always fascinated with the 10 million untreated patients worldwide, with an
idea of restoring any damaged tissue or organ. In additional 1.5 million new patients every year to
the Ancient Egypt, handmade prostheses were the waiting list [9].
made of hardwood or cartonnage to restore the Although the cornea is considered as an
function of lost toes [1]. Regarding corneal func- immune-privileged site of the body because of
tional restoration, the French ophthalmologist its avascularity, physiopathological changes such
Pellier de Quengsy was the first one proposing in as corneal neovascularization or inflammation
1789 a replacement of an opaque cornea using a disrupt this immune status, which subsequently
piece of glass surrounded by a silver ring [2]. increases the risk of graft rejection after perform-
However, the paradigm of corneal blindness ing a corneal transplant. Even in non-­vascularized
treatment does not change until 1905, when and non-inflamed host eyes (low-risk cases), one
Eduard Zirm performed the first corneal trans- in three of transplanted corneas eventually leads
plant to a patient implanting a donor cornea [3]. to rejection [10]. In high-risk cases, such as auto-
Corneal transplant is still the most used and reli- immune diseases, chemical burns and infections,
able treatment for some corneal diseases [4]. 50–70% transplants undergo rejection even with
Corneal diseases are one of the most important high doses of immunosuppressive drugs [11, 12].
causes of blindness in developing countries, Another major complication after corneal trans-
accounting for 4–8  million people that suffer plant is donor-derived infections [13]. Herpes sim-
bilateral corneal blindness [5, 6]. In 2015, only in plex virus type-1 (HSV-1) DNA found from donor
the USA, more than 48,792 corneal transplants corneas before and after corneal transplantation
were carried out, which was 53% more than the confirmed the spread of HSV-1 through the trans-
transplants performed in 2005 [7]. In the USA, plant from donor to recipient [14]. In this context,
donor corneas are readily available for transplan- the microbial testing, the administration and the
shipping of a donor cornea can cost $3000 in the
USA [15], becoming an unaffordable treatment
option for most of the people who live in develop-
M. M. Islam · R. Sharifi · M. Gonzalez-Andrades (*) ing nations. 90% of the visually impaired people
Massachusetts Eye and Ear and Schepens Eye
live in low-income countries, and most of them
Research Institute, Department of Ophthalmology,
Harvard Medical School, Boston, MA, USA (53% of the world population) do not have access
e-mail: miguel_gonzalez@meei.harvard.edu to transplantation facilities [16, 17].

© Springer Nature Switzerland AG 2019 23


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_3
24 M. M. Islam et al.

Therefore, there is a great need of finding new extracellular matrix (ECM) of the corneal stroma
therapeutic strategies to address the three major is mainly composed by collagen and proteogly-
drawbacks of corneal transplantation: the scarcity cans disposed in a highly specific arrangement.
of donors, the risk of rejection and the transmission This specific matrix arrangement grants the cornea
of infectious diseases after implantation into the its characteristic transparency and physical struc-
host. In this milieu, corneal tissue engineering (TE) ture, which allow an optimal vision and support
emerges with the ambition of generating artificial the intraocular pressure without deforming [23].
corneas or other types of tissue-engineered products ECM is also intimately associated with corneal
that lead to an optimal corneal regeneration, over- innervation. Corneal nerves cross through the cor-
coming those major disadvantages of allogeneic neal stroma toward the epithelium. Epithelial
corneal transplants. TE was defined in 1988, in a innervation plays a vital role in functional activi-
workshop on TE organized by the University of ties of the cornea such as preserving the viability
California, Los Angeles, as “the application of prin- and differentiation of the corneal epithelium, apart
ciples and methods of engineering and life sciences from their role in tear production and blinking
toward fundamental understanding of structure- [24]. Moreover, the optimal composition and
function relationships in normal and pathological porosity of the corneal stroma allow the diffusion
mammalian tissues and the development of biologi- of nutrients and other solutes from the posterior to
cal substitutes to restore, maintain, or improve tis- the anterior region of the cornea [25, 26]. Corneal
sue functions” [18–20]. Engineering effectual nutrition is complemented by the tears [27]. The
building blocks and assembling them to perform in local immunity of the cornea is conditioned to its
a unified architecture is the bedrock for the genera- avascularity. Antigen-presenting cells like den-
tion of fully functional biological substitutes. Such dritic cells are present in the cornea. These cells
constructs should also be able to communicate with are involved in T-cell-mediated immune responses
the other tissues and organs that surround it to coor- associated with corneal graft rejection. Natural
dinate a unified function [21]. This demands pro- killer cells also participate in the allograft rejection
found knowledge in material science, including [28, 29]. Graft rejection starts when the host
material interactions with cells and their microenvi- immune system is activated against antigens in the
ronment. Especially, this is crucial in ophthalmol- donor corneal tissue through different pathways
ogy where besides physical and chemical properties (for review see Refs. [30, 31]).
of the tissues, optical characteristics as well as
architectural design dictate the ultimate outcome.
3.3 Development of Tissue-­
Engineered Corneal
3.2 Corneal Structure-Function Substitutes
Relationships
Once we understand the structure-function rela-
Firstly, we need to understand the structure-­ tionships in the cornea, we can generate a tissue-­
function relationships in the cornea, in order to engineered corneal substitute to restore, maintain,
develop an optimal corneal substitute. The cornea or improve corneal functions, using different
is the outermost part of the eye and plays an impor- building blocks: cells, scaffolds and bioactive
tant role for vision by transmitting the light to the molecules.
retina while protecting the interior components of
the eye from external aggressions. The cornea is
composed of three primary cellular layers, the out- 3.3.1 Cells
ermost epithelium layer, a middle stroma contain-
ing keratocytes, and an innermost single layer of As it was discussed earlier, healthy cellular layers
endothelial cells called endothelium [22]. Two are necessary for the precise function of the
acellular layers separate these cellular layers: human cornea. In case of cell injury or loss, stem
Bowman’s layer and Descemet’s membrane. The cells need to regenerate and repopulate the dam-
3  Corneal Tissue Engineering 25

aged area. Stem cells are indispensable players Table 3.1  Human cells other than limbal stem cells eval-
uated for corneal epithelial regeneration
for the regeneration of any part of the body. In the
cornea, different types of stem cells reside in the In
vitro/in
limbus area with the capacity to regenerate the Source of cells vivo References
corneal epithelium [32] and the stroma [33, 34]. Conjunctival In vivo Ang et al. [55]; Sangwan
Limbal stem cells (LSCs) can differentiate into stem cell (human) et al. [56, 57];
corneal epithelial cells after isolating and cultur- Subramaniam et al. [58];
Tan et al. [59]
ing them from small biopsies of healthy limbal
Oral mucosal In vivo Burillon et al. [60];
areas [35–37]. Different substrates or carriers can epithelial cells (human) Inatomi et al. [61];
be used to culture and deliver LSCs in cases of Nakamura et al. [62];
limbal stem cell deficiency (LSCD) including Nishida et al. [63];
human amniotic membranes [38], fibrin sub- Takeda et al. [64];
Utheim [65]
strates [39] and collagen-based materials [40].
Nasal mucosal In vivo Chun et al. [44]; Kim
This therapeutic approach receives the name of epithelial cells (human) et al. [45]
cultured limbal epithelial transplantation (CLET). Dental pulp In vivo Gomes et al. [46];
This expansion of LSCs for transplantation stem cells (rabbit) Monteiro et al. [47]
requires certified good manufacturing practices Hair follicle In vivo Meyer-Blazejewska et al.
bulge-derived (mouse) [48]
facilities and procedures, which limit the expan- stem cells
sion of this therapeutic approach because of the Wharton’s jelly In vitro Garzon et al. [49]
high cost, especially in developing countries stem cells
[41]. Embryonic stem In vitro Ahmad et al. [66]; Zhang
In cases of suffering bilateral LSCD with no cells et al. [67]
Umbilical cord In vivo Reza et al. [50]; Reza
healthy limbal area to obtain an optimal biopsy, stem cells (rabbit) et al. [68]
only allogeneic limbal tissue can be used for per- Bone marrow-­ In vivo Ma et al. [20]; Rohaina
forming CLET.  To avoid the use of allogeneic derived MSC (Rat) et al. [52]
cultured cells and its inherent risk of immune Orbital In vivo Lin et al. [53]
rejection, other types of stem cells that do not fat-derived MSC (mouse)
Dermal In vitro Hayashi et al. [54]
reside in the limbal area are emerging as possible
fibroblast-­
future cell source for autologous CLET derived iPS cells
(Table  3.1). Conjunctival epithelial cells from Corneal limbal In vitro Hayashi et al. [54]
biopsies cultured and expanded in vitro on con- epithelial
tact lens were used to treat one patient, who cell-derived iPS
cells
improved visual acuity with no recurrence of cor-
neal vascularization [42]. The possibility of treat-
ing LSCD using in vitro cultivated oral mucosa epithelial-­like cells on fibrin-agarose-based stro-
autograft has been widely studied in at least 20 mal substitutes [49]. Transplantation of human
clinical trials in different countries. Accumulated umbilical cord stem cells in LSCD rabbit eyes
results from those clinical trials showed that 242 resulted in healthy corneal surface with positive
patients received this treatment with a success marker expression for corneal epithelial cells
rate of 72% [43]. Nasal mucosal epithelial cells [50]. Human bone marrow-derived mesenchymal
also showed promising results when transplanted, stem cells (MSC) were able to differentiate to
in two different clinical trials [44, 45]. Cultured corneal epithelial cells in vitro and in vivo, show-
human immature dental pulp stem cells recon- ing their capability to replace limbal epithelial
structed the eye surface in limbal stem cell-defi- stem cells [20, 51, 52]. Orbital fat-derived MSC
cient rabbits [46, 47]. Hair follicle bulge-derived also promoted corneal tissue regeneration
stem cells from transgenic mice also showed cor- through corneal epithelial differentiation [53].
nel epithelial cell differentiation in a LSCD Human adult dermal fibroblast-­derived induced
model [48]. Human Wharton’s jelly stem cells pluripotent stem (iPS) cells and human adult cor-
also showed potential to differentiate into corneal neal limbal epithelial cell-derived iPS cells were
26 M. M. Islam et al.

also tested for differentiation into corneal epithe- 3.3.2 Scaffolds


lial cells, which revealed that corneal epithelial
differentiation efficiency was higher in limbal- Engineered 3D scaffolds not only can substitute a
derived iPS cells [54]. damaged cornea, providing mechanical and
Keratocytes quiescently reside within colla- structural stability, but also provide the appropri-
gen lamellae in the stroma of a healthy cornea, ate microenvironment for the cells to regenerate
synthesizing ECM components, such as colla- the tissue. Conceptually, scaffold is an engi-
gen and proteoglycans [69, 70]. In the damaged neered template, which can mimic the ECM of
or injured cornea, keratocytes transform into the native tissue and imitate the in  vivo setting,
mitotically active fibroblasts [71, 72] and start supporting the cells to proliferate, migrate and
producing unorganized ECM which ultimately create their own microenvironment [83].
turn into fibrotic tissue, which might lead to The ideal corneal scaffold should (1) be trans-
vision loss [73]. Keratocytes can be isolated and parent for restoration of vision; (2) be biocompat-
cultured under specific conditions using corneal ible and support cellular adhesion, proliferation
biopsies, which can be digested or directly cul- and migration; (3) have similar biomechanical
tured applying an explant-based technique [74, properties to the human cornea to maintain its
75]. Keratocytes can also be obtained from shape, critical for an optimal vision, and har-
MSC isolated from limbal biopsies [33, 34]. monically respond to the intraocular pressure
They can synthetize aligned collagen and kera- fluctuations; (4) preserve its smooth surface in
tan sulfate proteoglycans, being able to recon- order to avoid scattering of light; (5) have bio-
stitute a fibrotic area in in vivo models, without degradation properties that match the time of
inducing inflammation, vascularization, or tissue remodeling and biointegration; (6) have a
rejection [76–78]. iPS cells can be also differen- refractive index similar to the cornea; (7) possess
tiated to neural crest cells and then cultured on appropriate porosity and diffusion for nutrients,
corneal tissue to promote keratocyte differentia- while serving as a microbial barrier; and (8) be
tion [79]. cost-effective in terms of manufacturing process
The corneal endothelium is a monolayer of and implementation [84].
cells that lines at the posterior corneal surface, The scaffolds explored in ophthalmology for
which are responsible for pumping out excess corneal substitution can be categorized into three
amount of water from the corneal stroma and classes: synthetic, natural-based and hybrid
prevent it from swelling [80, 81]. Its failure usu- materials. Polyethylene glycol [85], acrylate-­
ally requires a donor endothelial transplant based polymers [86], polyesters [87],
because of the very limited proliferative capac- polydimethylsiloxane [88], polyvinyl alcohol-
ity of these cells to self-regenerate the damaged based polymers [89] and polyamides [90] are the
area. However, under specific conditions, endo- main studied synthetic materials for corneal sub-
thelial cells can proliferate and cultured in vitro. stitutes. Although these synthetic polymers have
Shigeru Kinoshita and co-worker described a tunable chemical and mechanical properties that
new promising approach based on the inhibition can be matched to the medical needs, their bio-
of ROCK (Rho kinase), which enhances endo- mimetic properties required for cell adhesion,
thelial cell proliferation, promotes cell adhe- proliferation and effectual integration with the
sion, suppresses apoptosis and promotes wound host tissue need significant improvement before
healing. In 2013, they have started a clinical their translation in the clinical settings. In addi-
trial to evaluate cultured human endothelial tion to their non-biodegradable nature, their
cells in combination with a ROCK inhibitor as inability to carry cells and biointegrate during
treatment for corneal endothelial dysfunction. tissue healing and remodeling stands as their
Recently, they reported their initial results, sug- main challenge [91]. The continued progress in
gesting that this therapeutic option is safe and the engineering of novel biomaterials, along
effective [82]. with personalized modifications and the design
3  Corneal Tissue Engineering 27

of hybrid materials composed of synthetic and corneal native tissue in terms of not only physical
natural polymers, might address such shortcom- and chemical properties but also composition gra-
ings and can facilitate their widespread applica- dients, alignment, directionality and microar-
tions in the clinic. rangement manifested in the human corneal
On the other hand, natural-based biomateri- stroma. The emulation of such biomimetic char-
als present intrinsic biocompatibility and bio- acteristics is very crucial in tissue engineering and
degradability along with appreciable degree of dramatically dictates biointegration of artificial
biomimetic properties and biological functions. corneas with the host tissue and defines the ulti-
The most studied natural-based corneal scaffolds mate clinical outcome. Different strategies have
are protein- or polysaccharide-based scaffolds. been described in the literature to address some of
Collagen is one of the most studied protein- those challenges such as the use of self-­assembly
based scaffolds for artificial cornea. This stems or auto-generation of artificial matrixes in vitro.
from collagen’s biocompatibility, low toxicity, Peptide amphiphiles (PA) are engineered syn-
and well-studied structural, physical, chemi- thetic molecules constituted from a hydrophilic
cal and immunological characteristics alongside peptide sequence and hydrophobic long chain,
with maintaining arginine-glycine-aspartic acid which can self-assemble to generate nanofibers.
sequences in its structure that promotes cell adhe- The non-covalent interaction of such nanofibers
sion to the scaffold [92]. Although collagen-­based via intermolecular forces can lead to the forma-
scaffolds demonstrated such promising proper- tion of 3D networks. The immense programma-
ties [93–95], their biomimetic characteristics still bility of PA to hold different functional groups
require an enhancement to totally match those enables to generate 3D scaffolds, such as collagen
found in specific native tissues [96]. Although hydrogels, that, in principle, can mimic ordering
improving mechanical and biomimetic properties and complexity of native tissue [105–108].
of hydrogel is an ongoing challenge, engineering Integrating such self-assembled structures within
new hybrid scaffolds and integrating the biologi- various hydrogel offers a precious tool to intro-
cal cues might be a key to unlock their potential duce highly complex multifunctional hydrogels
as a true corneal substitute. Gelatin [97], fibrin for ophthalmic surgery. Another approach to
[98] and silk [99] are among the other protein- introduce such nanoscale organization is auto-­
based biopolymers that have also been explored generation. This concept is based on engineering
as possible candidates for corneal substitute. an in vitro culture system that stimulates synthesis
While each class possesses different characteris- of an in vivo-like stromal matrix that can lead to
tics, their mechanical properties are significantly the generation of highly organized collagen-based
inferior compared to those of the native cornea corneal stroma [109, 110]. Moreover, it allows to
and unable to support integrity of injured cornea. seed epithelial and endothelial cells on the synthe-
Polysaccharide-based materials (e.g., chitosan sized scaffold to create a functional, organotypic
[100], chondroitin sulfate [101], dextran [102], cornea. In this regard, human mesenchymal stem
hyaluronic acid [103], alginate [104], etc.) have cells derived from the limbal stroma were cul-
also been explored in corneal tissue engineer- tured in specific culture media, leading to rapid
ing. Despite their superior mechanical and opti- expansion and differentiation into keratocytes and
cal properties, they fall short in providing 3D ultimately generating an organized thick lamellar
microenvironment for effective cell adhesion and stroma-like tissue containing aligned collagen
proliferation. Therefore, polysaccharide-­based and keratan sulfate proteoglycans [76]. The con-
materials have not been yet able to offer an effec- structs synthesized by these cultured cells, how-
tive solution for corneal substitute. ever, also present poor mechanical properties and
Although both synthetic and natural-based need significant improvement prior to their appli-
scaffolds offer an initiative window to develop an cation in ophthalmic surgery.
effective artificial cornea, such scaffolds lack the Incompetence of bioengineered scaffolds as yet
complexity of the 3D microenvironment of the to fulfill the required properties of corneal substi-
28 M. M. Islam et al.

tute is the main driving force to study other strate- have been carried out to understand the control
gies in parallel such as the use of modified mechanisms of self-renewal and fate decision of
xenografts. Xenogeneic tissues and organs often LSCs. There are growing evidences supporting
contain cellular antigens, which can be recognized that LSCs are highly regulated by their stem cell
as foreign by the host tissue and consequently niche. LSC niche is a specific microenvironment
leads to an inflammatory response or an immune- that comprises cellular and noncellular compo-
mediated rejection [111]. Decellularization of nents that regulate the stem cell pluripotency,
donor tissue to remove the inhabiting cells and its proliferation, differentiation, survival and local-
cellular debris from the ECM of the tissue is a ization. LSC niche is located at the palisades of
practical strategy to obtain acellular scaffolds of Vogt of human corneoscleral limbus. Different
the ordinal tissue and bypass such adverse immune growth factors also play important role in the
response [112]. Besides intimate resemblance of differentiation of the stem cell to progeny. In
microarchitecture of xenogeneic corneas with this regard, insulin-like growth factor I (IGF-I)
humans, their availability, lower cost and compa- has been identified as the main factor respon-
rable optical and mechanical properties are the sible for LSC differentiation into mature cor-
main momentum to envisage their application for neal epithelial cells after injury. Furthermore,
human corneal substitution. Moreover, it can sup- some researchers have demonstrated that IGF-I
port construction and host-guest tissue remodeling showed synergistic effect with the neuropeptide
and bypass the stimulation of inflammation while substance P in proliferation and wound healing
avoiding scar tissue formation [113–116]. of corneal epithelium [120]. Corneal epithe-
Different animals have been used as source of cor- lium also produced fibroblast growth factor and
neal tissue for decellularization process. Due to epidermal growth factor (EGF) to support LSC
availability and the structural similarities between proliferation devoid of affecting differentiation
the porcine and human cornea, the domestic pig is [121]. EGF heparin-­binding EGF and amphireg-
the most commonly used animal to obtain decel- ulin have been also shown to stimulate epithe-
lularized corneal xenografts [117]. Although vari- lial wound repair by binding to a common EGF
ous chemical and physical techniques have been receptor [122]. Hepatocyte growth factor is also
explored to decellularized animal corneas, they an important factor expressed by epithelial cells
often alter the chemical, physical and biological and keratocytes after corneal epithelial injury,
properties of the ECM via cleaving the collagen which influences the proliferation, migration and
fibers and disrupting the matrix ultrastructure or apoptosis of corneal epithelial cells [123–125].
partially eliminating key matrix constituents such Keratinocyte growth factor also plays impor-
as glycosaminoglycans and growth factors and tant role in epithelial wound healing through
adversely affect its natural properties [112, 118]. MAP kinase and PI3K/p70 S6 signaling cascade
Such structural disruptions along with chemical, [126]. Moreover, transforming growth factor-β
mechanical and biological variations between por- (TGF-β) expressed by corneal epithelium and
cine and human cornea are the main challenges, stromal cells has mixed effect on corneal cells,
preventing their successful translation into the inhibiting ­epithelial cell proliferation [127] and
clinic [119]. stimulating fibroblast proliferation [128]. TGF-β
also showed to influence myofibroblast differen-
tiation of cultured primary keratocytes and cor-
3.3.3 Bioactive Molecules neal fibroblast cell line [129]. Platelet-derived
and Other Environmental growth factors expressed by differentiated cor-
Conditions neal epithelium in  vitro regulate the prolifera-
tion and migration of corneal fibroblast [130].
There are several soluble factors directly involved Nerve growth factor (NGF) is a neurotrophic
in the process of proliferation and differentiation factor expressed in the corneal epithelium that
of corneal cells. In this context, significant efforts promotes cell proliferation and wound healing.
3  Corneal Tissue Engineering 29

NGF improved epithelium restoration in patients 3.4 Clinical Experiences


with neurotrophic ulcers [131] and after cataract and International
surgery [132]. NGF also showed nerve regenera- Regulations
tion in a mechanical nerve injury mouse model
established by laser-assisted in situ keratomileu- Very few tissue-engineered products have been
sis [133]. Opioid growth factor (OGF) is another translated into the clinic. Some of the organs
growth factor expressed by basal and suprabasal where tissue-engineered substitutes have been
layers of epithelium that binds OGF receptor to successfully applied are the trachea [153], blood
inhibit DNA synthesis, cell migration and tissue vessels [154], the urinary bladder [155], and the
repair of the corneal epithelium [134]. Important cornea [4]. Regarding the cornea, some notable
growth factors with their key physiological func- mentions are as follows: (1) autologous limbal
tions are summarized in Table 3.2. stem cells were collected from the healthy con-
Not only growth factors but also culture con- tralateral eye and expanded on a fibrin substrate
ditions are key in the proliferation and differen- and finally transplanted in 112 patients with
tiation of corneal cells. Concentration of carbon LSCD. Restoration of a transparent cornea with
dioxide in culture conditions critically alters cell a restored corneal epithelium was achieved in
differentiation. It was shown that 7% CO2 in the 76.6% of eyes and 21 patients achieved perma-
culture positively influences the differentiation nent visual recovery of at least 0.6 [36]; (2) as a
of embryonic stem cell to corneal epithelial pro- phase 1 clinical trial, femtosecond laser cut ante-
genitor cells [67]. Hypoxic condition is also an rior corneal stroma was decellularized and trans-
important factor that influences the differentia- planted in patients with keratoconus. Four out
tion of limbal stem cells by downregulating of nine patients received a decellularized stroma
Polo-­like kinase 3 (Plk3) signaling activity at the seeded with autologous adipose-derived adult
transcription level [151]. Co-culturing condi- stem cells. Haze or scarring was not observed
tions also positively influence cell growth, as by 3-month postoperative follow-up, and
survival and proliferation of LSCs are promoted patients got visual improvement after 6 months
when these are co-cultured with bone marrow of the graft [156]; (3) recombinant human col-
MSC [152]. lagen (RHC)-based acellular artificial corneas

Table 3.2  List of growth factors that influence corneal regeneration


Growth factors Key function References
Epidermal growth Cell migration, proliferation and wound healing of Zieske et al. [122]; Nakamura et al. [135]
factor corneal epithelial cells
Hepatocyte Cell migration, proliferation and wound healing. It Wilson et al. [123]; Daniels et al. [124];
growth factor inhibits apoptosis of corneal epithelial cells Yanai et al. [136]
Keratinocyte Epithelial homeostasis and wound healing Chandrasekher et al. [137]
growth factor
Insulin-like Cell growth, energy metabolism, migration, Lee et al. [136]; Trosan et al. [138]; Yanai
growth factor differentiation, proliferation and survival of et al. [139]
corneal epithelial cells
Transforming Inhibition of corneal epithelial cell proliferation. It Pancholi et al. [127]; Haber et al. [128];
growth factor-β stimulates stromal fibroblast proliferation Andresen et al. [140]; Kay et al. [141]
Platelet-­derived Migration and proliferation of keratocytes Denk and Knorr [130]; Kamiyama et al.
growth factors [142]; Daniels and Khaw [143]
Thymosin-β4 Wound healing in corneal epithelial defects. It Sosne et al. [144, 145]; Dunn et al. [146]
decreases inflammation and inhibits apoptosis
Nerve growth Epithelial and stromal healing, anti-inflammatory Lambiase et al. [147]; Lambiase et al.
factor effect and recovery of corneal nerves [148]; Joo et al. [149]
Opioid growth Inhibitory effect on corneal epithelial cell Zagon IS et al. [150]
factor proliferation, migration, and tissue organization
30 M. M. Islam et al.

have been transplanted in a clinical trial on ten tered to human beings with a view to regenerat-
patients; nine of them had keratoconus and one ing, repairing or replacing a human tissue. A
patient with permanent mid-stromal scar. The tissue engineered product may contain cells or
implants promoted regeneration of corneal epi- tissues of human or animal origin, or both. The
thelium, stroma, and nerves from host cells. The cells or tissues may be viable or non-viable. It
transplanted cornea remained stable for 4 years may also contain additional substances, such as
without any rejection and without sustained cellular products, bio-molecules, biomaterials,
immune suppression. Implanted patients had a chemical substances, scaffolds or matrices.” This
4-year average corrected visual acuity of 0.37 definition together with other aspects contem-
[4, 157]; (4) acellular interpenetrating polymer plated in the European regulations is shown in
networks of RHC and 2-­methacryloyloxyethyl Fig. 3.1. Regarding the clinical trials previously
phosphorylcholine (MPC) have been trans- mentioned, only decellularized stromas seeded
planted in three patients with corneal ulcers and with autologous adipose-derived adult stem cells,
recurrent corneal erosions. The implants pro- autologous epithelial limbal stem cells cultured
vided relief from pain and discomfort, restored on fibrin scaffolds, and fibrin-agarose corneal
corneal integrity, and improved vision in two out substitutes with allogeneic cultured corneal cells
of three patients [158]. Another clinical trial has would be considered tissue-engineered products
been completed with this materials on January according to EMA guidelines.
2017 (CT.gov identifier:NCT02277054). The The Food and Drug Administration (FDA) is
result showed that all patients improved from responsible for the regulation of medical prod-
pain and discomfort within 1–2  weeks after ucts, including tissue-engineered products, in
transplantation. Corneal sensitivity regained, the USA.  FDA regulated medical products
and overall vision improved significantly in half under the separate categories of devices, bio-
of the study patients, and even if the vision was logics, and drugs. According to FDA, human
not enhanced, transplants made the cornea sta- tissue intended for transplantation, such as a
ble for further surgery to improve vision [159]; donor cornea, is regulated as a human cell, tis-
and (5) there is a randomized, controlled, open- sue, and cellular and tissue-based product or
label clinical trial going on in different Spanish HCT/P [162]. Tissue-­engineered products usu-
hospitals (CT.gov identifier:NCT01765244) ally consist in the combination of two or more
to test a fibrin-agarose corneal substitute com- components that belong to different categories
bined with allogeneic corneal epithelial cells in the FDA regulation, falling into the category
and keratocytes [160]. of combination products [163]. Tissue-
In the translation of tissue-engineered prod- engineered products based on biomaterials in
ucts into the clinic, the different regulatory agen- conjugation with cells would fall in this cate-
cies have created a regulatory framework that gory. Recently, FDA announced new guidance
controls and guarantees the correct and ethical documents for comprehensive regenerative
use of these therapies in humans, protecting not medicine policy, defining the regulatory
only the patients but also the clinicians who are requirements for devices used in the recovery,
applying the treatment. The European Medicines isolation and delivery of regenerative medicine
Agency (EMA) includes tissue-engineered prod- advanced therapies (RMATs), including combi-
ucts under the definition of advanced therapy nation products and the description of those
medicinal products, which are defined as medi- regenerative medicine therapies that may be
cines for human use that are based on genes, eligible for RMAT designation, including cell
cells, or tissue engineering [161]. According to therapies, therapeutic tissue-engineered prod-
European Parliament regulations and the EMA ucts, human cell and tissue products, and com-
guidelines [51], “Tissue engineered product bination products using any such therapies or
means a product that: contains or consists of products, as well as gene therapies that lead to
engineered cells or tissues, and is presented as a durable modification of cells or tissues
having properties for, or is used in or adminis- (including genetically modified cells) [164].
3  Corneal Tissue Engineering 31

- Regenerating,
Tissue engineered product Function** - Repairing or a human tissue
- Replacing

Composition
*Non-substantial manipulations: cutting;
- Human and/or
Origin grinding; shaping; centrifugation; soaking in
- Animal antibiotic or antimicrobial solutions; sterilization;
Engineered irradiation; cell separation; concentration or
purification; filtering; lyophilization; freezing;
cells and/or - Viable and cryopreservation; and vitrification.
tissues Viability - Non-viable

- Pharmacological,
Main mechanism of action - Immunological or
- Metabolic

Subjected to substantial manipulation*


to obtain its necessary properties to
achieve its function**
Requirements to be “engineered” And/or
Not intended to be used for the same
essential function or functions in the
recipient as in the donor
- Cellular products,
- Bio-molecules,
- Biomaterials,
Additional substances - Chemical substances,
- Scaffolds or
- Matrices

Fig. 3.1  Definition of tissue-engineered product according to the European regulations [51]

3.5 Conclusions References

Here we briefly highlight the corneal structure-­ 1. Finch J.  The ancient origins of prosthetic medicine.
Lancet. 2011;377:548–9.
function relationships and the principles to 2. Chirila TV, Hicks CR.  The origins of the artificial
develop a biological substitute of the human cor- cornea: Pellier de Quengsy and his contribution to
nea by tissue engineering, including some treat- the modern concept of keratoprosthesis. Gesnerus.
ment options for corneal diseases based on 1999;56:96–106.
3. Zirm EK.  Eine erfolgreiche totale Keratoplastik (a
specific tissue-engineering strategies. Moreover, successful total keratoplasty). 1906. Refract Corneal
we explained the concepts and regulations neces- Surg. 1989;5:258–61.
sary to understand the future clinical impact of 4. Fagerholm P, Lagali NS, Merrett K, et  al. A biosyn-
tissue engineering in ophthalmology. The next thetic alternative to human donor tissue for inducing
corneal regeneration: 24-month follow-up of a phase 1
chapters of this book will elaborately explain the clinical study. Sci Transl Med. 2010;2:46ra61.
use of cell and tissue-engineering therapies to be 5. Oliva MS, Schottman T, Gulati M.  Turning the
surgically applied to different corneal diseases. tide of corneal blindness. Indian J Ophthalmol.
2012;60:423–7.
6. Burton MJ.  Prevention, treatment and rehabilitation.
Compliance with Ethical Requirement Community Eye Health. 2009;22:33–5.
Statements  Mohammad Mirazul Islam, 7. Brunette I, Roberts CJ, Vidal F, et al. Alternatives to
Roholah Sharifi, and Miguel González-Andrades eye bank native tissue for corneal stromal replace-
declare that they have no conflict of interest. No ment. Prog Retin Eye Res. 2017;59:97–130.
8. Hara H, Cooper DK.  Xenotransplantation–the
human or animal studies were carried out by the future of corneal transplantation? Cornea. 2011;30:
authors for this article. 371–8.
32 M. M. Islam et al.

9. Whitcher JP, Srinivasan M, Upadhyay MP.  Corneal 28. Apte RS, Niederkorn JY.  Isolation and characteriza-
blindness: a global perspective. Bull World Health tion of a unique natural killer cell inhibitory factor
Organ. 2001;79:214–21. present in the anterior chamber of the eye. J Immunol.
10.
Abud TB, Di Zazzo A, Kheirkhah A, Dana 1996;156:2667–73.
R.  Systemic immunomodulatory strategies in high-­ 29. Niederkorn JY.  Role of innate immune system in

risk corneal transplantation. J Ophthalmic Vis Res. corneal transplantation. New Delhi: Jaypee Brothers
2017;12:81–92. Medical Publishers (P) Ltd; 2013.
11. Polisetti N, Islam MM, Griffith M. The artificial cor- 30. Foulsham W, Marmalidou A, Amouzegar A, Coco G,
nea. Methods Mol Biol. 2013;1014:45–52. Chen Y, Dana R. Review: the function of regulatory T
12. Pascolini D, Mariotti SP.  Global estimates of visual cells at the ocular surface. Ocul Surf. 2017;15:652–9.
impairment: 2010. Br J Ophthalmol. 2012;96:614–8. 31. Qazi Y, Hamrah P.  Corneal allograft rejection:

13.
O’Day DM.  Diseases potentially transmitted Immunopathogenesis to therapeutics. J Clin Cell
through corneal transplantation. Ophthalmology. Immunol. 2013;2013:pii: 006.
1989;96:1133–7; discussion 7–8. 32. Kruse FE. Stem cells and corneal epithelial regenera-
14. Remeijer L, Maertzdorf J, Doornenbal P, Verjans
tion. Eye. 1994;8(Pt 2):170–83.
GM, Osterhaus AD.  Herpes simplex virus 1 trans- 33. Du Y, Funderburgh ML, Mann MM, SundarRaj N,
mission through corneal transplantation. Lancet. Funderburgh JL.  Multipotent stem cells in human
2001;357:442. corneal stroma. Stem Cells. 2005;23:1266–75.
15. Miller TD, Maxwell AJ, Lindquist TD, Requard J 3rd. 34. Funderburgh ML, Du Y, Mann MM, SundarRaj N,
Validation of cooling effect of insulated containers for Funderburgh JL.  PAX6 expression identifies pro-
the shipment of corneal tissue and recommendations genitor cells for corneal keratocytes. FASEB J: Off
for transport. Cornea. 2013;32:63–9. Publication Fed Am Soc Exp Biol. 2005;19:1371–3.
16. Prevention of Blindness and Visual Impairment.
35. Pellegrini G, Traverso CE, Franzi AT, Zingirian M,
Accessed 11 Nov 2017, at http://www.who.int/ Cancedda R, De Luca M.  Long-term restoration of
blindness/causes/magnitude/en/. damaged corneal surfaces with autologous cultivated
17. Gain P, Jullienne R, He Z, et  al. Global survey of corneal epithelium. Lancet. 1997;349:990–3.
corneal transplantation and eye banking. JAMA 36. Rama P, Matuska S, Paganoni G, Spinelli A, De

Ophthalmol. 2016;134:167–73. Luca M, Pellegrini G.  Limbal stem-cell therapy
18. Langer R, Vacanti JP.  Tissue engineering. Science. and long-term corneal regeneration. N Engl J Med.
1993;260:920–6. 2010;363:147–55.
19. Nerem RM. Cellular engineering. Ann Biomed Eng. 37. Sangwan VS, Basu S, Vemuganti GK, et al. Clinical
1991;19:529–45. outcomes of xeno-free autologous cultivated l­imbal
20. Ma Y, Xu Y, Xiao Z, et al. Reconstruction of chemi- epithelial transplantation: a 10-year study. Br J
cally burned rat corneal surface by bone marrow-­ Ophthalmol. 2011;95:1525–9.
derived human mesenchymal stem cells. Stem Cells. 38.
Nakamura T, Inatomi T, Sotozono C, et  al.
2006;24:315–21. Transplantation of autologous serum-derived culti-
21. Leijten J, Seo J, Yue K, et al. Spatially and temporally vated corneal epithelial equivalents for the treatment
controlled hydrogels for tissue engineering. Mater Sci of severe ocular surface disease. Ophthalmology.
Eng R Rep. 2017;119:1–35. 2006;113:1765–72.
22. Griffith M, Osborne R, Munger R, et  al. Functional 39. Rama P, Bonini S, Lambiase A, et  al. Autologous
human corneal equivalents constructed from cell fibrin-cultured limbal stem cells permanently restore
lines. Science. 1999;286:2169–72. the corneal surface of patients with total limbal stem
23. Boote C, Dennis S, Newton RH, Puri H, Meek
cell deficiency. Transplantation. 2001;72:1478–85.
KM.  Collagen fibrils appear more closely packed 40. Dravida S, Gaddipati S, Griffith M, et  al. A biomi-
in the prepupillary cornea: optical and biome- metic scaffold for culturing limbal stem cells: a prom-
chanical implications. Invest Ophthalmol Vis Sci. ising alternative for clinical transplantation. J Tissue
2003;44:2941–8. Eng Regen Med. 2008;2:263–71.
24. Marfurt CF, Cox J, Deek S, Dvorscak L.  Anatomy 41. Sangwan VS, Basu S, MacNeil S, Balasubramanian
of the human corneal innervation. Exp Eye Res. D.  Simple limbal epithelial transplantation (SLET):
2010;90:478–92. a novel surgical technique for the treatment of uni-
25. Sweeney DF, Xie RZ, O'Leary DJ, et al. Nutritional lateral limbal stem cell deficiency. Br J Ophthalmol.
requirements of the corneal epithelium and anterior 2012;96:931–4.
stroma: clinical findings. Invest Ophthalmol Vis Sci. 42. Di Girolamo N, Bosch M, Zamora K, Coroneo MT,
1998;39:284–91. Wakefield D, Watson SL. A contact lens-based tech-
26. DiMattio J. In vivo entry of glucose analogs into lens nique for expansion and transplantation of autologous
and cornea of the rat. Invest Ophthalmol Vis Sci. epithelial progenitors for ocular surface reconstruc-
1984;25:160–5. tion. Transplantation. 2009;87:1571–8.
27. Holden BA, Mertz GW. Critical oxygen levels to avoid 43. Sehic A, Utheim OA, Ommundsen K, Utheim TP. Pre-
corneal edema for daily and extended wear contact clinical cell-based therapy for Limbal stem cell defi-
lenses. Invest Ophthalmol Vis Sci. 1984;25:1161–7. ciency. J Funct Biomater. 2015;6:863–88.
3  Corneal Tissue Engineering 33

44. Chun YS, Park IK, Kim JC. Technique for autologous severe ocular surface disorders: long-term survival
nasal mucosa transplantation in severe ocular surface analysis. Indian J Ophthalmol. 2013;61:202–7.
disease. Eur J Ophthalmol. 2011;21:545–51. 59. Tan DT, Ang LP, Beuerman RW.  Reconstruction of
45. Kim JH, Chun YS, Lee SH, et  al. Ocular surface the ocular surface by transplantation of a serum-free
reconstruction with autologous nasal mucosa in derived cultivated conjunctival epithelial equivalent.
cicatricial ocular surface disease. Am J Ophthalmol. Transplantation. 2004;77:1729–34.
2010;149:45–53. 60. Burillon C, Huot L, Justin V, et al. Cultured autolo-
46. Monteiro BG, Serafim RC, Melo GB, et  al. Human gous oral mucosal epithelial cell sheet (CAOMECS)
immature dental pulp stem cells share key charac- transplantation for the treatment of corneal limbal
teristic features with limbal stem cells. Cell Prolif. epithelial stem cell deficiency. Invest Ophthalmol Vis
2009;42:587–94. Sci. 2012;53:1325–31.
47. Gomes JA, Geraldes Monteiro B, Melo GB, et  al. 61. Inatomi T, Nakamura T, Kojyo M, Koizumi N,

Corneal reconstruction with tissue-engineered cell Sotozono C, Kinoshita S.  Ocular surface recon-
sheets composed of human immature dental pulp stem struction with combination of cultivated autologous
cells. Invest Ophthalmol Vis Sci. 2010;51:1408–14. oral mucosal epithelial transplantation and pen-
48. Meyer-Blazejewska EA, Call MK, Yamanaka O, et al. etrating keratoplasty. Am J Ophthalmol. 2006;142:
From hair to cornea: toward the therapeutic use of hair 757–64.
follicle-derived stem cells in the treatment of limbal 62. Nakamura T, Takeda K, Inatomi T, Sotozono C,

stem cell deficiency. Stem Cells. 2011;29:57–66. Kinoshita S.  Long-term results of autologous culti-
49. Garzon I, Martin-Piedra MA, Alfonso-Rodriguez C, vated oral mucosal epithelial transplantation in the
et al. Generation of a biomimetic human artificial cor- scar phase of severe ocular surface disorders. Br J
nea model using Wharton’s jelly mesenchymal stem Ophthalmol. 2011;95:942–6.
cells. Invest Ophthalmol Vis Sci. 2014;55:4073–83. 63. Nishida K, Yamato M, Hayashida Y, et  al. Corneal
50. Reza HM, Ng BY, Gimeno FL, Phan TT, Ang
reconstruction with tissue-engineered cell sheets
LP.  Umbilical cord lining stem cells as a novel and composed of autologous oral mucosal epithelium. N
promising source for ocular surface regeneration. Engl J Med. 2004;351:1187–96.
Stem Cell Rev. 2011;7:935–47. 64. Takeda K, Nakamura T, Inatomi T, Sotozono

51. Omoto M, Miyashita H, Shimmura S, et al. The use C, Watanabe A, Kinoshita S.  Ocular surface
of human mesenchymal stem cell-derived feeder cells ­reconstruction using the combination of autologous
for the cultivation of transplantable epithelial sheets. cultivated oral mucosal epithelial transplantation and
Invest Ophthalmol Vis Sci. 2009;50:2109–15. eyelid surgery for severe ocular surface disease. Am J
52. Rohaina CM, Then KY, Ng AM, et al. Reconstruction Ophthalmol. 2011;152:195–201. e1
of limbal stem cell deficient corneal surface with 65. Utheim TP.  Concise review: transplantation of cul-
induced human bone marrow mesenchymal stem tured oral mucosal epithelial cells for treating limbal
cells on amniotic membrane. Transl Res. 2014;163: stem cell deficiency-current status and future perspec-
200–10. tives. Stem Cells. 2015;33:1685–95.
53. Lin KJ, Loi MX, Lien GS, et al. Topical administra- 66. Ahmad S, Stewart R, Yung S, et al. Differentiation of
tion of orbital fat-derived stem cells promotes corneal human embryonic stem cells into corneal epithelial-­
tissue regeneration. Stem Cell Res Ther. 2013;4:72. like cells by in vitro replication of the corneal epithe-
54. Hayashi R, Ishikawa Y, Ito M, et  al. Generation of lial stem cell niche. Stem Cells. 2007;25:1145–55.
corneal epithelial cells from induced pluripotent stem 67. Zhang C, Du L, Pang K, Wu X.  Differentiation of
cells derived from human dermal fibroblast and cor- human embryonic stem cells into corneal epithelial
neal limbal epithelium. PLoS One. 2012;7:e45435. progenitor cells under defined conditions. PLoS One.
55. Ang LP, Tanioka H, Kawasaki S, et  al. Cultivated 2017;12:e0183303.
human conjunctival epithelial transplantation for total 68. Reza HM, Ng BY, Phan TT, Tan DT, Beuerman RW,
limbal stem cell deficiency. Invest Ophthalmol Vis Ang LP.  Characterization of a novel umbilical cord
Sci. 2010;51:758–64. lining cell with CD227 positivity and unique pat-
56.
Sangwan VS, Vemuganti GK, Singh S, tern of P63 expression and function. Stem Cell Rev.
Balasubramanian D.  Successful reconstruction of 2011;7:624–38.
damaged ocular outer surface in humans using lim- 69. Zieske JD. Corneal development associated with eye-
bal and conjunctival stem cell culture methods. Biosci lid opening. Int J Dev Biol. 2004;48:903–11.
Rep. 2003;23:169–74. 70. Meek KM, Knupp C. Corneal structure and transpar-
57. Sangwan VS, Vemuganti GK, Iftekhar G, Bansal AK, ency. Prog Retin Eye Res. 2015;49:1–16.
Rao GN. Use of autologous cultured limbal and con- 71. Carlson EC, Wang IJ, Liu CY, Brannan P, Kao CW,
junctival epithelium in a patient with severe bilateral Kao WW.  Altered KSPG expression by keratocytes
ocular surface disease induced by acid injury: a case following corneal injury. Mol Vis. 2003;9:615–23.
report of unique application. Cornea. 2003;22:478–81. 72. Jester JV, Petroll WM, Cavanagh HD.  Corneal stro-
58. Subramaniam SV, Sejpal K, Fatima A, Gaddipati S, mal wound healing in refractive surgery: the role
Vemuganti GK, Sangwan VS.  Coculture of autolo- of myofibroblasts. Prog Retin Eye Res. 1999;18:
gous limbal and conjunctival epithelial cells to treat 311–56.
34 M. M. Islam et al.

73. Fini ME.  Keratocyte and fibroblast phenotypes 88. Liu L, Sheardown H.  Glucose permeable poly
in the repairing cornea. Prog Retin Eye Res. (dimethyl siloxane) poly (N-isopropyl acrylamide)
1999;18:529–51. interpenetrating networks as ophthalmic biomateri-
74. Buss DG, Giuliano EA, Sharma A, Mohan als. Biomaterials. 2005;26:233–44.
RR.  Isolation and cultivation of equine corneal 89. Deng ZX, Zhang ZX, Li LH, Zhou
keratocytes, fibroblasts and myofibroblasts. Vet CR.  Biocompatibility evaluation of chitosan-g-­
Ophthalmol. 2010;13:37–42. polyvinylpyrrolidone. Di Yi Jun Yi Da Xue Xue Bao.
75. Patel DV, McKelvie J, Sherwin T, McGhee 2004;24:639–41.
C.  Keratocyte progenitor cell transplantation: a 90. Bae SR, Park C, Choi JC, Poo H, Kim CJ, Sung
novel therapeutic strategy for corneal disease. Med MH.  Effects of ultra high molecular weight poly-­
Hypotheses. 2013;80:122–4. gamma-­glutamic acid from Bacillus subtilis (chun-
76. Basu S, Hertsenberg AJ, Funderburgh ML, et  al. gkookjang) on corneal wound healing. J Microbiol
Human limbal biopsy-derived stromal stem Biotechnol. 2010;20:803–8.
cells prevent corneal scarring. Sci Transl Med. 91. Sharma CP.  Biointegration of medical implant
2014;6:266ra172. materials: science and design. New  York: Elsevier
77. Wu J, Rnjak-Kovacina J, Du Y, Funderburgh ML, Science; 2010.
Kaplan DL, Funderburgh JL.  Corneal stromal bio- 92. Li F, Griffith M, Li Z, et  al. Recruitment of mul-
equivalents secreted on patterned silk substrates. tiple cell lines by collagen-synthetic copolymer
Biomaterials. 2014;35:3744–55. matrices in corneal regeneration. Biomaterials.
78. Du Y, Sundarraj N, Funderburgh ML, Harvey SA, 2005;26:3093–104.
Birk DE, Funderburgh JL.  Secretion and organiza- 93. Duan X, Sheardown H.  Dendrimer crosslinked
tion of a cornea-like tissue in  vitro by stem cells collagen as a corneal tissue engineering scaffold:
from human corneal stroma. Invest Ophthalmol Vis mechanical properties and corneal epithelial cell
Sci. 2007;48:5038–45. interactions. Biomaterials. 2006;27:4608–17.
79. Naylor RW, McGhee CN, Cowan CA, Davidson 94. McLaughlin CR, Acosta MC, Luna C, et  al.
AJ, Holm TM, Sherwin T.  Derivation of corneal Regeneration of functional nerves within full
Keratocyte-like cells from human induced pluripo- thickness collagen-phosphorylcholine corneal
tent stem cells. PLoS One. 2016;11:e0165464. substitute implants in guinea pigs. Biomaterials.
80. Bahn CF, Falls HF, Varley GA, Meyer RF, 2010;31:2770–8.
Edelhauser HF, Bourne WM.  Classification of cor- 95. Liu W, Deng C, McLaughlin CR, et  al. Collagen-­
neal endothelial disorders based on neural crest ori- phosphorylcholine interpenetrating network
gin. Ophthalmology. 1984;91:558–63. hydrogels as corneal substitutes. Biomaterials.
81. Hatou S, Yamada M, Mochizuki H, Shiraishi A, Joko 2009;30:1551–9.
T, Nishida T.  The effects of dexamethasone on the 96. Duan X, McLaughlin C, Griffith M, Sheardown
Na,K-ATPase activity and pump function of corneal H.  Biofunctionalization of collagen for improved
endothelial cells. Curr Eye Res. 2009;34:347–54. biological response: scaffolds for corneal tissue
82. Kinoshita S, Koizumi N, Ueno M, Okumura N, Imai engineering. Biomaterials. 2007;28:78–88.
K, Tanaka H, Yamamoto Y, Nakamura T, Inatomi T, 97. Lai JY, Li YT. Functional assessment of cross-linked
Bush J, Toda M, Hagiya M, Yokota I, Teramukai S, porous gelatin hydrogels for bioengineered cell sheet
Sotozono C, Hamuro J.  Injection of cultured cells carriers. Biomacromolecules. 2010;11:1387–97.
with a ROCK inhibitor for bullous keratopathy. N 98. Alaminos M, Del Carmen Sanchez-Quevedo M,
Engl J Med. 2018;378(11):995–1003. https://doi. Munoz-Avila JI, et  al. Construction of a complete
org/10.1056/NEJMoa1712770. PMID:29539291. rabbit cornea substitute using a fibrin-agarose scaf-
83. Chan BP, Leong KW. Scaffolding in tissue engineer- fold. Invest Ophthalmol Vis Sci. 2006;47:3311–7.
ing: general approaches and tissue-specific consider- 99. Lawrence BD, Marchant JK, Pindrus MA, Omenetto
ations. Eur Spine J. 2008;17(Suppl 4):467–79. FG, Kaplan DL. Silk film biomaterials for cornea tis-
84. Grinstaff MW. Designing hydrogel adhesives for cor- sue engineering. Biomaterials. 2009;30:1299–308.
neal wound repair. Biomaterials. 2007;28:5205–14. 100. Grolik M, Szczubialka K, Wowra B, et al. Hydrogel
85. Rafat M, Li F, Fagerholm P, et  al. PEG-stabilized membranes based on genipin-cross-linked chitosan
carbodiimide crosslinked collagen-chitosan hydro- blends for corneal epithelium tissue engineering. J
gels for corneal tissue engineering. Biomaterials. Mater Sci Mater Med. 2012;23:1991–2000.
2008;29:3960–72. 101. Vrana NE, Builles N, Justin V, et  al. Development
86. Aldave AJ, Kamal KM, Vo RC, Yu F.  The Boston of a reconstructed cornea from collagen-­chondroitin
type I keratoprosthesis: improving outcomes and sulfate foams and human cell cultures. Invest
expanding indications. Ophthalmology. 2009;116: Opthalmol Vis Sci. 2008;49:5325–31.
640–51. 102. Maia J, Ribeiro MP, Ventura C, Carvalho RA,
87. Zorlutuna P, Tezcaner A, Kiyat I, Aydinli A, Hasirci Correia IJ, Gil MH.  Ocular injectable formulation
V.  Cornea engineering on polyester carriers. J assessment for oxidized dextran-based hydrogels.
Biomed Mater Res A. 2006;79:104–13. Acta Biomater. 2009;5:1948–55.
3  Corneal Tissue Engineering 35

103. Fiorica C, Senior RA, Pitarresi G, et  al. 116. Crapo PM, Gilbert TW, Badylak SF. An overview of
Biocompatible hydrogels based on hyaluronic acid tissue and whole organ decellularization processes.
cross-linked with a polyaspartamide derivative as Biomaterials. 2011;32:3233–43.
delivery systems for epithelial limbal cells. Int J 117. Gonzalez-Andrades M, de la Cruz Cardona J,
Pharm. 2011;414:104–11. Ionescu AM, Campos A, Del Mar Perez M, Alaminos
104. Tonsomboon K, Oyen ML.  Composite electrospun M. Generation of bioengineered corneas with decel-
gelatin fiber-alginate gel scaffolds for mechani- lularized xenografts and human keratocytes. Invest
cally robust tissue engineered cornea. J Mech Behav Ophthalmol Vis Sci. 2011;52:215–22.
Biomed. 2013;21:185–94. 118. Brown BN, Valentin JE, Stewart-Akers AM,
105. Hamley IW, Dehsorkhi A, Castelletto V, et al. Self-­ McCabe GP, Badylak SF.  Macrophage phenotype
assembly and collagen-stimulating activity of a pep- and remodeling outcomes in response to biologic
tide Amphiphile incorporating a peptide sequence scaffolds with and without a cellular component.
from Lumican. Langmuir. 2015;31:4490–5. Biomaterials. 2009;30:1482–91.
106. Gouveia RM, González-Andrades E, Cardona 119. Zhang MC, Liu X, Jin Y, Jiang DL, Wei XS, Xie
JC, González-Gallardo C, Ionescu AM, Garzon HT. Lamellar Keratoplasty treatment of fungal cor-
I, Alaminos M, González-Andrades M, Connon neal ulcers with acellular porcine corneal stroma.
CJ.  Controlling the 3D architecture of Self-Lifting Am J Transplant. 2015;15:1068–75.
Auto-generated Tissue Equivalents (SLATEs) for 120. Nishida T, Nakamura M, Ofuji K, Reid TW, Mannis
optimized corneal graft composition and stabil- MJ, Murphy CJ. Synergistic effects of substance P
ity. Biomaterials. 2017;121:205–19. https://doi. with insulin-like growth factor-1 on epithelial migra-
org/10.1016/j.biomaterials.2016.12.023. Epub 2016 tion of the cornea. J Cell Physiol. 1996;169:159–66.
Dec 23. PMID: 28092777. 121. Holan V, Trosan P, Krulova M, Zajicova
107. Islam MM, Ravichandran R, Olsen D, et  al. Self-­ A. Identification of factors regulating differentiation
assembled collagen-like-peptide implants as alterna- and growth of limbal stem cells for corneal surface
tives to human donor corneal transplantation. RSC regeneration. Acta Ophthalmologica. 2012;90:0.
Adv. 2016;6:55745–9. 122. Zieske JD, Takahashi H, Hutcheon AE, Dalbone
108. Jangamreddy JR, Haagdorens MKC, Mirazul
AC.  Activation of epidermal growth factor recep-
Islam M, et  al. Short peptide analogs as alter- tor during corneal epithelial migration. Invest
natives to collagen in pro-regenerative corneal Ophthalmol Vis Sci. 2000;41:1346–55.
implants. Acta Biomaterialia. 2018. pii: S1742– 123. Wilson SE, Walker JW, Chwang EL, He
7061(18)30022–9; https://doi.org/10.1016/j. YG. Hepatocyte growth factor, keratinocyte growth
actbio.2018.01.011. factor, their receptors, fibroblast growth factor recep-
109. Kumano Y, Sakamoto T, Egawa M, Tanaka M, tor-­2, and the cells of the cornea. Invest Ophthalmol
Yamamoto I.  Enhancing effect of 2-O-alpha-D- Vis Sci. 1993;34:2544–61.
glucopyranosyl-­ L-ascorbic acid, a stable ascorbic 124. Daniels JT, Limb GA, Saarialho-Kere U, Murphy
acid derivative, on collagen synthesis. Biol Pharm G, Khaw PT. Human corneal epithelial cells require
Bull. 1998;21:662–6. MMP-1 for HGF-mediated migration on collagen
110. Saika S.  Ultrastructural effect of L-ascorbic acid I. Invest Ophthalmol Vis Sci. 2003;44:1048–55.
2-phosphate on cultured keratocytes. Cornea. 125. Kakazu A, Chandrasekher G, Bazan HE. HGF pro-
1992;11:439–45. tects corneal epithelial cells from apoptosis by the
111. Cissell DD, Hu JC, Griffiths LG, Athanasiou PI-3K/Akt-1/Bad- but not the ERK1/2-mediated
KA. Antigen removal for the production of biome- signaling pathway. Invest Ophthalmol Vis Sci.
chanically functional, xenogeneic tissue grafts. J 2004;45:3485–92.
Biomech. 2014;47:1987–96. 126. Wilson SE, Li Q, Mohan RR, et al. Lacrimal gland
112. Gilbert TW, Sellaro TL, Badylak SF. growth factors and receptors: lacrimal fibroblastic
Decellularization of tissues and organs. Biomaterials. cells are a source of tear HGF. Adv Exp Med Biol.
2006;27:3675–83. 1998;438:625–8.
113. Vorotnikova E, McIntosh D, Dewilde A, et  al. 127. Haber M, Cao Z, Panjwani N, Bedenice D, Li
Extracellular matrix-derived products modulate WW, Provost PJ.  Effects of growth factors (EGF,
endothelial and progenitor cell migration and pro- PDGF-BB and TGF-beta 1) on cultured equine epi-
liferation in vitro and stimulate regenerative healing thelial cells and keratocytes: implications for wound
in vivo. Matrix Biol. 2010;29:690–700. healing. Vet Ophthalmol. 2003;6:211–7.
114. Barkan D, Green JE, Chambers AF.  Extracellular 128. Andresen JL, Ledet T, Ehlers N. Keratocyte migra-
matrix: a gatekeeper in the transition from dor- tion and peptide growth factors: the effect of PDGF,
mancy to metastatic growth. Eur J Cancer. 2010;46: bFGF, EGF, IGF-I, aFGF and TGF-beta on human
1181–8. keratocyte migration in a collagen gel. Curr Eye Res.
115. Petersen TH, Calle EA, Zhao L, et  al. Tissue-­ 1997;16:605–13.
engineered lungs for in vivo implantation. Science. 129. Jester JV, Huang J, Fisher S, et  al. Myofibroblast
2010;329:538–41. differentiation of normal human keratocytes and
36 M. M. Islam et al.

hTERT, extended-life human corneal fibroblasts. 143. Kamiyama K, Iguchi I, Wang X, Imanishi
Invest Ophthalmol Vis Sci. 2003;44:1850–8. J.  Effects of PDGF on the migration of rabbit cor-
130. Daniels JT, Khaw PT. Temporal stimulation of cor- neal fibroblasts and epithelial cells. Cornea. 1998;
neal fibroblast wound healing activity by differenti- 17:315–25.
ating epithelium in vitro. Invest Ophthalmol Vis Sci. 144. Sosne G, Siddiqi A, Kurpakus-Wheater
2000;41:3754–62. M.  Thymosin-beta4 inhibits corneal epithelial cell
131. Bonini S, Lambiase A, Rama P, Caprioglio G, apoptosis after ethanol exposure in  vitro. Invest
Aloe L.  Topical treatment with nerve growth fac- Ophthalmol Vis Sci. 2004;45:1095–100.
tor for neurotrophic keratitis. Ophthalmology. 145. Sosne G, Qiu P, Ousler GW 3rd, Dunn SP,
2000;107:1347–51; discussion 51–2. Crockford D.  Thymosin beta4: a potential novel
132. Cellini M, Bendo E, Bravetti GO, Campos EC. The dry eye therapy. Ann N Y Acad Sci. 2012;1270:
use of nerve growth factor in surgical wound healing 45–50.
of the cornea. Ophthalmic Res. 2006;38:177–81. 146. Dunn SP, Heidemann DG, Chow CY, et al. Treatment
133. Ma K, Yan N, Huang Y, Cao G, Deng J, Deng of chronic nonhealing neurotrophic corneal epithe-
Y. Effects of nerve growth factor on nerve regenera- lial defects with thymosin beta4. Ann N Y Acad Sci.
tion after corneal nerve damage. Int J Clin Exp Med. 2010;1194:199–206.
2014;7:4584–9. 147. Lambiase A, Bonini S, Micera A, Rama P, Bonini S,
134. McLaughlin PJ, Sassani JW, Klocek MS, Zagon Aloe L. Expression of nerve growth factor receptors
IS.  Diabetic keratopathy and treatment by modula- on the ocular surface in healthy subjects and dur-
tion of the opioid growth factor (OGF)-OGF recep- ing manifestation of inflammatory diseases. Invest
tor (OGFr) axis with naltrexone: a review. Brain Res Ophthalmol Vis Sci. 1998;39:1272–5.
Bull. 2010;81:236–47. 148. Lambiase A, Bonini S, Aloe L, Rama P, Bonini
135. Nakamura Y, Sotozono C, Kinoshita S.  The epi- S.  Anti-inflammatory and healing properties of
dermal growth factor receptor (EGFR): role in cor- nerve growth factor in immune corneal ulcers with
neal wound healing and homeostasis. Exp Eye Res. stromal melting. Arch Ophthalmol. 2000;118:
2001;72:511–7. 1446–9.
136. Yanai R, Yamada N, Inui M, Nishida T. Correlation 149. Joo M, Yuhan KR, Hyon J, et  al. The effect of
of proliferative and anti-apoptotic effects of nerve growth factor on corneal sensitivity after
HGF, insulin, IGF-1, IGF-2, and EGF in SV40-­ laserin situ keratomileusis. Arch Ophthalmol.
transformed human corneal epithelial cells. Exp Eye 2004;122:1338–41.
Res. 2006;83:76–83. 150. Zagon IS, Sassani JW, McLaughlin PJ.  Opioid
137. Chandrasekher G, Kakazu AH, Bazan HE.  HGF- growth factor modulates corneal epithelial out-
and KGF-induced activation of PI-3K/p70 s6 kinase growth in tissue culture. Am J Phys. 1995;268:
pathway in corneal epithelial cells: its relevance in R942–50.
wound healing. Exp Eye Res. 2001;73:191–202. 151. Wang L, Gonzalez S, Dai W, Deng S, Lu L. Effect
138. Lee JG, Kay EP.  FGF-2-induced wound healing in of hypoxia-regulated polo-like kinase 3 (Plk3) on
corneal endothelial cells requires Cdc42 activation human Limbal stem cell differentiation. J Biol
and Rho inactivation through the phosphatidylino- Chem. 2016;291:16519–29.
sitol 3-kinase pathway. Invest Ophthalmol Vis Sci. 152. Hu N, Zhang YY, Gu HW, Guan HJ. Effects of bone
2006;47:1376–86. marrow mesenchymal stem cells on cell prolifera-
139. Trosan P, Svobodova E, Chudickova M, Krulova M, tion and growth factor expression of limbal epithe-
Zajicova A, Holan V.  The key role of insulin-like lial cells in vitro. Ophthalmic Res. 2012;48:82–8.
growth factor I in limbal stem cell differentiation and 153. Gonfiotti A, Jaus MO, Barale D, et  al. The first
the corneal wound-healing process. Stem Cells Dev. tissue-engineered airway transplantation: 5-year
2012;21:3341–50. follow-up results. Lancet. 2014;383:238–44.
140. Pancholi S, Tullo A, Khaliq A, Foreman D, Boulton 154. Hibino N, McGillicuddy E, Matsumura G, et al. Late-
M.  The effects of growth factors and conditioned term results of tissue-engineered ­vascular grafts in
media on the proliferation of human corneal epi- humans. J Thorac Cardiovasc Surg. 2010;139:431–
thelial cells and keratocytes. Graefes Arch Clin Exp 6. 6.e1–2.
Ophthalmol. 1998;236:1–8. 155. Atala A, Bauer SB, Soker S, Yoo JJ, Retik AB. Tissue-­
141. Kay EP, Lee MS, Seong GJ, Lee YG.  TGF-beta s engineered autologous bladders for patients needing
stimulate cell proliferation via an autocrine produc- cystoplasty. Lancet. 2006;367:1241–6.
tion of FGF-2  in corneal stromal fibroblasts. Curr 156. Alió del Barrio JL, El Zarif M, Azaar A, et  al.
Eye Res. 1998;17:286–93. Corneal stroma enhancement with decellularized
142. Denk PO, Knorr M. The in vitro effect of platelet-­ stromal laminas with or without stem cell recellular-
derived growth factor isoforms on the proliferation ization for advanced keratoconus. Am J Ophthalmol.
of bovine corneal stromal fibroblasts depends on 2018;186:47–58.
cell density. Graefes Arch Clin Exp Ophthalmol. 157. Fagerholm P, Lagali NS, Ong JA, et al. Stable cor-
1997;235:530–4. neal regeneration four years after implantation of
3  Corneal Tissue Engineering 37

a cell-free recombinant human collagen scaffold. 161. Massey LK. Chapter 38 – acrylic polycarbonate alloy
Biomaterials. 2014;35:2420–7. (Acrylic PC). In: The effect of sterilization meth-
158. Buznyk O, Pasyechnikova N, Islam MM, Iakymenko ods on plastics and elastomers. 2nd ed. Norwich:
S, Fagerholm P, Griffith M.  Bioengineered cor- William Andrew Publishing; 2005. p. 241–4.
neas grafted as alternatives to human donor cor- 162. U.S.  FDA: Tissue & tissue products. Accessed
neas in three high-risk patients. Clin Transl Sci. 13 Dec 2017, at https://www.fda.gov/Biologics
2015;8:558–62. BloodVaccines/TissueTissueProducts/.
159. Islam MM, Buznyk O, Reddy JC, et al. Biomaterials-­ 163. U.S.  FDA: Combination product definition.
enabled cornea regeneration in patients at high risk Accessed 13 Dec 2017, at https://www.fda.gov/
for rejection of donor tissue transplantation. NPJ CombinationProducts/AboutCombinationProducts/
Regen Med. 2018;3:2. ucm118332.htm.
160. Gonzalez-Andrades M, Mata R, Gonzalez-Gallardo 164. FDA announces comprehensive regenerative
MDC, et  al. A study protocol for a multicentre medicine policy framework. November 16, 2017.
randomised clinical trial evaluating the safety and Accessed 11 Dec 2017, at https://www.fda.gov/
feasibility of a bioengineered human allogeneic NewsEvents/Newsroom/PressAnnouncements/
nanostructured anterior cornea in patients with ucm585345.htm.
advanced corneal trophic ulcers refractory to con-
ventional treatment. BMJ Open. 2017;7:e016487.
Part II
The Stem Cell
Stem Cells: Concept, Properties,
and Characterization
4
Natalia Escacena-Acosta, Javier Lopez-Beas,
Christian Claude Lachaud, Mehrdad Vakilian,
Juan Rigoberto Tejedo, Vivian Capilla-González,
Francisco Javier Bedoya, Franz Martin,
Abdelkrim Hmadcha, and Bernat Soria

4.1 Concept ium of these two properties is the capacity to colo-


nize and reconstitute a tissue (as in hematopoietic
The biology of stem cells (SCs) is one of the most SC transplantation). Currently, proliferation and dif-
important areas in current biomedical research. ferentiation use opposite intracellular pathways, and
Currently, they have an important role in regenera- most times forcing one of them will block the other.
tive medicine and many potential applications. Due
to involvement of SCs in development and their
4.2 Classification
potential to give rise to different cell types, SCs are
essential in the field of developmental biology. SCs
The SC can be classified according to their dif-
are characterized by the capacity to proliferate and
ferentiation potential or based on their source [1]:
generate daughter cells that keep the same phenotype
(self-renewal, symmetric division) and the capacity
to differentiate into other cell types (differentiation, 4.2.1 Source
asymmetric division). While the first property allows
reaching enough cell mass to fulfill a physiological (i) Embryonic
function, the former allows obtaining postmitotic At the first days of development, an outer
differentiated cell types or somatic cells. A corolar- layer of cells committed to becoming part

N. Escacena-Acosta · J. Lopez-Beas · C. C. Lachaud


Centro de Investigación Biomédica en Red de
M. Vakilian · V. Capilla-González
Diabetes y Enfermedades Metabólicas Asociadas
Department of Cell Regeneration and Advanced
(CIBERDEM), Madrid, Spain
Therapies, Andalusian Center of Molecular Biology
and Regenerative Medicine-CABIMER, Junta de University Pablo de Olavide, Seville, Spain
Andalucía-University of Pablo Olavide-University of e-mail: bernat.soria@cabimer.es
Seville-CSIC, Seville, Andalusia, Spain
A. Hmadcha (*)
J. R. Tejedo · F. J. Bedoya · F. Martin · B. Soria (*) Department of Cell Regeneration and Advanced
Department of Cell Regeneration and Advanced Therapies, Andalusian Center of Molecular Biology
Therapies, Andalusian Center of Molecular Biology and Regenerative Medicine-CABIMER, Junta de
and Regenerative Medicine-CABIMER, Junta de Andalucía-University of Pablo Olavide-University of
Andalucía-University of Pablo Olavide-University of Seville-CSIC, Seville, Andalusia, Spain
Seville-CSIC, Seville, Andalusia, Spain e-mail: karim.hmadcha@cabimer.es

© Springer Nature Switzerland AG 2019 41


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_4
42 N. Escacena-Acosta et al.

of the placenta (trophectoderm identity) 4.2.2 Potentiality


separates from the inner cell mass (ICM),
the part of the blastocyst that is fated to (iv) Totipotential
become the embryo, amnion, and yolk sac. Only the zygote or morula cells may be con-
These pluripotent cells from inner cell mass sidered totipotential since they may give rise
(ICM) of embryo, ICM-derived cells, are to either the inner cell mass (that will dif-
ES cells. ferentiate into any cell type of the fetus and
(ii) Placental and Umbilical Cord Stem Cells adult tissues) or the trophectoderm (which
The human placenta, cord blood, and umbil- contributes to the placental tissues) and gen-
ical cord are a rich source of hematopoietic erate a complete and viable organism.
stem cells (HSCs) and mesenchymal stem (v) Pluripotential
cells (MSCs), which possess enormous They can self-renew and differentiate into
regeneration potential. These SCs are multi- any of the three germ layers, ectoderm,
potent, meaning that they can differentiate endoderm, and mesoderm, from which all
into many different types of cells. tissues and organs develop. Embryonic stem
(iii)  Adult Stem Cells cells (ESCs) and induced pluripotent stem
Adult stem cells or tissue-specific stem cells (iPSCs) are currently the only two
cells are unspecialized cells that have been types known of pluripotent SCs.
found in a particular small area of many tis- (vi) Multipotential
sues. Adult stem cells are capable of long- They are self-renewing and developing into
term renewal and differentiation into closely related cell types but are more limited
specialized cell types. During the last years, than pluripotent cells. Mesenchymal stem cells
isolations of adult stem cells from different (MSCs) and hematopoietic stem cells (HSCs)
sources have been reported. Bone marrow- are a typical example of multipotent cells.
derived stem cells are still the most fre- (vii) Unipotential
quently investigated cell type. Mesenchymal They are progenitors with the capacity to
stem cells (MSCs) are an example of tissue divide and generate more cell mass of the
or “adult” stem cells. They are “multipo- same cell type (e.g., the hepatocyte).
tent,” meaning they can produce more than
one type of specialized cell of the body, but
not all types. MSCs make the different spe- 4.3 Properties
cialized cells found in the skeletal tissues. and Characterization
MSCs are considered clonogenic and non-
hematopoietic and are characterized by For decades, researchers have been studying the
their easy isolation and for presenting the biology of SCs and the molecular mechanisms
ability to differentiate into multilineage cell governing the pluripotency and differentiation of
types, including osteoblasts, chondrocytes, the SCs, by a complex interconnected relation-
endothelial cells, and even neural-like cells. ship between an intrinsic transcriptional program
MSCs constituted a heterogeneous cell and extrinsic signaling pathways.
population that contains progenitor cells at
different maturation stages. MSCs were
derived and characterized using different 4.3.1 Self-Renewal
methods; to best match a more uniform
characterization of MSCs, the Mesenchymal Self-renewal program of SCs is conferred by a
and Tissue Stem Cell Committee of the specific transcriptional network and cell cycle
International Society for Cellular Therapy regulation. Coordinated transcription factor net-
(ISCT) proposed minimal criteria to define works act as the master regulatory mechanisms
human MSC [2]. of SC pluripotency and differentiation. Three
4  Stem Cells: Concept, Properties, and Characterization 43

key pluripotent transcription factors OCT4, NANOG-OCT4-SOX2 transcriptional core in


SOX2, and NANOG constitute the essential tran- ESCs was shown to be essential to induce a sta-
scriptional core maintaining the self-renewing ble intrinsic pluripotency network to confer an
undifferentiated SCs [3]. The core pluripotency indefinite self-renewal.
circuitry is orchestrated to promote expression Several signaling pathways that can support
of genes that maintain pluripotency while pluripotent cell maintenance have been identi-
repressing genes inducing differentiation. The fied, including Wnt/β-catenin, FGF/ERK, TGF/
POU domain transcription factor OCT4 and SMAD, and PKC signaling [6]. SCs require
SRY-­ related HMG-box transcription factor extrinsic growth factors for the maintenance of
SOX2 are critical for the maintenance of pluripo- pluripotency in culture, and the signaling path-
tency. Although the homeodomain-containing ways controlled by these factors act to maintain
transcription factor NANOG is also essential for the intrinsic transcription factor network required
the establishment of SCs from blastocysts, it is for pluripotency (Fig.  4.1). Human SCs require
however not absolutely required for the mainte- basic fibroblast growth factor (bFGF) and insulin
nance of pluripotency under optimized culture or insulin-like growth factor (IGF) signaling to
conditions [4, 5]. The core of pluripotency regu- support self-renewal. bFGF activates the
lators works in a combinatorial manner to con- mitogen-­ activated protein kinase (MAPK) as
trol a pluripotent-specific program of gene well as the Activin/Nodal signaling pathways.
expression and repression to maintain pluripo- IGF activates the Ras and PI3K/AKT while sup-
tency. Recent studies have identified many addi- pressing MEK/ERK pathways, leading to high
tional transcription factors within the SC GSK3β activity and low β-catenin-mediated
pluripotency circuitry. Their interaction with the transcriptional activation to enhance propagation

bFGF→Activin/Nodal
WN
T
P P P P

P
P P P P
P Extracelular
lin

P P P P
P
su

P β-Catenin P P SSEAs
In

P
F/

SSEA-3/4
IG

P
MAPK P Cytoplasm
P P
P SMADs
P P
P P 2/3
GSK3
MEK/ERK

Nucleus Stem cells (SCs)


PI3K/AKT markers
Self-renewal
ESRRB
miR-302/367
P
SMADs
P
P

1/5/8 NANOG
P

P
P

IncRNA-ES1/ES2
P

P
BM

LINC-ROR
SOX2 OCT4
TRA-1-60
TRA-1-81
Pluripotency TF network

AP

Fig. 4.1  Overview of cell markers for identifying human phosphatase (AP), SSEAs (SSEA-3 and SSEA-4), TRA-­
stem cells. Schematic representation summarizing the 1-­60, and TRA-1-81) defining stem cell populations and
expression of specific pluripotency factors (OCT4, critical pathways involved in the maintenance of pluripo-
NANOG, and SOX2) and cell-surface markers (alkaline tency and self-renewal
44 N. Escacena-Acosta et al.

and survival of human ESCs and prevent their entiation protocols to derive a specific lineage or
differentiation [7]. Therefore, modulation of key cell type progenitor that is required. Therefore,
extrinsic pathways such as TGF-β/Activin/ protocols have been developed for directly dif-
Nodal/SMAD2/3 and the BMP (bone morphoge- ferentiating SCs toward a specific germ layer. At
netic protein) signaling pathway is important to this point, it is possible to promote the efficient
maintain SC identity [8]. and reproducible development of the cell type
of interest. Protocols should recapitulate in vitro
the global developmental program taking place
4.3.2 Differentiation during the early embryogenesis. Recombinant
growth factors and small-molecule compounds
The cross talk between different signaling path- are commonly used to mimic the signaling path-
ways occurs constantly in SCs and results in a ways which are activated during embryonic
cooperative regulation of self-renewal and/or dif- development. Finally, differentiated cells that
ferentiation. The cell fate decision is accom- have been developing must display functional
plished through intricate mechanisms, involving properties that are characteristic of mature
a network of cross talk between signaling path- cell populations both in culture and when
ways, externa/internal cues, gene expression reg- transplanted.
ulation, and epigenetic modifications.
In recent years, a large number of growth fac-
tors have been identified, which depending on 4.3.3 Immunomodulation
their biological activity are capable of differenti- and Immunogenicity
ating SCs selectively toward mesodermal, endo-
dermal, or ectodermal lineages. As such, inducer Prior to clinical application, translational research
factors can play important roles during embryo- will be necessary in order to ensure that differen-
genesis and can be used to direct or support the tiated progenies generated are functional and
differentiation of SCs. All of these approaches stable without the ability to dedifferentiate, at the
increasingly underline the need to identify same time that is evaluated one of the major con-
inducer agents that can regulate the differentia- cerns for the clinical application of SC-derived
tion of SCs and reliably direct the differentiation grafts, such as the risks of immune rejection and
of SCs toward specific cell types, which offer the tumorigenicity. For a long time, SCs were con-
possibility of a renewable source of replacement sidered to have immune-privileged properties
cells and tissues to treat a myriad of diseases [9]. due to their low expression of major histocom-
Hence, during in  vitro differentiation experi- patibility complex (MHC) class I, MHC class II,
ments, a common approach is to treat the cells by and costimulatory molecules [1]. However, the
proper differentiation factors that promote an exit induction of SCs to become more differentiated
from the self-renewal program and drive cell dif- cells in vitro can result in their loss of immuno-
ferentiation. Under appropriate conditions, SCs logical privilege.
generate a progeny consisting of derivatives of Indeed, transplanted allogeneic and/or xeno-
the three embryonic germ layers: mesoderm, geneic SCs and their derivatives are not immune-­
endoderm, and ectoderm. privileged. iPSCs would provide a source of
Some critical stages common to all protocols pluripotent cells for autologous cell therapy
of differentiation should be considered when without the concern of immune rejection.
­
using the SC model for lineage-specific differ- However, iPSC-derived cells may express mole-
entiation. First, the earliest event in the devel- cules of embryonic origin associated with repro-
opment is the formation of three germ layers: gramming and other abnormalities, so that even
ectoderm, mesoderm, and endoderm. The poten- autologous iPSCs may be targets of immune
tial use of SCs to replace a functional loss of rejection [9, 10]. Hence, immunogenicity has
­particular tissues may depend on efficient differ- emerged as a significant problem not only with
4  Stem Cells: Concept, Properties, and Characterization 45

allogeneic SCs but also potentially with autolo- and are downregulated upon differentiation [14–
gous SC sources. 16]. TRA-1-60 and TRA-1-81 antigens expressed
Previous studies have demonstrated that tradi- on the surface of human pluripotent SCs are
tional immune suppression can prolong the sur- widely used as markers for identifying pluripo-
vival of graft of SC derivatives but not prevent tency [14, 15, 17]. Besides this, “stemness” of
immune rejection for a long-lasting transplanted cells can also be identified by functional assays.
cell survival [10–12]. In addition, the immune There are several additional marker systems
suppressants are highly toxic for patients and available to identify SCs; these include specific
have serious collateral damage such as increased enzyme activities, such as alkaline phosphatase
risk for cancer and infection. Therefore, it is (AP) and telomerase. The ubiquitous alkaline
important to develop safe and effective strategies phosphatase (AP) is a hydrolase enzyme respon-
to induce immune tolerance of SC-derived cells. sible for dephosphorylating molecules such as
To address this bottleneck, recent studies are nucleotides and proteins. As such, the pluripotent
focusing on addressing the immunomodulatory status of SCs can be characterized by a high level
capability of MSCs, due to its ability to prevent of AP expression [14, 18]. In the same way, SCs
immune rejection after implantation of both display high levels of telomerase activity and
autologous and allogeneic MSCs, through their hTERT expression, both of which are rapidly
intrinsic low immunogenicity and immunomodu- downregulated during differentiation [14, 19,
latory properties. Over the last few years, research 20]. Hence, telomerase activity and telomere
has been able to propose other cell types with maintenance are considered a general signature
immunomodulatory properties such as mesothe- of SCs.
lial cells capable of producing a wide number of OCT4 and NANOG were the first proteins
cytokines, growth factors, and extracellular identified as essential for both early embryo
matrix components possessing anti-inflammatory development and pluripotency maintenance in
and immunosuppressive properties [13]. SCs [15, 21]. Besides what we discussed so far,
OCT4-NANOG-SOX2 constitute the core pluri-
potency factors required to sustain SC self-­
4.3.4 Characterization: SC renewal and pluripotency. Emerging studies have
Signature identified additional molecular markers of pluri-
potency, including SALL4, DAX1, ESSRB, TBX3,
Once a set of defined culture conditions is estab- TCL1, RIF1, NAC1, and ZFP281 [22].
lished, SCs can continue to proliferate and repli-
cate themselves indefinitely while maintaining
the capacity to generate any type of cell in the 4.3.6 N
 oncoding RNAs (lncRNAs
body. The identification of established markers of and miRNAs)
pluripotency in SCs ensures that downstream cell
proliferation and differentiation studies are con- Recent identification of specific combinations of
ducted on high-quality, undifferentiated starting noncoding RNA (ncRNA) associated with pluri-
cell populations [14]. potency has highlighted the importance of long
and small RNA regulators for the maintenance
of pluripotency and controlling cell fate [23, 24].
4.3.5 C
 ell Membrane Markers In particular, microRNAs (miRNAs) and long
and Intracellular Markers noncoding RNAs (lncRNAs) have emerged as
eminent players to drive major cellular processes
Stage-specific embryonic antigens (SSEAs) are in SCs (Fig. 4.1). Over the past few years, miR-
cell-surface molecules used to describe the SC NAs are emerging as novel regulators of small
differentiation state; SSEA-3 and SSEA-4 are endogenous RNA of about 21–25 nucleotides in
highly expressed in undifferentiated human SCs length that play essential roles in modulating SC
46 N. Escacena-Acosta et al.

pluripotency maintenance, differentiation, and as serum albumin, hydrolysates, growth fac-


reprogramming of somatic cells to an SC-like tors, hormones, carrier proteins, and attachment
state. The miR-302/367 cluster is essential for factors.
the self-­renewal maintenance of human SCs Chase et al. described a serum-free MSC cul-
and particularly in the primed state of pluripo- ture medium that supports robust expansion and
tency. These clusters belong to the embryonic phenotype preservation of human MSCs
SC-specific cell cycle-regulating (ESCC) fam- (hMSCs) as compared with cells expanded in tra-
ily of miRNAs, related to cell cycle regulation ditional serum-containing medium [30]. Later
[25, 26]. SCs exhibit an unusually short G1 cell on, these authors developed a xeno-free medium
cycle and high proportion of cells in the S-phase based on a set of growth factors, including
to maintain their proliferation ability and pluri- platelet-­derived growth factor-BB (PDGF-BB),
potency [27]. This short duration is associated basic fibroblast growth factor (bFGF), and trans-
with a unique mechanism of cell cycle regula- forming growth factor beta (TGF-β1) that have
tion, which is highlighted by a lack of MAPK, been shown to support the expansion of hMSCs.
cyclin D, and pRB control and has been shown Several experiments were done using hMSCs
miR-302-367 cluster (Fig. 4.1) holds the ability isolated from bone marrow and adipose tissue.
of regulating cyclin D1 [28]. In the last decade, This medium revealed a promoting effect on cell
expression studies have identified newly dis- proliferation, maintaining the major functional
covered members of gene regulatory networks, and phenotype properties of the tested cells.
pluripotent long noncoding RNAs (lncRNAs), Ishikawa et  al. performed a proliferation com-
which are endogenous cellular RNAs of >200 parative study with three types of medium
nucleotides in length and involved in the main- (DMEM, MSCGM, STK2) used to culture
tenance and induction of SC pluripotency. It has hMSCs and found a remarkably higher prolifera-
become apparent that lncRNAs can transcrip- tion in hMSCs cultured in STK2, which is a novel
tionally or posttranscriptionally regulate gene serum-free medium developed for hMSC expan-
expression by diverse molecular mechanisms. sion [31].
A number of lncRNAs, such as lncRNA-ES1, On the other hand, and for clinical application
lncRNA-ES2, and Linc-ROR, are involved in of hMSCs, it is critical to develop an effective
the maintenance of pluripotency of SCs or iPSCs cryopreservation medium, to optimize protocols
[29]. The miRNA-lncRNA cross talk contrib- to preserve hMSCs for clinical use, and to ensure
utes to modulate gene expression patterns on a cell backup for repeated transplantations. An
all levels (transcriptional, posttranscriptional, ideal cryoprotectant should be nontoxic and non-­
and posttranslational) to regulate the induction, immunogenic for both the cells and the recipient,
maintenance, and directed differentiation of SCs preserving cell viability and characteristics,
and iPSCs. being chemically inert and highly water soluble
in low temperature, having a low molecular
weight, and being affordable.
4.4 Stem Cell Culture Al-Saqi et  al. reported the cryopreserva-
tion of hMSCs obtained from the bone marrow
Serum-free media (SFM) presents an alterna- and adipose tissue in a SFM, showing that the
tive to serum-containing media and offers sev- cells maintained a high viability and expressed
eral advantages, which include better control specific surface markers and have the ability to
of media composition, reduced cost, and avoid differentiate to mesodermal lineages [32]. Skog
the risk of contamination by possible infectious et al. developed a chondrogenic xeno-free differ-
agents found in serum. SFM is considered as a entiation medium for human bone marrow MSCs
chemically defined medium, characterized by [33]. These authors compared the success of this
the absence of animal serum; however it could medium to commercially available chondro-
contain undefined animal-derived products, such genic differentiation media and confirmed that
4  Stem Cells: Concept, Properties, and Characterization 47

the xeno-free differentiation medium was more pended in medium to be re-plated for more
efficient than the commercially available chon- passages or in PBS to be analyzed.
drogenic medium. Finally, Iwamoto et  al. dem- • Cryopreservation: An adequate cryopreserva-
onstrated that liver fibrosis was reduced by bone tion medium is crucial to maintain MSCs – the
marrow-derived cells cultured in SFM, improv- harvest cells are cryopreserved in CryoTubes
ing liver function in cirrhotic mice [34]. with the SFM and 10% DMSO and transferred
Based on these observations, we describe to −80 °C with a slow-freezing cryopreserva-
below the crucial steps for improving the culture tion process. After 24  h the cells are trans-
of adipose tissue-derived MSCs using a serum-­ ferred to liquid nitrogen.
free medium (SFM): • Quality Analysis of Mesenchymal Stem Cells
Cultured in a SFM: Human MSCs (hMSCs)
• Isolation: Adipose tissue-derived MSCs are cultured in SFM have been isolated and
obtained from subcutaneous adipose tissue, expanded; they need to pass a quality control
which is acquired from selective liposuc- to assure that these cells maintain MSC char-
tion or abdominoplasty surgeries. The tis- acteristics and can be useful for therapeutic
sues come from patients with medium age, applications. To determine the quality of the
with good health, and with no diabetes or cultured SFM human MSCs, their viability,
other complications. The tissue is minced proliferation, morphology, phenotype, and
and washed with sterile Hanks’ Balanced multilineage differentiation potential are ana-
Salt Solution (HBSS) until the eluent is clear lyzed, comparing these parameters with
and then digested with collagenase type II hMSCs cultured in SCM.
for 45  min at 37  °C with intermittent shak- • Cell Viability and Cell Proliferation: The via-
ing. After centrifugation, three layers are bility and cell proliferation are essential and
obtained; the upper yellow layer corresponds can be determined by a counting cell prolifer-
to the fat, the middle red layer to the blood ation assay, expecting an equal or higher den-
cells, and the pellet cells to the adipocytes, sity and proliferation for hMSCs cultured in
which are resuspended in medium. (For SFM than for those cultured in SCM.
example, for human bone marrow MSC iso- • Cell Morphology: MSC cultured in SFM
lation, 10 to 20  ml of bone marrow is aspi- should be successfully expanded during sev-
rated from the iliac crest of normal donors eral passages and maintain their morphological
in a medium age. Heparinized bone marrow characteristics, demonstrating a homogenous
is mixed with PBS, centrifuged, and layered morphology with no gross changes in respect
on with Percoll solution. After centrifugation with the MSC cultured in control medium.
the interface is collected, washed with PBS, • Immunophenotype Analysis: It is crucial that
and resuspended in medium). hMSCs expanded in SFM retain their pheno-
• Expansion: The cells obtained are cultured typic characteristics. For that, the surface
according to the manufacturer instructions for expression markers are characterized by flow
SFM use. Often, the culture dish needs to be cytometry, using monoclonal FITC or PE con-
treated with substrates such as fibronectin or jugated antibodies: CD3, CD14, CD34, CD45,
vitronectin. The cells are plated and main- CD80, CD14, HLA, CD73, CD90, and
tained at 37  °C in a humidified environment CD105. Thus, cells should be positive in more
containing 5% CO2, changing the medium than 95% for CD73, CD90, CD105, and
every 3  days until they reached 80% conflu- HLA-I and negative in less than 5% for CD3,
ence. These cells are defined as “passage 0.” CD34, CD14, CD45, CD80, and HLA-II.
• Harvesting: The harvest protocol replaces • Potential of Differentiation: The multilin-
trypsin-EDTA with TrypLE for the serum-free eage differentiation potential is investigated
culture, to ensure the process is animal post-­harvest by inducing hMSCs to differen-
component-­ free. Then the cells are resus- tiate toward the adipogenic, chondrogenic,
48 N. Escacena-Acosta et al.

and osteogenic lineages. After 14–21  days as the safety and efficacy in the clinical phases.
in lineage-­ specific differentiation media, Cell therapy drugs are the only ones in which the
hMSCs cultured in SFM should maintain active component is living cells, and therefore,
their capability for osteogenic differentiation, the cells are the product and not what they pro-
being positive for alkaline phosphatase; have duce. It is difficult to define their pharmacologi-
the potential for adipogenic differentiation, cal characteristics since they produce a variable
forming lipid vacuoles positive for Oil Red and often unknown quantity of bioactive mole-
staining; and demonstrate chondrogenic dif- cules. Therefore, the definition of the composi-
ferentiation ability, being positive for Alcian tion of a cellular product, its mechanisms of
Blue staining. action, pharmacokinetics, toxicity, and efficacy
represent challenges which have never been
anticipated before in conventional pharmacology
4.5 Advanced Therapies [35, 36]. For this reason, the transfer of basic
with Stem Cells research to clinic is subject to requirements on
which good manufacturing practices (GMP),
The advances in cellular and molecular biotech- clean room facilities, qualified personnel, quality
nology have introduced a new era of innovative controls, and aseptic process are involved. In this
medicines with complex structures, and the active way, the holder of an authorization to market
pharmaceutical element does not only refer to medicinal products for advanced therapy must
inert chemical synthesis substances. In recent guarantee a system that ensures that each product
years, it has increased the interest in products that and its components can be traced throughout the
use a biological approach to restore, maintain, or process, including procurement, manufacturing,
improve tissue function, due to their potential to packaging, storage, transportation, and delivery.
treat diverse chronic diseases or conditions with European Medicines Agency (EMA) tracks
unmet medical needs. In this regard, products research into the use of stem cells in medicines
and/or therapies derived from biological or bio- thoroughly and is responsible for assessing mar-
technological approaches should not be confused keting authorization applications for medicines
with those derived from advanced therapies. containing stem cells.
Advanced therapy medicinal products The development of ATMPs is an active
(ATMPs) are those that incorporate living cel- and fast-growing field and also requires new
lular component, tissues, or genes, and their approaches in terms of regulatory aspects.
characteristics clearly differentiate them from Considered as medicinal products ATMPs should
conventional drugs obtained by chemical synthe- be subjected to the same regulatory principles
sis. These differences range from early stages of applied on any other drugs. It is important for
development to its clinical use. Therefore, many all linked professionals to familiarize themselves
of the concepts we have about conventional drugs with the rules, standards, directives, regulations,
cannot be extrapolated to the advanced therapy and guidelines related with these pharmaceuti-
medicines and require that they be considered, at cal innovative products. The regulation builds
all levels, as a very different class of drugs. on directives (Directive 2001/20/EC; Directive
Currently, most of the medicinal production of 2001/83/EC; Directive 2012/26/EU) concerning
advanced therapies is conducted by academics, medicinal products for human use [37], qual-
academic spin-offs, not-for-profit organizations, ity and safety standards for human tissues and
and small- and medium-sized enterprises (SMEs). cells [37, 38], clinical trials, as well as regula-
Even so, ATMPs must follow the same develop- tions concerning centralized marketing authori-
ment and requirements as for pharmacological zation procedures [39, 40]. On 30 October 2007,
compounds, which consists of several phases the European Council formally adopted the spe-
including preclinical and clinical studies. Toxicity cific regulation on ATMPs (Regulation (EC) No
and the biological activity are evaluated, as well 1394/2007) [41], which came into force in the
4  Stem Cells: Concept, Properties, and Characterization 49

GLPC
Regulation 1394/2007/EC
Experimental observations Active substance
Pre-clinical research
Directive 2004/23/EC

Directive 2006/17/EC
GCP/GMP
Directive 2001/83/EC
Investigational
Clinical research Phase I-II
medicinal product
Clinical research Phase III Directive 2009/120/EC
(IMP)

Directive 2003/94/EC
GMP/GVP
Authorization Regulation (EU) No 536/2014
Marketing Registered product

Fig. 4.2  Compliance with GMP is mandatory for all GMP. GLPC, good laboratory practices compliance; GCP,
medicinal products that have been granted a marketing good clinical practices; GMP, good manufacturing prac-
authorization. Likewise, the manufacture of investiga- tices; GVP, good pharmacovigilance practices
tional medicinal products must be in accordance with

European Union on 30 December 2008. It rep- Based on these definitions, it is clear that a
resents a regulatory framework to achieve har- classification of a medicinal product as a GTMP
monized market availability within the European depends on the addition of a recombinant nucleic
Union (Fig. 4.2). acid sequence and that its mode of action is
related with this gene sequence. Viral and nonvi-
ral vectors (which includes plasmid DNA), as
4.6 Definitions well as genetically modified viruses and cells, are
all examples of this type of product.
Due to the innovative concept of these products, a SCTMP and TEP are two types of cell-based
new Committee for Advanced Therapies (CAT) medicinal products, which may contain similar
was established at the EMA. The CAT is respon- components. To ensure your product falls into
sible for all regulatory procedures concerning some of the above definitions, the following cri-
ATMP in the EU, among which the referred prod- teria must be taken into account:
uct falls within definition of ATMP, applicable to:
(i) Degree of “engineering” to which cells are
• Somatic cell therapy medicinal product subjected in order to achieve their intended
(SCTMP) mode of action in humans.
• Gene therapy medicinal product (GTMP) It is important to know if your manu-
• Tissue-engineered product (TEP) facturing process includes some type of
• Combined advanced medicinal product substantial manipulation of the cells and
tissues, which can alter their biological
The definitions of a gene and a somatic cell characteristics, physiological functions,
therapy medicinal product according to Directive or structural properties. For example, cell
2001/83/EC, Annex I, Part IV, as amended culturing leading to expansion is consid-
(implementing Directive 2009/120/EC) [42] are ered a substantial manipulation. Annex I of
presented below. A legal definition of TEP was the ATMP Regulation listed manipulations
for the first time introduced with the Regulation which shall not be considered as substantial
(EC) No 1394/2007. manipulations.
Table 4.1 summarizes the principal character- (ii) Intended to be used (same essential function
istics that biological medicinal products have to or nonhomologous use in the recipient and
fulfill. the donor).
50 N. Escacena-Acosta et al.

Table 4.1  Principal characteristic of different biological medicinal products


Biological
medicinal
products Characteristics Observations
Gene therapy (a) It contains an active substance which contains or consists Gene therapy medicinal
medicinal product of a recombinant nucleic acid used in or administered to products should not include
(GTMP) human beings with a view to regulating, repairing, vaccines against infectious
replacing, adding, or deleting a genetic sequence diseases
(b) Its therapeutic, prophylactic, or diagnostic effect relates
directly to the recombinant nucleic acid sequence it contains
or to the product of genetic expression of this sequence
Somatic cell (a) Contains or consists of cells or tissues that have been
therapy medicinal subject to substantial manipulation so that biological
product (SCTMP) characteristics, physiological functions, or structural
properties relevant for the intended clinical use have been
altered or of cells or tissues that are not intended to be
used for the same essential function(s) in the recipient and
the donor
(b) I s presented as having properties for or is used in or
administered to human beings with a view to treating,
preventing, or diagnosing a disease through the
pharmacological, immunological, or metabolic action of
its cells or tissues
Tissue-­engineered Tissue-engineered product means a product that Article 2, 1(b)
products (TEP)  Contains or consists of engineered cells or tissues
 Is presented as having properties for or is used in or
administered to human beings with a view to regenerating,
repairing, or replacing a human tissue
A tissue-engineered product may contain cells or tissues of
human or animal origin or both. The cells or tissues may be
viable or nonviable. It may also contain additional substances,
such as cellular products, biomolecules, biomaterials,
chemical substances, scaffolds, or matrices
Products containing or consisting exclusively of nonviable
human or animal cells and/or tissues, which do not contain
any viable cells or tissues and which do not act principally by
pharmacological, immunological, or metabolic action, shall
be excluded from this definition

Combined It must incorporate, as an integral part of the product, one or Article 2 1(d)
advanced therapy more medical devices within the meaning of Article 1(2)(a) of
medicinal product Directive 93/42/EEC or one or more active implantable
medical devices within the meaning of Article 1(2)(c) of
Directive 90/385/EEC, and its cellular or tissue part must
contain viable cells or tissues, or its cellular or tissue part
containing nonviable cells or tissues must be liable to act
upon the human body with action that can be considered as
primary to that of the devices referred to

Same essential function is related with the (such as cornea) is regarded as a use for the
preservation of the original function in the same essential function. Conversely, cells or
same anatomical or histological environment tissues applied in a new physiological micro-
in the donor and the recipient. Replacement environment will be considered a nonhomol-
of a tissue as its whole or functional unit ogous use.
4  Stem Cells: Concept, Properties, and Characterization 51

Since in many cases the classification of these in Europe and the USA due to the novelty, com-
products is not an easy task, the CAT can be con- plexity, and extreme diversity of these innovative
sulted and will issue a classification recommen- products. An optimal and current knowledge of
dation within 60 days after receiving the request, regulatory issues is necessary to advance in the
according to Article 17 of the ATMP Regulation. intricate path of the development of a cellular
medicine.

4.7  oloclar®, the First Stem Cell


H
Product Approved 4.8 Stem Cells and Cancer
in the European Union
The capacities of SCs to self-renew and to dif-
According to the previous definitions, stem cells ferentiate into multiple lineages make them a
are categorized as ATMPs when these cells powerful tool in regenerative medicine.
undergo a substantial manipulation or are used Unfortunately, these properties also make stem
for a different essential function. They can be cells capable of tumorigenesis, representing an
SCTMPs if they treat, prevent, or diagnose a dis- obstacle to the safe use of stem cell-based
ease (pharmacological, immunological, meta- therapies.
bolic action) or TEPS to regenerate, repair, or Normal SCs share many characteristics with
replace a human tissue. cancer stem cells (CSCs), a population of cells
Although stem cell treatment is not a new con- within the tumor capable to replicate the original
cept, Holoclar® (ex vivo expanded autologous tumor after transplantation into a mouse model
human corneal epithelial cells containing stem [45]. CSCs were first identified in human acute
cells) is the first ATMP to be approved and for- myeloid leukemia (AML) after their transplanta-
mally registered in the Western world. It is a cell-­ tion into immunodeficient mice [46]. Lapidot and
based tissue-engineered therapy that replaces colleagues demonstrated that the peripheral
epithelial cells in a damaged cornea. Adult blood of patients with AML contains approxi-
patients with moderate to severe limbal stem cell mately 1:250,000 CSCs, which are capable of
deficiency (LSCD) due to corneal surface injury, engraftment in the bone marrow of mice after tail
such as after chemical burns, may benefit from vein injection and initiating a human AML.  A
this innovative medicine. In 1997, the research subsequent study showed that CSCs presented a
team of Holoclar reported the successful ocular more undifferentiated phenotype than mature
surface regeneration in two patients with LSCD blood cells, being more closely related to hema-
after transplantation of an in  vitro engineered topoietic SCs [47]. Thus, it was postulated that
neo-epithelium from autologous limbal stem the population of cancer cells with stem cell-like
cells [43]. In 2010, 112 patients with corneal features gives rise to the majority of the tumor
damage were included in a study obtaining satis- cells with more differentiated phenotypes. In the
factory effects in 77% of the cases [44]. In 2013, last decade, the cancer stem cell theory has
the team published further results from 152 gained increasing attention from researchers, and
patients treated with Holoclar for burn-related evidences have been provided for the existence of
eye damage. In February 2015, the commercial CSCs in most, if not all, tumors [48–53].
use of Holoclar was authorized within the The theory that cancer arises from normal
European Union. Therefore, the release of this SCs uncovered the tumorigenic property of SCs.
product to the market has taken 25 years. For instance, human ESCs spontaneously dif-
The pathway to regulatory approval requires a ferentiate into an array of various tissue types
considerable amount of expertise, time, and of the embryo when transplanted into immuno-
investment. The EMA’s specialist experts in the suppressed mice, forming complex teratomas
CAT rigorously evaluate quality, safety, and effi- [54, 55]. First evidences were found in 1967,
cacy of product. Few ATMPs have been approved when normal germinal SCs from the gonadal
52 N. Escacena-Acosta et al.

ridge were transplanted into the testicles of adult 4.9 Conclusions


male mice and developed teratomas [56, 57].
Pluripotency of iPSCs is also associated with The use of living cells as a medicinal product for
tumorigenesis [58]. Recently, a study found that clinical application has become increasingly more
iPSCs survive and differentiate into three neural realistic. SCs possess the capacity of self-­renewal
lineages when transplanted into injured spinal and differentiate into other cell types, being a valu-
cord mice. Despite initial recovery of motor func- able tool in regenerative medicine. However, the
tion, grafted mice developed tumors and exhib- potential tumorigenic property of SCs has ques-
ited gradual deterioration of hind limb movement tioned their safety in cell therapy. For this reason,
during long-term observation [59]. Studies have clinical trials to validate the safety of using SCs for
provided evidences that adult SCs may also trans- health problems need to be carried out before
form into oncogenic cells. For example, neural treating patients, not only to assure that SCs do not
stem cells within the subventricular zone, a main induce a graft-versus-host disease but also that
SC compartment in the adult brain, are known cells do not form tumors. In this context, one chal-
to proliferate in response to the platelet-­derived lenge is to develop strategies to control the mecha-
growth factor (PDGF), generating glioma-­ like nisms inducing SC oncogenicity, thus increasing
masses [60]. In accordance with this study, Chen the safety when using these cells for regenerative
and colleagues identified a subset of cells within medicine. The efforts already achieved in this field
the subventricular zone (i.e., neural SCs) as a will definitely improve the clinical use of SCs;
source of new cancer cells responsible for tumor probably we are reaching a new era of a safe and
growth. The chemical depletion of this subset of more efficient stem cell-based therapy.
cells with ganciclovir and temozolomide impedes
tumor development [61]. Acknowledgments Authors are supported by the non-
Understanding the mechanisms responsible profit Fundación Progreso y Salud, Consejería de Salud,
Junta de Andalucía; FEDER cofunded grants from Instituto
of SC tumorigenesis has become a primary de Salud Carlos III and the Ministry of Economy, Industry
goal in the field of regenerative medicine. In and Competitiveness (Red TerCel: RD12/0019/0028 and
this context, the interaction of genetic and epi- RD16/00259; CIBERDEM: CB07/08/0006; PI14/01015,
genetic mechanisms seems to play important PI16/00259, PI17/02104, and CD16/00118); and Junta de
Andalucía (PAI-BIO311, CTS-576, CTS 11-727, PI-0109-
roles in the tumorigenic potential of SCs [62, 2014, PI0007/2016, and PI0272/2017). CIBERDEM is an
63]. Genetic modifications are defined as initiative of the Instituto de Salud Carlos III.
changes in the genomic DNA sequence, includ-
ing point mutation, single nucleotide polymor- Competing Interests  The authors declare no
phism, or copy number variability. Point conflict of interest.
mutations are the main DNA changes that can
accumulate in somatic cells. In most cases, Informed Consent  No human studies were car-
processes of base mispairing during DNA rep- ried out by the authors for this article.
lication cause these point mutations. On the
other hand, epigenetic changes regulate gene
Animal Studies  No animal studies were carried
expression without modifying the underlying
out by the authors for this article.
DNA sequence. The epigenetic changes include
DNA methylation, histone modifications, chro-
matin remodeling, and changes in noncoding
RNAs. A study of a subset of cancer methyl- References
ated genes revealed that these genes are fre-
quently methylated in human ESCs, indicating
1. Hmadcha A, Domínguez-Bendala J, Wakeman J,
et  al. The immune boundaries for stem cell based
a role for DNA methylation in the control of therapies: problems and prospective solutions. J
gene expression in human stem cells and sug- Cell Mol Med. 2009;13(8 A):1464–75. https://doi.
gesting a role in tumorigenesis [64]. org/10.1111/j.1582-4934.2009.00837.x.
4  Stem Cells: Concept, Properties, and Characterization 53

2. Dominici M, Le Blanc K, Mueller I, Slaper-­ Nat Biotechnol. 2007;25(7):803–16. https://doi.


Cortenbach I, Marini F, Krause D, et al. Minimal cri- org/10.1038/nbt1318.
teria for defining multipotent mesenchymal stromal 16. Kannagi R, Cochran NA, Ishigami F, et  al. Stage-­
cells. The International Society for Cellular Therapy specific embryonic antigens (SSEA-3 and -4) are
position statement. Cytotherapy. 2006;8(4):315–7. epitopes of a unique globo-series ganglioside iso-
3. Boyer LA, Lee TI, Cole MF, et al. Core transcriptional lated from human teratocarcinoma cells. EMBO J.
regulatory circuitry in human embryonic stem cells. 1983;2(12):2355–61.
Cell. 2005;122(6):947–56. https://doi.org/10.1016/j. 17. Andrews PW, Banting G, Damjanov I, Arnaud D,
cell.2005.08.020. Avner P.  Three monoclonal antibodies defining dis-
4. Zhang M, Leitch HG, Tang WWC, et al. Esrrb com- tinct differentiation antigens associated with different
plementation rescues development of Nanog-null high molecular weight polypeptides on the surface
germ cells. Cell Rep. 2018;22(2):332–9. https://doi. of human embryonal carcinoma cells. Hybridoma.
org/10.1016/j.celrep.2017.12.060. 1984;3(4):347–61.
5. Schwarz BA, Bar-Nur O, Silva JCR, Hochedlinger 18. Štefková K, Procházková J, Pacherník J.  Alkaline
K. Nanog is dispensable for the generation of induced phosphatase in stem cells. Stem Cells Int.
pluripotent stem cells. Curr Biol. 2014;24(3):347–50. 2015;2015:628368. https://doi.org/10.1155/2015/
https://doi.org/10.1016/j.cub.2013.12.050. 628368.
6. Hassani S-N, Totonchi M, Gourabi H, Schöler 19. Hiyama E, Hiyama K.  Telomere and telomerase in
HR, Baharvand H.  Signaling roadmap modulat- stem cells. Br J Cancer. 2007;96(7):1020–4. https://
ing naive and primed pluripotency. Stem Cells doi.org/10.1038/sj.bjc.6603671.
Dev. 2014;23(3):193–208. https://doi.org/10.1089/ 20. Marión RM, Blasco MA.  Telomeres and telomerase
scd.2013.0368. in adult stem cells and pluripotent embryonic stem
7. Vallier L, Alexander M, Pedersen RA. Activin/Nodal cells. Adv Exp Med Biol. 2010;695:118–31. https://
and FGF pathways cooperate to maintain pluripotency doi.org/10.1007/978-1-4419-7037-4_9.
of human embryonic stem cells. J Cell Sci. 2005;118. 21. Nichols J, Zevnik B, Anastassiadis K, et al. Formation
(Pt 19:4495–509. https://doi.org/10.1242/jcs.02553. of pluripotent stem cells in the mammalian embryo
8. Ying QL, Nichols J, Chambers I, Smith A.  BMP depends on the POU transcription factor Oct 4. Cell.
induction of Id proteins suppresses differentiation and 1998;95(3):379–91.
sustains embryonic stem cell self-renewal in collabo- 22. Lee KC, Wong WK, Feng B.  Decoding the pluripo-
ration with STAT3. Cell. 2003;115(3):281–92. tency network: the emergence of new transcription
9. Soria B, Montanya E, Martín F, Hmadcha A. A role factors. Biomedicine. 2013;1(1):49–78. https://doi.
for the host in the roadmap to diabetes stem cell org/10.3390/biomedicines1010049.
therapy. Diabetes. 2016;65(5):1155–7. https://doi. 23. Ghosal S, Das S, Chakrabarti J.  Long noncoding

org/10.2337/dbi16-0003. RNAs: new players in the molecular mechanism for
10. Sasaki H, Wada H, Baghdadi M, et  al. New immu- maintenance and differentiation of pluripotent stem
nosuppressive cell therapy to prolong survival of cells. Stem Cells Dev. 2013;22(16):2240–53. https://
induced pluripotent stem cell–derived allografts. doi.org/10.1089/scd.2013.0014.
Transplantation. 2015;99(11):2301–10. https://doi. 24. Mathieu J, Ruohola-Baker H.  Regulation of stem

org/10.1097/TP.0000000000000875. cell populations by microRNAs. Adv Exp Med
11. Swijnenburg R-J, Schrepfer S, Govaert JA, et  al.
Biol. 2013;786:329–51. https://doi.org/10.1007/978-
Immunosuppressive therapy mitigates immunological 94-007-6621-1_18.
rejection of human embryonic stem cell xenografts. 25.
Judson RL, Babiarz JE, Venere M, Blelloch
Proc Natl Acad Sci U S A. 2008;105(35):12991–6. R.  Embryonic stem cell-specific microRNAs
https://doi.org/10.1073/pnas.0805802105. promote induced pluripotency. Nat Biotechnol.
12. Pearl JI, Lee AS, Leveson-Gower DB, et al. Short-term 2009;27(5):459–61. https://doi.org/10.1038/nbt.1535.
immunosuppression promotes engraftment of embry- 26. Zhang Z, Hong Y, Xiang D, et al. MicroRNA-302/367
onic and induced pluripotent stem cells. Cell Stem cluster governs hESC self-renewal by dually regu-
Cell. 2011;8(3):309–17. https://doi.org/10.1016/j. lating cell cycle and apoptosis pathways. Stem Cell
stem.2011.01.012. Rep. 2015;4(4):645–57. https://doi.org/10.1016/j.
13. Lachaud CC, Rodriguez-Campins B, Hmadcha A,
stemcr.2015.02.009.
Soria B. Use of mesothelial cells and biological matri- 27. Becker KA, Ghule PN, Therrien JA, et  al. Self-­

ces for tissue engineering of simple epithelium surro- renewal of human embryonic stem cells is supported
gates. Front Bioeng Biotechnol. 2015;3:117. https:// by a shortened G1 cell cycle phase. J Cell Physiol.
doi.org/10.3389/fbioe.2015.00117. 2006;209(3):883–93. https://doi.org/10.1002/jcp.
14. Thomson JA, Itskovitz-Eldor J, Shapiro SS, et  al. 20776.
Embryonic stem cell lines derived from human blas- 28. Wang Y, Baskerville S, Shenoy A, Babiarz JE,

tocysts. Science. 1998;282(5391):1145–7. Baehner L, Blelloch R. Embryonic stem cell-specific
15. Adewumi O, Aflatoonian B, Ahrlund-Richter L,
microRNAs regulate the G1-S transition and promote
et  al. Characterization of human embryonic stem rapid proliferation. Nat Genet. 2008;40(12):1478–83.
cell lines by the International Stem Cell Initiative. https://doi.org/10.1038/ng.250.
54 N. Escacena-Acosta et al.

29. Fatica A, Bozzoni I.  Long non-coding RNAs: new COUNCIL of 16 April 2014 on clinical trials on
players in cell differentiation and development. Nat medicinal products for human use, and repealing
Rev Genet. 2014;15(1):7–21. https://doi.org/10.1038/ Directive 2001/20/EC. Off J Eur Communities. 2014;L
nrg3606. 158:1–76. https://eur-lex.europa.eu/legal-content/EN/
30. Chase LG, Lakshmipathy U, Solchaga LA, Rao MS, TXT/PDF/?uri=CELEX:32014R0536&from=en.
Vemuri MC.  A novel serum-free medium for the 41. REGULATION (EC) No 1394/2007 OF THE
expansion of human mesenchymal stem cells. Stem EUROPEAN PARLIAMENT AND OF THE
Cell Res Ther. 2010;1(1):8. COUNCIL of 13 November 2007 on advanced
31. Ishikawa I, Sawada R, Kato Y, Tsuji K, Shao J,
therapy medicinal products and amending Directive
Yamada T, et  al. Effectivity of the novel serum-free 2001/83/EC and Regulation (EC) No 726/2004. Off J
medium STK2 for proliferating human mesenchymal Eur Communities. 2007;L 324:121–137. https://eur-
stem cells. Yakugaku Zasshi. 2009;129(3):381–4. lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CEL
32. Al-Saqi SH, Saliem M, Quezada HC, Ekblad A,
EX:32007R1394&from=EN.
Jonasson AF, Hovatta O, et  al. Defined serum- and 42. COMMISSION DIRECTIVE 2009/120/EC of 14
xeno-free cryopreservation of mesenchymal stem September 2009 amending Directive 2001/83/EC of
cells. Cell Tissue Bank. 2015;16(2):181–93. the European Parliament and of the Council on the
33. Skog M, Muhonen V, Nystedt J, Narcisi R, Kontturi Community code relating to medicinal products for
LS, Urtti A, et al. Xeno-free chondrogenesis of bone human use as regards advanced therapy medicinal
marrow mesenchymal stromal cells: towards clinical-­ products. Off J Eur Communities. 2009;L 242:3–12.
grade chondrocyte production. Cytotechnology. https://ec.europa.eu/health/sites/health/files/files/
2015;67(5):905–19. eudralex/vol-1/dir_2009_120/dir_2009_120_en.pdf.
34. Iwamoto T, Terai S, Hisanaga T, Takami T, Yamamoto 43. Pellegrini G, Traverso CE, Franzi AT, Zingirian M,
N, Watanabe S, et al. Bone-marrow-derived cells cul- Cancedda R, De Luca M.  Long-term restoration of
tured in serum-free medium reduce liver fibrosis and damaged corneal surfaces with autologous culti-
improve liver function in carbon-tetrachloride-treated vated corneal epithelium. Lancet. 1997;349(9057):
cirrhotic mice. Cell Tissue Res. 2013;351(3):487–95. 990–3. https://doi.org/10.1016/S0140-6736(96)
35.
Escacena N, Quesada-Hernández E, Capilla-­ 11188-0.
Gonzalez V, Soria B, Hmadcha A.  Bottlenecks 44. Rama P, Matuska S, Paganoni G, Spinelli A, De
in the efficient use of advanced therapy medici- Luca M, Pellegrini G.  Limbal stem-cell therapy
nal products based on mesenchymal stromal cells. and long-term corneal regeneration. N Engl J Med.
Stem Cells Int. 2015;2015:895714. https://doi. 2010;363(2):147–55. https://doi.org/10.1056/NEJM
org/10.1155/2015/895714. oa0905955.
36. Gálvez P, Clares B, Hmadcha A, Ruiz A, Soria 45. Clarke MF, Dick JE, Dirks PB, et al. Cancer stem cells-

B. Development of a cell-based medicinal product: regu- -perspectives on current status and future directions:
latory structures in the European Union. Br Med Bull. AACR Workshop on cancer stem cells. Cancer Res.
2013;105:85–105. https://doi.org/10.1093/bmb/lds036. 2006;66(19):9339–44. https://doi.org/10.1158/0008-
37. DIRECTIVE 2001/83/EC OF THE EUROPEAN
5472.CAN-06-3126.
PARLIAMENT AND OF THE COUNCIL of 6 46. Lapidot T, Sirard C, Vormoor J, et  al. A cell initiat-
November 2001 on the Communitycode relat- ing human acute myeloid leukaemia after transplanta-
ing to medicinal products for human use. Off J Eur tion into SCID mice. Nature. 1994;367(6464):645–8.
Communities. 2001;L 311:67–128. https://eurlex. https://doi.org/10.1038/367645a0.
europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX: 47. Bonnet D, Dick JE.  Human acute myeloid leuke-
32001L0083&from=EN. mia is organized as a hierarchy that originates from
38. Directive 2012/26/EU of the European Parliament
a primitive hematopoietic cell. Nat Med. 1997;3(7):
and of the Council of 25 October 2012 amending 730–7.
directive 2001/83/EC as regards pharmacovigilance. 48. Visvader JE, Lindeman GJ. Cancer stem cells in solid
Off J Eur Communities. 2012;L 299:1–4. https:// tumours: accumulating evidence and unresolved ques-
ec.europa.eu/health/sites/health/files/files/eudralex/ tions. Nat Rev Cancer. 2008;8(10):755–68. https://
vol-1/dir_2012_26/dir_2012_26_en.pdf. doi.org/10.1038/nrc2499.
39. DIRECTIVE 2001/20/EC OF THE EUROPEAN
49. Al-Hajj M, Wicha MS, Benito-Hernandez A,
PARLIAMENT AND OF THE COUNCIL of 4 April Morrison SJ, Clarke MF. Prospective identification of
2001 on the approximation of the laws, regulations tumorigenic breast cancer cells. Proc Natl Acad Sci
and administrative provisions of the Member States U S A. 2003;100(7):3983–8. https://doi.org/10.1073/
relating to the implementation of good clinical prac- pnas.0530291100.
tice in the conduct of clinical trials on medicinal prod- 50. O’Brien CA, Pollett A, Gallinger S, Dick JE. A human
ucts for human use. Off J Eur Communities. 2001;L colon cancer cell capable of initiating tumour growth in
121:34–44. https://eurlex.europa.eu/legal-content/EN/ immunodeficient mice. Nature. 2007;445(7123):106–
TXT/PDF/?uri=CELEX:32001L0020&from=EN. 10. https://doi.org/10.1038/nature05372.

40. REGULATION (EU) No 536/2014 OF THE 51. Singh SK, Hawkins C, Clarke ID, et al. Identification
EUROPEAN PARLIAMENT AND OF THE of human brain tumour initiating cells. Nature.
4  Stem Cells: Concept, Properties, and Characterization 55

2004;432(7015):396–401. https://doi.org/10.1038/ 59. Nori S, Okada Y, Nishimura S, et  al. Long-term



nature03128. safety issues of iPSC-based cell therapy in a spi-
52. Li C, Heidt DG, Dalerba P, et  al. Identification
nal cord injury model: oncogenic transformation
of pancreatic cancer stem cells. Cancer Res. with epithelial-mesenchymal transition. Stem Cell
2007;67(3):1030–7. https://doi.org/10.1158/0008- Rep. 2015;4(3):360–73. https://doi.org/10.1016/j.
5472.CAN-06-2030. stemcr.2015.01.006.
53. Collins AT, Berry PA, Hyde C, Stower MJ, Maitland 60. Jackson EL, Garcia-Verdugo JM, Gil-Perotin S,

NJ. Prospective identification of tumorigenic prostate et al. PDGFR alpha-positive B cells are neural stem
cancer stem cells. Cancer Res. 2005;65(23):10946– cells in the adult SVZ that form glioma-like growths
51. https://doi.org/10.1158/0008-5472.CAN-05- in response to increased PDGF signaling. Neuron.
2018. 2006;51(2):187–99. https://doi.org/10.1016/j.neuron.
54. Blum B, Benvenisty N. The tumorigenicity of human 2006.06.012.
embryonic stem cells. Adv Cancer Res. 2008;100:133– 61. Chen J, Li Y, Yu T-S, et al. A restricted cell popula-
58. https://doi.org/10.1016/S0065-230X(08)00005-5. tion propagates glioblastoma growth after chemo-
55. Ben-David U, Benvenisty N.  The tumorigenicity of therapy. Nature. 2012;488(7412):522–6. https://doi.
human embryonic and induced pluripotent stem cells. org/10.1038/nature11287.
Nat Rev Cancer. 2011;11(4):268–77. https://doi. 62. Tung P-Y, Knoepfler PS.  Epigenetic mechanisms of
org/10.1038/nrc3034. tumorigenicity manifesting in stem cells. Oncogene.
56. Stevens LC.  Experimental production of testicu-
2015;34(18):2288–96. https://doi.org/10.1038/
lar teratomas in mice. Proc Natl Acad Sci U S A. onc.2014.172.
1964;52:654–61. 63. Blokzijl F, de Ligt J, Jager M, et  al. Tissue-specific
57. Stevens LC.  Origin of testicular teratomas from
mutation accumulation in human adult stem cells dur-
primordial germ cells in mice. J Natl Cancer Inst. ing life. Nature. 2016;538(7624):260–4. https://doi.
1967;38(4):549–52. org/10.1038/nature19768.
58. Lee AS, Tang C, Rao MS, Weissman IL, Wu
64. Calvanese V, Horrillo A, Hmadcha A, et  al. Cancer
JC. Tumorigenicity as a clinical hurdle for pluripotent genes hypermethylated in human embryonic stem
stem cell therapies. Nat Med. 2013;19(8):998–1004. cells. PLoS One. 2008;3(9):e3294. https://doi.
https://doi.org/10.1038/nm.3267. org/10.1371/journal.pone.0003294.
Corneal Stem Cells: Identification
and Methods of Ex Vivo Expansion
5
Christian Claude Lachaud, Abdelkrim Hmadcha,
and Bernat Soria

5.1 Introduction severe visual impairment can only be corrected


by transplantation of a donated corneal tissue.
The cornea is the clear convex membrane at the Due to the shortage of donors, there is a growing
front of the eye. The maintenance of its integrity interest in developing alternative strategies such
and transparency is critical for an optimal light as a tissue engineering of artificial corneal equiv-
transmission and refraction into the internal eye. alents created from stem cells and scaffolds. In
The integrity and functionality of the corneal epi- this way, among the different types of stem cells
thelial, stromal, and endothelial layers are main- proposed so far, are found corneal layer-specific
tained in a very large extent by their cellular stem/progenitor cells.
components [1, 2]. Massive injuries or diseases The limbal epithelial stem cells (LESC) were
affecting any of those corneal layers lead ineluc- the first type of stem/progenitor cells discovered
tably to a loss of corneal function, tissue scarring, in the cornea. Since far, their use has revolution-
and corneal opacification [3–6]. This is particu- ized the field of ocular surface regenerative
larly true for the stromal and endothelial layers medicine [7–9]. The best example is the suc-
due to their almost inexistent regenerative capac- cessful regeneration of unilateral epithelial
ities [1, 7]. To date, most corneal disorders with defects with the transplantation of a patch of
denuded amniotic membrane (AM) re-epitheli-
alized with limbal tissue biopsies collected from
C. C. Lachaud · A. Hmadcha the contralateral healthy eye [10]. A further
Department of Cell Regeneration and Advanced refinement of this application using an AM car-
Therapies, Andalusian Center of Molecular Biology
and Regenerative Medicine-CABIMER, Junta de rier with cultivated limbal epithelial stem cells
Andalucía-University of Pablo Olavide-University of (LESC) has been possible, thanks to progresses
Seville-CSIC, Seville, Andalusia, Spain in the procedures permitting their isolation and
e-mail: christian.lachaud@cabimer.es ex  vivo expansion [9, 11]. More recently were
B. Soria (*) identified a population of ABCG2+/PAX6+ pro-
Department of Cell Regeneration and Advanced genitors with mesenchymal stem cell (MSC)
Therapies, Andalusian Center of Molecular Biology
and Regenerative Medicine-CABIMER, Junta de characteristics, clonal growth capacity, and ker-
Andalucía-University of Pablo Olavide-University of atocyte differentiation potential in the periph-
Seville-CSIC, Seville, Andalusia, Spain eral corneal stroma [12]. Converging evidence
Centro de Investigación Biomédica en Red de suggest that these corneal stromal stem cells
Diabetes y Enfermedades Metabólicas Asociadas (CSSC) might be natural ­keratocyte progenitors
(CIBERDEM), Madrid, Spain
and thus are considered useful for corneal
e-mail: bernat.soria@cabimer.es

© Springer Nature Switzerland AG 2019 57


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_5
58 C. C. Lachaud et al.

stroma regeneration. In a similar way, it is also increase as corneal epithelial cells (CEC) become
the identification of immature corneal endothe- more mature in the central cornea and upper epi-
lial cells (CEnC) in the transitional zone thelium layers [28]. Of special interest, Schermer
between the peripheral corneal endothelium and et al. by analyzing the CK3 expression pattern in
the trabecular meshwork [13, 14]. Consistent the rabbit corneal epithelium and stratified cor-
with their expression of stem and progenitor neal epithelial cells in culture led to the conclu-
markers, these putative corneal endothelial pro- sion that CK3-negative cells identify a population
genitor cells (CEnPC) also display superior pro- of immature epithelial cells which is mostly con-
liferation capacities than CEnC collected from centrated into the basal layer of the peripheral
the central cornea [14–17]. Different laborato- limbal zone and that are responsible for the pro-
ries are actually focusing efforts on their ex vivo duction of more mature CK3-positive epithelial
expansion and use in experimental animal mod- cells [29]. The identification of the limbal basal
els of corneal endothelial injury [14, 18–24]. epithelium as the niche of corneal epithelial stem
Below we review the different findings on the cells niche was rapidly confirmed by “pulse-­
molecular and biological properties of corneal chase” experiments with tritiated thymidine (3H-­
layer-specific stem/progenitor cells. We detail for T) or bromodeoxyuridine (BrdU) onto ex  vivo
limbal epithelial stem cells, corneal stromal stem cultured whole cornea [30, 31]. By using this
cells, and corneal endothelial progenitor cells approach, the authors could detect a population
which are the isolation procedures used and also of label-retaining cells in the limbal basal epithe-
review the different approaches of cultures lium. These low-cycling small rounded cells with
reported so far, the culture systems used for their high nuclear cytoplasmic ratio were found to pro-
expansion and their experimental or/and clinical liferate in response to corneal surface wounding,
applications. thus suggesting they correspond to a population
Figure 5.1 shows a schematic representation of epithelial stem cells for corneal epithelium
of a cross section of the peripheral human renewal [30, 31]. Additionally, these experiments
cornea also evidenced the slow centripetal migration of
LESC toward the central cornea [30]. LESC were
shown to be located into invaginations of the
5.2  imbal Epithelial Stem Cell
L basal epithelium, the limbal crypts found between
(LESC) the palisades of Vogt. This limbal niche was
shown to provide a unique protective environ-
The first evidence of the re-epithelialization ment for LESC with an underlying stroma highly
capacities of the peripheral corneal epithelium vascularized and innervated [32].
arose from studies performed in rabbits, where Stem cell properties of LESC, including their
after the debridement of the entire corneal epithe- markers’s signature [33], their asymmetric divi-
lium, epithelial cells of suggested “conjunctival sion kinetic under normal homeostasis or experi-
origin” were shown to propagate onto the entire mental injury models [29–31, 34], extracellular
corneal surface and regenerate the transparent matrix composition of their basal membrane
corneal epithelium [25, 26]. The regenerative niche [35–37], and proliferative responsiveness
capacity of the corneal epithelium was further to growth factors [38], have also been object of
found to be attributed to epithelial cells located in intense research.
the limbus, the transitional border area between Of particular interest are studies which have
the cornea and the conjunctiva [27]. provided significant advances on the identifica-
The existence of a population of epithelial tion of intracellular and surface markers that
stem cells in the human limbus was first sug- are preferentially or specifically expressed in
gested in the mid early 1980s on the basis of a LESC.  Pellegrini et  al. demonstrated that the
differential expression of paired cytokeratins 3 transcription factor p63 is restricted to basal
and 12 (CK3 and CK12), which expression cells of the limbal epithelium, but not to transient
5  Corneal Stem Cells: Identification and Methods of Ex Vivo Expansion 59

EA
RN
CONJUNCTIVA CO

ANTERIOR
CHAMBER

SCLERA
S
IRI

LENS

Conjunctiva Limbus Peripheral cornea Central cornea

Epithelium

Bowman’s layer

Stroma

Descemet’s membrane
Endothelium

Anterior chamber

Transitional zone
Trabecular
meshwork

Terminally differentiated cell (squamous) Limbal stromal MSC Blood vessel

Post mitotic cell (wing cell) Keratocyte Nerve

Transit amplifying cell (TA) Corneal endothelial cell (CEnC) Collagen fibril

Limbal epithelial Stem cell (LESC) Corneal endothelial progenitor (CEnPC)

Fig. 5.1  Schematic representation of a cross section of Underneath the limbal epithelial niche, in the limbal
the peripheral human cornea. The limbus at the periphery stroma, are residing corneal stromal stem cells (CSSC).
of the cornea is the niche for lineage layer’s specific cor- Immature corneal endothelial progenitor cells (CEnPC)
neal stem cells (CSC) and progenitors. Limbal epithelial with expression of neural crest markers were identified in
stem cells (LESC) reside in the basal layer of the limbal the transitional zone, a small area between the peripheral
crypt, an invagination of the limbal epithelium. corneal endothelium and the trabecular meshwork

amplifying (TA) cells in suprabasal corneal epi- epithelial cells of the limbal epithelium, but not
thelium layers, thus suggesting to be a specific into basal cells of the central corneal epithelium
marker for epithelial stem cells [39]. In a simi- [40]. In the same year, Watanabe et al. indicated
lar fashion, Chen et al. further reported that the that the human limbal epithelium contains side
ATP-­ binding cassette subfamily G member 2 population (SP) cells expressing the cell mem-
(ABCG2), a member of the ATP-binding cas- brane marker ABCG2 [41], being SP+ cells a
sette (ABC) transporters, was expressed in basal subset of tissue cells which can efficiently efflux
60 C. C. Lachaud et al.

the dye Hoechst 33342 via the multidrug trans- l­imbal epithelium from the stromal layer by dis-
porter ABCG2, a characteristic which is almost solving basement membrane proteins, mainly
restricted to somatic and cancer stem cells [42]. laminin-5 and also in minor extent collagen IV
Soon after, de Paiva et al. reported that ABCG2 [45, 46, 48–50]. Alternatively, the use of trypsin
is a marker identifying clonogenic limbal epithe- was also compared to dispase II, showing how-
lial cells [43]. Of special interest, Ksander et al. ever higher effectiveness with the use of dispase
recently reported that the ATP-binding cassette, II in terms of epithelial cell recovery, viability,
subfamily B, member 5 (ABCB5) is a gene criti- and colony-forming efficiency [51]. In a similar
cal for corneal epithelium development, homeo- fashion, a comparative study proved a better
stasis, and regeneration and which expression is preservation, integrity, and recovery of the limbal
restricted to LESC [44]. epithelium sheets with dispase II with respect to
Taking advantages on these findings, Kim collagenase [49]. Nonetheless, it was however
and colleagues recently demonstrated the pos- found that collagenase treatment leads to the
sibility of purifying LESC from highly prolif- recovery of greater numbers of epithelial progen-
erative limbal epithelium-derived cell cultures itor cells together with closely associated mesen-
by fluorescent activated cell sorting (FACS) of chymal cells, providing cultures with higher
ABCG2+/ABCB5+ double-positive cells. These numbers of holoclones (stem cells derived) and
purified ABCG2+/ABCB5+ LESC could effi- meroclones (transient amplifying cells derived)
ciently differentiate into CK3+ CEC once cul- on 3T3 feeder fibroblasts [49, 52].
tured into CnT30 medium [45]. Of interest, they From this isolation stage, different procedures
reported culture media conditions and the sub- for isolating or obtaining LESC from limbal epi-
strate (mixture of matrigel and fibronectin) thelial tissue have been reported, being mostly
favoring their expansion, thus providing signifi- culture explant systems or adherent culture of
cant advances for studying in vitro their biologi- limbal epithelial cells suspension within a media
cal mechanisms or factors maintaining their conditioned by feeder layer cells. Alternatively,
stemness or promoting their epithelial differen- LESC were also co-cultured separately with
tiation. Additionally, these novel insights lay the inactivated 3T3 feeder cells in a transwell system
basis toward the definition of an optimized (see Fig.  5.2 for summary techniques and
ex vivo expansion protocol of LESC for corneal approaches for LESC isolation and expansion).
diseases associated with LSCD.

5.3.2 C
 ulture Media and the Use
5.3 Methods of Isolation of Feeder Cells
and Expansion
Media and culture conditions currently used for
5.3.1 Limbal Epithelium Tissue expanding limbal epithelial cells have been pro-
Preparation gressively optimized with the objective of
improving their long-term proliferation and
By majority, laboratories isolate limbal tissue favoring clonal LESC outgrowth. Of particular
from remaining corneoscleral rims proceeding importance was the introduction of lethally irra-
from donor cornea transplantation. Corneoscleral diated 3T3 feeder layer cells as a critical culture
rims are obtained after excision or trephination of component supporting CEC and LESC expan-
the corneal button used for keratoplasty. Limbal sion [53, 54].
rim tissue is obtained after removal of excess Most laboratories have used as basal expan-
sclera, conjunctiva, iris remnants, and endothe- sion media a mixture of DMEM and F12 (either
lium [46, 47]. Limbal rim tissues are usually 1:1 or 2:1) supplemented with 10% of either
enzymatically digested with dispase II for several fetal calf serum (FCS) [55, 56], human serum
hours at 4  °C to allow the detachment of the (HS) [46], or knockout serum replacement
5  Corneal Stem Cells: Identification and Methods of Ex Vivo Expansion 61

Corneal button for


Explant culture
keratoplasty
onto 3T3 feeder fibroblasts

Air lifting Submerged

corneoscleral scrapping
donor’s eye rim endothelium Surgical isolation
Limbal tissue
Cell
suspension
Limbal
epithelium
Enzymatic Mechanical
digestion separation
(trypsin, colagenase)
Partial enzymatic
Expansion ? digestion
LESC
(dispase II)
Explant culture
Sorting ABCG2+/ABCB5+ LESC Limbal stroma

LESC

Fig. 5.2  Procedures for obtaining human limbal epithe- dispase II and/or collagenase to digest basement mem-
lial stem cells (LESC). Corneoscleral disc removed from brane ECM molecules and help separate limbal epithe-
donor’s eye is further subjected to the excision (trephina- lium and stroma layers. Full enzymatic digestion of limbal
tion) of the central cornea button (used for transplanta- epithelium tissue allows obtaining limbal epithelial cells
tion). The resulting corneoscleral rim is further cleaned suspension, which can be further cultured at low density
off residual scleral, conjunctival, iris, and endothelial tis- onto 3T3-inactivated mouse fibroblasts to stimulate their
sues. The corneoscleral rim is then cut into small pieces proliferation and the outgrowth of LESC clones. ABCG2+/
that can be directly subjected to explant culture (airlifting ABCB5+ highly proliferating LESC can be ultimately
or submerged) or further enzymatically processed with purified by FACS [45]

(KSR) [56]. Media are usually supplemented s­ upplementation of culture media with 10 ng/ml
with insulin, transferrin, and selenium. Limbal of keratinocyte growth factor (KGF) and 10 μM
epithelial cell proliferation is additionally stimu- of the ROCK inhibitor Y-27632 allowed a long-
lated with 1–10  ng/ml epidermal growth factor term expansion of LESC [60].
(EGF), and hydrocortisone is incorporated as an CEC are also prone to differentiate and
important agent inhibiting epithelial-to-mesen- undergo a stratification process upon culture in a
chymal transition (EMT) [57]. Additionally, cul- media containing FBS [61]. This issue has stimu-
ture media have incorporated several other lated a special interest in developing serum-free
important compounds stimulating LESC prolif- media formulations able to sustain CEC prolifer-
eration: basic fibroblast growth factor (bFGF or ation in a more stable epithelial progenitor phe-
FGF-2), adenine, cholera toxin, leukemia inhib- notype. In this way, Castro-Muñozledo et  al.
iting factor (LIF), and triiodothyronine (T3) [54, formulated a complete culture media in which
55, 57]. Of particular importance is the concen- FBS was substituted with a low content of serum
tration of calcium in the media, which was albumin and supplemented with TFG-α and EGF
shown to be preferentially low to medium (0.03– and that could sustain the growth of CEC onto
0.4 mM), as limbal epithelial cells cultured with 3T3 feeder layer cells. Parallely, different labora-
high calcium concentrations undergo rapid dif- tories also successfully cultured human limbal
ferentiation and stratification [58, 59]. Recently, epithelial cells in a keratinocyte serum-free
Miyashita et  al. reported that a combined medium (KSFM) supplemented with bovine
62 C. C. Lachaud et al.

pituitary extract (BPE) and EGF [55, 62]. Seeber approach in term of LESC recovery [55]. In
et  al. reported that human CEC cultured under agreement with Kim et  al.’s results, Li et  al.
serum-free media supplemented with ascorbic recently reported almost similar outcomes with
acid, calcium, hydrocortisone, and retinoic acid both approaches in terms of LESC purity and
retains barrier properties and native epithelium-­ clonal growth capacity, thus reinforcing the use-
like morphology [63]. fulness of the explant culture for LESC expan-
Of particular importance is the development sion [77].
of efficient culture systems avoiding potential
animal-derived infectious agents. In this direc-
tion, recent advances have been made with the 5.3.4 Single-Cell Suspension
introduction of commercial human fibroblast cell Culture
lines able to efficiently substitute 3T3 feeder
layer cells [51, 64] and also of synthetic culture The number of laboratories using the cell suspen-
media compounds [51, 65] able to sustain human sion culture of limbal epithelial cells has
CEC proliferation. Yokoo et al. reported the pos- increased notably in recent years. This tendency
sibility of growing corneal limbal epithelial cells may be largely explained by progresses made in
in a serum-, feeder cells-, and BPE-free culture their enzymatic isolation procedure and the
medium containing B27 supplements and EGF, development of optimized culture media promot-
making of this culture system a valuable approach ing and sustaining their growth potential. By
for future clinical use [66] . overall majority, single-cell suspension were
obtained by trypsinization (0.25% trypsin/EDTA)
of purified limbal epithelium tissues at 37 °C for
5.3.3 Explant Culture incubation times ranging from 10 min to 2 h [46,
48, 78]. Of note, the report by Kim et  al. that
The explant culture was the first type of culture LESC can be purified from limbal epithelial cells
introduced in the laboratory. It is the simplest culture by FACS of ABCG2+/ABCB5+ cells [45]
way of culturing cells and offers the advantage of opens thus the possibility of using this approach
preserving much of the morphostructural and to obtain FACS-purified ABCG2+ and/or ABCB5+
biochemical properties of tissues and of the stem single-cell suspensions of freshly isolated LESC
cell niche [67, 68]. In this way, this approach was and perform their ex  vivo expansion in an
proved very valuable for in vitro culture of limbal improved way.
epithelium explants and LESC expansion [45, 46,
54, 55, 69–77]. Indeed, cultured limbal explants
onto denuded AM are currently applied for re-­ 5.3.5 Sphere-Forming Assay
epithelializing the ocular surface in patients with
total LSCD [8, 9]. Different laboratories have Alternatively, the suspension culture approach
compared the efficiency of the explant culture has also been used to evaluate the sphere-forming
approach with single-cell suspension culture sys- capacity of CEC [79, 80]. In this way, Mimura
tems in term of recovery of LESC, growth poten- et al. cultured CEC suspensions from the limbus
tial, morphology, and stemness markers [54, 55, and central cornea under serum-free and feeder
77]. Although the explant system was reported to layer-free culture conditions in a medium con-
require higher amount of time to reach conflu- taining methylcellulose gel matrix to impede
ence, both approaches however produced quite CEC reaggregation and sphere formation at the
similar numbers of slow-cycling BrdU label-­ clonal level. Authors reported a higher rate of pri-
retaining cells and percentages of small cells mary and secondary sphere-forming capacities
expressing progenitor epithelial cells markers from epithelial cells of the limbal region
[54]. By contrast, Zhang et  al. reported better compared to those of the central cornea.
­
outcomes using the single-cell suspension Consistent with their stemness, cells in primary
5  Corneal Stem Cells: Identification and Methods of Ex Vivo Expansion 63

spheres expressed mitogenesis (BrdU incorpora- with basal cells retaining LESC phenotype (p63,
tion) and stem/progenitor (p63, p75(NTR), and CK15, integrin α6) and suprabasal cells express-
nestin) markers [79]. By using a similar approach, ing the differentiated CECs marker CK3 [84].
Chang et al. also reported similar sphere-forming
capacities and efficiencies of CEC isolated from
both the limbal and central corneal region [80]. 5.3.7 Amniotic Membrane

The use of cryopreserved amniotic membrane


5.3.6 C
 ulture onto ECM-Coated (AM) as temporary patch for ocular surface
Surface, Carrier, and Scaffolds emergencies such as thermal and chemical burns
has proved to be extremely useful to prevent the
The incorporation of components of the limbal occurrence of limbal stem cell deficiency (LSCD)
epithelial niche in culture systems has been by preserving remaining healthy limbal stem
address to evaluate their mitogenic potential onto cells and accelerating corneal re-­epithelialization,
cultured LESC.  Of particular importance are thus allowing to regain a healthy ocular surface
extracellular matrix (ECM) molecules compo- ocular [10, 85, 86]. In patients with poor prog-
nents of the limbal epithelium basal membrane nostic indicators because of a massive loss of the
(BM) and their interactions with their cognate native LESC pool, the sole use of AM is however
LESC-surface receptors. The underlying BM of insufficient and requires the co-transplantation of
the corneal epithelium as well as those from other autologous LESC from limbal biopsies collected
epithelial tissues is basically composed by a mix- from the contralateral healthy cornea or from
ture of collagens, laminins, and heparan sulfate donor’s limbal biopsies for treating patients with
proteoglycans [81]. Laminins are important BM total bilateral LSCD [87, 88]. Further improve-
components and are composed of three distinct ment of this approach using tissue-engineered
types of chains, alpha, beta, and gamma. Different corneal epithelial equivalents created by in vitro
genetic variants of laminin chains produce a large re-epithelialization of the denuded AM with cul-
variety of heterotrimeric isoforms with tissue-­ tured limbal epithelial cell suspension or limbal
specific distribution pattern [82]. Ebihara et  al. biopsy explants was shown to provide successful
evaluated the growth capacity of human CECs in outcomes in animal transplantation models [66,
response to laminin-5, a major component of the 71, 73] and in humans [89–91]. Progresses made
corneal BM. Coating of plastic dishes with exog- toward the purification of LESC and of the pres-
enous laminin-5 was shown to increase CEC ervation of their stemness in culture are sug-
adhesion but however failed to improve their pro- gested advantageous for tissue engineering of
liferation [83]. Of particular interest, Polisetti more functional AM-based corneal epithelial
et al. performed qPCR and confocal microscopy equivalents.
image analysis of laminin chains expression on
microdissected LEPC clusters and in human lim-
bal epithelium sections, respectively [84]. They 5.3.8 Silk Fibroin
could identify the expression pattern of different
laminin isoforms with the laminin-α5 specific to Laminar fibrous scaffolds or porous films made of
epithelial progenitor cells. From this analysis, the silk fibroin (SF) are suggested very promising bio-
authors tested the activities of laminin-511 and logic substrates for tissue engineering of corneal
laminin-521 onto cultured LESC and show how epithelial equivalents and a possible alternative to
both of them could significantly increase their the use of AM in ocular surface surgery [92]. SF
adhesion and proliferation. Incorporation of lam- used in experimental tissue engineering is usually
inin-­511 to fibrin-based hydrogels could effi- obtained from the silkworm (Bombyx mori). This
ciently improve LESC adhesion and led to the protein is suggested to be extremely useful for
establishment of stratified epithelium-like tissue ­ corneal regeneration, because of its high
64 C. C. Lachaud et al.

b­iocompatibility, biodegradability, transparency, c­entral cornea [103]. Other studies also con-
and mechanical strength, among other properties firmed the presence of cells with CSSC charac-
(for specific review see [92, 93]). The production teristics in the peripheral corneal stroma and
of nano-/micro-SF fibers by electrospinning and provided additional evidences of their stemness
the possibility to control their deposition pattern in characteristics, multilineage differentiation
diverse fibrous arrangements and thickness allow toward adipocytes and osteocytes, and of their
the creation of customizable laminar fibrous mesh- keratocyte differentiation potential [104–107].
works, with adequate topography and porosity Hasmani et al. analyzed with flow cytometry the
favoring LESC adhesion, migration, and expan- co-expression of CD34 and CD105  in cultured
sion [94]. In most cases, the SF-based scaffolds limbal and peripheral stromal cells. They show
tested experimentally for the adhesion and expan- how the CD34+/CD105+ cells subset displays the
sion of LESC have been thin porous films or mem- highest stemness characteristics and transdiffer-
branes obtained by pouring SF solutions into a entiation ability into CK3+ and CK19+ epithelial
mold [95–100]. Of particular importance, thin SF cells phenotypes, suggesting the presence of
films are transparent, and the possibility of incor- bipotent mesenchymal and epithelial stem cells
porating poly(ethylene glycol) (PEG) in SF solu- in the limbal and peripheral stromal layer [108].
tions, followed by its removal from the films,
allows creating pores for metabolites diffusion
[77, 99, 101]. Consistent with its high biocompat- 5.4.1 M
 ethods of Obtaining Corneal
ibility, rabbit and human limbal epithelial cells Stromal Stem Cells
grown onto porous SF films were shown to be able
to form a stratified epithelial sheet [99, 101]. This Different methods for obtaining human CSSC
construct was shown to regenerate the corneal epi- have been reported so far [12, 104, 105, 107–
thelial surface in a rabbit model of LSCD [77]. 109]. By majority, laboratories have isolated
hCSSC from corneoscleral rims, which include
limbal and peripheral cornea stromal tissue.
5.4  orneal Stromal Stem Cell
C Limbal stromal tissue was obtained after scrap-
(CSSC) ping off the epithelial and endothelial cells. Small
pieces of limbal stroma were then enzymatically
The identification by Du et al. in 2005 of a popu- digested with collagenase [65, 104, 109, 110]. In
lation of cells in the peripheral corneal stroma their work, Wu et  al. isolated hCSSC from
with MSC characteristics and keratocyte differ- collagenase-­ digested human limbal tissue.
entiation potential has opened up the promise of Primary cultures of limbal stromal cells were
using these corneal stromal stem cells (CSSC) in established at clonal low density. With this
tissue engineering of a corneal stroma equivalent approach, cultures produced outgrowth colonies
[12]. These CSSC were shown to express the of small polygonal cells consisting in CSSC,
stem and progenitor cell markers ABCG2 and which were further subcultured in stem cell
PAX6, being this later a transcription factor growth medium to foster their expansion [107].
required for embryonic ocular development [12, Hashmani et  al. indicated that hCSSC can be
102]. Culture of CSSC with basic fibroblast directly purified from heterogeneous corneal
growth factor (bFGF or FGF-2) was found to pro- stromal cells by FACS of the CD34+/CD105+
mote their differentiation toward cells up express- cells subset [108]. In their work, Du et al. demon-
ing the keratocyte markers: keratocan, aldehyde strated that hCSSC can be sorted based on their
dehydrogenase 3A1, and keratan sulfate [12, ability to efflux the dye Hoechst 33342, thus rep-
102]. Supporting these findings, Yamagami et al. resenting a demarcated population termed “side
reported how stromal cells of the peripheral population or SP.” Cultured SP+ cells were shown
human corneal stroma had a fourfold increased to display clonogenic proliferation capacity and
sphere-forming capacity than those from the to express the stem/progenitor markers ABCG2
5  Corneal Stem Cells: Identification and Methods of Ex Vivo Expansion 65

Limbal epithelium
Explant culture

stroma Enzymatic
scraping Limbal stromal
digestion cells
(collagenase)
endothelium

Limbal tissue

Plastic adherent Culture onto


culture scaffolds
Partial enzymatic digestion
Suspension
(dispase II) culture

Fig. 5.3  Schematic summary of methods for isolating rior mechanical detachment of the epithelium and endo-
and expanding corneal stromal stem cells (CSSC). Limbal thelium layers. The resulting limbal stromal tissue can be
tissue is first excised from whole cornea or corneoscleral then directly placed in explant culture or fully disaggre-
rim. Limbal stromal tissue is then obtained after scrapping gated with collagenase to release stromal cells suspen-
off the epithelium and endothelium layers. Alternatively, sions containing CSSC. Cell suspension of limbal stromal
limbal stroma can also be obtained through partial enzy- cells have been cultured using three different methods: (i)
matic digestion of limbal tissue which dispase II to digest plastic adherent culture, (ii) suspension culture for spheres
basement membrane proteins, thus facilitating the poste- formation, or (iii) culture onto scaffolds (carrier)

and PAX6. When cultured in media supple- thus allow initiating subcultures with highly
mented with bFGF, SP+ cells could efficiently enriched hCSSC [12, 102, 107, 118]. In most
upregulate their expression of several keratocyte cases, hCSSC expansion has been performed in
markers including keratocan, aldehyde dehydro- culture growth media supplemented with fetal
genase 3A1, and keratan sulfate [12, 105]. bovine serum (FBS) and growth factors [12, 105].
Figure 5.3 is a schematic summary of methods Du et al. stimulated hCSSC clonal growth in a low
for isolating and expanding corneal stromal stem FBS-containing media (2%) containing epidermal
cells. growth factor (EGF), platelet-­derived growth fac-
tor BB (PDGF), and leukemia inhibitory factor
(LIF) among other media additives [12].
5.4.2 Methods of Expanding Funderburgh et  al. used similar growth medium
Corneal Stromal Stem Cells incorporating basic fibroblast growth factor
(bFGF) [102]. By contrast, other studies used high
5.4.2.1 Plastic Adherent Culture FBS-containing media (10–20%) without growth
CSSC are in nature similar to mesenchymal stem factors supplementation [104, 108]. Interestingly,
cells (MSCs), which possess high adherence Sidney et  al. compared the ability of different
capacities [12]. As such, the culture of corneal media formulations to sustain hCSSC growth and
stromal cells of murine, rabbit, bovine, and human concluded that a stem cell media formulation con-
origin has been mostly performed into classic sisting in a DMEM-F12 basal media supplemented
plastic-treated tissue culture dishes to allow CSSC with 20% serum replacement, basic fibroblast
expansion [12, 65, 102, 104–117]. The capacity of growth factor, and LIF is the most appropriate for
hCSSC to undergo clonal growth under low-den- the clinical production of hCSSC [110].
sity culture has been indicated to be helpful to
select rapidly proliferating outgrowth hCSSC col- 5.4.2.2 Explant Culture
onies (either by colony picking or in limited dilu- The possibility of obtaining CSSC from explant
tion cloning culture) from primary cultures and culture of stromal tissue pieces is also attractive
66 C. C. Lachaud et al.

as this approach does not require proteolysis (not et al. also use the sphere-forming assay to evalu-
stressful for the cells) and thus preserve their ate the expansion capacity of hCSSC. They ini-
niche environment. Factors released from the tially generated floating hCSSC spheres under
explanted tissue are also suggested important for feeder-free conditions using the xeno-free E8
stimulating the posterior outgrowth of spreading culture media and show their retention of prolif-
cells [68]. An adequate tissue culture explant sys- erative capacities upon serial subculture. They
tem allowing the fixation of tissue pieces on the suggest this method to be convenient for the pro-
substrate is a critical issue, necessary for cell duction of hCSSC for therapeutic purposes [65].
adhesion and spreading. This is particularly true
for tissues with a high ECM content and harbor-
ing a low amount of cells, which the case for the 5.5 Corneal Endothelial
corneal stroma. In many instances, this initial Progenitor Cell (CEnPC)
step of culture is challenging and may require
surface coating of plastic surface, using small A major function of corneal endothelial cells
pieces of tissue in a minimum amount of media (CEnC) is to regulate the hydration state of the
to cover the pieces and increase their stability cornea to maintain an optimal corneal stromal
[68]. In that way, Ghoubay-Benallaoua et  al. matrix organization and transparency [1]. This
reported the successful establishment of corneal function is principally assumed by a sodium- and
stromal fibroblastic cells through explant culture potassium-dependent adenosine triphosphatase
of human limbal stromal tissue [65]. (Na+/K+ATPase) pump which actively remove an
excess water content inside the stroma into the
5.4.2.3 Suspension Culture anterior chamber. A normal pump function of the
The suspension culture or sphere-forming assay corneal endothelium is primarily linked to its cel-
of corneal stromal cells has been addressed by lularity or density, which in healthy subjects
several laboratories in order to quantify their con- decrease from 6000 CEnC/mm2 soon after birth
tent in CSSC.  A Japanese group focused great to 2300 CEnC/mm2 in 85  years old persons [1,
attention on this approach, showing how human 122]. CEnC are arrested in the G1  cycle, with
corneal stromal cells could efficiently formed limited evidence of in vivo proliferation capaci-
spheres containing cells expressing mesenchy- ties in healthy or injured corneas. Instead, the
mal and neural markers [19, 119]. Of special corneal endothelium compensates its gradual loss
interest, and consistent with the report by Du of cellularity along life by hypertrophy of adja-
et al., the authors also reported how cells isolated cent cells to restore a continuous functional
from the peripheral cornea stroma have signifi- endothelium monolayer [1, 123–126]. This bio-
cantly higher sphere-forming capacities than logic phenomenon is however critically accentu-
those from the central cornea [19, 103, 119–121]. ated in patients suffering from corneal endothelial
The media formulation used in their studies con- dystrophies, such as the Fuchs endothelial cor-
sisted in a DMEM/F12 supplemented with the neal dystrophy (FECD) where the abnormally
growth supplements B27, 20  ng/ml EGF, and high rate of CEnC death leads to a rapid decline
40 ng/ml bFGF and the addition of methylcellu- of corneal endothelium cellularity to a point
lose to create a gel matrix preventing cell reag- (below 500 cells/mm2) at which it does not sup-
gregation, thereby ensuring clonogenic port a normal function, thereby causing corneal
sphere-forming capacity. Basu et al. reported the edema and opacity [127–131].
isolation of limbal biopsy-derived stromal cells Different studies have provided cumulating
(LBSC) from human cadaveric corneoscleral evidences supporting the existence of immature
rims. These LBSC cultured in low-attachment CEnC with progenitor/stem cell characteris-
plates were able to generate spheres expressing tics within the Schwalbe’s ring region, a small
stem cells (ABCG2, Nestin, NGFR, Oct4, PAX6, transitional area located between the periph-
and Sox2) genes [109]. Ghoubay-Benallaoua eral corneal endothelium and the trabecular
5  Corneal Stem Cells: Identification and Methods of Ex Vivo Expansion 67

meshwork [13, 14, 23]. These stem/progeni- 5.5.1 Methods of Obtaining


tor cells have been suggested to originate cells and Culturing CEnC
of both the trabecular meshwork (TM) and
corneal endothelium. Two studies conducted 5.5.1.1 Adherent Culture of Single Cells
onto healthy human corneas shortly cultured The first approach is used to isolate and culture
in vitro reported the presence of cells express- hCEnC consisted in digesting the entire cornea
ing nestin and displaying alkaline phosphatase with trypsin or collagenase. With this method, a
(AP) activity and bromodeoxyuridine (BrdU) heterogeneous population of epithelial, stromal,
incorporation in the trabecular meshwork (TM) and endothelial cells was brought in cultures,
and transitional zone, but not in the central which inevitably developed outgrowth stromal
corneal endothelium [13, 14]. Interestingly, fibroblasts [133]. The development of novel tech-
and by comparison, in corneas experimentally niques allowing a selective collection of the cor-
wounded cells of the trabecular meshwork neal endothelium by endothelial scrapping or
started to express the stem cell marker Oct-3/ Descemet’s membrane stripping allowed obtain-
Oct-4. In addition, CEnC of the peripheral cor- ing purified hCEnC single-cell suspensions and
neal endothelium also started to express stem cultures with a very low content of contaminating
cells differentiation markers (Pax-6, Sox-2), fibroblasts outgrowth [125, 133–135]. To date,
two findings supporting the presence of stem/ the different in vitro culture protocols reported to
progenitor cells at the posterior limbus [14]. promote hCEnC expansion have however reached
In line of this concept, different studies also success with diverse degrees among laboratories.
reported how CEnC isolated from the periph- These divergent outcomes appear to be largely
eral area possess higher sphere-forming capaci- linked to their capacity of inhibiting hCEnC to
ties than CEnC of the central cornea [132]. Of undergo an endothelial-to-mesenchymal transi-
particular interest, a comparative microanatomy tion (EndMT), a mechanism triggering their phe-
study of flat-mounted corneas from young and notypic conversion into fibroblastic cells (for
old donors provided evidences of the occur- specific review see [136]). Among reported
rence of a continuous slow centripetal migra- EndMT-inducing factors for hCEnC are their
tion of CEnC.  Limbal peripheral endothelial repeated trypsin-EDTA enzymatic disruption
cells were found to be less differentiated than along sequential subculture steps. Another criti-
central CEnC and to display higher expression cal factor is fetal bovine serum (FBS) or culture
of the stem cell markers nestin and telomerase media containing pro-fibrotic growth factors,
[15]. Recently, Katikireddy et  al. reported the such as the transforming growth factor-beta
presence in the healthy human corneal endothe- (TGF-β) and also in lower extent to basic fibro-
lium of a subpopulation of cells with distinc- blast growth factor (bFGF) [136, 137]. The
tive higher proliferating capacities in vitro and EndMT activity of FBS may be largely explained
expressing stem cells (SOX2, OCT4, LGR5, by its high concentrations in the growth factors
p63) and neural crest (PSIP1, p75(NTR), nes- bFGF, EGF, PDGF, and IGF-1 [138]. In support
tin, PAX3, SOX9, AP2B1 (AP-2beta)) genes. of this premise is a previous report indicating
This subset of neural crest-derived progenitors how hCEnC cultured in media with bFGF
could differentiate into Na+/K+ ATPase CEnC increase significantly their EndMT [139]. Of par-
cells [16]. ticular interest, Zhu et al. reported that an experi-
Thus, converging evidence points to the pres- mental cell-cell junction rupture of monolayered
ence of a reservoir of corneal endothelial progen- hCEnC cultures with trypsin-EDTA treatment
itor cells (CEnPC) at the boundary of the corneal followed by their culture with either FGF,
endothelium. These findings have propitiated ­TGF-­β1, or EGF are two factors synergistically
hope on the possibility of using these peripheral inducing hCEnC EndMT [17, 140].
CEnPC for tissue engineering of corneal endo- A major issue in the process of hCEnC cul-
thelial equivalents. ture is also their arrest of division once they
68 C. C. Lachaud et al.

have reached confluence, a phenomenon termed 5.5.1.3 Suspension Culture: Sphere-­


“contact inhibition.” In this way, and by using Forming Assay
interfering siRNA, Zhu et al. demonstrated that The modality of culturing cells under suspension
the knockdown of p120 catenin (p120) into into anti-/low-adherent culture plates or semi-
confluent hCEnC cultures was able to unlock solid media offers several advantages over adher-
their contact inhibition by inducing trafficking ent plastic surface cultures as rigid matrices were
of p120 to their nucleus and activation of reported to affect stem cell behaviors [147]. In
RhoA-ROCK signaling. With this method, that sense, and taking in account the neural crest
monolayered hCEnC cultures increased signifi- origin of the corneal endothelium, several labora-
cantly their cellularity [140]. Later on, Zhu tories logically came to testing the growth capac-
et  al. demonstrated that a combined siRNA ity and stemness characteristics of hCEnC under
knockdown of p120 and Kaiso in cultured the sphere-forming assay [18, 22, 148]. Of par-
hCEnC monolayers could induce even more ticular interest, it was shown that hCEnC
efficiently their loss of contact inhibition, gen- expanded under this condition retained higher
erating significantly higher density cultures telomerase activity and telomere length and
than by knockdowning only p120 [17, 141]. lower senescence that their plastic adherent coun-
Using this method, the authors report tissue terparts, thus indicating that this approach may
engineering of eight corneal endothelial equiv- enrich their content in hCEnC progenitors and
alents from hCEnC isolated from one corneo- also maintain or/and improve their progenitor
scleral rim [17]. characteristics [148].

5.5.1.2 Explant Culture 5.5.1.4 Culture onto Biological


Techniques such endothelial cells scrapping or and Biosynthetic Substrates
Descemet’s membrane stripping have permitted The limited proliferation capacities of hCEnC
obtaining purified corneal endothelium, enabling usually occurring after few rounds of subculture
thus the possibility of using its explant culture for have stimulated the search for novel culture sub-
expanding nearly pure populations of CEnC strates able to sustain their proliferation in a more
[142–145]. With this method, pieces of corneal efficient way. In line with this idea, different lab-
endothelium that could successfully adhere onto oratories cultured hCEnC onto tissue cultures
the substrate initiate their spreading within few plates coated with biologic polymers, natural
days and start to produce early outgrowth cells components of the animal tissues extracellular
migrating centrifugally to generate a surrounding matrix (ECM). The different results obtained
dense cell monolayer [142]. This approach offers could be interpreted as somehow disappointing
the advantages of avoiding the use of digestive in this way, as although collagens I and IV and
enzymes which substantially degrade cell surface fibronectin could significantly increase hCEnC
molecules (adhesion proteins, receptors), thereby adhesion; they however failed to produce any sig-
damaging cells and making them potentially nificant proliferation activity [20, 149, 150].
more susceptible to undergo EndMT. The explant Only laminin V coating could moderately
culture is therefore recognized by many research- improve hCEnC proliferation [151]. Conversely,
ers as a more natural approach of cell culture. ECM-derived fibrous macroproteins might be
The case report by Walshe et al. showing that a useful to improve the adhesion of hCEnC onto
serial explant culture was able to provide conflu- biologic or synthetic scaffolds. In this way, Kim
ent hCEnC monolayer over seven subsequent et  al. reported an improved adhesion of rabbit
subculture steps, for a total period of 6  months CEnC onto SF films coated with collagen type I
[146], suggests that future improvements of this and their retention of normal polygonal morphol-
procedure of culture will allow to expand hCEnC ogy [152]. Additionally, different laboratories
into clinically relevant numbers in reduced peri- evaluated the capacities of hCEnC to adhere onto
ods of time. natural biological scaffolds, potential surrogates
5  Corneal Stem Cells: Identification and Methods of Ex Vivo Expansion 69

of the Descemet membrane, and to form on their Table 5.1  Markers expressed in corneal stem/progenitor
cells
surface a corneal endothelium-like monolayer.
These include non-ocular decellularized tissues SC
type Markers Exp° Ref
such as the human amniotic membrane [153] and
LESC CK3/CK12 − [28]
corneal tissues such as whole corneas devoid of CK14 + [48]
their corneal endothelium (for specific review see CK15 + [48,
[154]). Additionally extracorneal ocular tissues 159]
such as the decellularized anterior lens capsule P63 + [39]
were also used [142, 155]. ABCG2 + [40, 43,
48]
ABCB5 + [44, 45]
5.5.1.5 The Use of Cell-Conditioned CXCL12, ISL1, COLA2, + [48]
Media for CEnC Expansion NCAM1, ACAN, FOXA2,
The application of cell-conditioned media for CX32, MSX1
stimulating the growth of hCEnC has also CSSC ABCG2 + [12,
105]
attracted interest. Among the different cell types
Pax6 + [12,
from which were obtained conditioned media 102]
showing significant mitogenic activity onto CD34, CD105 + [108]
hCEnC are found embryonic stem cells [21], CEnPC Alkaline phosphatase + [13, 14]
human amniotic epithelial cells [156], corneal Telomerase + [14, 15]
stromal cells [157], bone marrow-derived endo- Nestin + [14–17]
thelial progenitor cells [157], and bone marrow- SOX2, OCT4, LGR5, p63, + [16]
p75(NTR), PAX3, SOX9,
derived mesenchymal stem cells [157, 158]. AP2B1
Interestingly, the different cell type-conditioned cMyc, KLF4, Nanog, Oct4, + [17]
media tested onto hCEnC were also reported to Rex1, Sox2, SSEA4, AP2α,
preserve their functional phenotype. AP2β, FOXD3, HNK1,
MSX1, P75NTR, Sox9
Table 5.1 shows markers expressed in corneal
Abbreviations: CEnPC corneal endothelial progenitor
stem/progenitor cells
cell, CK cytokeratin, CSSC corneal stromal stem cell,
CEnPC corneal endothelial progenitor cell, Exp expres-
sion, LESC limbal epithelial stem cell
5.6 Conclusions
a challenge which is under intense research.
This chapter has focused on recovering past and Other issue of relevant interest is also the devel-
current findings on the identification of corneal opment of efficient xeno-free media for amplify-
stem/progenitor cells in the corneal epithelium, ing corneal stem/progenitor cells in a safe way.
stroma, and endothelium layers. We recovered Other important focus of efforts is around the
the different procedures which have been used characterization and optimization of scaffolds
for their isolation and ex vivo expansion, much of mimicking the native properties of corneal base-
them being still at the bench, and few others ment membranes, Bowman’s layer, and the
based on the use LESC being successfully trans- Descemet membrane. Solving these issues is of
lated to human corneal surface applications. major importance to progress toward the manu-
Despite clear progresses and successes in this facture of a full-thickness corneal graft.
direction, many challenges remain on the way
until experimental applications using corneal Acknowledgments Authors are supported by the non-
stromal stem cells (CSSC) and corneal endothe- profit Fundación Progreso y Salud, Consejería de Salud,
and Junta de Andalucía; FEDER co-funded grants from
lial progenitor cells (CEnC) can be traduced in Instituto de Salud Carlos III and the Ministry of Economy,
effective clinical therapies. Controlling the Industry and Competitiveness (Red TerCel:
in vitro expansion of corneal endothelial progeni- RD12/0019/0028 and RD16/00259; CIBERDEM:
tor cells and the long-term expansion of LESC is CB07/08/0006; PI14/01015, PI16/00259, PI17/02104 and
70 C. C. Lachaud et al.

CD16/00118); Junta de Andalucía (PAI-BIO311, CTS-­ 1 4. McGowan SL, Edelhauser HF, Pfister RR, Whikehart
576, CTS 11-727, PI-0109-2014, PI0007/2016 and DR. Stem cell markers in the human posterior limbus
PI0272/2017). CIBERDEM is an initiative of the Instituto and corneal endothelium of unwounded and wounded
de Salud Carlos III. corneas. Mol Vis. 2007;13:1984–2000.
15. He Z, Campolmi N, Gain P, Ha Thi BM, Dumollard
JM, Duband S, et al. Revisited microanatomy of the
Competing Interests  The authors declare no corneal endothelial periphery: new evidence for con-
conflict of interest. tinuous centripetal migration of endothelial cells in
humans. Stem Cells. 2012;30(11):2523–34.
Informed Consent  No human studies were car- 16. Katikireddy KR, Schmedt T, Price MO, Price FW,
Jurkunas UV.  Existence of neural crest-derived
ried out by the authors for this article. progenitor cells in normal and Fuchs endothe-
lial dystrophy corneal endothelium. Am J Pathol.
Animal Studies  No animal studies were carried 2016;186(10):2736–50.
out by the authors for this article. 17. Zhu YT, Li F, Han B, Tighe S, Zhang S, Chen SY,
et  al. Activation of RhoA-ROCK-BMP signaling
reprograms adult human corneal endothelial cells. J
Cell Biol. 2014;206(6):799–811.
References 18. Yokoo S, Yamagami S, Yanagi Y, Uchida S, Mimura
T, Usui T, et al. Human corneal endothelial cell pre-
1. Bourne WM. Biology of the corneal endothelium in cursors isolated by sphere-forming assay. Invest
health and disease. Eye (Lond). 2003;17(8):912–8. Ophthalmol Vis Sci. 2005;46(5):1626–31.
2. Torricelli AA, Wilson SE.  Cellular and extracellular 19. Amano S, Yamagami S, Mimura T, Uchida S, Yokoo
matrix modulation of corneal stromal opacity. Exp S.  Corneal stromal and endothelial cell precursors.
Eye Res. 2014;129:151–60. Cornea. 2006;25(10 Suppl 1):S73–7.
3. Klintworth GK. Corneal dystrophies. Orphanet J Rare 20. Choi JS, Williams JK, Greven M, Walter KA, Laber
Dis. 2009;4:7. PW, Khang G, et  al. Bioengineering endothelialized
4. Tuft SJ, Coster DJ.  The corneal endothelium. Eye neo-corneas using donor-derived corneal endothelial
(Lond). 1990;4 .( Pt 3:389–424. cells and decellularized corneal stroma. Biomaterials.
5. Kruse FE. Stem cells and corneal epithelial regenera- 2010;31(26):6738–45.
tion. Eye (Lond). 1994;8(Pt 2):170–83. 21. Lu X, Chen D, Liu Z, Li C, Liu Y, Zhou J, et  al.
6. Nishida T. Commanding roles of keratocytes in health Enhanced survival in  vitro of human corneal endo-
and disease. Cornea. 2010;29(Suppl 1):S3–6. thelial cells using mouse embryonic stem cell condi-
7. Sun TT, Lavker RM.  Corneal epithelial stem cells: tioned medium. Mol Vis. 2010;16:611–22.
past, present, and future. J Investig Dermatol Symp 22. Parikumar P, John S, Senthilkumar R, Manjunath S,
Proc. 2004;9(3):202–7. Baskar S, Haraguchi K, et  al. Successful transplan-
8. Dua HS, Miri A, Elalfy MS, Lencova A, Said tation of in  vitro expanded human corneal endothe-
DG. Amnion-assisted conjunctival epithelial redirec- lial precursors to corneal endothelial surface using
tion in limbal stem cell grafting. Br J Ophthalmol. a nanocomposite sheets. J Stem Cells Regen Med.
2017;101(7):913–9. 2011;7(2):94.
9. Atallah MR, Palioura S, Perez VL, Amescua 23. Yu WY, Sheridan C, Grierson I, Mason S, Kearns V,
G. Limbal stem cell transplantation: current perspec- Lo AC, et al. Progenitors for the corneal endothelium
tives. Clin Ophthalmol. 2016;10:593–602. and trabecular meshwork: a potential source for per-
10. Fernandes M, Sridhar MS, Sangwan VS, Rao
sonalized stem cell therapy in corneal ­ endothelial
GN.  Amniotic membrane transplantation for ocu- diseases and glaucoma. J Biomed Biotechnol.
lar surface reconstruction. Cornea. 2005;24(6): 2011;2011:412743.
643–53. 24. Peh GS, Lee MX, Wu FY, Toh KP, Balehosur D,
11. Shortt AJ, Secker GA, Notara MD, Limb GA,
Mehta JS. Optimization of human corneal endothelial
Khaw PT, Tuft SJ, et  al. Transplantation of ex  vivo cells for culture: the removal of corneal stromal fibro-
cultured limbal epithelial stem cells: a review of blast contamination using magnetic cell separation.
techniques and clinical results. Surv Ophthalmol. Int J Biomater. 2012;2012:601302.
2007;52(5):483–502. 25. Shapiro MS, Friend J, Thoft RA.  Corneal re-­

12. Du Y, Funderburgh ML, Mann MM, SundarRaj N, epithelialization from the conjunctiva. Invest
Funderburgh JL.  Multipotent stem cells in human Ophthalmol Vis Sci. 1981;21(1 Pt 1):135–42.
corneal stroma. Stem Cells. 2005;23(9):1266–75. 26. Friedenwald JS.  Growth pressure and metaplasia of
13. Whikehart DR, Parikh CH, Vaughn AV, Mishler K, conjunctival and corneal epithelium. Doc Ophthalmol
Edelhauser HF. Evidence suggesting the existence of Adv Ophthalmol. 1951;5-6:184–92.
stem cells for the human corneal endothelium. Mol 27. Kruse FE, Chen JJ, Tsai RJ, Tseng SC. Conjunctival
Vis. 2005;11:816–24. transdifferentiation is due to the incomplete removal
5  Corneal Stem Cells: Identification and Methods of Ex Vivo Expansion 71

of limbal basal epithelium. Invest Ophthalmol Vis Sci. clonogenic human limbal epithelial cells. Stem Cells.
1990;31(9):1903–13. 2005;23(1):63–73.
28. Kurpakus MA, Stock EL, Jones JC. Expression of the 44. Ksander BR, Kolovou PE, Wilson BJ, Saab KR, Guo
55-kD/64-kD corneal keratins in ocular surface epithe- Q, Ma J, et  al. ABCB5 is a limbal stem cell gene
lium. Invest Ophthalmol Vis Sci. 1990;31(3):448–56. required for corneal development and repair. Nature.
29. Schermer A, Galvin S, Sun TT. Differentiation-­related 2014;511(7509):353–7.
expression of a major 64K corneal keratin in vivo and 45. Kim EK, Lee GH, Lee B, Maeng YS. Establishment of
in culture suggests limbal location of corneal epithe- novel limbus-derived, highly proliferative ABCG2(+)/
lial stem cells. J Cell Biol. 1986;103(1):49–62. ABCB5(+) limbal epithelial stem cell cultures. Stem
30. Cotsarelis G, Cheng SZ, Dong G, Sun TT, Lavker Cells Int. 2017;2017:7678637.
RM. Existence of slow-cycling limbal epithelial basal 46. Lopez-Paniagua M, Nieto-Miguel T, de la Mata A,
cells that can be preferentially stimulated to pro- Dziasko M, Galindo S, Rey E, et al. Comparison of
liferate: implications on epithelial stem cells. Cell. functional limbal epithelial stem cell isolation meth-
1989;57(2):201–9. ods. Exp Eye Res. 2016;146:83–94.
31. Lehrer MS, Sun TT, Lavker RM.  Strategies of epi- 47. Parekh M, Ferrari S, Di Iorio E, Barbaro V,

thelial repair: modulation of stem cell and transit Camposampiero D, Karali M, et al. A simplified tech-
amplifying cell proliferation. J Cell Sci. 1998;111(Pt nique for in situ excision of cornea and evisceration
19):2867–75. of retinal tissue from human ocular globe. J Vis Exp:
32. Lavker RM, Tseng SC, Sun TT.  Corneal epithelial JoVE. 2012;64:e3765.
stem cells at the limbus: looking at some old problems 48. Nieto-Miguel T, Calonge M, de la Mata A, Lopez-­
from a new angle. Exp Eye Res. 2004;78(3):433–46. Paniagua M, Galindo S, de la Paz MF, et  al. A
33.
Shanmuganathan VA, Foster T, Kulkarni BB, comparison of stem cell-related gene expression in
Hopkinson A, Gray T, Powe DG, et al. Morphological the progenitor-rich limbal epithelium and the dif-
characteristics of the limbal epithelial crypt. Br J ferentiating central corneal epithelium. Mol Vis.
Ophthalmol. 2007;91(4):514–9. 2011;17:2102–17.
34. Zieske JD. Perpetuation of stem cells in the eye. Eye 49. Chen SY, Hayashida Y, Chen MY, Xie HT, Tseng
(Lond). 1994;8(Pt 2):163–9. SC.  A new isolation method of human limbal pro-
35. Funderburgh JL, Funderburgh ML, Du Y. Stem cells genitor cells by maintaining close association
in the limbal stroma. Ocul Surf. 2016;14(2):113–20. with their niche cells. Tissue Eng Part C Methods.
36. Schlotzer-Schrehardt U, Dietrich T, Saito K, Sorokin 2011;17(5):537–48.
L, Sasaki T, Paulsson M, et  al. Characterization 50. Espana EM, Romano AC, Kawakita T, Di Pascuale M,
of extracellular matrix components in the limbal Smiddy R, Tseng SC.  Novel enzymatic isolation of
epithelial stem cell compartment. Exp Eye Res. an entire viable human limbal epithelial sheet. Invest
2007;85(6):845–60. Ophthalmol Vis Sci. 2003;44(10):4275–81.
37. Polisetti N, Zenkel M, Menzel-Severing J, Kruse FE, 51. Stasi K, Goings D, Huang J, Herman L, Pinto F,
Schlotzer-Schrehardt U. Cell adhesion molecules and Addis RC, et al. Optimal isolation and xeno-free cul-
stem cell-niche-interactions in the limbal stem cell ture conditions for limbal stem cell function. Invest
niche. Stem Cells. 2016;34(1):203–19. Ophthalmol Vis Sci. 2014;55(1):375–86.
38. Kruse FE, Volcker HE.  Stem cells, wound healing, 52. Pellegrini G, Golisano O, Paterna P, Lambiase

growth factors, and angiogenesis in the cornea. Curr A, Bonini S, Rama P, et  al. Location and clonal
Opin Ophthalmol. 1997;8(4):46–54. analysis of stem cells and their differentiated
39. Pellegrini G, Dellambra E, Golisano O, Martinelli progeny in the human ocular surface. J Cell Biol.
E, Fantozzi I, Bondanza S, et  al. p63 identifies 1999;145(4):769–82.
keratinocyte stem cells. Proc Natl Acad Sci U S A. 53. Sun TT, Green H.  Cultured epithelial cells of cor-
2001;98(6):3156–61. nea, conjunctiva and skin: absence of marked intrin-
40. Chen Z, de Paiva CS, Luo L, Kretzer FL, Pflugfelder sic divergence of their differentiated states. Nature.
SC, Li DQ.  Characterization of putative stem cell 1977;269(5628):489–93.
phenotype in human limbal epithelia. Stem Cells. 54. Kim HS, Jun Song X, de Paiva CS, Chen Z,

2004;22(3):355–66. Pflugfelder SC, Li DQ.  Phenotypic characterization
41. Watanabe K, Nishida K, Yamato M, Umemoto T,
of human corneal epithelial cells expanded ex  vivo
Sumide T, Yamamoto K, et  al. Human limbal epi- from limbal explant and single cell cultures. Exp Eye
thelium contains side population cells expressing the Res. 2004;79(1):41–9.
ATP-binding cassette transporter ABCG2. FEBS Lett. 55. Zhang X, Sun H, Tang X, Ji J, Li X, Sun J, et  al.
2004;565(1–3):6–10. Comparison of cell-suspension and explant cul-
42. Ding XW, Wu JH, Jiang CP.  ABCG2: a potential ture of rabbit limbal epithelial cells. Exp Eye Res.
marker of stem cells and novel target in stem cell and 2005;80(2):227–33.
cancer therapy. Life Sci. 2010;86(17–18):631–7. 56. Xie HT, Chen SY, Li GG, Tseng SC.  Isolation and
43. de Paiva CS, Chen Z, Corrales RM, Pflugfelder SC, expansion of human limbal stromal niche cells. Invest
Li DQ. ABCG2 transporter identifies a population of Ophthalmol Vis Sci. 2012;53(1):279–86.
72 C. C. Lachaud et al.

57. Yu M, Bojic S, Figueiredo GS, Rooney P, de


70. Kethiri AR, Basu S, Shukla S, Sangwan VS, Singh
Havilland J, Dickinson A, et  al. An important role V.  Optimizing the role of limbal explant size
for adenine, cholera toxin, hydrocortisone and triio- and source in determining the outcomes of lim-
dothyronine in the proliferation, self-renewal and dif- bal transplantation: an in  vitro study. PLoS One.
ferentiation of limbal stem cells in vitro. Exp Eye Res. 2017;12(9):e0185623.
2016;152:113–22. 71. Li W, Hayashida Y, He H, Kuo CL, Tseng SC.  The
58. Meyer-Blazejewska EA, Kruse FE, Bitterer K,
fate of limbal epithelial progenitor cells during
Meyer C, Hofmann-Rummelt C, Wunsch PH, et  al. explant culture on intact amniotic membrane. Invest
Preservation of the limbal stem cell phenotype by Ophthalmol Vis Sci. 2007;48(2):605–13.
appropriate culture techniques. Invest Ophthalmol Vis 72. Luznik Z, Hawlina M, Malicev E, Bertolin M,

Sci. 2010;51(2):765–74. Kopitar AN, Ihan A, et  al. Effect of cryopreserved
59. Kruse FE, Tseng SC. Proliferative and differentiative amniotic membrane orientation on the expression of
response of corneal and limbal epithelium to extra- limbal mesenchymal and epithelial stem cell mark-
cellular calcium in serum-free clonal cultures. J Cell ers in prolonged limbal explant cultures. PLoS One.
Physiol. 1992;151(2):347–60. 2016;11(10):e0164408.
60. Miyashita H, Yokoo S, Yoshida S, Kawakita T,
73. Pathak M, Olstad OK, Drolsum L, Moe MC,

Yamagami S, Tsubota K, et  al. Long-term mainte- Smorodinova N, Kalasova S, et al. The effect of cul-
nance of limbal epithelial progenitor cells using rho ture medium and carrier on explant culture of human
kinase inhibitor and keratinocyte growth factor. Stem limbal epithelium: a comparison of ultrastructure,
Cells Transl Med. 2013;2(10):758–65. keratin profile and gene expression. Exp Eye Res.
61. Yanez-Soto B, Leonard BC, Raghunathan VK,
2016;153:122–32.
Abbott NL, Murphy CJ.  Effect of stratification on 74. Valdetaro GP, Aldrovani M, Padua IR, Cristovam
surface properties of corneal epithelial cells. Invest PC, Gomes JA, Laus JL.  Supra-organization and
Ophthalmol Vis Sci. 2015;56(13):8340–8. optical anisotropies of the extracellular matrix in
62. Kawakita T, Espana EM, He H, Yeh LK, Liu CY, the amniotic membrane and limbal stroma before
Tseng SC.  Calcium-induced abnormal epidermal-­ and after explant culture. Biomed Opt Express.
like differentiation in cultures of mouse corneal-­ 2016;7(12):4982–94.
limbal epithelial cells. Invest Ophthalmol Vis Sci. 75.
Zamudio A, Wang Z, Chung SH, Wolosin
2004;45(10):3507–12. JM.  Inhibition of TGFbeta cell signaling for limbal
63.
Seeber JW, Zorn-Kruppa M, Lombardi-Borgia explant culture in serumless, defined xeno-free condi-
S, Scholz H, Manzer AK, Rusche B, et  al. tions. Exp Eye Res. 2016;145:48–57.
Characterisation of human corneal epithelial cell cul- 76. Zhang ZH, Liu HY, Liu K, Xu X.  Comparison of
tures maintained under serum-free conditions. Altern explant and enzyme digestion methods for ex  vivo
Lab Anim: ATLA. 2008;36(5):569–83. isolation of limbal epithelial progenitor cells. Curr
64. Lu R, Bian F, Lin J, Su Z, Qu Y, Pflugfelder SC, et al. Eye Res. 2016;41(3):318–25.
Identification of human fibroblast cell lines as a feeder 77. Li Y, Yang Y, Yang L, Zeng Y, Gao X, Xu H.

layer for human corneal epithelial regeneration. PLoS Poly(ethylene glycol)-modified silk fibroin membrane
One. 2012;7(6):e38825. as a carrier for limbal epithelial stem cell transplan-
65. Ghoubay-Benallaoua D, de Sousa C, Martos R,
tation in a rabbit LSCD model. Stem Cell Res Ther.
Latour G, Schanne-Klein MC, Dupin E, et  al. Easy 2017;8(1):256.
xeno-free and feeder-free method for isolating and 78. Kim MK, Lee JL, Shin KS, Jung GA, Wee WR,

growing limbal stromal and epithelial stem cells of the Lee JH, et  al. Isolation of putative corneal epithe-
human cornea. PLoS One. 2017;12(11):e0188398. lial stem cells from cultured limbal tissue. Korean J
66. Yokoo S, Yamagami S, Usui T, Amano S, Araie
Ophthalmol. 2006;20(1):55–61.
M.  Human corneal epithelial equivalents for ocu- 79. Mimura T, Yamagami S, Uchida S, Yokoo S, Ono K,
lar surface reconstruction in a complete serum-free Usui T, et al. Isolation of adult progenitor cells with
culture system without unknown factors. Invest neuronal potential from rabbit corneal epithelial cells
Ophthalmol Vis Sci. 2008;49(6):2438–43. in serum- and feeder layer-free culture conditions.
67.
Resau JH, Sakamoto K, Cottrell JR, Hudson Mol Vis. 2010;16:1712–9.
EA, Meltzer SJ.  Explant organ culture: a review. 80. Chang CY, McGhee JJ, Green CR, Sherwin T.

Cytotechnology. 1991;7(3):137–49. Comparison of stem cell properties in cell popula-
68. Hendijani F.  Explant culture: an advantageous
tions isolated from human central and limbal corneal
method for isolation of mesenchymal stem cells from epithelium. Cornea. 2011;30(10):1155–62.
human tissues. Cell Prolif. 2017;50(2). https://doi. 81. Tuori A, Uusitalo H, Burgeson RE, Terttunen J,

org/10.1111/cpr.12334. Epub 2017 Feb 1. Virtanen I.  The immunohistochemical composition
69. Gurdal M, Barut Selver O, Baysal K, Durak
of the human corneal basement membrane. Cornea.
I.  Comparison of culture media indicates a role for 1996;15(3):286–94.
autologous serum in enhancing phenotypic preser- 82. Simon-Assmann P.  The laminin family: founding

vation of rabbit limbal stem cells in explant culture. members of the basement membrane. Cell Adhes
Cytotechnology. 2018;70:687–700. Migr. 2013;7(1):44–7.
5  Corneal Stem Cells: Identification and Methods of Ex Vivo Expansion 73

83. Ebihara N, Mizushima H, Miyazaki K, Watanabe Y, 98. Bray LJ, George KA, Ainscough SL, Hutmacher DW,
Ikawa S, Nakayasu K, et al. The functions of exoge- Chirila TV, Harkin DG.  Human corneal epithelial
nous and endogenous laminin-5 on corneal epithelial equivalents constructed on Bombyx mori silk fibroin
cells. Exp Eye Res. 2000;71(1):69–79. membranes. Biomaterials. 2011;32(22):5086–91.
84. Polisetti N, Sorokin L, Okumura N, Koizumi N, 99. Higa K, Takeshima N, Moro F, Kawakita T,
Kinoshita S, Kruse FE, et al. Laminin-511 and -521-­ Kawashima M, Demura M, et al. Porous silk fibroin
based matrices for efficient ex  vivo-expansion of film as a transparent carrier for cultivated cor-
human limbal epithelial progenitor cells. Sci Rep. neal epithelial sheets. J Biomater Sci Polym Ed.
2017;7(1):5152. 2011;22(17):2261–76.
85. Sabater AL, Perez VL. Amniotic membrane use for 100. Hogerheyde TA, Suzuki S, Walshe J, Bray LJ,
management of corneal limbal stem cell deficiency. Stephenson SA, Harkin DG, et  al. Optimization of
Curr Opin Ophthalmol. 2017;28(4):363–9. corneal epithelial progenitor cell growth on Bombyx
86. Shimazaki J, Yang HY, Tsubota K. Amniotic mem- mori silk fibroin membranes. Stem Cells Int.
brane transplantation for ocular surface reconstruc- 2016;2016:8310127.
tion in patients with chemical and thermal burns. 101. Suzuki S, Dawson RA, Chirila TV, Shadforth AM,
Ophthalmology. 1997;104(12):2068–76. Hogerheyde TA, Edwards GA, et  al. Treatment
87. Kenyon KR, Tseng SC. Limbal autograft transplan- of silk fibroin with poly(ethylene glycol) for the
tation for ocular surface disorders. Ophthalmology. enhancement of corneal epithelial cell growth. J
1989;96(5):709–22; discussion 22–3. Funct Biomater. 2015;6(2):345–66.
88. Tsai RJ, Tseng SC.  Human allograft limbal trans- 102. Funderburgh ML, Du Y, Mann MM, SundarRaj N,
plantation for corneal surface reconstruction. Funderburgh JL.  PAX6 expression identifies pro-
Cornea. 1994;13(5):389–400. genitor cells for corneal keratocytes. FASEB J.
89. Tsai RJ, Li LM, Chen JK.  Reconstruction of dam- 2005;19(10):1371–3.
aged corneas by transplantation of autologous limbal 103. Yamagami S, Yokoo S, Mimura T, Takato T,
epithelial cells. N Engl J Med. 2000;343(2):86–93. Araie M, Amano S.  Distribution of precursors in
90. Schwab IR, Reyes M, Isseroff RR. Successful trans- human corneal stromal cells and endothelial cells.
plantation of bioengineered tissue replacements Ophthalmology. 2007;114(3):433–9.
in patients with ocular surface disease. Cornea. 104. Branch MJ, Hashmani K, Dhillon P, Jones DR, Dua
2000;19(4):421–6. HS, Hopkinson A.  Mesenchymal stem cells in the
91. Grueterich M, Espana EM, Touhami A, Ti SE, Tseng human corneal limbal stroma. Invest Ophthalmol Vis
SC.  Phenotypic study of a case with successful Sci. 2012;53(9):5109–16.
transplantation of ex  vivo expanded human limbal 105. Du Y, Sundarraj N, Funderburgh ML, Harvey SA,
epithelium for unilateral total limbal stem cell defi- Birk DE, Funderburgh JL.  Secretion and organiza-
ciency. Ophthalmology. 2002;109(8):1547–52. tion of a cornea-like tissue in  vitro by stem cells
92. Harkin DG, George KA, Madden PW, Schwab from human corneal stroma. Invest Ophthalmol Vis
IR, Hutmacher DW, Chirila TV.  Silk fibroin Sci. 2007;48(11):5038–45.
in ocular tissue reconstruction. Biomaterials. 106. Lu JM, Zhou ZY, Zhang XR, Li XL, Wang HF,
2011;32(10):2445–58. Song XJ. A preliminary study of mesenchymal stem
93. Kundu B, Rajkhowa R, Kundu SC, Wang X.  Silk cell-like cells derived from murine corneal stroma.
fibroin biomaterials for tissue regenerations. Adv Graefe’s Archive Clin Exp Ophthalmol = Albrecht
Drug Deliv Rev. 2013;65(4):457–70. von Graefes Archiv fur klinische und experimentelle
94. Biazar E, Baradaran-Rafii A, Heidari-keshel S, Ophthalmologie. 2010;248(9):1279–85.
Tavakolifard S.  Oriented nanofibrous silk as a nat- 107. Wu J, Du Y, Mann MM, Funderburgh JL, Wagner
ural scaffold for ocular epithelial regeneration. J WR.  Corneal stromal stem cells versus corneal
Biomater Sci Polym Ed. 2015;26(16):1139–51. ­fibroblasts in generating structurally appropriate cor-
95. Bray LJ, Suzuki S, Harkin DG, Chirila neal stromal tissue. Exp Eye Res. 2014;120:71–81.
TV.  Incorporation of exogenous RGD peptide and 108. Hashmani K, Branch MJ, Sidney LE, Dhillon PS,
inter-species blending as strategies for enhanc- Verma M, McIntosh OD, et  al. Characterization of
ing human corneal limbal epithelial cell growth corneal stromal stem cells with the potential for
on Bombyx mori silk fibroin membranes. J Funct epithelial transdifferentiation. Stem Cell Res Ther.
Biomater. 2013;4(2):74–88. 2013;4(3):75.
96. Lawrence BD, Pan Z, Liu A, Kaplan DL, Rosenblatt 109. Basu S, Hertsenberg AJ, Funderburgh ML, Burrow
MI. Human corneal limbal epithelial cell response to MK, Mann MM, Du Y, et al. Human limbal biopsy-­
varying silk film geometric topography in vitro. Acta derived stromal stem cells prevent corneal scarring.
Biomater. 2012;8(10):3732–43. Sci Transl Med. 2014;6(266):266ra172.
97. Liu J, Lawrence BD, Liu A, Schwab IR, Oliveira 110. Sidney LE, Branch MJ, Dua HS, Hopkinson
LA, Rosenblatt MI.  Silk fibroin as a biomaterial A.  Effect of culture medium on propagation and
substrate for corneal epithelial cell sheet generation. phenotype of corneal stroma-derived stem cells.
Invest Ophthalmol Vis Sci. 2012;53(7):4130–8. Cytotherapy. 2015;17(12):1706–22.
74 C. C. Lachaud et al.

111. Kureshi AK, Funderburgh JL, Daniels JT. Human cor- 126. Murphy C, Alvarado J, Juster R, Maglio M. Prenatal
neal stromal stem cells exhibit survival capacity fol- and postnatal cellularity of the human corneal endo-
lowing isolation from stored organ-culture corneas. thelium. A quantitative histologic study. Invest
Invest Ophthalmol Vis Sci. 2014;55(11):7583–8. Ophthalmol Vis Sci. 1984;25(3):312–22.
112. Kureshi AK, Dziasko M, Funderburgh JL, Daniels 127. Shearer TR, Chamberlain WD, Fujii A, Azuma
JT. Human corneal stromal stem cells support limbal M.  Selecting Fuchs patients for drug trials involv-
epithelial cells cultured on RAFT tissue equivalents. ing endothelial cell proliferation. Eur J Ophthalmol.
Sci Rep. 2015;5:16186. 2016;26(6):536–9.
113. Nakatsu MN, Gonzalez S, Mei H, Deng SX. Human 128. Guell JL, El Husseiny MA, Manero F, Gris O, Elies
limbal mesenchymal cells support the growth of D.  Historical review and update of surgical treat-
human corneal epithelial stem/progenitor cells. ment for corneal endothelial diseases. Ophthalmol
Invest Ophthalmol Vis Sci. 2014;55(10):6953–9. Therapy. 2014;3(1–2):1–15.
114. Soleimanifar F, Mortazavi Y, Nadri S, Soleimani 129. Aldave AJ, Han J, Frausto RF. Genetics of the cor-
M.  Conjunctiva derived mesenchymal stem cell neal endothelial dystrophies: an evidence-based
(CJMSCs) as a potential platform for differentia- review. Clin Genet. 2013;84(2):109–19.
tion into corneal epithelial cells on bioengineered 130. Zaniolo K, Bostan C, Rochette Drouin O,
electrospun scaffolds. J Biomed Mater Res A. Deschambeault A, Perron MC, Brunette I, et  al.
2017;105(10):2703–11. Culture of human corneal endothelial cells isolated
115. Shen T, Shen J, Zheng QQ, Li QS, Zhao HL, Cui L, from corneas with Fuchs endothelial corneal dystro-
et al. Cell viability and extracellular matrix synthe- phy. Exp Eye Res. 2012;94(1):22–31.
sis in a co-culture system of corneal stromal cells 131. Van den Bogerd B, Dhubhghaill SN, Koppen C,
and adipose-derived mesenchymal stem cells. Int J Tassignon MJ, Zakaria N. A review of the evidence
Ophthalmol. 2017;10(5):670–8. for in  vivo corneal endothelial regeneration. Surv
116. Mukhey D, Phillips J, Daniels JT, Kureshi Ophthalmol. 2018;63:149–65.
AK.  Controlling human corneal stromal stem cell 132. Mimura T, Yamagami S, Yokoo S, Araie M, Amano
contraction to mediate rapid cell and matrix organi- S. Comparison of rabbit corneal endothelial cell pre-
zation of real architecture for 3-dimensional tissue cursors in the central and peripheral cornea. Invest
equivalents. Acta Biomater. 2018;67:229–37. Ophthalmol Vis Sci. 2005;46(10):3645–8.
117. Hertsenberg AJ, Shojaati G, Funderburgh ML, 133. Peh GS, Beuerman RW, Colman A, Tan DT, Mehta
Mann MM, Du Y, Funderburgh JL.  Corneal stro- JS.  Human corneal endothelial cell expansion for
mal stem cells reduce corneal scarring by mediating corneal endothelium transplantation: an overview.
neutrophil infiltration after wounding. PLoS One. Transplantation. 2011;91(8):811–9.
2017;12(3):e0171712. 134. Li W, Sabater AL, Chen YT, Hayashida Y, Chen
118. Pinnamaneni N, Funderburgh JL.  Concise review: SY, He H, et  al. A novel method of isolation,
stem cells in the corneal stroma. Stem Cells. preservation, and expansion of human corneal
2012;30(6):1059–63. endothelial cells. Invest Ophthalmol Vis Sci.
119. Uchida S, Yokoo S, Yanagi Y, Usui T, Yokota C, 2007;48(2):614–20.
Mimura T, et  al. Sphere formation and expres- 135. Engelmann K, Friedl P. Optimization of culture con-
sion of neural proteins by human corneal stro- ditions for human corneal endothelial cells. In Vitro
mal cells in  vitro. Invest Ophthalmol Vis Sci. Cell Dev Biol. 1989;25(11):1065–72.
2005;46(5):1620–5. 136. Roy O, Leclerc VB, Bourget JM, Theriault M, Proulx
120. Mimura T, Amano S, Yokoo S, Uchida S, Usui T, S. Understanding the process of corneal endothelial
Yamagami S.  Isolation and distribution of rabbit morphological change in  vitro. Invest Ophthalmol
keratocyte precursors. Mol Vis. 2008;14:197–203. Vis Sci. 2015;56(2):1228–37.
121. Mimura T, Amano S, Yokoo S, Uchida S, Yamagami 137. Li C, Dong F, Jia Y, Du H, Dong N, Xu Y, et al. Notch
S, Usui T, et al. Tissue engineering of corneal stroma signal regulates corneal endothelial-to-mesenchymal
with rabbit fibroblast precursors and gelatin hydro- transition. Am J Pathol. 2013;183(3):786–95.
gels. Mol Vis. 2008;14:1819–28. 138. Gstraunthaler G.  Alternatives to the use of fetal
122. Nucci P, Brancato R, Mets MB, Shevell SK. Normal bovine serum: serum-free cell culture. ALTEX.
endothelial cell density range in childhood. Arch 2003;20(4):275–81.
Ophthalmol. 1990;108(2):247–8. 139. Lee JG, Kay EP.  FGF-2-mediated signal transduc-
123. Senoo T, Joyce NC.  Cell cycle kinetics in corneal tion during endothelial mesenchymal transforma-
endothelium from old and young donors. Invest tion in corneal endothelial cells. Exp Eye Res.
Ophthalmol Vis Sci. 2000;41(3):660–7. 2006;83(6):1309–16.
124. Joyce NC. Proliferative capacity of the corneal endo- 140. Zhu YT, Chen HC, Chen SY, Tseng SC.  Nuclear
thelium. Prog Retin Eye Res. 2003;22(3):359–89. p120 catenin unlocks mitotic block of contact-­
125. Zhu C, Joyce NC.  Proliferative response of cor- inhibited human corneal endothelial monolayers
neal endothelial cells from young and older donors. without disrupting adherent junctions. J Cell Sci.
Invest Ophthalmol Vis Sci. 2004;45(6):1743–51. 2012;125(Pt 15):3636–48.
5  Corneal Stem Cells: Identification and Methods of Ex Vivo Expansion 75

141. Zhu YT, Han B, Li F, Chen SY, Tighe S, Zhang S, 151. Yamaguchi M, Ebihara N, Shima N, Kimoto M,
et  al. Knockdown of both p120 catenin and Kaiso Funaki T, Yokoo S, et al. Adhesion, migration, and
promotes expansion of human corneal endothelial proliferation of cultured human corneal endothe-
monolayers via RhoA-ROCK-noncanonical BMP-­ lial cells by laminin-5. Invest Ophthalmol Vis Sci.
NFkappaB pathway. Invest Ophthalmol Vis Sci. 2011;52(2):679–84.
2014;55(3):1509–18. 152. Kim EY, Tripathy N, Cho SA, Joo CK, Lee D, Khang
142. Lachaud CC, Soria F, Escacena N, Quesada-­ G.  Bioengineered neo-corneal endothelium using
Hernandez E, Hmadcha A, Alio J, et al. Mesothelial collagen type-I coated silk fibroin film. Colloids Surf
cells: a cellular surrogate for tissue engineering of B Biointerfaces. 2015;136:394–401.
corneal endothelium. Invest Ophthalmol Vis Sci. 153. Ishino Y, Sano Y, Nakamura T, Connon CJ, Rigby
2014;55(9):5967–78. H, Fullwood NJ, et  al. Amniotic membrane as a
143. Glassman AB, Coles WH, Bennett CE.  Corneal carrier for cultivated human corneal endothelial
endothelium: a modified method for cultivation. In cell transplantation. Invest Ophthalmol Vis Sci.
Vitro. 1979;15(11):873–6. 2004;45(3):800–6.
144. Gao Y, Zhou Q, Qu M, Yang L, Wang Y, Shi W. In 154. Navaratnam J, Utheim TP, Rajasekhar VK,
vitro culture of human fetal corneal endothelial cells. Shahdadfar A.  Substrates for expansion of cor-
Graefe’s Archive Clin Exp Ophthalmol = Albrecht neal endothelial cells towards bioengineering of
von Graefes Archiv fur klinische und experimentelle human corneal endothelium. J Funct Biomater.
Ophthalmologie. 2011;249(5):663–9. 2015;6(3):917–45.
145. Choi JS, Kim EY, Kim MJ, Khan FA, Giegengack 155. Yoeruek E, Saygili O, Spitzer MS, Tatar O, Bartz-­
M, DE’gostino R Jr, et al. Factors affecting success- Schmidt KU, Szurman P.  Human anterior lens
ful isolation of human corneal endothelial cells for capsule as carrier matrix for cultivated human
clinical use. Cell Transplant. 2014;23(7):845–54. corneal endothelial cells. Cornea. 2009;28(4):
146. Walshe J, Harkin DG.  Serial explant culture pro- 416–20.
vides novel insights into the potential location and 156. Sha X, Liu Z, Song L, Wang Z, Liang X.  Human
phenotype of corneal endothelial progenitor cells. amniotic epithelial cell niche enhances the func-
Exp Eye Res. 2014;127:9–13. tional properties of human corneal endothelial cells
147. Engler AJ, Sen S, Sweeney HL, Discher DE. Matrix via inhibiting P53-survivin-mitochondria axis. Exp
elasticity directs stem cell lineage specification. Eye Res. 2013;116:36–46.
Cell. 2006;126(4):677–89. 157. Zhu MY, Yao QK, Chen JZ, Shao CY, Yan CX,
148. Mimura T, Yamagami S, Yokoo S, Usui T, Amano Ni N, et  al. Effects of corneal stromal cell- and
S.  Selective isolation of young cells from human bone marrow-­ derived endothelial progenitor cell-­
corneal endothelium by the sphere-forming assay. conditioned media on the proliferation of corneal
Tissue Eng Part C Methods. 2010;16(4):803–12. endothelial cells. Int J Ophthalmol. 2016;9(3):
149. Underwood PA, Bennett FA.  The effect of extra- 332–9.
cellular matrix molecules on the in  vitro behav- 158. Nakahara M, Okumura N, Kay EP, Hagiya M,
ior of bovine endothelial cells. Exp Cell Res. Imagawa K, Hosoda Y, et  al. Corneal endothelial
1993;205(2):311–9. expansion promoted by human bone marrow mes-
150. Choi JS, Kim EY, Kim MJ, Giegengack M, Khan enchymal stem cell-derived conditioned medium.
FA, Khang G, et al. In vitro evaluation of the inter- PLoS One. 2013;8(7):e69009.
actions between human corneal endothelial cells 159. Figueira EC, Di Girolamo N, Coroneo MT, Wakefield
and extracellular matrix proteins. Biomed Mater. D.  The phenotype of limbal epithelial stem cells.
2013;8(1):014108. Invest Ophthalmol Vis Sci. 2007;48(1):144–56.
Corneal Epithelial Stem Cells:
Methods for Ex Vivo Expansion
6
Gustavo S. Figueiredo, Hardeep Singh Mudhar,
Majlinda Lako, and Francisco C. Figueiredo

6.1 Introduction during homeostasis but can become highly pro-


liferative in periods of injury.
The corneal epithelium is the only layer in the Fuchs et  al. (2000) were among the first to
cornea that is able to regenerate in vivo, and cor- describe the idea of stem cells being found in a
neal integrity and function are dependent on this. “niche” or a specific location within an organ in
The corneal epithelium is a stratified squamous which the microenvironment supports stem cell
epithelium, where a few layers of flattened epi- viability and provides protection to them [2]. The
thelial cells lie on top of a basement membrane. limbal niche is a term which describes a protec-
It is thought that this ability of the epithelium to tive environment containing a reservoir of LSCs,
regenerate relies on a small population of stem ready to migrate from the peripheral limbus cen-
cells, located in the basal region of the limbus. tripetally towards the centre of the cornea. This is
Limbal epithelial stem cells (LSCs) are found a continuous process but is exaggerated during
within the limbus, at the corneoscleral junction. periods of stress to the cornea, such as during a
These cells (LSCs) share many features with wound healing response, guided by signals gen-
other adult stem cells including their small size erated by the limbal niche.
and high nuclear to cytoplasm ratio [1]. They also A previous study from Dua et al. in 2005 and
do not express markers of differentiation such as subsequently confirmed by a further study from
cytokeratins 3 (CK3) and 12 (CK12). Limbal Tseng’s group in 2007 both demonstrated that the
epithelial stem cells are generally slow cycling LSC niche is located within the palisades of Vogt
at the corneoscleral limbus [3, 4]. Due to their
ridge-like structure, vascularity and pigmenta-
G. S. Figueiredo · M. Lako tion, the palisades of Vogt have long been thought
Newcastle University, Institute of Genetic Medicine,
International Centre for Life, to play an important role in the renewal and
Newcastle upon Tyne, Tyne and Wear, UK regeneration of the corneal epithelium [5].
H. S. Mudhar The stem cell niche provides an environment
Royal Hallamshire Hospital, Department of that protects the stem cells from stimuli that may
Histopathology, Sheffield, South Yorkshire, UK adversely promote differentiation or apoptosis.
F. C. Figueiredo (*) Tseng’s group also showed as an example that
Newcastle University, Institute of Genetic Medicine, melanocytes present in the limbal basal layer
International Centre for Life, help produce and transport melanin pigments
Newcastle upon Tyne, Tyne and Wear, UK
into epithelial cells to minimise damage by ultra-
Royal Victoria Infirmary, Newcastle upon Tyne, UK violet irradiation [4]. Thus, in a normal eye, LSCs
e-mail: francisco.figueiredo@newcastle.ac.uk

© Springer Nature Switzerland AG 2019 77


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_6
78 G. S. Figueiredo et al.

are reasonably protected from external insult. As cell division occurs, the epithelial cells
However, with damage to the palisades of Vogt (TAC) migrate centripetally and in a superficial
following chemical or thermal burn injuries, the direction (i.e. horizontally and vertically),
stem cells fail or become dysfunctional, and a becoming more differentiated in the process, to
sequelae of symptoms and signs ensues, often eventually form terminally differentiated cells
accompanied by inflammation and clinically (TDC). Transient amplifying cells have the
defined as limbal stem cell deficiency (LSCD). capacity to divide exponentially when the epithe-
The notion that the corneal epithelium is able lium is wounded. Terminally differentiated cells
to regenerate was first hypothesised by Thoft and form the majority of the corneal epithelium.
Friend in 1983, when they explained their “X, Y, Lavker et al. showed in 1991 that the most super-
Z hypothesis of corneal epithelial maintenance” ficial layers of the corneal epithelium contain
[6]. They proposed that the proliferation of basal postmitotic cells and are therefore unable to
cells (X), added together with the centripetal regenerate through mitotic cell division [15].
migration of cells (Y), was equal to the epithelial These cells (TDCs) are constantly sloughed off
cell loss from the surface (Z). Subsequently, the ocular surface through blinking and normal
Sharma and Coles indicated in 1989 that corneal wear and tear, and as they are incapable of divid-
epithelial cell mass could be renewed by cells ing, this stimulates the cycle of cell division,
from the limbal epithelium alone [7]. migration and differentiation [1]. Corneal epithe-
At the limbus, mitotic cell division occurs in lial cells have been shown to have a rapid turn-
those cells in contact with the basement mem- over of between 1 week and 10 days [16].
brane. These putative LSCs can divide symmetri-
cally to regenerate and asymmetrically to produce
daughter cells, termed transient amplifying cells 6.2 Limbal Stem Cell Deficiency
(TAC), which are greater in number and retain a
high proliferative ability in health, to replace Limbal stem cell deficiency (LSCD) is a painful,
superficial epithelial cells that have been natu- blinding, chronic condition characterised by a
rally desquamated. Aside from the vertical and failure of the limbus in its function as a stem cell
horizontal movement of cells that characterises niche and also as a physical barrier to prevent
the corneal epithelium, centripetal movement of vascularised conjunctival tissue from migrating
cells from the limbus to the centre of the cornea into the cornea. Consequently, failure of limbal
has also been previously demonstrated as stem cells can occur due to either a loss of limbal
described above [6]. Early clinical observations stem cells or their dysfunction [17]. It is a disease
involving migration of limbal pigment post- therefore that results in conjunctivalisation of the
trauma [8], epithelial microcysts at the site of corneal surface with migration of a thick, vascu-
corneal sutures [9] and limbal cells marked with larised, conjunctival tissue (pannus) leading to
India ink demonstrated a centripetal movement scarring and vascularisation of the cornea and
of epithelial cells from the limbus towards the subsequent permanent visual impairment,
centre of the cornea [10] and evidence of epithe- together with recurrent or persistent epithelial
lial regeneration following limbal transplantation erosions and ulcerations that result in stromal
[11]. This led to the long-established hypothesis opacity and chronic pain and discomfort.
that the limbus contains an important driving
force towards epithelial regeneration and repair
[5, 12]. Evidence of this centripetal movement is 6.2.1 Aetiology
also seen clinically in the healing of central cor-
neal abrasions, where a leading edge of three or LSCD can occur as part of a hereditary group of
more sheets of epithelium arising from an intact conditions, such as aniridia, among others, or as
peripheral epithelium migrates centrally until an acquired condition, including chemical and
they meet to close the epithelial defect [13, 14]. thermal burns (the most frequent causes of
6  Corneal Epithelial Stem Cells: Methods for Ex Vivo Expansion 79

LSCD), inflammatory diseases of the anterior France and 0.87 cases per million population in
segment (e.g. ocular cicatricial pemphigoid, Germany [25].
Stevens-Johnson syndrome and chronic limbitis) The leading causes of LSCD in the UK are
or extensive microbial infections. It can also chemical/thermal injury, SJS, aniridia and ocular
result from potentially damaging ocular treat- mucous membrane pemphigoid. The combined
ments such as extensive cryotherapy; radiation; incidence of these conditions is 1:25000  in the
surgery at the limbus; subconjunctival or topical UK, and Shortt et al. stipulated that assuming that
anti-proliferative agents, e.g. mitomycin C 10% of these patients develop LSCD, this results
(MMC) for ocular tumours and adjunct therapy in 240 new cases of LSCD per year in the UK
in glaucoma surgery; and long-term use of pre- alone [26].
served topical medications (e.g. for glaucoma). In view of its low incidence and prevalence
LSCD has also been shown to result from pro- but significant debilitation from the condition,
longed contact lens use, although this is rare LSCD is designated an orphan disease by the
[17]. In the absence of an obvious aetiology, it is European Medicines Agency. This status pro-
a condition that may be described as idiopathic, vides incentives for trial sponsors to develop
or of an unknown cause, although this is medicines that would otherwise fail a test of eco-
extremely rare. In most cases of acquired LSCD, nomic viability. In 2015, the European Medicines
it is likely that both LSCs and their niche are Agency recommended for Holoclar® (Chiesi
affected [18]. Farmaceutici S.p.A., Parma, Italy) to be given
approval for the marketing authorisation for the
production of an autologous limbal stem cell
6.2.2 Epidemiology product for patients with moderate to severe
LSCD due to ocular surface burns (see
The incidence and prevalence of LSCD are not Sect. 6.4.5).
well understood. As has already been described,
there are a variety of conditions known to cause
LSCD, and these conditions are largely 6.2.3 Semiology
uncommon.
Among the hereditary conditions, the reported Limbal stem cell deficiency may affect patients
prevalence of aniridia ranges from 1 in 40,000 to of all ages, depending on the original aetiology.
1 in 100,000 people [19]. Patients typically present with discomfort or pain
Globally, chemical burns account for 7–18% and decreased vision. They may also have suf-
and thermal burns for 16% of eye injuries leading fered from recurrent breakdowns in their corneal
to an emergency visit [20]. However, it is epithelium resulting in recurrent corneal erosions
unknown what proportion of these patients will or even corneal ulceration and consequent sec-
develop LSCD.  The UK national incidence of ondary infection, resulting in corneal thinning or
severe chemical corneal injuries is unclear [21, even perforation. These features, over time, lead
22]. A recent national British Ophthalmological to permanent conjunctivalisation of the epithe-
Surveillance Unit (BOSU) study estimated the lium and may also lead to corneal neovasculari-
incidence of severe chemical injuries, with the sation at the deeper stromal layer, often
potential to lead to total LSCD, as 12 per 60 mil- accompanied by scar formation. The irregular
lion or 0.02 per 100,000 per year [23]. conjunctival epithelial surface over the cornea
The incidence of Stevens-Johnson Syndrome also leads to an unstable tear film, reduced visual
(SJS) is 2.6 to 7.1 cases per 1 million person-­ acuity and photosensitivity; it may produce fila-
years in the USA and 1.1 cases per 1 million ments and further corneal erosions. Epithelial
person-years in Germany [24]. The incidence of keratinisation may also occur in the presence of
ocular mucous membrane pemphigoid is esti- an adverse ocular surface environment, including
mated at 1.16 cases per million population in tear deficiency.
80 G. S. Figueiredo et al.

Limbal stem cell deficiency can be subclassi- appearance on slit lamp biomicroscopy.
fied into partial or total LSCD (TLSCD), depend- Sometimes a clear area of corneal thinning may
ing on whether the whole limbus or only a be observed, along with other areas of corneal
segment of the limbus is involved, or unilateral or scarring, which may be superficial (anterior stro-
bilateral, depending on whether both eyes are mal) or deep (mid-deep stromal) or both. The
affected [27]. palisades of Vogt at the limbus are easily observed
Limbal stem cell deficiency presents with typ- through slit lamp biomicroscopy, in normal eyes,
ically classical clinical features, which are depen- as “fronds” of tissue at the limbus; however, they
dent on the severity of the condition (i.e. partial become much more disorganised and less obvi-
or total LSCD). With an abnormal (dysfunc- ous or flattened in cases of LSCD (Fig. 6.2).
tional) limbus, conjunctivalisation of the cornea In total limbal stem cell deficiency (TLSCD),
occurs and is the hallmark of the condition, as there is complete obliteration of the limbus 360°
described above. Dua et al. showed in 2003 that around the cornea, resulting in conjunctivalisa-
the most significant clinical manifestation of con- tion of the entire corneal surface (Fig.  6.3). In
junctivalisation of the cornea is a persistent epi- such cases, the characteristic areas and demarca-
thelial defect [28]. This can lead to other tion line of late epithelial staining seen in cases of
complications such as corneal ulceration and partial LSCD are not observed. Chronic inflam-
even perforation. In LSCD, late fluorescein stain- mation and redness of the anterior segment of the
ing of the corneal epithelium can be observed, eye often result.
corresponding to the area of conjunctivalisation As described above, limbal stem cell defi-
as a result of partial or total LSCD (Fig. 6.1). ciency can be classified into partial or total
This characteristic finding is the result of the LSCD, unilateral or bilateral, and this classifica-
fluorescein dye seeping through to the basement tion defines further management of the patient.
membrane of the cornea due to the loss of cell-­ Total LSCD is defined as the complete absence of
cell tight junctions in the conjunctivalised cor- normal corneal epithelial cells clinically by slit
neal epithelium. Late staining of these cells lamp biomicroscopy but also on histopathologi-
lacking in these tight junctions represents either a cal examination (corneal impression cytology,
total replacement of the corneal epithelium by CIC); in contrast, partial LSCD contains some
conjunctival-type epithelium or a mosaic pattern normal corneal epithelial cells as previously
of corneal and conjunctival epithelial cells. A described.
clear demarcation line between corneal and con-
junctival phenotypic epithelial cells can some-
times be clearly observed in cases of partial 6.2.4 Management
LSCD as conjunctival cells stain with fluorescein
much more easily (Fig.  6.1). This stippled, late The management of partial LSCD is essen-
fluorescein staining on the cornea can occasion- tially medical, including the use of inten-
ally be seen to form “whorls”, giving a typical sive preservative-­
free ocular surface topical

Fig. 6.1 Colour Well-defined area of


Partial
photograph showing delayed fluorescein
conjunctivalisation
clinical signs of partial staining with a clear
of corneal surface
limbal stem cell demarcation line
deficiency not involving
the visual axis

Vascularised, Area of clear cornea


opaque section of in partial LSCD
cornea
6  Corneal Epithelial Stem Cells: Methods for Ex Vivo Expansion 81

a b

Fig. 6.2  High magnification colour photographs showing presence of palisades of Vogt (arrows) in a healthy eye (a)
and absence of palisades of Vogt (arrow) in a limbal stem cell deficiency eye caused by chemical burn (b)

a drops (AlloSE) [27, 29, 30]. These patients often


do not require surgical treatment if the visual axis
is not affected. In severe cases, as the limbal area
is more affected, the visual axis is often involved,
which leads to a very low vision. In addition,
these patients suffer from pain and photopho-
bia, due to recurrent epithelial defects, chronic
ocular surface inflammation, corneal conjucti-
valisation, central corneal opacity and superficial
vascularisation, which make the patient function-
ally blind [18]. If the visual axis is affected with
b consequent reduced visual acuity, partial LSCD
can be treated by surgically excising the abnor-
mal epithelium (i.e. epitheliectomy) to allow the
denuded cornea to be re-epithelialised from the
remaining intact limbal epithelium [13, 14, 31].
This may be combined with an amniotic mem-
brane transplantation to help reduce the risk of
recurrence of conjunctivalisation [32]. In addi-
tion, topical preservative-­free lubricants, steroid
eye drops and ASE/AlloSE are often used [26].
Total LSCD has traditionally been more com-
Fig. 6.3  Colour photograph showing clinical signs of
total limbal stem cell deficiency, resulting in conjuncti- plex to manage, with a variety of conservative
valisation of the entire corneal surface involving the visual measures aimed at controlling symptoms and
axis (a) and total late epithelial staining without demarca- maintaining the integrity of the ocular surface,
tion line (b)
which as a result may also produce some form of
l­ubrication and the judicial use of topical steroid visual improvement. These measures included
therapy (often as pulse therapy), and, in some copious use of topical preservative-free lubricants
cases, these can be combined with autologous and ASE/AlloSE [26, 33–35]. More challenging
serum eye drops (ASE) or allogeneic serum eye cases, with a persistent epithelial defect that is
82 G. S. Figueiredo et al.

unresponsive to medical treatment, a few addi- expanded limbal stem cells in patients with uni-
tional treatment options such as tarsorrhaphy [26, lateral TLSCD, where a much smaller biopsy
36–41], botulinum toxin-induced ptosis [42–45], was taken from the healthy fellow eye and
bandage contact lenses [26, 46–50] or human expanded ex  vivo in culture before being trans-
amniotic membrane transplantation [51–55] could planted into the affected eye with negligible or no
be used to temporarily protect the diseased ocular risk to the donor eye [69]. They described the
surface. In most cases of LSCD, patients are often cases of two patients who had suffered severe
managed in more specialised tertiary centres with alkali burns in one eye, with the fellow eye
treatment including a number of these options in remaining healthy. A 1  mm2 biopsy was taken
combination. Lamellar (partial thickness) or pen- from the limbus of the uninjured eye, fragmented
etrating keratoplasty (PKP, full thickness) corneal and treated with trypsin (0.05% trypsin and
transplantation involves partial or total replace- 0.91% edetic acid). The cells were then plated on
ment of the central cornea, preserving the limbus. mitotically inactivated 3T3-J2 mouse fibroblast
Therefore, although it replaces an opaque cornea cells and cultured in 5% carbon dioxide in a
with a clear cornea, it does not restore LSC func- medium that had been used previously to culture
tion, and, in time, the corneal transplant also skin epithelial cells. After 16 or 19 days, the cells
becomes opaque and fails due to persistent LSCD were released from the plastic dish using a neu-
and consequent recurrent conjunctivalisation. This tral protease, Dispase II, and placed either on a
is therefore not a suitable treatment option for this petrolatum gauze or a soft contact lens. The
condition unless the LSCs are replaced prior to injured eye was then prepared for transplantation
corneal transplantation [56]. by surgically removing the diseased conjuncti-
Where there is a lack of healthy, functional valised epithelial surface by superficial keratec-
stem cells such as in total LSCD, the ideal treat- tomy (i.e. dissection with a blunt knife or
ment would be to replace them, so that they can scissors), with a 360° conjunctival peritomy,
then proliferate, repopulate their niche and extending at least 2 mm beyond the limbus. This
provide the natural LSC resource capable of
­ protocol is still in use, although there have been
maintaining normal corneal epithelial homeosta- some important modifications since, as described
sis. Replacement of limbal stem cells was first by the Newcastle Limbal Stem Cell Group [70].
proposed by Kenyon and Tseng in 1989, by per- Current procedures involve the use of an animal-­
forming whole limbal tissue grafts [11, 57]. free culture method that adheres to Good
Whole tissue grafts were taken either from a Manufacturing Practice (GMP) principles, using
patient’s contralateral healthy eye (autograft) in human autologous serum (HAS) in place of fetal
unilateral cases of LSCD or from a living related bovine serum (FBS); transferrin and glutamine
or cadaveric donor (allograft) in cases of bilateral have been removed from the culture medium; and
LSCD.  In allogeneic tissue transplantations, the 1  mm2 limbal biopsy is immediately plated
post-operative systemic immunosuppression was onto human amniotic membrane (epithelium side
necessary, which had its own drawbacks particu- up), instead of being fragmented, trypsinised and
larly in terms of the risk of opportunistic infec- cultured on 3T3-J2 mouse fibroblast cells.
tions and iatrogenic malignancy [11, 58–66]. Following transplantation of the human amniotic
Additionally, they are associated with poor long-­ membrane (HAM) with the ex  vivo expanded
term survival outcomes, as they often fail due to stem cells, a further HAM is sutured in place on
allograft rejection [19, 67]. The large amount of top of the transplanted stem cells and a bandage
tissue required to be transplanted in these opera- contact lens placed on top of the second HAM for
tions poses enormous risk to the donor eye in protection [70].
autografts and living-related allografts in terms Two-year follow-up and analysis in the
of developing LSCD [68]. Pellegrini case series showed these autologous
In 1997, Pellegrini et al. described the first two limbal stem cell transplants (ALSCT) to have
cases of autologous transplantation of ex  vivo been successful, and this report remains the
6  Corneal Epithelial Stem Cells: Methods for Ex Vivo Expansion 83

benchmark against which all other limbal stem and to allow the eye to settle. During this period,
cell transplants are based on. In the following the eye is treated with intensive eye drops in
years, a number of studies have refined or adapted the form of preservative-free ocular lubricants,
this technique and reported on their outcomes. preservative-­free topical steroid drops and
This was extensively reviewed by Baylis et al. in preservative-­free topical antibiotic drops, often in
2010, who reported an overall success rate of combination with ASE. This treatment is gradu-
77% for autografts and 73% for allografts [56]. ally tapered over the initial few weeks to months,
They also showed that failures appeared to hap- aiming to stop all topical steroid treatment by
pen in the first 1–2  years post-transplant, based 12 months [70].
on studies by Sangwan et  al. (2006), Pauklin Whilst ex vivo expanded stem cell or whole
et  al. (2010) and Rama et  al. (2010) [71–73]. tissue transplantation alone may improve vision
Significantly, Rama et al. reported that those LSC in these patients, a combination with corneal
cultures that had more than 3% of cells staining transplantation is often necessary to achieve
positive for p63 bright resulted in 78% success optimal visual recovery. Corneal transplanta-
clinically, whereas those with less than 3% cells tion without prior regeneration of the corneal
staining p63 positive were only 11% successful epithelium by stem cell transplantation inevi-
clinically [73]. tably fails in patients with severe or TLSCD
Prior to successful ALSCT, especially in [74]. Following successful ALSCT, the deeply
patients who have suffered a thermal injury to scarred cornea can be subsequently replaced by
the eye, it is important to ensure the eyelids are corneal transplantation (penetrating or lamel-
well positioned and apposed, without any lag- lar keratoplasty) often as part of a two-stage
ophthalmos or any exposure of the corneal sur- approach [75, 76]. Basu et  al. demonstrated a
face, and that Schirmer’s test of tears production two-stage approach as having favourable cor-
is within normal limits to maximise success of neal transplantation survival outcomes com-
ALSCT. It is greatly important that preoperative pared with a single-stage approach [76].
­inflammation is managed with topical steroids,
whether pulsed with high doses or on a low-
maintenance dose, and that the ocular surface is 6.3  orneal Epithelial Cell
C
well lubricated using preservative-free topical Culture
therapy, particularly (although not essential)
with ASE in advance of the ALSCT. It is essen- Corneal epithelial culture can be carried out either
tial that a careful assessment of the “unaffected” in a single cell suspension culture with an inacti-
other eye in unilateral cases is made to ensure vated 3T3-J2 mouse fibroblast feeder layer or as
that it is in fact “healthy”. As a minimum, CIC an ex  vivo expansion of limbal explant cells on
and, if available, IVCM should be performed in HAM. The single cell suspension method allows for
both eyes to confirm the TLSCD diagnosis in the cells to be counted so a specific number of cells can
affected eye and to exclude any evidence of lim- be used, and experiments are therefore better con-
bal stem cell damage in the unaffected contralat- trolled, improving reproducibility and comparison
eral healthy eye so as to not iatrogenically induce in scientific experiments. It is therefore a method
limbal stem cell failure post-biopsy, although widely used in laboratory science. However, the
this has not yet been reported since the introduc- single cell suspension method involves the use of
tion of small limbal biopsies by Pellegrini et al. animal feeder cells which potentially carry the risk
in 1997. of disease transmission. The ex vivo expansion
Following successful ALSCT, it is important method of limbal explants on HAM that we devel-
to maintain a bandage contact lens for a suit- oped is a xeno-­free culture system that eliminates
able period of time, usually 8–12  weeks post-­ all animal feeders and components, which therefore
operatively, to protect the ocular surface, the satisfies MHRA regulations for animal-free prod-
donor LSCs and the HAM covering the LSCs ucts to be used in human clinical trials as described
84 G. S. Figueiredo et al.

in a review by the World Health Organization and cells by trypsinisation as described above, cen-
the Advisory Committee on the Safety of Blood, trifuging the cell suspension to a pellet and
Tissues and Organs (SaBTO) of the donation of resuspending the cell pellet in 1  ml of 90%
starting material for cell-based advanced therapies FBS, 10% dimethyl sulphoxide (DMSO,
[77–80]. The ex vivo expansion of limbal explant Sigma-Aldrich). The 3T3 fibroblasts are then
culture system is therefore more widely used in frozen at −80 °C.
human clinical trials [71, 81].
6.3.1.4 Inactivation by Irradiation
of 3T3 Cells and Preparation
6.3.1 M
 ethod of 3T3 Fibroblast of Feeder Layers
Coculture (Cell Suspension) For successful coculture of LSCs, the 3T3
mouse fibroblast layer must first be mitotically
6.3.1.1 Initiation of 3T3 Cultures inactivated to prevent their proliferation/over-
from Frozen Stock growth and subsequent suffocation of LSCs in
A Cryovial of frozen mouse strain J2-3T3 fibro- culture. To do this, the cells are irradiated at
blast stock is thawed in a water bath at 37 °C. Once 60  Gy. The medium is then removed and the
defrosted, the 3T3 fibroblasts are suspended in cells rinsed twice with DPBS before their
5 ml of 3T3 medium and centrifuged at 1000 RPM release by trypsinisation as described above.
for 3 min. The cell pellet is then resuspended in The suspended cells are then transferred to a
10 ml 3T3 medium (Dulbecco’s Modified Eagle centrifuge tube and the trypsin inactivated by
Medium [DMEM] with 10% FBS) in a 75  cm2 the addition of 3T3 medium. The resulting sus-
(T75) cell culture flask. pension is then centrifuged at 1000  RPM for
3  min at room temperature. The resulting cell
6.3.1.2 Maintenance and Subculture pellet is resuspended in 3T3 medium, and cell
of 3T3 Feeder Layer count and viability are assessed. The cells are
The 3T3-J2 fibroblast cultures are examined then plated in a 9.6 cm2 tissue culture well at a
every day and the medium exchanged on alter- final density of 2.4  ×  104 cells per cm2 with
nate days. Once 60–75% confluent, the cells are 0.5  ml 3T3 medium and incubated for 24  h at
passaged by firstly aspirating the 3T3 medium 37 °C to allow for their adhesion to the culture
from the flask and gently rinsing the cells twice well before the addition of LSCs.
with Dulbecco’s Phosphate-Buffered Saline
(DPBS). The 3T3 cells are then released from the
base of the cell culture flask by incubation with 6.3.2 L
 imbal Explant Culture
trypsin 0.05% for 10 min at 37 °C. After ensuring on a Human Amniotic
the cells are free in suspension, the cells are Membrane Substrate
removed into a centrifuge tube, and the trypsin is
inactivated by adding 3T3 medium. The resulting 6.3.2.1 Preparation of Limbal Epithelial
suspension is then centrifuged at 1000 RPM for Culture Medium
3 min at room temperature. The resulting pellet is Limbal epithelial culture medium is prepared
then resuspended, and cell count and viability are 2  days prior to the limbal biopsy. Dulbecco’s
assessed. One hundred thousand viable cells are Modified Eagle Medium (DMEM, low glucose
then added to 10 ml 3T3 medium and the suspen- with GlutaMAX and pyruvate) is used as the
sion transferred to a new 75 cm2 cell culture flask. main basic ingredient in a 3:1 mixture with 25%
Ham’s F12 and GlutaMAX nutrient medium (i.e.
6.3.1.3 Preparation of Frozen Stocks 125 ml Ham’s F12 was added to 375 ml DMEM).
To maintain a supply of 3T3 mouse fibroblasts, From this mixture, 1  ml is removed, and seven
frozen stocks are prepared from the early pas- supplements (all Sigma-Aldrich, UK) are added
sages. This is done by releasing subconfluent in sequence as per Table 6.1.
6  Corneal Epithelial Stem Cells: Methods for Ex Vivo Expansion 85

Table 6.1  Table showing supplementation of basic lim- can be stored at 4  °C and leftover autologous
bal epithelial culture medium
serum stored at −20 °C in case further complete
Supplement Concentration Volume LSC medium is required during the culture
Hydrocortisone 0.4 μg/ml 500 μl process.
Insulin 5 μg/ml 250 μl
Triiodothyronine 1.4 ng/ml 35 μl
Adenine hydrochloride 24 μg/ml 300 μl 6.3.2.2 Preparation of HAM Construct
Cholera toxin 8.4 ng/ml 84 μl Large HAMs (3 × 3 cm) suitable for clinical use
Epidermal growth factor 10 ng/ml 50 μl are obtained in the UK from NHS Blood and
Transplant (NHSBT) Tissue Services based at
Speke, Liverpool, UK, and stored at −80 °C. The
The mixture is then filtered in a 0.2 μm pore HAM construct consists of a sheet of 3 × 3 cm
size 500 ml fast PAS Nalgene® sterile filter unit HAM that is trimmed to size and wrapped around
(Sigma-Aldrich, UK) to make up the basic LSC a sterile (autoclaved) glass coverslip, epithelial
medium. side up with the edges trapped between a second
For the preparation of complete LSC medium, coverslip underneath (Fig. 6.4).
human autologous serum is then prepared using A wash solution of 49.5  ml DPBS with
the patient’s own blood that has been obtained 0.5  ml penicillin-streptomycin is made, and
24 h previously and allowed to clot overnight at 2 ml of the wash solution is then added to four
4 °C. To do this, the blood is initially collected (at wells of a six-­well plate, and 2 ml of complete
least 60 mls) in sterile 50 ml centrifuge tubes and LSC medium is added to the remaining two
centrifuged at 1600 × g at room temperature for wells of the six-well plate (Fig. 6.5a). Two ster-
10 min. The supernatant (serum) is then carefully ile (autoclaved) glass coverslips are carefully
removed into two new sterile 50  ml centrifuge taken and washed in sequence in the prepared
tubes and centrifuged again at 1600 × g at room wells of the six-well plate so that each glass
temperature for 10 min. The supernatant (serum) coverslip is rinsed twice in the wash solution
is again carefully removed into a new sterile and once in complete LSC medium. The glass
50 ml centrifuge tube. Eleven microlitres of the coverslips are then placed on a flat surface
prepared autologous serum is then added to 99 ml inside the laminar flow hood, ready to receive
of the previously prepared basic LSC medium the HAM.  The HAM comes frozen wrapped
(i.e. 10% autologous serum), and the mixture is round a nitrocellulose paper and is prepared by
filtered once again in a 0.2 μm pore size 500 ml thawing at room temperature immediately prior
fast PAS Nalgene® sterile filter unit (Sigma-­ to use before washing it sequentially in a similar
Aldrich, UK) to make up the complete LSC manner (twice in the wash solution and once in
medium. The complete LSC medium is then ali- LSC medium) in the six-well plate. The HAM is
quoted into 15 ml centrifuge tubes. There are two then carefully unwrapped from its nitrocellulose
10  ml LSC medium aliquots to which 100  μl backing (Fig. 6.5b) and stretched out epithelial
penicillin-streptomycin (or 50  μl gentamicin in side up, over a pre-washed glass coverslip on a
penicillin-allergic patients) is added, and these flat surface carefully using sterile suture tying
are used for plating the HAM construct and sub- ophthalmic micro forceps (Duckworth & Kent,
sequent plating the limbal biopsy. A further 13
aliquots of 6 ml LSC medium each are prepared Human amniotic
without antibiotics as per MHRA requirements to membrane (HAM)
be used for the medium exchanges. Three micro-
biological samples are sent per LSC medium Sterile glass
preparation: a sample from the basic LSC coverslips
medium, autologous serum and the complete
Fig. 6.4  Diagram showing human amniotic membrane
LSC medium. Complete LSC medium aliquots (HAM) construct with HAM trapped between two sterile
are stored at 4 °C. Leftover LSC basic medium glass coverslips
86 G. S. Figueiredo et al.

a b

c d

Fig. 6.5  Colour photographs showing preparation of from its nitrocellulose paper backing, (c) stretching out
human amniotic membrane (HAM) construct. (a) Washing and trimming edges of HAM, (d) folding edges of HAM
HAM in DPBS and LEC medium, (b) unravelling HAM around glass coverslip

UK) so that its edges just overhang the edges of gently l­ifting it, maintaining the same orienta-
the coverslip. HAM orientation (i.e. epithelial tion, so that the HAM edges neatly wrap around
side up) is checked several times with a ster- the edges of the coverslip (Fig. 6.5d). The HAM
ile lint-free surgical spear (EyeTec, UK). The is then stretched out over the edges of the cov-
edges of the HAM are then trimmed with an erslip so that it lies taught over the coverslip.
overhang of 2–3 mm from the edges of the glass Care is always taken to prevent the HAM from
coverslip (Fig. 6.5c). The glass coverslip is then drying out by moistening it with complete LSC
lifted off the flat surface by carefully sliding medium as required throughout the procedure.
the blade of a scalpel with a number 22 surgi- Two or three drops of complete LSC medium
cal scalpel blade under the glass coverslip and are then added to the surface of the second glass
6  Corneal Epithelial Stem Cells: Methods for Ex Vivo Expansion 87

coverslip, and the HAM/first coverslip construct the palisades of Vogt, and the biopsy is then
is slowly lowered from one edge to the opposite placed in a well of a sterile six-well plate, coated
edge of the second coverslip so that there are no in ocular viscoelastic device (Healon GV®,
trapped air bubbles between the two coverslips Abbott Medical Optics Inc., Sweden) to prevent
and the edges of the HAM are neatly tucked it from desiccation during immediate transpor-
between the coverslips. Capillary suction from tation to the GMP facility [70].
the drops of complete LSC medium adheres the
two glass coverslips together. If any of the HAM 6.3.2.4 Limbal Explant Culture
edges are loose, the whole procedure is repeated Once the limbal explant biopsy was received in
as the coverslips are separated and the HAM/first the GMP facility, it is immediately transferred to
coverslip is gently lowered onto the second cov- the clean room, where it would be plated onto
erslip again until all edges are neatly tucked in. the HAM construct. The HAM construct is
The HAM/coverslip construct is then placed in a inspected a final time, and medium is removed
well of a new sterile six-well plate and incubated from the culture well. The six-well plate contain-
in 2 ml complete LSC medium in an incubator at ing the HAM construct is tilted approximately
37  °C and 5% CO2. A 24  h check of the HAM 15° to allow all of the medium to be drained
construct is made the following day, prior to the from the surface of the HAM, whilst the limbal
limbal biopsy scheduled for the day after, and if explant biopsy is washed in complete LSC
any HAM edges are loose, it is replated. medium. To do this, the biopsy is carefully
picked up with sterile non-toothed ophthalmic
6.3.2.3 Limbal Explant Biopsy micro-tying forceps (Duckworth & Kent, UK),
The limbal biopsy is a small, 1 by 2 mm partial preventing any potential damage to the biopsy
thickness autologous biopsy of the limbus from and maintaining its orientation throughout (epi-
the patient’s healthy contralateral eye in cases of thelial side up). The biopsy is washed thor-
unilateral LSCD, superiorly at the 12 o’clock oughly, with complete LSC medium, to facilitate
position or inferiorly at the 6 o’clock position, adhesion to the HAM. It is then dried by gently
where it is known there are more LSCs [82]. “running” it on a sterile, flat surface, until com-
The limbal explant biopsy is performed in a pletely dry. The six-­well plate containing the
dedicated ophthalmic theatre under local or gen- HAM construct is then laid flat, and the explant
eral anaesthetic (i.e. patient preference) as day biopsy carefully placed in the centre of the HAM
case procedure. It involves exposing the limbus and allowed to rest for 2  min before 1.3  ml of
superiorly, followed by performing a linear complete LSC medium is gently added. The con-
superficial keratotomy (150  μm deep) using a struct is then carefully inspected to make sure
diamond knife (Duckworth & Kent, UK) paral- that limbal biopsy is completely adherent to the
lel to the limbus (either side, close to 12 o’clock HAM before it is incubated. The culture medium
superiorly) in an area previously determined (by is exchanged on the third culture day and every
the surgeon performing the operation) on slit second/third day thereafter until the culture
lamp assessment as being rich in well-character- reaches >90% confluence. The explant culture is
ised palisades of Vogt. Then, a careful lamellar incubated at 37 °C in 5% CO2.
dissection posteriorly is performed using a
2.25 mm, 60° circular edge head blade (Beaver-
Visitec International Ltd., UK) going beyond 6.3.3 T
 he Use of Human Amniotic
the posterior edge of the limbus. This is fol- Membrane as a Substrate
lowed by two relieving incisions at the two lat- for the Culture of Limbal
eral edges of the biopsy and finally a posterior Stem Cells
dissection of the limbal biopsy that is accompa-
nied by a peritomy of the conjunctiva. The pos- Human amniotic membrane (HAM) has long
terior and superior corner edge of the biopsy is been used in ophthalmic procedures to treat ocu-
marked with a surgical marker pen away from lar surface diseases and burns [55], due to its
88 G. S. Figueiredo et al.

ability to promote re-epithelialisation via growth Decellularisation of human tissues has been
factors (such as EGF, KGF and HGF) and its extensively investigated and applied to a wide
basement membrane [83], and to inhibit fibrosis array of tissues for different clinical applications.
through suppression of TGFβ signalling [84]. A wide range of agents have been used to decel-
Fresh HAM consists of an epithelial layer (devi- lularise tissues, including sodium dodecyl sul-
talised by the freezing process), a stroma and a phate (SDS) [99], dispase, thermolysin, trypsin,
thick basement membrane. The HAM stroma ethylenediamine tetra-acetic acid (EDTA) and
exhibits anti-inflammatory effects by suppress- ethanol, among others [95, 105]. In 2013,
ing interleukin 1α and interleukin 1β in cultured Saghizadeh et  al. described a new, simpler and
limbal epithelial cells [85] and preventing poly- faster decellularisation method that involved
morphonuclear cells infiltrating into the corneal using a cotton bud soaked in sodium hydroxide
stroma [86]. In addition, the basement mem- 0.5  M solution to rub the epithelial side of the
brane and extracellular matrix components of HAM to debride its epithelial cells in a period of
HAM, when used as a substrate, have shown 5–10 s [95]. It was demonstrated that the HAM
similar properties to conjunctival and corneal retained the native ECM structure following
epithelium [87, 88]. decellularisation with sodium hydroxide; how-
Human LSCs have commonly been expanded ever, LSCs were not cultured on the substrate.
ex vivo on human amniotic membrane for both Decellularisation may better define HAM;
clinical and research purposes [70, 89, 90]. however, the risk of disease transmission remains
HAM shows low or no immunogenicity [91–94]. in a non-sterile substrate. With this in mind, Hogg
It is not a defined substrate and has some poten- et al. demonstrated in 2015 that γ-irradiation is an
tial disadvantages, most importantly its biologi- effective method of terminal sterilisation in the
cal variability and the potential to carry or production of decellularised skin dermis for
transmit infections, but despite this, HAM direct allogeneic transplantation [106].
remains the gold standard and most widely used A characteristic of HAM is its biological vari-
substrate for the expansion of LSCs in vitro and ability, both in thickness and, consequently, elas-
in clinical trials [56, 95, 96]. Consequently, there ticity [107, 108]. In 2012, Jones et  al. [109]
has been a significant drive to better characterise applied a theory first demonstrated in 2006 by
HAM in order to develop a better substrate for two separate groups that showed that human
the in  vitro expansion of LSCs [96–98]. One mesenchymal stem cells are sensitive to substrate
such approach has been to decellularise HAM, rigidity and matrix elasticity which when altered
as this process can remove all cellular, poten- can give rise to specific stem cell lineages [110,
tially immunogenic components from a tissue, 111]. Jones et  al. demonstrated that LSCs cul-
whilst preserving key extracellular matrix tured on uncompressed (i.e. with reduced stiff-
(ECM) components and the basement mem- ness) collagen gels were less differentiated than
brane, ensuring cell attachment and expansion. those cultured on compressed collagen gels
The decellularisation process may create a more [109]. The same group then showed the differ-
consistent, defined substrate that does not elicit ences in corneal stiffness, where in a healthy,
an adverse immunological response in vivo [99– clear cornea, the central cornea is stiffer than the
101]. Several studies have suggested that LSCs peripheral cornea, which drives the migration
cultured on decellularised HAM expand more and differentiation of LSCs centripetally towards
quickly but are also more differentiated com- the centre [112]. Molladavoodi et  al. supported
pared to fresh frozen HAM [102–104]. The this finding when demonstrating that human cor-
improved cell proliferation has been attributed to neal epithelial cells had lower rates of migration
better cell adhesion to the substrate compared to in compliant tissues [113]. It has therefore been
fresh frozen HAM, due to exposure of the base- hypothesised that the mechanical properties
ment membrane, unobstructed by a devitalised (stiffness) of the corneal anterior surface repre-
epithelial cell layer. sent a major factor regulating corneal epithelium
6  Corneal Epithelial Stem Cells: Methods for Ex Vivo Expansion 89

homeostasis [112]. In particular, it has been pro- rate. Nevertheless, it must also be recognised
posed that the mechanical environment of the that the degree of stiffening achieved with
limbus (i.e. being more compliant) may be funda- γ-irradiation may not be sufficient to influence
mental for the maintenance of LSC stemness, the behaviour of LSCs in culture. SDS decellula-
whereas the stiffer matrix of the central cornea is rised HAM may be more efficacious as a sub-
instrumental in driving centripetal cell migration strate for the ex  vivo expansion of limbal
through durotaxis and inducing epithelial cell epithelial cells for use in clinical trials, and in
differentiation [112]. Lepert et al. used Brillouin particular, the use of a γ-irradiated decellularised
microscopy as a reliable method of measuring HAM allows the user to start the manufacturing
corneal stiffness and demonstrated similar find- process with a sterile substrate, making it poten-
ings [114]. Chen et  al. demonstrated that tially safer [106, 117, 118].
γ-irradiation can cross-link collagen-chitosan
scaffolds [115]. Mi et al. showed that UVA cross-­
linked plastically compressed collagen (PCC) 6.3.4 Assessment of Culture Growth
gels had a greater breaking force than uncross-­
linked PCC [116]. 6.3.4.1 Histology and Immunostaining
In our own experiment [117], comparing the Histological analysis includes an assessment of
ex vivo culture method described here using four both the macroscopic and microscopic appear-
different substrates, namely, fresh HAM, NaOH ances of the HAM and the ex vivo expanded epi-
decellularised HAM, SDS decellularised HAM thelial cells with haematoxylin and eosin staining,
and γ-irradiated SDS decellularised HAM, we including whether a monolayer of epithelium is
were able to show that there was a significantly present, with a quantitative description of the
more efficient rate of ex vivo LSC expansion on thickness of the epithelium and the presence or
decellularised HAM compared with fresh HAM, absence of goblet cells. It also includes whether
with cultures grown on SDS decellularised HAM there were any abnormal features relating to the
reaching confluence fastest overall. Cultures on size of the cells and their nucleus or cytoplasm.
SDS decellularised HAM preserved a greater The cells are stained by immunocytochemistry for
expression of putative stem cell (ΔNp63 and putative limbal stem cell markers ΔNp63 (Dako
ABCG2) and cell proliferation markers (Ki67) Ready-to-Use pre-diluted [Agilent Technologies,
and lower expression of markers of cell differen- Glostrup, Denmark]; antigen retrieval using Dako
tiation (CK12 and CK13) than the limbal high pH  8 target retrieval solution) and ABCG2
explants cultured on fresh HAM or NaOH decel- (clone BXP-21 [Chemicon, EMD Millipore,
lularised HAM.  Gamma-irradiation enables us Billerica MA, USA], used at a dilution of 1:50
to start with a sterile substrate and did not appear with pH  6 retrieval using Bond ER1 [Leica
to affect either the rate of ex vivo expansion from Biosystems, Wetzlar, Germany]), a marker of cor-
limbal explants, the preservation of stem cell neal epithelium CK12 ([Abgent, San Diego CA,
properties in the cultured cells or the stiffness of USA], diluted 1:100  in 0.2% calcium chloride
the substrate. Using Brillouin shift measure- buffer; trypsin retrieval at pH 7.8), two markers of
ments, we have shown that decellularised HAM conjunctival epithelium CK13 ([Abcam,
without γ-irradiation was found to have slightly Cambridge, UK], dilution 1:50, trypsin antigen
reduced stiffness compared to fresh HAM and retrieval at pH 7.8) and MUC5AC (clone CLH2
that cells cultured on decellularised HAM had [Novocastra, Newcastle-upon-Tyne, UK] dilution
greater expression of stem cell markers. HAM of 1:100, with pH  6 retrieval using Bond ER1
stiffness is not influenced by our decellularisa- [Leica Biosystems, Wetzlar, Germany]) and a
tion or sterilisation methods and, in turn, that marker of stem cell proliferation Ki67 (Dako
LSC differentiation does not appear to be influ- Ready-to-Use pre-diluted [Agilent Technologies,
enced primarily by substrate stiffness; however, Glostrup, Denmark]; antigen retrieval using Dako
HAM stiffness did appear to affect cell ­migration low pH  6 Dako target retrieval solution).
90 G. S. Figueiredo et al.

Visualisation of staining is with avidin-biotin expanded epithelial cells covering at least 90% of
complex (ABC) (Vector Labs, Peterborough, UK) the HAM surface area was observed microscopi-
with diaminobenzidine brown reaction product. cally, the epithelial cells can be dissociated from
Taking the longest section of tissue, 100 nuclei the HAM for counting and further investigation
are then counted in the centre of the section, and with histological or polymerase chain reaction
at either end (i.e. three fields in total), a number of (PCR) techniques. For this, the limbal explant is
immuno-positive cells are noted and a percentage mechanically removed from the HAM using
of positive cells calculated for each section. An ­sterile non-toothed ophthalmic micro-tying for-
average of the percentage positive cells is then ceps (Duckworth & Kent, UK), and the ex vivo
calculated and the proportion of cells expressing expanded epithelial cells and the underlying
each marker determined. HAM are washed twice with 2 mls DPBS, trans-
ferred to a 15 ml centrifuge tube and incubated
6.3.4.2 Measuring Explant Outgrowths for 10 mins at 37 °C with 4 mls of trypsin 0.05%
Limbal explant outgrowth expansion is marked solution (Gibco, USA). The trypsin is then inacti-
on the underside of the culture well every vated with 8 mls of LSC medium and the tryp-
2–3 days, at every medium exchange (Fig. 6.6). sinised culture centrifuged for 3  min at 1300
At the termination of the culture, the sequential RPM at room temperature. The inactivated tryp-
outgrowth markings are transferred to an acetate sin is then aspirated and the cells resuspended in
sheet, which is then placed over a graph paper to 1 ml of DPBS. Twenty microlitres of the cell sus-
calculate the area covered by the limbal explant pension is transferred to a 1.5  ml cryo tube
outgrowth by counting the number of small together with 180 μl 0.04 % Trypan blue solution
squares and multiplying by 4 to give the total sur- for cell counting.
face area in mm2 covered by expanded cells on The number of viable cells in 10 μl is counted
the HAM. This is then plotted as a graph of the in a dual-chamber haemocytometer (Scientific
total surface area of the outgrowth (in mm2) Laboratory Supplies) under a light microscope.
against time in culture [119]. Cell viability is assessed by the Trypan blue
exclusion test, where a 10 μl sample of the cell
6.3.4.3 Cell Counting and Cell Viability suspension is mixed with 25  μl Trypan blue
In cultures that are not to be used for transplanta- 0.04% (Sigma-Aldrich) and 15 μl DPBS (Gibco).
tion, as soon as a confluent monolayer of ex vivo Cells that absorb Trypan blue solution do not
have intact cell membranes and are therefore
considered non-viable; viable cells have intact
cell membranes and therefore exclude the dye.
Blue, non-viable cells are excluded from the final
cell count.

6.3.4.4 Colony-Forming Efficiency


Assays
In cultures that are not to be used for transplanta-
tion, a colony-forming efficiency (CFE) assay is
a method of determining the ability of limbal epi-
thelial cells to form colonies, where a predeter-
mined number of cells are suspended in culture
over a fixed period of time and the final total
number of colonies derived from those initial
Fig. 6.6  Colour photograph showing limbal explant in
culture on human amniotic membrane construct.
cells is counted.
Outgrowth expansion was marked on the underside of the Cells from confluent limbal explant cultures
culture well (arrow) are separated from the HAM by trypsinisation, as
6  Corneal Epithelial Stem Cells: Methods for Ex Vivo Expansion 91

described above. Five hundred cells are used in Medicinal Products was adopted [120]. The
CFE assays. regulation imposed strict rules on the develop-
First, mitotically inactivated 3T3-J2 mouse ers of ATMPs to ensure the production of ATMPs
fibroblasts are suspended in 3T3 medium and adheres to Good Manufacturing Practice (GMP)
plated in a 9.6 cm2 tissue culture well at a final principles. The World Health Organisation
density of 2.4 × 104 cells per cm2 and placed in (WHO) version of GMP guidelines is widely
a tissue culture incubator at 37  °C overnight to used in the pharmaceutical industry and phar-
allow the establishment of a 3T3-J2 feeder layer. maceutical regulators. This set of practice guide-
The following day, 500 viable limbal epithelial lines encompasses the manufacturing, testing and
cells from each explant culture are plated onto quality assurance of a product, to ensure that it is
the prepared 3T3-J2 feeder layer along with 2 ml safe for human consumption.
of LSC medium. The culture is then placed in Good Manufacturing Practice is thus defined,
an incubator at 37 °C and the medium changed as “a system for ensuring that products are con-
every 2–3  days. The CFE is measured on the sistently produced and controlled according to
12th day of culture by removal of epithelial quality standards. It is designed to minimize the
medium followed by two brief irrigations with risks involved in any pharmaceutical production
DPBS.  The culture is then fixed in 3.7% form- that cannot be eliminated through testing the
aldehyde (VWR International, UK) in DPBS for final product”. GMP covers all aspects of pro-
10 min at room temperature. The formaldehyde duction from the starting materials, premises and
solution is then removed and the culture washed equipment to the training and personal hygiene
with PBS. Colonies are then stained by incuba- of operators. Detailed, written procedures are
tion with 1% Rhodamine B (Sigma-Aldrich, UK) essential for each process that could affect the
in methanol for 10  min at room temperature. quality of the finished product. All materials,
Colonies are then counted under a dissecting production environments and relevant documen-
microscope (SMZ645, Nikon, Japan), and CFE tation must be quality controlled, and there must
is calculated by the formula as follows: (number be systems to provide documented proof that
of colonies/number of cells plated) × 100. correct procedures are consistently followed at
each step in the manufacturing process every
time a product is made.
6.4 The Manufacturing
of a Corneal Epithelial Cell
Product for Autologous 6.4.2 Quality Control
Transplantation
There are three phases of quality control for the
6.4.1 Good Manufacturing Practice autologous limbal stem cell product.

In the past two decades, medicine has moved 1. Phase I


more and more towards the development of (a) Preparation of media: a 2 ml sample from
advanced therapy medicinal products (ATMPs), the basic limbal epithelial culture medium,
with new, experimental products being devel- a 1 ml sample from the autologous serum
oped, based on genes (gene therapy), cells (cell used in medium preparation and a 1  ml
therapy) and tissues (tissue engineering). sample from the complete limbal epithe-
With this drive towards the development of lial culture medium are sent in blood cul-
ATMPs came challenges as to how these prod- ture bottles (BacT/ALERT®PF Plus,
ucts should be regulated to ensure public safety. bioMerieux, France) to a local microbiol-
In view of these challenges, in 2007, Regulation ogy department.
(EC) No. 1394/2007 of the European Parliament (b) Preparation of HAM: the prepared HAM
and of the Council on Advanced Therapy is checked 24 h after plating to ensure it
92 G. S. Figueiredo et al.

remains attached at all edges and also to transplantation into the patient’s diseased ocular
ensure the medium remains clear, without surface with TLSCD:
any sign of contamination.
2 . Phase II 1. ALSC must be shown to be at least 90% con-
(a) Cell culture: at day 0, the day of the lim- fluent by visual assessment.
bal biopsy, the culture medium is removed 2. ALSC must be morphologically sound (an

from the HAM in culture, and 0.5  ml is evenly distributed layer of epithelial pheno-
sent for mycology screening, whilst the type, i.e. small and regular with primitive
remaining medium is sent in a blood cul- cytoplasm and large nuclei) macroscopically.
ture bottle as before to a local Department 3. The donor must have negative results for all
of Microbiology. Finally, a further 1  ml infectious diseases screening from a blood
medium sample is sent in a blood culture sample taken on the day of limbal biopsy.
bottle 72 h prior to the product release. 4. Day 0 medium must be negative for aerobes
(b) A macroscopic and microscopic examina- and anaerobes after 5  days BacT/ALERT
tion of the LSC culture system is per- culture.
formed 24 h prior to the expected release 5. Day 0 medium mycology screen must be

of the product to ensure the product meets negative.
the release criteria and that the medium is 6. A medium sample taken within 72 h of release
clear and free of contamination. must be negative after 48  h BacT/ALERT
(c) A further 1  ml of culture medium is culture.
taken immediately prior to the release
of the product for retrospective quality
analysis by an independent microbiology 6.4.4 A
 utologous Limbal Stem Cell
­laboratory for European Pharmacopoeia Transplantation
standard sterility testing (Ph Eur 2.6.1.).
3. Phase III Whilst the explant culture remains in the incubator
(a) At the time of transplantation in the oper- in the GMP facility after having been released for
ating theatre, the posterior edge excess transplantation, the procedure for ALSC transplan-
product is sent to an external and indepen- tation begins in the ophthalmic operating theatre.
dent histopathology laboratory for histo- This is often performed under general anaesthesia.
logical assessment (haematoxylin and The initial part of the procedure involves perform-
eosin staining and transmission electron ing a complete superficial keratectomy to excise
microscopy), in addition to testing with the superficial conjunctivalised scar tissue (pannus)
immunocytochemistry for the expression covering the patient’s own cornea and going well
of putative LSC markers p63 and ABCG2, beyond the limbus 360°. This excised corneal pan-
cell proliferation marker Ki67 and low nus is analysed histologically for the presence of
expression of CK12 (corneal) or CK13 vascularisation and inflammation, goblet cells,
(conjunctival) differentiation markers. keratinisation and metaplasia and is immunos-
(b) A growth curve demonstrating the culture tained for CK12 and CK13 to confirm that the epi-
growth is recorded for every culture thelial cells present are conjunctival in nature.
process. Once the pannus has been excised, the explant
culture is removed from the incubator in the GMP
facilities, carefully packed and transported by
6.4.3 Release of the Manufactured hand to the ophthalmic operating theatre. There,
Product the product construct, consisting of HAM covered
in the epithelial monolayer, is unpacked and
There are a set number of criteria the product has handed over in sterile condition to the operating
to satisfy before it is deemed ready for release for surgeon and carefully removed from the culture
6  Corneal Epithelial Stem Cells: Methods for Ex Vivo Expansion 93

well and separated from the glass coverslips. It is the UK, this product was subsequently given
then placed onto the eye (epithelial side up), care- approval by the National Institute for Health and
fully lining up the explant biopsy at the 12 o’clock Care Excellence (NICE) in August 2017 [121].
position at the limbus of the patient’s diseased eye.
It is then sutured in place using four separate Compliance with Ethical Requirements Gustavo
Ethilon® nylon 10-0 suture (Ethicon, Scotland, S.  Figueiredo, Hardeep S.  Mudhar, Majlinda Lako and
Francisco C. Figueiredo declare that they have no conflict
UK), and the excess product posteriorly is care- of interest. All procedures followed were in accordance
fully dissected and sent for histological analysis as with the ethical standards of the responsible committee on
a secondary quality control measure to confirm the human experimentation (institutional and national) and
transplantation of non-­conjunctival corneal epithe- with the Helsinki Declaration of 1975, as revised in 2000.
Informed consent was obtained from all patients for being
lium containing LSCs. A second HAM (stroma included in the study. No animal studies were performed
facing down) is placed over the culture HAM/ by the authors for this chapter.
LSCs as a bandage, in effect sandwiching the
expanded epithelial monolayer between two HAM
sheets [70]. The second HAM is then sutured in References
place using a combination of separate 10-0 nylon
1. Secker GA, Daniels JT. Limbal epithelial stem cells
sutures (Ethicon, UK) and a continuous 11-0 nylon of the cornea 2009. In: StemBook [Internet]. http://
suture (Ethicon, UK), and a large bandage contact www.stembook.org/node/588.
lens (22 × 9.0 mm; Bausch & Lomb, UK) is placed 2. Fuchs E, Segre JA. Stem cells: a new lease on life.
on the eye to protect the ALSC transplant and Cell. 2000;100(1):143–55.
3. Dua HS, Shanmuganathan VA, Powell-Richards
HAM for at least 6–8  weeks. This method was AO, Tighe PJ, Joseph A.  Limbal epithelial crypts:
originally described by Kolli et al. [70]. a novel anatomical structure and a putative lim-
All patients receive post-operative frequent bal stem cell niche. Br J Ophthalmol. 2005;89(5):
preservative-free topical steroid eye drops (pred- 529–32.
4. Li W, Hayashida Y, Chen YT, Tseng SC. Niche regu-
nisolone acetate 1%), preservative-free antibiotic lation of corneal epithelial stem cells at the limbus.
eye drops (chloramphenicol 0.5% preservative-­ Cell Res. 2007;17(1):26–36.
free) and preservative-free ASE (or AlloSE) 5. Dua HS, Azuara-Blanco A.  Limbal stem cells
drops. All drops are gradually reduced and subse- of the corneal epithelium. Surv Ophthalmol.
2000;44(5):415–25.
quently stopped within 12  months, with the 6. Thoft RA, Friend J. The X, Y, Z hypothesis of cor-
exception of ASE (or AlloSE) that is kept indefi- neal epithelial maintenance. Investig Ophthalmol
nitely at least four times per day. Vis Sci. 1983;24(10):1442–3.
7. Sharma A, Coles WH.  Kinetics of corneal epi-
thelial maintenance and graft loss. A popula-
tion balance model. Investig Ophthalmol Vis Sci.
6.4.5 European Marketing 1989;30(9):1962–71.
Authorisation 8. Davanger M, Evensen A.  Role of the pericorneal
for the Production of Corneal papillary structure in renewal of corneal epithelium.
Nature. 1971;229(5286):560–1.
Epithelial Stem Cells 9. Chen JJ, Tseng SC.  Abnormal corneal epithe-
lial wound healing in partial-thickness removal of
In 2015, the European Medicines Agency recom- limbal epithelium. Investig Ophthalmol Vis Sci.
mended for Holoclar® (Chiesi Farmaceutici S.p.A., 1991;32(8):2219–33.
10. Cotsarelis G, Cheng SZ, Dong G, Sun TT, Lavker
Parma, Italy) to be given approval for the marketing RM.  Existence of slow-cycling limbal epithelial
authorisation for the production of an autologous basal cells that can be preferentially stimulated to
limbal stem cell product for patients with moderate proliferate: implications on epithelial stem cells.
to severe LSCD due to ocular surface burns, defined Cell. 1989;57(2):201–9.
11. Kenyon KR, Tseng SC. Limbal autograft transplan-
as eyes with corneal neovascularisation in at least tation for ocular surface disorders. Ophthalmology.
two quadrants, with central corneal involvement 1989;96(5):709–22; discussion 22–3.
and severely impaired visual acuity. This is the first 12. Dua HS, Kulkarni B, Singh R. Quest for limbal stem
approved stem cell therapy product in Europe. In cells. Clin Exp Ophthalmol. 2006;34(1):1–2.
94 G. S. Figueiredo et al.

13. Dua HS, Forrester JV.  Clinical patterns of cor- out limbal allografts for corneal surface reconstruc-
neal epithelial wound healing. Am J Ophthalmol. tion in patients with limbal stem cell deficiency.
1987;104(5):481–9. Arch Ophthalmol. 1998;116(4):431–41.
14. Dua HS, Forrester JV.  The corneoscleral limbus 33. Poon AC, Geerling G, Dart JK, Fraenkel GE, Daniels
in human corneal epithelial wound healing. Am J JT. Autologous serum eyedrops for dry eyes and epi-
Ophthalmol. 1990;110(6):646–56. thelial defects: clinical and in vitro toxicity studies.
15. Lavker RM, Dong G, Cheng SZ, Kudoh K, Cotsarelis Br J Ophthalmol. 2001;85(10):1188–97.
G, Sun TT. Relative proliferative rates of limbal and 34. Geerling G, Maclennan S, Hartwig D.  Autologous
corneal epithelia. Implications of corneal epithe- serum eye drops for ocular surface disorders. Br J
lial migration, circadian rhythm, and suprabasally Ophthalmol. 2004;88(11):1467–74.
located DNA-synthesizing keratinocytes. Investig 35. Young AL, Cheng AC, Ng HK, Cheng LL, Leung
Ophthalmol Vis Sci. 1991;32(6):1864–75. GY, Lam DS.  The use of autologous serum tears
16. Hanna C, Bicknell DS, O'Brien JE.  Cell turn- in persistent corneal epithelial defects. Eye.
over in the adult human eye. Arch Ophthalmol. 2004;18(6):609–14.
1961;65(5):695–8. 36. Anderson C, Moretti S, Gieser RG.  The effect
17. Ahmad S.  Concise review: limbal stem cell defi- of tarsorrhaphy on normal healing of corneal
ciency, dysfunction, and distress. Stem Cells Transl epithelial defects in a rabbit model. Cornea.
Med. 2012;1(2):110–5. 1991;10(6):478–82.
18. Sejpal KBP, Deng SX.  Presentation, diagnosis and 37. Robinson C, Tantri A, Shriver E, Oetting
management of limbal stem cell deficiency. Middle T.  Temporary eyelid closure applique. Arch
East Afr J Ophthalmol. 2013;20(1):5–10. Ophthalmol. 2006;124(4):546–9.
19. Secretariat MA. Limbal stem cell transplantation: an 38. Wagoner MD, Steinert RF. Temporary tarsorrhaphy
evidence-based analysis. Ont Health Technol Assess enhances reepithelialization after epikeratoplasty.
Ser. 2008;8(7):1–58. Arch Ophthalmol. 1988;106(1):13–4.
20. Solano J. RCL. Ocular Burns 2013 [updated 25 Jun 39. Cosar CB, Cohen EJ, Rapuano CJ, Maus M,
2013. Available from: http://emedicine.medscape. Penne RP, Flanagan JC, et  al. Tarsorrhaphy: clini-
com/article/798696-overview. cal experience from a cornea practice. Cornea.
21. Morgan SJ.  Chemical burns of the eye: causes 2001;20(8):787–91.
and management. Br J Ophthalmol. 1987;71(11): 40. Pakarinen M, Tervo T, Tarkkanen A.  Tarsorraphy
854–7. in the treatment of persistent corneal lesions. Acta
22. Davis AR, Ali QH, Aclimandos WA, Hunter Ophthalmol Suppl. 1987;182:69–73.
PA.  Topical steroid use in the treatment of ocular 41. Pfister RR.  Clinical measures to promote cor-
alkali burns. Br J Ophthalmol. 1997;81(9):732–4. neal epithelial healing. Acta Ophthalmol Suppl.
23. Macdonald ECA, Cauchi PA, Azuara-Blanco A, Foot 1992(202):73–83.
B. Surveillance of severe chemical corneal injuries 42. Kasaee A, Musavi MR, Tabatabaie SZ, Hashemian
in the UK. Br J Ophthalmol. 2009;93(9):1177–80. MN, Mohebbi S, Khodabandeh A, et al. Evaluation
24. Foster CS, Ba-Abbad R, Letko E, Parrillo of efficacy and safety of botulinum toxin type A
SJ.  Stevens-Johnson syndrome 2013 [updated 12 injection in patients requiring temporary tarsor-
Aug 2013]. Available from: http://emedicine.med- rhaphy to improve corneal epithelial defects. Int J
scape.com/article/1197450-overview. Ophthalmol. 2010;3(3):237–40.
25. Freiman A.  Cicatricial pemphigoid 2013 [updated 43. Ellis MF, Daniell M.  An evaluation of the safety
24 Jan 2013]. Available from: http://emedicine.med- and efficacy of botulinum toxin type A (BOTOX)
scape.com/article/1062534-overview. when used to produce a protective ptosis. Clin Exp
26. Shortt AJ, Tuft SJ, Daniels JT. Corneal stem cells in Ophthalmol. 2001;29(6):394–9.
the eye clinic. Br Med Bull. 2011;100:209–25. 44. Lee C, Kikkawa DO, Pasco NY, Granet
27. Osei-Bempong C, Figueiredo FC, Lako M. The lim- DB.  Advanced functional oculofacial indica-
bal epithelium of the eye--a review of limbal stem tions of botulinum toxin. Int Ophthalmol Clin.
cell biology, disease and treatment. BioEssays. 2005;45(3):77–91.
2013;35(3):211–9. 45. Wuebbolt GE, Drummond G.  Temporary tarsor-
28. Dua HS, Joseph A, Shanmuganathan VA, Jones rhaphy induced with type A botulinum toxin. Can J
RE. Stem cell differentiation and the effects of defi- Ophthalmol. 1991;26(7):383–5.
ciency. Eye. 2003;17(8):877–85. 46. Rosenthal P, Cotter JM, Baum J. Treatment of per-
29. Nugent RB, Lee GA. Ophthalmic use of blood-derived sistent corneal epithelial defect with extended wear
products. Surv Ophthalmol. 2015;60(5):406–34. of a fluid-ventilated gas-permeable scleral contact
30. Chiang CC, Chen WL, Lin JM, Tsai YY. Allogeneic lens. Am J Ophthalmol. 2000;130(1):33–41.
serum eye drops for the treatment of persistent cor- 47. Kanpolat A, Ucakhan OO.  Therapeutic use of
neal epithelial defect. Eye. 2009;23(2):290–3. Focus Night & Day contact lenses. Cornea.
31. Dua HS, Gomes JA, Singh A.  Corneal epithelial 2003;22(8):726–34.
wound healing. Br J Ophthalmol. 1994;78(5):401–8. 48. Lim L, Tan DT, Chan WK.  Therapeutic use of
32. Tseng SC, Prabhasawat P, Barton K, Gray T, Meller Bausch & Lomb PureVision contact lenses. CLAO
D. Amniotic membrane transplantation with or with- J. 2001;27(4):179–85.
6  Corneal Epithelial Stem Cells: Methods for Ex Vivo Expansion 95

49. Arora R, Jain S, Monga S, Narayanan R, Raina UK, 65. Tsubota K, Toda I, Saito H, Shinozaki N, Shimazaki
Mehta DK. Efficacy of continuous wear PureVision J.  Reconstruction of the corneal epithelium by
contact lenses for therapeutic use. Cont Lens limbal allograft transplantation for severe ocular
Anterior Eye. 2004;27(1):39–43. surface disorders. Ophthalmology. 1995;102(10):
50. Blackmore SJ. The use of contact lenses in the treat- 1486–96.
ment of persistent epithelial defects. Cont Lens 66. Tsubota K, Satake Y, Ohyama M, Toda I, Takano Y,
Anterior Eye. 2010;33(5):239–44. Ono M, et al. Surgical reconstruction of the ocular
51. Sangwan VS, Burman S, Tejwani S, Mahesh SP, surface in advanced ocular cicatricial pemphigoid
Murthy R.  Amniotic membrane transplantation: and Stevens-Johnson syndrome. Am J Ophthalmol.
a review of current indications in the management 1996;122(1):38–52.
of ophthalmic disorders. Indian J Ophthalmol. 67. Ilari L, Daya SM. Long-term outcomes of keratolim-
2007;55(4):251–60. bal allograft for the treatment of severe ocular surface
52. Saw VP, Minassian D, Dart JK, Ramsay A, disorders. Ophthalmology. 2002;109(7):1278–84.
Henderson H, Poniatowski S, et al. Amniotic mem- 68. Jenkins C, Tuft S, Liu C, Buckley R. Limbal trans-
brane transplantation for ocular disease: a review plantation in the management of chronic contact-­
of the first 233 cases from the UK user group. Br J lens-­associated epitheliopathy. Eye. 1993;7:629–33.
Ophthalmol. 2007;91(8):1042–7. 69. Pellegrini G, Traverso CE, Franzi AT, Zingirian M,
53. Tejwani S, Kolari RS, Sangwan VS, Rao GN. Role Cancedda R, De Luca M. Long-term restoration of
of amniotic membrane graft for ocular chemical and damaged corneal surfaces with autologous culti-
thermal injuries. Cornea. 2007;26(1):21–6. vated corneal epithelium. Lancet. 1997;349(9057):
54. Kruse FE, Cursiefen C.  Surgery of the cornea: 990–3.
corneal, limbal stem cell and amniotic mem- 70. Kolli S, Ahmad S, Lako M, Figueiredo F. Successful
brane transplantation. Dev Ophthalmol. 2008;41: clinical implementation of corneal epithelial stem
159–70. cell therapy for treatment of unilateral limbal stem
55. Meller D, Pires RT, Mack RJ, Figueiredo F, cell deficiency. Stem Cells. 2010;28(3):597–610.
Heiligenhaus A, Park WC, et  al. Amniotic mem- 71. Sangwan VS, Matalia HP, Vemuganti GK, Fatima
brane transplantation for acute chemical or thermal A, Ifthekar G, Singh S, et  al. Clinical outcome of
burns. Ophthalmology. 2000;107(5):980–9. discus- autologous cultivated limbal epithelium transplanta-
sion 90 tion. Indian J Ophthalmol. 2006;54(1):29–34.
56. Baylis O, Figueiredo F, Henein C, Lako M, Ahmad 72. Pauklin M, Fuchsluger TA, Westekemper H, Steuhl
S. 13 years of cultured limbal epithelial cell ther- KP, Meller D. Midterm results of cultivated autolo-
apy: a review of the outcomes. J Cell Biochem. gous and allogeneic limbal epithelial transplanta-
2011;112(4):993–1002. tion in limbal stem cell deficiency. Dev Ophthalmol.
57. Cauchi PA, Ang GS, Azuara-Blanco A, Burr JM. A 2010;45:57–70.
systematic literature review of surgical interventions 73. Rama P, Matuska S, Paganoni G, Spinelli A, De
for limbal stem cell deficiency in humans. Am J Luca M, Pellegrini G.  Limbal stem-cell therapy
Ophthalmol. 2008;146(2):251–9. and long-term corneal regeneration. N Engl J Med.
58. Copeland RA Jr, Char DH. Limbal autograft recon- 2010;363(2):147–55.
struction after conjunctival squamous cell carci- 74. Maguire MG, Stark WJ, Gottsch JD, Stulting
noma. Am J Ophthalmol. 1990;110(4):412–5. RD, Sugar A, Fink NE, et  al. Risk factors for cor-
59. Coster DJ, Aggarwal RK, Williams KA.  Surgical neal graft failure and rejection in the collabora-
management of ocular surface disorders using con- tive corneal transplantation studies. Collaborative
junctival and stem cell allografts. Br J Ophthalmol. Corneal Transplantation Studies Research Group.
1995;79(11):977–82. Ophthalmology. 1994;101(9):1536–47.
60. Frucht-Pery J, Siganos CS, Solomon A, Scheman 75. Sangwan VS, Matalia HP, Vemuganti GK, Ifthekar G,
L, Brautbar C, Zauberman H.  Limbal cell auto- Fatima A, Singh S, et al. Early results of penetrating
graft transplantation for severe ocular surface keratoplasty after cultivated limbal epithelium trans-
disorders. Graefes Arch Clin Exp Ophthalmol. plantation. Arch Ophthalmol. 2005;123(3):334–40.
1998;236(8):582–7. 76. Basu S, Mohamed A, Chaurasia S, Sejpal K,
61. Holland EJ.  Epithelial transplantation for the man- Vemuganti GK, Sangwan VS. Clinical outcomes of
agement of severe ocular surface disease. Trans Am penetrating keratoplasty after autologous cultivated
Ophthalmol Soc. 1996;94:677–743. limbal epithelial transplantation for ocular surface
62. Morgan S, Murray A. Limbal autotransplantation in burns. Am J Ophthalmol. 2011;152(6):917–24. e1
the acute and chronic phases of severe chemical inju- 77. Organization WH. Recommendations for the evalua-
ries. Eye. 1996;10 (Pt 3:349–54. tion of animal cell cultures as substrates for the man-
63. Tan DT, Ficker LA, Buckley RJ. Limbal transplanta- ufacture of biological medicinal products and for the
tion. Ophthalmology. 1996;103(1):29–36. characterization of cell banks. WHO Press. 2010.
64. Theng JT, Tan DT.  Combined penetrating kera- 78. Group C-bATW.  Donation of starting material for
toplasty and limbal allograft transplantation for cell-based advanced therapies: a SaBTO review.
severe corneal burns. Ophthalmic Surg Lasers. Advisory Committee on the Safety of Blood, Tissues
1997;28(9):765–8. and Organs (SaBTO). 2014.
96 G. S. Figueiredo et al.

79. Kolli S.  Corneal epithelial stem cell biology and 92. Hori J, Wang M, Kamiya K, Takahashi H,
its therapeutic application: PhD Thesis. Newcastle Sakuragawa N.  Immunological characteristics of
University. 2009. amniotic epithelium. Cornea. 2006;25(10 Suppl
80. Baylis O. The safety and efficacy of ex vivo expanded 1):S53–8.
autologous limbal stem cells for the treatment of uni- 93. Akle CA, Adinolfi M, Welsh KI, Leibowitz S,
lateral total limbal stem cell deficiency: PhD Thesis. McColl I. Immunogenicity of human amniotic epi-
Newcastle University. 2014. thelial cells after transplantation into volunteers.
81. Zakaria N, Possemiers T, Dhubhghaill SN, Leysen Lancet. 1981;2(8254):1003–5.
I, Rozema J, Koppen C, et  al. Results of a phase 94. Akle C, McColl I, Dean M, Adinolfi M, Brown
I/II clinical trial: standardized, non-xenogenic, S, Fensom AH, et  al. Transplantation of amniotic
­cultivated limbal stem cell transplantation. J Transl epithelial membranes in patients with mucopoly-
­
Med. 2014;12:58. saccharidoses. Exp Clin Immunogenet. 1985;2(1):
82. Shortt AJ, Secker GA, Munro PM, Khaw PT, Tuft 43–8.
SJ, Daniels JT.  Characterization of the limbal epi- 95. Saghizadeh M, Winkler MA, Kramerov AA,
thelial stem cell niche: novel imaging techniques Hemmati DM, Ghiam CA, Dimitrijevich SD, et al.
permit in  vivo observation and targeted biopsy of A simple alkaline method for decellularizing human
limbal epithelial stem cells. Stem Cells. 2007;25(6): amniotic membrane for cell culture. PLoS One.
1402–9. 2013;8(11):e79632.
83. Koizumi NJ, Inatomi TJ, Sotozono CJ, Fullwood NJ, 96. Menzel-Severing J, Kruse FE, Schlötzer-Schrehardt
Quantock AJ, Kinoshita S.  Growth factor mRNA U.  Stem cell–based therapy for corneal epithe-
and protein in preserved human amniotic membrane. lial reconstruction: present and future. Can J
Curr Eye Res. 2000;20(3):173–7. Ophthalmol. 2013;48(1):13–21.
84. Tseng SCG, Li D-Q, Ma X.  Suppression of trans- 97. Levis H, Daniels JT.  New technologies in lim-
forming growth factor-beta isoforms, TGF-β recep- bal epithelial stem cell transplantation. Curr Opin
tor type II, and myofibroblast differentiation in Biotechnol. 2009;20(5):593–7.
cultured human corneal and limbal fibroblasts 98. Levis HJ, Brown RA, Daniels JT.  Plastic com-
by amniotic membrane matrix. J Cell Physiol. pressed collagen as a biomimetic substrate for
1999;179(3):325–35. human limbal epithelial cell culture. Biomaterials.
85. Solomon A, Rosenblatt M, Monroy D, Ji Z, 2010;31(30):7726–37.
Pflugfelder SC, Tseng SCG.  Suppression of inter- 99. Wilshaw S-P, Kearney JN, Fisher J, Ingham
leukin 1α and interleukin 1β in human limbal epithe- E.  Production of an acellular amniotic membrane
lial cells cultured on the amniotic membrane stromal matrix for use in tissue engineering. Tissue Eng.
matrix. Br J Ophthalmol. 2001;85(4):444–9. 2006;12(8):2117–29.
86. Park WC, Tseng SCG. Modulation of acute inflam- 100. Keane T, Londono R, Turner N, Badylak
mation and keratocyte death by suturing, blood, and S.  Consequences of ineffective decellularization of
amniotic membrane in PRK. Invest Ophthalmol Vis biologic scaffolds on the host response. Biomaterials.
Sci. 2000;41(10):2906–14. 2012;33(6):1771–81.
87. Endo K, Nakamura T, Kawasaki S, Kinoshita 101. Keane T, Swinehart I, Badylak S.  Methods of tis-
S. Human amniotic membrane, like corneal epithe- sue decellularization used for preparation of bio-
lial basement membrane, manifests the alpha5 chain logic scaffolds and in  vivo relevance. Methods.
of type IV collagen. Investig Ophthalmol Vis Sci. 2015;84:25–34.
2004;45(6):1771–4. 102. Koizumi N, Rigby H, Fullwood N, Kawasaki S,
88. Fukuda K, Chikama T, Nakamura M, Nishida Tanioka H, Koizumi K, et al. Comparison of intact
T. Differential distribution of subchains of the base- and denuded amniotic membrane as a substrate
ment membrane components type IV collagen and for cell-suspension culture of human limbal epi-
laminin among the amniotic membrane, cornea, and thelial cells. Graefes Arch Clin Exp Ophthalmol.
conjunctiva. Cornea. 1999;18(1):73–9. 2007;245(1):123–34.
89. Meller D, Pires RTF, Tseng SCG.  Ex vivo preser- 103. Grueterich M, Espana EM, Tseng SCG.  Connexin
vation and expansion of human limbal epithelial 43 expression and proliferation of human limbal epi-
stem cells on amniotic membrane cultures. Br J thelium on intact and denuded amniotic membrane.
Ophthalmol. 2002;86(4):463–71. Investig Ophthalmol Vis Sci. 2002;43(1):63–71.
90. Koizumi N, Fullwood N, Bairaktaris G, Inatomi 104. Shortt AJ, Secker GA, Lomas RJ, Wilshaw SP,
T, Kinoshita S, Quantock A.  Cultivation of cor- Kearney JN, Tuft SJ, et  al. The effect of amniotic
neal epithelial cells on intact and denuded human membrane preparation method on its ability to
amniotic membrane. Investig Ophthalmol Vis Sci. serve as a substrate for the ex-vivo expansion of
2000;41(9):2506–13. limbal epithelial cells. Biomaterials. 2009;30(6):
91. Kubo M, Sonoda Y, Muramatsu R, Usui 1056–65.
M.  Immunogenicity of human amniotic mem- 105. Riau AK, Beuerman RW, Lim LS, Mehta
brane in experimental xenotransplantation. Investig JS.  Preservation, sterilization and de-­
Ophthalmol Vis Sci. 2001;42(7):1539–46. epithelialization of human amniotic membrane for
6  Corneal Epithelial Stem Cells: Methods for Ex Vivo Expansion 97

use in ocular surface reconstruction. Biomaterials. Brillouin spectro-microscopy. Faraday Discuss.


2010;31(2):216–25. 2016;187(0):415–28.
106. Hogg P, Rooney P, Leow-Dyke S, Brown C, Ingham 115. Chen Z, Du T, Tang X, Liu C, Li R, Xu C, et  al.
E, Kearney JN.  Development of a terminally ster- Comparison of the properties of collagen-chitosan
ilised decellularised dermis. Cell Tissue Bank. scaffolds after gamma-ray irradiation and carbodi-
2015;16(3):351–9. imide cross-linking. J Biomater Sci Polym Ed.
107. Dua HS, Maharajan VS, Hopkinson A. Controversies 2016;27(10):937–53.
and limitations of amniotic membrane in ophthal- 116. Mi S, Khutoryanskiy VV, Jones RR, Zhu X, Hamley
mic surgery. In: Reinhard T, Larkin DFP, editors. IW, Connon CJ.  Photochemical cross-linking of
Cornea and external eye disease. Berlin/Heidelberg: plastically compressed collagen gel produces an
Springer, Berlin Heidelberg; 2006. p. 21–33. optimal scaffold for corneal tissue engineering. J
108. Chen B, Jones RR, Mi S, Foster JW, Alcock SG, Biomed Mater Res A. 2011;99((1):1–8.
Hamley IW, et  al. The mechanical properties of 117. Figueiredo GS, Bojic S, Rooney P, Wilshaw SP,
amniotic membrane influence its effect as a bio- Connon CJ, Gouveia RM, et  al. Gamma-irradiated
material for ocular surface repair. Soft Matter. human amniotic membrane decellularised with
2012;8(32):8379–87. sodium dodecyl sulfate is a more efficient substrate
109. Jones RR, Hamley IW, Connon CJ. Ex vivo expan- for the ex vivo expansion of limbal stem cells. Acta
sion of limbal stem cells is affected by substrate Biomater. 2017;61:124–33.
properties. Stem Cell Res. 2012;8(3):403–9. 118. Rooney P, Eagle M, Hogg P, Lomas R, Kearney
110. Guo WH, Frey MT, Burnham NA, Wang J. Sterilisation of skin allograft with gamma irradia-
YL.  Substrate rigidity regulates the forma- tion. Burns. 2008;34(5):664–73.
tion and maintenance of tissues. Biophys J. 119. Baylis O, Rooney P, Figueiredo F, Lako M, Ahmad
2006;90(6):2213–20. S. An investigation of donor and culture parameters
111. Engler AJ, Sen S, Sweeney HL, Discher DE. Matrix which influence epithelial outgrowths from cultured
elasticity directs stem cell lineage specification. human cadaveric limbal explants. J Cell Physiol.
Cell. 2006;126(4):677–89. 2013;228(5):1025–30.
112. Foster JW, Jones RR, Bippes CA, Gouveia RM, 120. Regulation (EC) No 1394/2007 of the European
Connon CJ. Differential nuclear expression of Yap in Parliament and of the Council of 13 November 2007
basal epithelial cells across the cornea and substrates on advanced therapy medicinal products and amend-
of differing stiffness. Exp Eye Res. 2014;127:37–41. ing Directive 2001/83/EC and Regulation (EC)
113. Molladavoodi S, Kwon HJ, Medley J, Gorbet No726/2004., L324 (2007).
M.  Human corneal epithelial cell response to sub- 121. (NICE) NIfHaCE. Holoclar for treating limbal stem
strate stiffness. Acta Biomater. 2015;11:324–32. cell deficiency after eye burns NICE Guidance2017.
114. Lepert G, Gouveia RM, Connon CJ, Paterson Available from: https://www.nice.org.uk/guidance/
C.  Assessing corneal biomechanics with ta467.
Corneal Stromal Stem Cell:
Methods for Ex Vivo Expansion
7
Olena Al-Shymali, Jorge L. Alió del Barrio,
and James L. Funderburgh

7.1 Introduction: The Cornea, The corneal stroma is the centermost tissue


Corneal Stroma, and Stem that makes up approximately 90% of the corneal
Cells volume [1]. The corneal stroma is composed of
keratocytes interspersed in a unique extracellular
The cornea is an important barrier between the eye matrix (ECM) made of multiple lamellae of long
and the outer world; it represents a tough and well- parallel collagen fibrils which spread throughout
organized connective tissue. The transparent cor- the full diameter of the cornea. The uniform size,
nea is a significant refractive element providing organization, and tight packing of the stromal
two-thirds of the eye’s total refractive power. The collagen ensures corneal transparency. The latter
cornea consists of three cellular layers and three relies as well on the collagen fibril spacing regu-
basement membranes. The stratified squamous epi- lated by tissue-specific keratan sulfate and chon-
thelium is the outermost layer of the cornea limited droitin sulfate proteoglycans in the ECM [2–6].
posteriorly by the Bowman’s membrane. The Three percent of the corneal stromal volume is
innermost layer is represented by a single layer of occupied by the keratocytes, neural crest-derived
metabolically active endothelial cells in direct con- mesenchymal cells. The number of dividing kera-
tact with the anterior chamber aqueous and is sepa- tocytes decreases after birth and eventually these
rated by the Descemet’s membrane from the cells withdraw from the cell cycle and become
corneal stroma. Although each layer of the cornea quiescent in adult mammals [7, 8]. However, fol-
plays an indispensable role in its transparency, lowing wounding, keratocytes adjacent to the site
organization, and function, the content of our chap- go through apoptosis, and keratocytes peripheral
ter will be mostly about the corneal stroma. to the wound become mobile, mitotically active
fibroblasts [9–11] that repair the damage by
secreting scar tissue [12]. The fibrotic ECM
found in stromal scars cause incident light scat-
O. Al-Shymali tering, thus resulting in visual impairment.
Vissum Corporation, Alicante, Spain
Millions of people worldwide suffer blindness or
J. L. Alió del Barrio deterioration of their visual function due to cor-
University Miguel Hernandez, Vissum-Instituto
neal scarring [13–15].
Oftalmologico de Alicante, Alicante, Spain
Nowadays, keratoplasty in its different options
J. L. Funderburgh (*)
is the most commonly applied treatment for cor-
University of Pittsburgh, Department of
Ophthalmology, Pittsburgh, PA, USA neal damage and scarring, especially in the devel-
e-mail: jlfunder@pitt.edu oping world [16]. However, its limitations such

© Springer Nature Switzerland AG 2019 99


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_7
100 O. Al-Shymali et al.

as immune response reactions, donor tissue avail- adult stem cell marker ABCG2 [21, 22]. The lat-
ability, and central size-restricted transplants ter is an ATP-binding cassette transporter G fam-
open the way for the development of new proce- ily member and represents a specific marker for
dures such as the use of stem cells  which different kinds of stem cells such as neural,
have gained a lot of attention in the recent years. hematopoietic, cardiac, muscle, mesenchymal,
islet, and keratinocyte stem cells [23]. This trans-
porter protein is able to efflux the fluorescent dye
7.2 Characteristics of Corneal Hoechst 33342; thus adult stem cells expressing
Stromal Stem Cells (CSSC) ABCG2 are isolated as a “side population” (SP)
by fluorescence-activated cell sorting (FACS)
7.2.1 Progenitor Potential [24]. FACS works by separating a population of
of Stromal Cells cells into subpopulations depending on certain
fluorescence characteristics and patterns. In order
Adult keratocytes of the corneal stroma typically to have a sufficient amount of human stromal
adopt a fibroblastic morphology and produce a cells for FACS analysis, the primary cultures
scar-like ECM during their expansion in  vitro were expanded in a stem cell growth medium
[17]. However, it was shown recently that early-­ (SCGM), in which stem cells can proliferate
passage stromal cells mimic a nonfibrotic repair without losing their differential potential [23].
phenotype and have the potential to re-express the Eventually FACS allows the separation of an
characteristics of a differentiated keratocyte [18]. ABCG2-expressing cell population, corneal stro-
Funderburgh et al. [19] showed in an adult bovine mal stem cells (CSSC), from a heterogenous
model that corneal stromal cells have different population of cells.
capacities of transferring a keratocyte phenotype
after extensive proliferation and only 3% of them
were detected to grow clonally. This population of 7.2.3 Embryonic Origin of CSSC
cells shared some genes typical to mesenchymal
stem cells but not to keratocytes. Nevertheless, Although corneal epithelium is a derivative of
upon culturing in a low-mitogen media, these embryonic ectoderm, the endothelium and cor-
clonal cells developed dendritic morphology and neal stroma are derived from the neural crest.
expressed molecular markers of differentiated The expression of genes typical for descendants
keratocytes such as keratocan, aldehyde dehydro- of the neural ectoderm such as PAX6, Six2,
genase 3A1 (ALDH3 A1), and keratan sulfate. In Six3, and Notch1 by CSSC suggests a neural
addition to the elevated expression of mRNA for crest lineage for these cells. However, some
several gene characteristics of stem cells, the pro- authors proposed a bone marrow origin for all
genitor cells expressed elevated mRNA for genes, adult MSCs as bone marrow-derived cells have
expressed during ocular development including been detected in the corneal stroma [25]. Other
PAX6, Six2, Six3, and Notch1 [20]. The stable authors suggest a perivascular origin (pericytes)
progenitor phenotype of these stem-like cells has for MSC and other adult progenitor cells [26].
been proven by their sustention of keratocyte dif- This hypothesis is difficult to test in human
ferentiation potential after their expansion through CSSC because of the stem cell phenotypic plas-
more than 50 cumulative population doublings ticity. However, in a mouse model, Yoshida et al.
during consecutive subculture. [27] proved a neural crest origin for the multipo-
tent progenitor cell population in the stromal
limbus. These cells, analogous to human CSSC,
7.2.2 Primary Isolation of CSSC are CD45(−) supposing that they do not have a
in Humans bone marrow origin; however, they express
embryonic neural crest markers Twist, Slug,
The isolation of adult stem cells from the corneal Snail, and Sox9 [27]. As the CSSC and mouse
stroma was done using the well-documented stromal progenitor cells have similar character,
7  Corneal Stromal Stem Cell: Methods for Ex Vivo Expansion 101

we can insinuate that these stromal stem cells are Notch1, in addition to genes present in early cor-
not of perivascular or bone marrow origin, but neal development, PAX6 and Six2. CSSC
they are derivatives of ocular neural crest. also expressed genes associated with pluripotent
cells, SSEA4, SOX2, REX1, NANOG, KLF4,
and OCT4, and genes associated with neural
7.2.4 Mesenchymal Stem development such as NESTIN, NGFR, and
Cell Phenotype of CSSC CDH2 [20]. Yet, upon differentiation of CSSC,
they expressed high levels of keratocan,
The term mesenchymal stem cells (MSCs) was ALDH3A1, PTDGS, PDK4, and CXADR, all
introduced by Caplan in 1991 as a collective term genes highly expressed in keratocytes [35].
for nonhematopoietic stem cell populations [28].
Afterward, the Mesenchymal and Tissue Stem
Cell Committee of the International Society for 7.2.7 CSSC Differentiation
Cellular Therapy (ISCT) suggested minimal cri-
teria to define human MSC [29]. First, when 7.2.7.1 Differentiation into Non-ocular
maintained in standard culture conditions, MSC Cells
must be plastic-adherent. The second criterion is Human CSSC as with most MSC populations
the trilineage differentiation potential in  vitro. demonstrate a multipotent differentiation capac-
Lastly, MSC expressing stem cell markers ity. Upon culturing these cells in chondrogenic
CD105, CD90, and CD73, should lack the media, mRNAs for collagen II, aggrecan, and
expression of CD14, CD19, CD34, CD45, cartilage oligomeric protein (COMP), all
CD11b, CD79alpha, and HLA-DR surface mol- cartilage-­specific ECM molecules were detected.
ecules. MSC characteristics of cells isolated from Under these chondrogenic conditions, a signifi-
the human limbal stroma were reported by some cant amount of ECM was deposited and stained
authors [30–32]. Branch et al. [33] identified cor- strongly with toluidine blue, a characteristic
neal stromal cells according to the ISCT criteria resulting from proteoglycan accumulation typi-
as a source of MSCs. During their clonal expan- cal of cartilage. In a like manner, when incubated
sion, most of the CSSC expressed MSC markers in a medium reported to induce neurogenesis,
such as CD73, CD90, and CD166 and did not CSSC showed mRNA upregulation of glial fibril-
express the hematopoietic progenitor cell antigen lary acidic protein (GFAP) and neurofilament
CD34 [20]. protein [23]. The ability of CSSC to differentiate
to osteoblast, chondroblast, and adipocyte lin-
eages was demonstrated by conventional staining
7.2.5 CSSC Properties [33]. When CSSC were stimulated for adipogen-
esis, the transcriptions of the adipocyte markers
Like any MSC, CSSC exhibit clonal growth, self-­ PPARG, PLIN, and FASN were all significantly
renewal properties, and a potential for differenti- upregulated. Osteogenic stimulation of CSSC
ation into multiple distinct tissue types [23]. upregulated significantly the following markers
BMP-4 and BMP-6 and insignificantly
OPG. Similarly, upon chondrogenic stimulation,
7.2.6 CSSC Gene Expression CSSC demonstrated a significant upregulation of
COMP and ACAN.
Du et al. [34] used gene array technology to dis-
cover genes useful in distinguishing keratocytes 7.2.7.2 Differentiation into Keratocytes
and CSSC.  They identified a panel of genes
highly expressed in CSSC that are weakly In Vitro
expressed in keratocytes and vice versa. The Keratocytes when activated in acute wound heal-
CSSC included genes specific to MSC such as ing transfer to fibroblasts and secrete contractile
ABCG2, CXCR4, BMi1, CD166, cKIT, and ECM components such as α-smooth muscle actin
102 O. Al-Shymali et al.

[36]. However, CSSC maintain the ability 7.2.8 C


 orneal Limbal Stem Cell
to acquire a keratocyte phenotype. After incuba- Niche
tion in a keratocyte differentiation medium
(KDM) that is a serum-free medium containing 7.2.8.1 Localization of CSSC
fibroblast growth factor 2 (FGF2), CSSC upregu- Based on the immunostaining for ABCG2 and
lated expression of mRNA for keratocan, a pro- PAX6 proteins, CSSC were observed in the lim-
teoglycan core protein that in adults is uniquely bus, a transitional zone between the cornea and
expressed in the corneal stroma. The same applies sclera [23]. These stained cells were found subja-
for keratan sulfate, a glycosaminoglycan unique cent to the epithelial basement membrane in the
in abundance in the corneal stroma, and anterior stroma, in regions where the basement
ALDH3A1, a protein present in elevated amounts membrane has ripples and folds (Fig. 7.1). These
in keratocytes in vivo [23]. anatomical features, named the palisades of Vogt,
CSSC can be cultured in  different media form a niche for limbal epithelial stem cells
resulting in differentiation into keratocytes with (LESC), a population of slow-cycling stem cells
some variations in the produced ECM character- [39]. The limbal stroma beside the palisades has
istics depending on the cultures used (see Sect. particular characteristics such as a blood supply
7.3.2). and melanocytes (Fig. 7.2). LESCs maintain the
self-renewing corneal epithelium. LESC defi-
In Vivo ciency drives in-growth of conjunctival cells
One study demonstrated the ability of CSSC to across the cornea leading to its opacity and visual
adopt keratocyte phenotype in vivo in a murine loss. However, corneal transparency could be
model [37]. After the direct injection of human restored in some individuals by the transplanta-
CSSC into a mouse corneal stroma, human tion of corneal LESC either from collateral eyes
corneal-­specific ECM accumulated, including or from other individuals [40–42].
the proteoglycans keratocan and lumican, and
replaced mouse matrix components. The
injected human stem cells remained viable for
months, apparently having permanently become
quiescent keratocytes. Consequently, these
results suggest that keratocyte represents the
default lineage for the CSSC. The consequence
of this is that CSSC keep their stem/progenitor
character in vivo due to their limbal microenvi-
ronment; however when they enter the stroma,
the CSSC automatically differentiate into
keratocytes.
In a recent study, scaffold-free tissue engi-
neering methods were used, in which human
CSSC were cultured on substrates with aligned
microgrooves, which directed parallel cell align-
ment and matrix organization, similar to a native
corneal stromal lamella [38]. These cells pro-
duced cellular and collagenous tissue sheets that Fig. 7.1  Localization of the corneal stromal stem cells in
were afterward transplanted into mouse corneal the limbus. PAX6-positive (red) and ABCG2-positive
stromal pockets. Subsequently, these engineered (green) cells were observed in the human anterior stroma.
Arrows show stromal cell staining with both markers near
corneal stromal tissues became transparent, and the folded region of the epithelial basement membrane
the human CSSC persisted expressing human known as palisades of Vogt. The white bar represents
corneal stromal matrix molecules [38]. 50 μm (Image from Du et al. [23], with permission)
7  Corneal Stromal Stem Cell: Methods for Ex Vivo Expansion 103

CSSC in terms of their expression of stem cell


genes, clonal growth, formation of spheres, and
differentiation ability to functional keratocytes.
Consequently, this study showed that in vitro, the
anterior limbal mesenchymal “niche cells” could
not be distinguished from the previously charac-
terized CSSC [51].

Fig. 7.2  Anatomical and cellular features of the stromal 7.2.8.3 Function of CSSC
stem cell niche. The section shows the anterior region of According to several studies, in the limbal niche
the limbus. Epithelium is thickened over regions of undu-
lations in the epithelial basement membrane known as the in vivo, CSSC ensure the ideal support system for
palisades of Vogt. Limbal epithelial stem cells (LESCs) maintaining the active population of LESC [32,
are localized in limbal basal epithelium. Unlike the central 43, 44, 52–57]. Upon the co-isolation of CSSC
cornea, limbal stroma is vascularized, containing melano- and LESC by collagenase digestion, LESC showed
cytes (black) and mesenchymal keratocytes (blue).
Corneal stromal stem cells (green) are located subjacent improved expansion and clonogenicity [43] as
to the basement membrane near LESCs (Image from well as the formation of more holoclones in the
Pinnamaneni and Funderburgh [35] with permission) presence of CSSC [44]. The disruption of the
reunion of these co-isolated cells led to reduced
sphere and holoclone formation by LESC [58].
7.2.8.2 LSSC-CSSC Interactions Furthermore, LESC showed better expansion
Several studies have confirmed the presence of when cultured with limbal CSSC than with mesen-
limbal mesenchymal cells with stem cell proper- chymal cells from the central stroma [44]. These
ties that are closely associated in vivo with limbal studies and the anatomical closeness of these two
epithelial cells [32, 35, 43–48]. These  two cell cell populations lend support to the idea that the
types were co-isolated in collagenase digests of principal role of CSSC in vivo is the homeostatic
tissue from the limbal region [32, 44]. Both cell maintenance of the LESC in the limbal niche.
types express stem cell genes but exhibit different
protein phenotypes, both in vivo and ex vivo [32,
44]. It is hypothesized that MSCs in this zone 7.2.9 Immune Reactions of CSSC
(termed “niche cells”) help preserve the stem
phenotype of the LESC [30]. Different studies The immune-modulatory properties of MSC and
demonstrated that LESC and limbal MSC express their therapeutic effectiveness in immune-­
N-cadherin, suggesting this cell-cell junction incompatible individuals were demonstrated in
protein provides interaction between these cell some studies and clinical trials [59–61]. In addi-
populations [43, 45, 49, 50]. This interaction was tion to exhibiting characteristic of adult MSC,
documented by Dziasko et  al. in a three-­ CSSC can modify the cellular immune response.
dimensional reconstruction of the LESC niche Mouse corneal stroma in vivo was injected with
using serial block-face scanning electron micros- human CSSC on one hand and with human cor-
copy (SBFSE) [48]. These images showed long neal fibroblasts on the other. One month after
processes of stromal cells extending through injection, the corneas injected with CSSC were
focal basement membrane interruptions and transparent, while those injected with fibroblasts
forming contact with basal epithelial cells. had a slight haze [37]. CSSC in the mouse cornea
The identity of the LESC-associated anterior did not experience a xenogeneic T-cell-mediated
stromal cells as CSSC was confirmed by Basu immune response rejection; however, the injected
et al., as they isolated the epithelial-­mesenchymal fibroblasts produced a notable increase in CD45+
cell aggregates from mock biopsies with collage- cells after 1 week. Similarly, CD3+ T cells were
nase treatment [51]. Afterward, they compared detected in fibroblast-injected corneas, but not in
the mesenchymal cells from these isolates to CSSC-injected corneas, 2  weeks after injection
104 O. Al-Shymali et al.

[37]. Cytometric analysis of immune cells in the serum (FBS) or human serum (HS). CSSC in HS
cornea showed a fast influx of neutrophils into had significantly larger colonies compared to
the stroma of the mouse after both injections. Yet, clones in FBS [51]. CSSC cultured as a pellet in
it was transient in the cornea injected with CSSC a serum-free medium containing FGF2 differen-
and the tissue cleared from the cells within tiate to keratocytes exhibiting a characteristic
1  week. Still, in mouse cornea injected with gene expression pattern and accumulation of
human fibroblasts, the population of inflamma- ECM with tracts of aligned collagen fibrils, like
tory cells increased significantly [37]. Moreover, those seen in the stroma in vivo [34]. When cul-
limbal MSC cultured from both human and rab- tured on a substrate of aligned poly(ester-­
bit corneas showed the ability to suppress T-cell urethane-­ urea) (PEUU) nanofibers, CSSC
proliferation in  vitro [55]. These studies and produced layer of highly parallel collagen fibrils
results support an immunomodulatory function with uniform diameter and regular interfibrillar
for CSSC and the potential use of allogeneic spacing similar to that of native stroma [65].
CSSC in different cell-based or tissue-engineered CSSC incubation in PEUU in the presence of
therapeutic applications. both fibroblast growth factor 2 (FGF2) and trans-
forming growth factor-beta 3 (TGF-β3) resulted
in a more abundant multilayered lamellae with
7.3 CSSC Isolation and Culturing orthogonally oriented collagen fibrils, in a pattern
mimicking the human corneal stromal tissue [66,
7.3.1 CSSC Isolation from Limbal 67]. Moreover, CSSC showed to secrete highly
Biopsy organized stroma-like ECM by responding to a
specific pattern of topographical cues when cul-
Nowadays, limbal stem cells can be obtained tured in a silk or polycarbonate substrata with
from limbal biopsies of living human patients shallow parallel grooves [68–71].
[40, 62–64]. After obtaining the tissue samples, In a late study, authors compared the properties
they can be manipulated in two ways in order to of cultured CSSC in seven different media formu-
obtain the MSC (CSSC) [51]. The first consists of lations: two FBS-containing media, DMEM with
an initial removal of epithelial cells using the 10% FBS and M199 with 20% FBS; stem cell
neutral protease Dispase, and then stromal cell medium (SCM) with basic fibroblast growth fac-
dispersion is achieved by collagenase digestion. tor (b-FGF) and leukemia inhibitory factor (LIF);
The second approach consists of cell isolation a medium developed for the growth of endothelial
using only collagenase in which cells of both epi- cells (EGM); a serum-free medium (K-SFM); and
thelial and mesenchymal morphologies are pro- two media designed for the culture of hematopoi-
duced in the primary cultures. The epithelial cells etic cells, one semisolid (MethoCult) and one liq-
are lost after expansion, and cultures of homoge- uid (StemPro-34) [72]. The SCM medium was
nous mesenchymal cells are obtained. In fact, considered the most potential for cell therapy
limbal stem cells proliferated much faster while applications. Yet, as it was not completely free of
using collagenase-only isolation than the dispase/ animal-derived products (xeno-free), it was still
collagenase isolation [51]. not perfect to produce CSSC for clinical use [72].
However, in a more recent study, Matthyssen
et al. used DMEM as a basal medium for human-
7.3.2 CSSC Culturing isolated CSSC cultivation and recruited it with
four supplements: FBS, XerumFree (XF), human
After their isolation, CSSC can be expanded AB (HAB) serum, and human platelet lysate
without loss of differentiation potential in a low-­ (HPL) [73]. The supplements were added at three
glucose culture medium containing ascorbate, different concentrations, thus giving rise to 13 dif-
EGF, PDGF, insulin, and dexamethasone in addi- ferent culture conditions. The rate of expansion of
tion to a low level of serum, either fetal bovine CSSC was demonstrated to be significantly higher
7  Corneal Stromal Stem Cell: Methods for Ex Vivo Expansion 105

with 10% HPL medium compared to the others, borders. Subjacent to the basal cells,  dendritic
as it generated the best growth kinetics for CSSC cells with bright cell bodies and processes were
proliferation, in the same time preserving their noted. At a mean depth of 50.2 ± 8.7 μm in the
viability, cellular phenotype, and differentiation anterior limbal stroma clusters of hyperreflective
potential. The xeno-free HPL supplement sur- structures were revealed subjacent to the basal
passed HAB, XerumFree, and FBS for human epithelium [74]. Up to a mean depth of 98 ± 12.8,
CSSC expansion [73]. these clusters continued in an individual manner
to form eventually a continuous nonhomoge-
neous structure with different brightness intensi-
7.4 Confocal Microscopy ties (Fig.  7.3a). Deeper, they appeared to
of Limbal Niche surround blood vessels. Continuous hyperreflec-
tive linear structures were found in the interpali-
Recently, Mathews et  al. characterized the sade regions, extending from the clusters in the
microarchitecture of the limbal niche in healthy anterior stroma. No such hyperreflective struc-
individuals using in vivo confocal microscopy tures were observed upon the examination of the
(IVCM) [74]. The limbal superficial epithelial central stroma using IVCM [74].
cells appeared as polygonal cells with a bright On examination of tangential sections of
nuclei and cytoplasm and a perinuclear dark cadaver limbus stained with hematoxylin-eosin,
halo. Wing cells in the suprabasal layer, in con- highly compact clusters and a large number of
trast, displayed highly reflective well-demar- individual cells were seen in the anterior stroma,
cated cell borders and dark intracellular extending into the interpalisade region [74].
regions.  Basal  cells in the palisades of Vogt Upon immunostaining for MSC-specific ­markers,
appeared hyperreflective with indistinguishable CD90- and CD105-positive cells were observed

a b

Fig. 7.3  Comparison of (a) in vivo confocal microscopic logical features of these HR structures were similar to the
image of superior limbus in a healthy subject with (b) clusters of CD90- and CD105-positive mesenchymal stem
confocal microscopic image of immunostained [CD90-­ cells observed in donor limbal sections. LE limbal epithe-
FITC (green), CD105-Alexa 633 (red), propidium iodide lium, St stroma. IVCM image is 400 × 400 μm (Courtesy
(blue)] tangential cryosection of limbus from a deceased of Saumi Mathews, Jaya Devi Chidambaram, Shruti
donor. Unique hyperreflective (HR) structures were Lanjewar, Jeena Mascarenhas, Namperumalsamy
observed in the anterior limbal stroma subjacent to the Venkatesh Prajna, Veerappan Muthukkaruppan, Gowri
basal epithelial cells by IVCM. The location and morpho- Priya Chidambaranathan)
106 O. Al-Shymali et al.

subjacent to the limbal basal epithelium and 2. Funderburgh J. Keratan sulfate: structure, biosynthe-
sis, and function. Glycobiology. 2000;10(10):951–8.
between the interpalisade ridges during the 3. Hassell J, Birk D.  The molecular basis of corneal
microscopic analysis of the tangential sections transparency. Exp Eye Res. 2010;91(3):326–35.
[74] (Fig. 7.3b). Depending on the similarity in 4. Kao W, Liu C. Roles of lumican and keratocan on cor-
the location, organization, and distribution of neal transparency. Glycoconj J. 2002;19(4–5):275–85.
5. Lewis P, Pinali C, Young R, Meek K, Quantock
both MSC clusters and the hyperreflective clus- A, Knupp C.  Structural interactions between col-
ters, it may be concluded that CSSC form these lagen and proteoglycans are elucidated by three-­
hyperreflective clusters in the anterior limbal dimensional electron tomography of bovine cornea.
stroma (Fig.  7.3). In patients with limbal stem Structure. 2010;18(2):239–45.
6. Parfitt G, Pinali C, Young R, Quantock A, Knupp
cell deficiency, during IVCM analysis of their C.  Three-dimensional reconstruction of collagen-­
limbal stroma beyond the basal epithelium, the proteoglycan interactions in the mouse corneal
hyperreflective niche structures were substituted stroma by electron tomography. J Struct Biol.
by bright fibrotic structures [75]. 2010;170(2):392–7.
7. Zieske J. Corneal development associated with eyelid
opening. Int J Dev Biol. 2004;48(8–9):903–11.
8. Jester J, Lee Y, Huang J, Houston J, Adams B,
7.5 Conclusions Cavanagh H, et al. Postnatal corneal transparency, ker-
atocyte cell cycle exit and expression of ALDH1A1.
Investig Ophthalmol Vis Sci. 2007;48(9):4061–9.
Corneal stromal stem cells are a recently discov- 9. Wilson S. Analysis of the keratocyte apoptosis, kera-
ered source for autologous stem cells. They are tocyte proliferation, and myofibroblast transformation
mesenchymal cells of neural crest origin. They are responses after photorefractive keratectomy and laser
located in the limbal niche where they insure the in situ keratomileusis. Trans Am Ophthalmol Soc.
2002;100:411–33.
homeostatic maintenance of the limbal epithelial 10. Mohan R, Hutcheon A, Choi R, Hong J, Lee J, Mohan
stem cells. CSSC are clonal growing cells with R, et al. Apoptosis, necrosis, proliferation, and myo-
self-renewal properties and a potential for differ- fibroblast generation in the stroma following LASIK
entiation into multiple distinct tissue types. In and PRK. Exp Eye Res. 2003;76(1):71–87.
11. Szentmary N, Nagy Z, Resch M, Szende B, Suveges
vitro, CSSC differentiate into functional kerato- I. Proliferation and apoptosis in the corneal stroma in
cytes that produce extracellular matrix compo- longterm follow-up after photorefractive keratectomy.
nents similar to that of the human cornea. Since Pathol Res Pract. 2005;201(5):399–404.
CSSC display immunosuppressive properties, 12. Funderburgh J, Mann M, Funderburgh M. Keratocyte
phenotype mediates proteoglycan structure: a role
their rejection risk should be minimal. Therefore, for fibroblasts in corneal fibrosis. J Biol Chem.
they have a promising future in different cell-­based 2003;278(46):45629–37.
or tissue-engineered therapeutic applications. 13. Johnson G. Vision 2020: the right to sight: report on
the sixth general assembly of the International Agency
for the Prevention of Blindness (IAPB). Community
Conflict of Interest Olena Al-Shymali: no conflict of
Eye Health. 1999;12(32):59–60.
interest
14. Whitcher J, Srinivasan M, Upadhyay M.  Corneal

Jorge L. Alio del Barrio: no conflict of interest
blindness: a global perspective. Bull World Health
James L. Funderburgh: no conflict of interest
Organ. 2001;79(3):214–21.
15. Pascolini D, Mariotti S, Pokharel G, Pararajasegaram
Informed Consent  No human studies were carried out R, Etya’ale D, Negrel A, et al. 2002 global update of
by the authors for this article. available data on visual impairment: a compilation
of population-based prevalence studies. Ophthalmic
Epidemiol. 2004;11(2):67–115.
Animal Studies  No animal studies were carried out by
16. Dandona L, Ragu K, Janarthanan M, Naduvilath

the authors for this article.
T, Shenoy R, Rao G.  Indications for penetrat-
ing keratoplasty in India. Indian J Ophthalmol.
1997;45(3):163–8.
References 17. Long C, Roth M, Tasheva E, Funderburgh M, Smit R,
Conrad G, et al. Fibroblast growth factor-2 promotes
keratan sulfate proteoglycan expression by kerato-
1. Levin L, Nilsson S, Ver Hoeve J, Wu S, Kaufman P,
cytes in vitro. J Biol Chem. 2000;275(18):13918–23.
Alm A.  Adler’s physiology of the eye: expert con-
18. Ren R, Hutcheon A, Guo X, Saeidi N, Melotti S,
sult - Online and Print. 11th ed. Edingburg: Saunders/
Ruberti J, et  al. Human primary corneal fibroblasts
Elsevier; 2011.
7  Corneal Stromal Stem Cell: Methods for Ex Vivo Expansion 107

synthesize and deposit proteoglycans in long-term 36. West-Mays J, Dwivedi D.  The keratocyte: corneal
3-D cultures. Dev Dyn. 2008;237(10):2705–15. stromal cell with variable repair phenotypes. Int J
19. Funderburgh M, Du Y, Mann M, SundarRaj N,
Biochem Cell Biol. 2006;38(10):1625–31.
Funderburgh J.  PAX6 expression identifies pro- 37. Du Y, Carlson E, Funderburgh M, Birk D, Pearlman
genitor cells for corneal keratocytes. FASEB J. E, Guo N, et  al. Stem cell therapy restores trans-
2005;19(10):1371–3. parency to defective murine corneas. Stem Cells.
20. Funderburgh J, Funderburgh M, Du Y. Stem cells in 2009;27(7):1635–42.
the limbal stroma. Ocul Surf. 2016;14(2):113–20. 38. Syed-Picard F, Du Y, Hertsenberg A, Palchesko R,
21. Golebiewska A, Brons N, Bjerkvig R, Niclou
Funderburgh M, Feinberg A, et  al. Scaffold-free tis-
S.  Critical appraisal of the side population assay in sue engineering of functional corneal stromal tissue. J
stem cell and cancer stem cell research. Cell Stem Tissue Eng Regen Med. 2018;12(1):59–69.
Cell. 2011;8(2):136–47. 39. Shortt A, Secker G, Munro P, Khaw P, Tuft S, Daniels
22. Hertsenberg A, Funderburgh J. Stem cells in the cor- J.  Characterization of the limbal epithelial stem cell
nea. Prog Mol Biol Transl Sci. 2015;134:25–41. niche: novel imaging techniques permit in vivo obser-
23. Du Y, Funderburgh M, Mann M, SundarRaj N,
vation and targeted biopsy of limbal epithelial stem
Funderburgh J. Multipotent stem cells in human cor- cells. Stem Cells. 2007;25(6):1402–9.
neal stroma. Stem Cells. 2005;23(9):1266–75. 40. Sangwan V, Basu S, Vemuganti G, Sejpal K,

24. Goodell M, Brose K, Paradis G, Conner A, Mulligan Subramaniam S, Bandyopadhyay S, et  al. Clinical
R.  Isolation and functional properties of murine outcomes of xeno-free autologous cultivated lim-
hematopoietic stem cells that are replicating in vivo. bal epithelial transplantation: a 10-year study. Br J
J Exp Med. 1996;183(4):1797–806. Ophthalmol. 2011;95(11):1525–9.
25. Sosnova M, Bradl M, Forrester J.  CD34+ cor-
41. Rama P, Matuska S, Paganoni G, Spinelli A, De

neal stromal cells are bone marrow-derived and Luca M, Pellegrini G.  Limbal stem-cell therapy
express hemopoietic stem cell markers. Stem Cells. and long-term corneal regeneration. N Engl J Med.
2005;23(4):507–15. 2010;363(2):147–55.

26. Corselli M, Chen C, Crisan M, Lazzari L, Peault 42. Shortt A, Tuft S, Daniels J. Corneal stem cells in the
B. Perivascular ancestors of adult multipotent stem cells. eye clinic. Br Med Bull. 2011;100:209–25.
Arterioscler Thromb Vasc Biol. 2010;30(6):1104–9. 43. Xie H, Chen S, Li G, Tseng S. Isolation and expan-
27. Yoshida S, Shimmura S, Nagoshi N, Fukuda K,
sion of human limbal stromal niche cells. Investig
Matsuzaki Y, Okano H, et al. Isolation of multipotent Ophthalmol Vis Sci. 2012;53(1):279–86.
neural crest-derived stem cells from the adult mouse 44. Chen S, Hayashida Y, Chen M, Xie H, Tseng S. A new
cornea. Stem Cells. 2006;24(12):2714–22. isolation method of human limbal progenitor cells by
28. Caplan A.  Adult mesenchymal stem cells for tissue maintaining close association with their niche cells.
engineering versus regenerative medicine. J Cell Tissue Eng Part C Methods. 2011;17(5):537–48.
Physiol. 2007;213(2):341–7. 45. Hayashi R, Yamato M, Sugiyama H, Sumide T, Yang
29. Dominici M, Le Blanc K, Mueller I, Slaper-­
J, Okano T, et  al. N-Cadherin is expressed by puta-
Cortenbach I, Marini F, Krause D, et al. Minimal cri- tive stem/progenitor cells and melanocytes in the
teria for defining multipotent mesenchymal stromal human limbal epithelial stem cell niche. Stem Cells.
cells. The International Society for Cellular Therapy 2007;25(2):289–96.
position statement. Cytotherapy. 2006;8(4):315–7. 46. Higa K, Kato N, Yoshida S, Ogawa Y, Shimazaki J,
30. Polisetty N, Fatima A, Madhira S, Sangwan V,
Tsubota K, et al. Aquaporin 1-positive stromal niche-­
Vemuganti G. Mesenchymal cells from limbal stroma like cells directly interact with N-cadherin-positive
of human eye. Mol Vis. 2008;14:431–42. clusters in the basal limbal epithelium. Stem Cell Res.
31. Choong P, Mok P, Cheong S, Then K. Mesenchymal 2013;10(2):147–55.
stromal cell-like characteristics of corneal kerato- 47. Massie I, Dziasko M, Kureshi A, Levis H, Morgan
cytes. Cytotherapy. 2007;9(3):252–8. L, Neale M, et al. Advanced imaging and tissue engi-
32. Li G, Zhu Y, Xie H, Chen S, Tseng S. Mesenchymal neering of the human limbal epithelial stem cell niche.
stem cells derived from human limbal niche cells. Methods Mol Biol. 2015;1235:179–202.
Investig Ophthalmol Vis Sci. 2012;53(9):5686–97. 48. Dziasko M, Amer H, Levis H, Shortt A, Tuft S,

33. Branch M, Hashmani K, Dhillon P, Jones D, Dua H, Daniels J.  Localisation of epithelial cells capable of
Hopkinson A. Mesenchymal stem cells in the human holoclone formation in  vitro and direct interaction
corneal limbal stroma. Investig Ophthalmol Vis Sci. with stromal cells in the native human limbal crypt.
2012;53(9):5109–16. PLoS One. 2014;9(4):e94283.
34. Du Y, SundarRaj N, Funderburgh M, Harvey S,
49. Higa K, Shimmura S, Miyashita H, Kato N, Ogawa
Birk D, Funderburgh J.  Secretion and organiza- Y, Kawakita T, et  al. N-cadherin in the mainte-
tion of cornea-like tissue in vitro by stem cells from nance of human corneal limbal epithelial pro-
human corneal stroma. Investig Ophthalmol Vis Sci. genitor cells in  vitro. Investig Ophthalmol Vis Sci.
2007;48(11):5038–45. 2009;50(10):4640–5.
35. Pinnamaneni N, Funderburgh J.  Concise review:
50. Omoto M, Miyashita H, Shimmura S, Higa K,

stem cells in the corneal stroma. Stem Cells. Kawakita T, Yoshida S, et  al. The use of human
2012;30(6):1059–63. mesenchymal stem cell-derived feeder cells for the
108 O. Al-Shymali et al.

c­ ultivation of transplantable epithelial sheets. Investig 64. Lal I, Panchal B, Sangwan V.  In-vivo expan-

Ophthalmol Vis Sci. 2009;50(5):2109–15. sion of autologous limbal stem cell using simple
51. Basu S, Hertsenberg A, Funderburgh M, Burrow M, limbal epithelial transplantation for treatment
Mann M, Du Y, et  al. Human limbal biopsy-derived of limbal stem cell deficiency. BMJ Case Rep.
stromal stem cells prevent corneal scarring. Sci Transl 2013;2013:bcr2013009247.
Med. 2014;6(266):266ra172. 65. Wu J, Du Y, Watkins S, Funderburgh J, Wagner

52. Zhang X, Sun H, Yuan X, Zhang L, Zhao S.
W. The engineering of organized human corneal tis-
Utilization of human limbal mesenchymal cells as sue through the spatial guidance of corneal stromal
feeder layers for human limbal stem cells cultured stem cells. Biomaterials. 2012;33(5):1343–52.
on amniotic membrane. J Tissue Eng Regen Med. 66. Wu J, Du Y, Mann M, Yang E, Funderburgh J, Wagner
2010;4(1):38–44. W.  Bioengineering organized, multilamellar human
53. Ainscough S, Linn M, Barnard Z, Schwab IR, Harkin corneal stromal tissue by growth factor supplemen-
D. Effects of fibroblast origin and phenotype on the tation on highly aligned synthetic substrates. Tissue
proliferative potential of limbal epithelial progenitor Eng Part A. 2013;19(17–18):2063–75.
cells. Exp Eye Res. 2011;92(1):10–9. 67. Wu J, Mann M, Funderburgh J, Wagner W.  Corneal
54. Hashmani K, Branch M, Sidney L, Dhillon P, Verma stromal stem cells versus corneal fibroblasts in gener-
M, McIntosh O, et al. Characterization of corneal stro- ating structurally appropriate corneal stromal tissue.
mal stem cells with the potential for epithelial trans- Exp Eye Res. 2014;120:71–81.
differentiation. Stem Cell Res Ther. 2013;4(3):75. 68. Wu J, Rnjak-Kovacina J, Du Y, Funderburgh M,

55. Bray L, Heazlewood C, Munster D, Hutmacher D, Kaplan D, Funderburgh J.  Corneal stromal bio-
Atkinson K, Harkin D. Immunosuppressive properties equivalents secreted on patterned silk substrates.
of mesenchymal stromal cell cultures derived from Biomaterials. 2014;35(12):3744–55.
the limbus of human and rabbit corneas. Cytotherapy. 69. Karamichos D, Funderburgh M, Hutcheon A, Zieske
2014;16(1):64–73. J, Du Y, Wu J, et al. A role for topographic cues in the
56. Katikireddy K, Dana R, Jurkunas U.  Differentiation organization of collagenous matrix by corneal fibro-
potential of limbal fibroblasts and bone marrow mes- blasts and stem cells. PLoS One. 2014;9(1):e86260.
enchymal stem cells to corneal epithelial cells. Stem 70. Ghezzi C, Marelli B, Omenetto F, Funderburgh

Cells. 2014;32(3):717–29. J, Kaplan D. 3D functional corneal stromal tis-
57. Huang M, Wang B, Wan P, Liang X, Wang X, Liu Y, sue equivalent based on corneal stromal stem cells
et  al. Roles of limbal microvascular net and limbal and multi-layered silk film architecture. PLoS One.
stroma in regulating maintenance of limbal epithelial 2017;12(1):e0169504.
stem cells. Cell Tissue Res. 2015;359(2):547–63. 71. Gosselin E, Torregrosa T, Ghezzi C, Mendelsohn A,
58. Xie H, Chen S, Li G, Tseng S. Limbal epithelial stem/ Gomes R, Funderburgh J, et  al. Multi-layered silk
progenitor cells attract stromal niche cells by SDF-1/ film coculture system for human corneal epithelial
CXCR4 signaling to prevent differentiation. Stem and stromal stem cells. J Tissue Eng Regen Med.
Cells. 2011;29(11):1874–85. 2018;12(1):285–95.
59. Knaan-Shanzer S. Concise review: the immune status 72. Sidney L, Branch M, Dua H, Hopkinson A.  Effect
of mesenchymal stem cells and its relevance for thera- of culture medium on propagation and phenotype
peutic application. Stem Cells. 2014;32(3):603–8. of corneal stroma-derived stem cells. Cytotherapy.
60. Atoui R, Chiu R.  Immune responses after mesen- 2015;17(12):1706–22.
chymal stem cell implantation. Methods Mol Biol. 73. Matthyssen S, Ni Dhubhqhaill S, Van Gerwen

2013;1036:107–20. V, Zakaria N.  Xeno-free cultivation of mes-
61. Stagg J, Galipeau J. Mechanisms of immune modula- enchymal stem cells from the corneal stroma.
tion by mesenchymal stromal cells and clinical trans- Investig Ophthalmol Vis Sci. 2017;58(5):
lation. Curr Mol Med. 2013;13(5):856–67. 2659–65.
62. Basu S, Fernandez M, Das S, Gaddipati S, Vemuganti 74. Mathews S, Chidambaram J, Lanjewar S, Mascarenhas
G, Sangwan V.  Clinical outcomes of xeno-free allo- J, Prajna N, Muthukkaruppan V, et al. In vivo confocal
geneic cultivated limbal epithelial transplantation for microscopic analysis of normal human anterior limbal
bilateral limbal stem cell deficiency. Br J Ophthalmol. stroma. Cornea. 2015;34(4):464–70.
2012;96(12):1504–9. 75. Chidambaranathan G, Mathews S, Panigrahi A,

63. Basu S, Ali H, Sangwan V.  Clinical outcomes of
Mascarenhas J, Prajna N, Muthukkaruppan V.  In
repeat autologous cultivated limbal epithelial trans- vivo confocal microscopic analysis of limbal stroma
plantation for ocular surface burns. Am J Ophthalmol. in patients with limbal stem cell deficiency. Cornea.
2012;153(4):643–50. 2015;34(11):1478–86.
Corneal Endothelial Cells: Methods
for Ex Vivo Expansion
8
Stephen Wahlig, Matthew Lovatt,
Gary Swee-­Lim Peh, and Jodhbir S. Mehta

8.1 Introduction rogenic trauma, for example, the fragile dynam-


ics of corneal hydration may be disrupted. This
The human cornea consists of three main cell weakened functional capacity can result in stro-
types: outer multilayered corneal epithelial cells, mal edema and decreased transparency that will
a stromal layer containing keratocytes embedded eventually progress to corneal blindness if left
within a highly ordered array of collagen fibrils, untreated [7, 8]. Corneal endothelial cells
and a one-cell-layered corneal endothelium. (CEnCs) have limited ability to regenerate in vivo
Descemet’s membrane, synthesized by the endo- in response to cell loss and instead utilize cell
thelium, is situated between the stroma and endo- enlargement and migration to maintain func-
thelial layer. The monolayered corneal tional integrity [9]. For this reason many endo-
endothelium is composed of a mosaic of hexago- thelial pathologies must be treated with
nal endothelial cells, which serves a critical func- transplantation, making the cornea the most
tion in modulating fluid and nutrient transport transplanted tissue in the human body [10].
into the corneal stroma [1, 2]. The endothelium Despite corneal transplants having a high initial
also serves as a “pump,” generating an ionic gra- success rate greater than 90% at 1  year, this
dient to promote movement of excess fluid from decreases to 74% and 64% at 5- and 10-year
the stroma to the anterior chamber, thereby regu- intervals, respectively [11]. In addition, the
lating stromal hydration. Endothelial cell density global supply of donor corneal graft material is
decreases with age, at an approximate rate of insufficient to meet the growing demand. From
0.6% per year for adults [3]. Despite this attri- 2011 to 2015, the number of total US corneal
tion, corneal endothelium function is spared until grafts increased 17%, while donation number
a critical cell density is reached, between 500 and only increased 14.6% [12]. The most recent inter-
1000 cells/mm2 [4–6]. When excess endothelial national data suggests there are 70 individuals
cells are lost, from hereditary dystrophies or iat- requiring surgery for every 1 transplant recipient,
with 12.7 million people currently awaiting
transplantation [13]. The need for alternative
S. Wahlig · M. Lovatt · G. S.-L. Peh therapies is readily apparent as donor shortages
Tissue Engineering and Stem Cell Group, The become more pronounced in a rapidly aging
Academia, Singapore Eye Research Institute,
Singapore, Singapore global population.
Cellular and regenerative therapies are an
J. S. Mehta (*)
Singapore National Eye Centre, Department of attractive substitute for cadaveric transplants.
Cornea and External Disease, Singapore, Singapore Although human corneal endothelial cells

© Springer Nature Switzerland AG 2019 109


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_8
110 S. Wahlig et al.

(HCEnCs) were initially thought to lack EMT in HCEnC culture, it is essential to monitor
­proliferative capacity, a report in 1979 first dem- cellular phenotype during expansion. Beyond
onstrated in  vitro endothelial cell proliferation standard microscopic observations of morphol-
using an explant culture methodology [14]. Since ogy, endothelial phenotype is classically deter-
that discovery there have been steady improve- mined initially through the expression of
ments in our ability to culture HCEnCs in vitro function-associated markers such as the Na+/K+
while maintaining an appropriate endothelial ATPase and ZO-1 tight junction protein [28],
phenotype [7–18]. Preliminary analysis estimates although more specific cell markers such as cell
that a pair of donor corneas could yield up to 80 surface peroxiredoxin 6 (sPrdx-6) [29] have been
tissue-­engineered endothelial constructs, gener- recently described and will be discussed in more
ating significant economic advantages over cur- detail later in this chapter.
rent cadaveric transplants [19, 20]. In addition to the deleterious effects of EMT,
Hence, a viable treatment alternative to reverse there are significant known variations in the cul-
the effects of corneal endothelial dysfunction ture yield based on donor characteristics. One
will first require a robust approach for the isola- particular donor factor is age: young donor cells
tion and subsequent expansion of primary generally possess a greater proliferative capacity
HCEnCs. This chapter will review the biology of than their older counterparts [25, 30]. Young
HCEnCs and the potential for in vitro expansion donor cells are also less prone to undergo EMT
of these cells. We will then discuss the develop- during culture [15, 31]. One recent study was
ment of in  vitro culture protocols, including able to successfully culture endothelial cells from
HCEnC isolation techniques, culture conditions, 61% of donors younger than 30 but only 18% of
and characterization of endothelial cells after donors older than 60 [15]. This observation has
expansion. been partly explained by increased replicative
senescence in cells from older donors [32]. Other
causes of interdonor variability include pre-­
8.2  uman Corneal Endothelial
H existing pathology, drug use, cause of death, and
Cells graft storage conditions [33–35]. Due to this vari-
ability, even a highly robust culture system will
8.2.1 N
 ative Human Corneal yield significantly different quantities of HCEnCs
Endothelial Cells based on available donor corneas.

HCEnCs are arrested in G1 phase and do not


undergo active mitosis within the eye, due mainly 8.2.2 Progenitor Cells
to inhibition from cell-cell contacts [21–23].
However, current culture protocols, using a vari- 8.2.2.1 Existence of Endothelial
ety of media and growth factors, are able to induce Progenitor Cells
in vitro proliferation of isolated HCEnCs. The pri- HCEnCs are normally nonproliferative in  vivo,
mary challenge with in vitro endothelial culture is instead compensating for cell loss with migration
a tendency for HCEnCs to develop a fibroblastic and enlargement. While corneal limbal epithelial
morphology, a process termed endothelial-­ progenitor cells have been well-described and
mesenchymal transition (EMT) [24, 25]. These used for regenerative therapies [36, 37], such a
fibroblastic-looking cells lack the regular hexago- stem or progenitor cell population has not been
nal architecture and pump functionality of healthy clearly identified for the corneal endothelium. Yet
endothelium [4]. Multiple biochemical pathways some data suggests the existence of peripherally
have been noted to induce EMT across cell types, located HCEnCs with the ability to proliferate
including Wnt, Notch, and TGFβ; activation of in  vivo. In 1982 a unique cell subpopulation,
Wnt signaling, in particular, is correlated with termed “Schwalbe’s line cells,” was discovered in
EMT in HCEnCs [26, 27]. Given the relevance of the anterior trabecular meshwork (TM) bordering
8  Corneal Endothelial Cells: Methods for Ex Vivo Expansion 111

the corneal periphery [38]. Later work demon- onic stem cells (ESCs) [45, 48, 49], mesenchy-
strated that laser stimulation of the TM caused a mal stem cells (MSCs) [50], and induced
fourfold increase in cellular division, with 60% pluripotent stem cells (iPSCs) [51] appear to pos-
of the dividing cells located in the area of sess the capacity for differentiation into HCEnC-­
Schwalbe’s cells [39]. Following this discovery like cells. These stem cells have been isolated
of anatomically distinct cells in the peripheral and are already being utilized in a variety of
cornea, a series of in  vitro studies offered evi- human clinical trials in the fields of neurology,
dence of possible HCEnC progenitor cells. Stem rheumatology, and cardiology, among others
cell markers telomerase, alkaline phosphatase, [52], unlike the undefined corneal endothelial
and nestin have been observed in the endothelial progenitor cells. They are also significantly eas-
periphery, while BrdU fluorescence assays dem- ier to produce in large quantities compared to pri-
onstrated baseline cell division activity near the mary endothelial cells, avoiding the difficulties
anterior TM [40–42]. From these observations, it of ex  vivo HCEnC culture. The challenge with
has been hypothesized that progenitor cells in the stem cells lies in defining optimal conditions to
corneal periphery could potentially proliferate in direct differentiation of stem cells toward a cor-
response to endothelial damage, producing prog- neal endothelial phenotype. Use of lens epithelial
enies that migrate centrally toward the wounded cell-conditioned medium (LECCM) induces
areas. This is supported by anatomical analyses human umbilical cord blood MSCs to form an
describing unique cell clusters in the extreme HCEnC-like phenotype, migrating toward dam-
periphery with adjacent cells positioned in cen- aged endothelial cells and forming a monolayer
tripetal rows [40]. In addition to this peripheral [50]. However, this MSC-derived monolayer
cell subpopulation, neural crest-like cells have does not resemble the regular hexagonal form of
drawn significant interest recently. Corneal endo- native HCEnCs, highlighting the difficulty in
thelial cells are developmentally derived from inducing a specific phenotype from multipotent
neural crest [43]. One potential progenitor cells. One solution is to generate neural crest
HCEnC subpopulation defined by expression of cells, the embryological progenitor of HCEnCs
neural crest stem cell markers HNK-1 and p75 [43], as an intermediate cell type during the dif-
neurotropin receptor (p75NTR) has been discov- ferentiation process. Transformation of human
ered in the transitional zone between peripheral ESCs into neural crest-like cells can be induced
endothelium and trabecular meshwork [44, 45]. with SMAD inhibition, yielding cells positive for
Morphological analysis of ex vivo HCEnCs has the neural crest marker PAX6 [53]. Using this
identified another p75NTR-positive HCEnC sub- SMAD inhibition technique, other groups have
population, which expresses stem markers LGR5 developed protocols for further differentiating
and PAX2 [46, 47]. However, since each of these these intermediary neural crest-like cells into
studies utilizes a different method for selection of HCEnC-like cells that express endothelial mark-
neural crest-like cells, there may be several dis- ers ZO-1, Na+/K+ ATPase, and collagen 8A1 [48,
tinct p75NTR+ HCEnC subpopulations within 54]. Human ESCs have also been utilized in vivo;
the corneal endothelium. Additionally, while a two-phase method using corneal stromal cell
these cells express stem cell markers and are co-culture and LECCM differentiated human
capable of proliferation in  vitro, their function ESCs into HCEnC-like cells, capable of forming
in  vivo is still unknown. To date, each of these a functional endothelial monolayer in a rabbit
potential progenitor cells have yet to be isolated, model [49]. However, the restricted availability
defined definitively, or clearly characterized. and ethical controversies surrounding human
ESCs may limit future applications of these pro-
8.2.2.2 Pluripotent/Multipotent Stem tocols [55].
Cells Induced pluripotent stem cells (iPSCs) lack
In addition to precursors derived from corneal these ethical concerns and may reduce the likeli-
tissue, other cell sources including human embry- hood of posttransplantation immune rejection
112 S. Wahlig et al.

compared to autologous donor cells. Recently, ferred. Since selective media is rarely used, isola-
human iPSCs have been shown to differentiate tion of the DM/endothelial layer must be
into neural crest cells, which then assumed an performed by an experienced technician with
endothelial phenotype upon ROCK and TGFβ skillful, gentle manipulation to avoid stromal
inhibition [51]. Although promising, this human contamination.
iPSC approach still requires validation in an ani-
mal model. Use of iPSCs may also be hindered
by the risk of tumorigenesis [56, 57] as well as 8.3 Endothelial Cell Culture
potential contamination with undesired cell types
through off-target differentiation [58]. 8.3.1 Growth Media
Optimization of iPSC programming and differen-
tiation to address these issues is necessary before Selecting the appropriate culture conditions for
clinical applications can be achieved. expansion of isolated HCEnCs is a difficult task
as evidenced by the numerous studies attempting
to find an ideal culture medium [15, 16, 24].
8.2.3 HCEnC Isolation Protocols Inducing HCEnC expansion is a fine balance
between maximizing proliferation and maintain-
The first step in establishing an HCEnC culture is ing a functional polygonal monolayer with
isolation of the corneal endothelial cells from appropriate cell-cell connections and pump
donor tissue. One approach is the explant tech- activity. The different stages of in vitro HCEnC
nique. The endothelium and attached Descemet’s growth are illustrated in Fig.  8.3. Comparisons
membrane (DM) are cut into smaller pieces and of various culture media derived from different
placed endothelial side-up in culture dishes, basal media identified OptiMEM-I and Ham’s
wherein cells gradually migrate from the explant F12/M199-­ based formulations as superior for
to the culture surface [14, 59]. Another method facilitation of HCEnC proliferation [24]. A criti-
utilizes a scalpel or other surgical instruments to cal component of these culture media is addi-
directly scrape cells off the DM [60, 61]. HCEnCs tives: the OptiMEM-I based media is
can also be isolated by a peel-and-digest method. supplemented by nerve growth factor (NGF),
The DM and adhered endothelium are removed epidermal growth factor (EGF), chondroitin sul-
from the cornea (Fig.  8.1) and then digested fate, and bovine pituitary extract (BPE), while
either nonenzymatically in EDTA or enzymati- the Ham’s F12/M199 media is supplemented by
cally in collagenase, dispase, or trypsin to isolate basic fibroblast growth factor (bFGF), insulin,
the HCEnCs (Fig.  8.2) [7]. It should be noted transferrin, and selenite [25, 65, 66]; both media
here that the DM/endothelium peeling process are supplemented with ascorbic acid and serum.
must be approached with care as improper tech- Analysis of these individual growth factors does
nique can result in detachment of stromal fibers, not always demonstrate a significant benefit;
which may in turn release keratocytes into cul- NGF and EGF appear to have minimal effect on
ture during enzymatic digestion. Since kerato- endothelial proliferation [25, 65, 67]. BPE mod-
cytes have a higher proliferative index than erately stimulates HCEnC growth, but this affect
HCEnCs, they will overwhelm the culture with is blunted in older donor corneas [25, 67].
fibroblastic cells. A vacuum suction holder can Additionally, pituitary extract is composed of
be used to assist during the DM peeling, improv- unknown quantities of individual factors, which
ing successful culture yield [62, 63]. Use of is undesirable when attempting to create a well-
L-valine-free media to selectively inhibit kerato- defined, robust culture platform. Although bFGF
cyte growth has also been described [64]. does augment endothelial cell proliferation [68,
However, given the difficulty in eliminating 69], it also functions as a differentiation factor
L-valine from the complex media needed for that contributes to EMT through activation of
HCEnC culture, this method is not generally pre- Wnt signaling [4, 27].
8  Corneal Endothelial Cells: Methods for Ex Vivo Expansion 113

a b

c d

e f

Fig. 8.1 Isolation of the Descemet’s membrane and cells can be identified by blue staining (white arrow-
endothelium. (a) Cadaveric corneoscleral rim is placed in heads). (d) DM is carefully peeled off the stroma and
a suction holder for the peeling process. (b) Trypan blue is spontaneously scrolls. (e) Trypan blue stains exposed
used to identify the margins of the endothelium; the blue stroma, illustrating the circle of removed DM. (f) The
ring at the periphery is transition zone/extreme peripheral removed DM/endothelium is collected in Liberase for
endothelium that is not removed. (c) Dead endothelial digestion

Among the nongrowth factor additives, chon- [59, 70]. Ascorbic acid, specifically L-ascorbic
droitin sulfate is a proteoglycan with known pro-­ acid 2-phosphate (Asc-2P), both inhibits EMT
proliferative effects on vascular endothelial cells and promotes HCEnC proliferation [69, 71]. The
that also enhances HCEnC growth and adhesion benefits of Asc-2P are thought to be due to
114 S. Wahlig et al.

a b

c d

Fig. 8.2  Digestion of the Descemet’s membrane/endo- cells have separated from the DM, as indicated by the near
thelium. (a) Microscopic views of the scrolled DM shows transparent DM. Many of the floating CEnCs can be seen
the monolayer of hexagonal endothelial cells. (b) After clustering together to form larger clumps (arrows). (d)
45  min of Liberase digestion, the endothelial cells are HCEnC clumps are partially broken up with TrypLE and
beginning to separate from the DM (marked by white are subsequently seeded onto an adherent tissue culture
arrows). (c) By the fourth hour of digestion, nearly all plate. Scale bar = 100 μm

p­ rotection from oxidative DNA damage [69] in additives can enhance HCEnC proliferation, this
addition to upregulation of hepatocyte growth benefit is often offset by a loss of endothelial
factor production [71]. Insulin, transferrin, and functionality through EMT [72].
selenite each have a modest proliferative benefit In addition to growth factors, small molecule
and allow for lower concentrations of animal media additives can be used for targeted modula-
serum to be used; insulin in particular acts by tion of biochemical pathways thought to control
synergistically improving the growth-promoting EMT. The most significant development has been
effect of bFGF [66]. However, while these media the use of the Rho-associated kinase (ROCK)
8  Corneal Endothelial Cells: Methods for Ex Vivo Expansion 115

a b

c d

Fig. 8.3  Morphology of cultured primary human endo- P4). (c) Confluent HCEnC culture, demonstrating regular
thelial cells. (a) Elongated HCEnC phenotype during polygonal monolayer (Donor age 3, P3). (d) Evidence of
early proliferation, 3 days after passaging (Donor age 3, replicative senescence, including intracellular vacuoliza-
Passage 3). (b) Subconfluent HCEnC culture, partially tion and irregular morphology (Donor age 13, P3). Scale
assuming a flattened polygonal morphology (Donor age 3, bar = 100 μm

inhibitor Y-27632 to maintain HCEnC morphology An alternative method of balancing endothelial


during expansion. ROCK inhibition has been dem- proliferation and EMT is a dual-media expansion
onstrated to enhance cell adhesion to prevent EMT protocol. This method initially uses a proliferative,
while simultaneously increasing proliferation and growth factor-supplemented media to expand the
suppressing apoptosis; this is thought to be due to HCEnCs. Once 80–90% confluence is reached, a
intracellular cytoskeleton modulation [17, 73]. growth factor-free maintenance media is used to
Another major biochemical target for HCEnC prevent EMT prior to passaging [18]. Even after
expansion is TGFβ, which is known to inhibit cor- the third passage, the HCEnCs continue to form a
neal endothelial proliferation [74, 75] and promote hexagonal/polygonal monolayer and expressed
EMT, likely through Smad2/Smad3 signaling [76, high levels of endothelial markers Na+/K+ ATPase
77]. The specific TGFβ receptor kinase inhibitor pump, ZO-1, GPC4, and CD200 [18, 78]. This
SB431542 suppresses in vitro fibroblastic transfor- dual-media solution can be further optimized with
mation of HCEnCs while maintaining expression addition of a ROCK inhibitor, increasing HCEnC
of ZO-1 and Na+/K+ ATPase [76]. yield by two- to threefold [79].
116 S. Wahlig et al.

Translating these propagated HCEnCs into ment of corneal endothelial cells. A variety of
clinical applications will require development of natural and synthetic substrates have been trialed,
a culture system adherent to good manufacturing including collagen IV, laminin, Matrigel, and
practice (GMP) guidelines. Research-grade FNC Coating Mix, a proprietary reagent contain-
reagents like collagenase and trypsin must be ing fibronectin, collagen, and albumin [25, 79,
exchanged for GMP-compliant alternatives such 85, 86]. All of these substances appear compati-
as Liberase™ and recombinant TrypLE™ Select. ble with HCEnC culture, although direct com-
In addition to substituting individual components parison between collagen IV, Matrigel, laminin,
with GMP-grade equivalents, a continued con- and fibronectin noted that collagen IV is most
cern is the use of reagents that are animal-derived conducive to HCEnC adherence and expansion
and/or of undefined composition, specifically [25]. A similar study demonstrated superior hex-
bovine serum. Use of animal products increases agonal morphology as well as ZO-1 and actin
the risk of microbial contamination as well as expression with a collagen IV substrate [87].
immunogenic complications, a potential obstacle More recently, the specific laminins in DM have
for applications in human transplantation [80]. been identified as laminin-511 and laminin-521
Xeno-free growth conditions have been [88]. These laminin isotypes induced superior
described: one study comparing human serum HCEnC adhesion, proliferation, and maintenance
with bovine serum reported no differences in pro- of endothelial phenotype compared to collagen I
liferation of cultured HCEnCs [81], although and fibronectin [88]. As collagen IV and laminin-
morphological variability has been noted [19]. ­511 and laminin-521 are both major components
Protein concentrates released from activated of the human DM, these observations support the
human platelets can also serve as a bovine serum hypothesis that materials similar to native DM
substitute, producing morphologically normal produce desirable culture results [89]. In addition
HCEnCs albeit with a reduced proliferative to the surface proteins used, adjusting the elastic
capacity [82]. However, these xeno-free alterna- modulus of culture substrates to mimic the
tives have yet to surpass bovine serum in the mechanical properties of Descemet’s membrane
majority of published HCEnC culture protocols. also improves endothelial morphology [87].
Further modifications to eliminate serum entirely Many of these substrates are animal-derived,
would also be desirable, as serum contains com- and as mentioned above, this increases the risk
plex and highly variable mixtures of assorted for infectious and immunogenic complications in
proteins that can differ tremendously from batch clinical applications [80]. Recombinant proteins
to batch [83]. Serum can also introduce microbi- like laminin-511 and laminin-521 avoid this con-
ological contaminants into cell culture [84]. cern, but are prohibitively expense [19]. One
Despite these concerns, serum has proven so crit- potential xeno-free culture substrate is a pericel-
ical to endothelial cell culture that it remains a lular matrix from human decidua-derived mesen-
necessary component of HCEnC culture systems, chymal cells (PCM-DM). This PCM-DM,
although use of GMP-grade serum products like composed of collagen IV and fibronectin among
EquaFetal can minimize the risks described other proteins, has been shown to produce
above [19]. HCEnC monolayers at a density even higher than
collagen IV or fibronectin alone [90]. Although
its derivation from human fetal membranes poses
8.3.2 Culture Surfaces a challenge for upscaling and reproducibility, the
potency and xeno-free nature of PCM-DM is
In addition to culture media, there is a growing attractive for endothelial culture.
interest around the role of culture substrates and A further refinement of the culture substrate
their role in modulating cell growth. Ideally, the has been the use of nano-topographical patterns
culture substrate should mimic Descemet’s mem- in the substrate surface, intended to mimic the
brane in order to replicate the in  vivo environ- biophysical interaction between endothelial cells
8  Corneal Endothelial Cells: Methods for Ex Vivo Expansion 117

and the DM. Standard tissue culture polystyrene (ALCAM), is highly expressed among vascular
(TCPS) patterned with 1 μm or 250 nm structures endothelial cells [101], while GPC4 augments the
has been shown to demonstrate increased cell Wnt5a pathway that contributes to endothelial cell
density, regular morphology, and upregulated migration [102, 103].
ZO-1 and Na+/K+ ATPase expression [91]. A more recent comprehensive transcriptomic
Specifically, the 1  μm pillar pattern produced a analysis highlights the importance of markers,
threefold increase in proliferation, an impressive reporting significant differences in marker levels
improvement given the low-risk nature of pat- between cultured HCEnCs and primary endothe-
terning culture substrates [91]. lial cells [104]. For example, CD200 and GPC4
were mainly expressed in primary HCEnCs, while
SLC4A11 was present at significantly higher lev-
8.3.3 Functional Cell Markers els in ex vivo cultured HCEnCs [104]. The differ-
ence in gene expression between cultured and
Typically, cultured HCEnCs have been identified primary HCEnCs highlights the potential for cul-
through detection of ZO-1 tight junction protein ture protocols to influence endothelial phenotype
and Na+/K+ ATPase expression in addition to direct markers. Numerous variables including culture
morphological visualization [28]. However, these media, donor age, and passage number may con-
markers are primarily expressed while in a conflu- tribute to differential marker expression among
ent monolayer and regress when the monolayer is functional endothelial cells. An exhaustive explo-
disrupted throughout the culture process [92]. ration of the factors contributing to differential
Detection of bona fide endothelial cells is useful marker expression is not realistic, so determina-
both for quantifying the quality of propagated tion of a broader panel of endothelial markers
HCEnCs and potentially for purifying cultures of would be helpful for identifying functional
contaminating fibroblasts as discussed earlier. For HCEnCs across culture protocols.
this reason, recent studies have attempted to gener- Along with discovery of endothelial markers,
ate a panel of defining HCEnC markers [78, 93, there has also been progress in developing spe-
94]. A number of endothelial surface markers have cific antibodies to these markers to isolate and
been reported, including CD200 and GPC4 [78], enrich HCEnCs. Monoclonal antibodies
CD 166 [29, 94, 95], CD98 and CD340 [94], neu- TAG-1A3 and TAG-2A12 have been found to
ral cell adhesion molecule (NCAM) [96], and identify HCEnCs, targeting CD166/ALCAM and
sPrdx-6 [29, 96]. Gene expression analyses have sPrdx-6, respectively [29]. Notably, TAG-2A12
also uncovered a set of genes overexpressed in bound only to HCEnCs (Fig. 8.4) and not to other
HCEnCs, such as SLC4A11, COL8A2, and cells screened, including lung fibroblasts, human
CYYR1 [93] and CLRN1, MRGPRX3, HTR1D, ESCs, and neural crest cells. Unlike TAG-2A12,
GRIP1, and ZP4 [97]. Certain marker sets provide commercially available Prdx-6 antibodies dem-
additional phenotypic information; for example, onstrate cross-reactivity with stromal fibroblasts,
CD9, CD49e, CD44, and CD73 are reportedly highlighting the importance of the targeted epit-
associated with fibroblastic HCEnCs [94], while a ope and specific antibody used when attempting
signature of CD166+, CD105−, and CD44− is to isolate HCEnCs using cell surface markers
thought to mark an endothelial subpopulation that [29]. While these results are encouraging, there is
is particularly resistant to EMT [98]. These “nega- still not a definitive array of HCEnC markers. As
tive” HCEnC markers are consistent with prior mentioned above, the specific functional HCEnC
characterizations, as CD73 is associated with lung markers will likely vary depending on the culture
fibrosis, while CD44 is used to differentiate mature protocol. Standardized markers should be devel-
fibroblasts from iPSCs [99, 100]. This also holds oped for each particular cell therapy protocol to
true for the positive HCEnC markers; CD166, or facilitate HCEnC enrichment and ensure quality
activated leukocyte cell adhesion molecule control.
118 S. Wahlig et al.

a b

Fig. 8.4  Immunofluorescence histology of marker proteins in endothelial cells. Primary cultured HCEnCs stain with
TAG-2A12 antibody (a) and express proteins ZO-1 (b) and Na+/K+ ATPase (c). Scale bar = 50 μm

8.4 Conclusion enticing prospect of endothelial-like tissue


derived from stem cells, avoiding many of the
The cornea is one of the most transplanted tissues difficulties with primary HCEnC culture, is also
in the world, and demand continues to outpace an object of future research. With continuous
the supply of donor tissue. A cellular therapy, improvement in these areas of in  vitro culture,
composed of human corneal endothelial cells this is an especially exciting time for corneal
expanded in vitro, is an intriguing potential solu- endothelial cell therapies.
tion to this problem. While human endothelial
cells are usually non-mitotic, newly developing Conflict of Interest  No conflicting relationship exists for
culture techniques can induce proliferation while any author.
maintaining the native phenotype necessary for
function in vivo. Although there is not yet a stan- Financial Disclosure  No financial disclosures.
dardized protocol for isolation and expansion of
HCEnCs, advances in cell extraction, culture Compliance Statements  J. S. Mehta, M. Lovatt,
media, and specially designed culture substrates G. Peh, and S. Wahlig declare that they have no
have significantly improved culture yields. The conflict of interest. All procedures performed by
8  Corneal Endothelial Cells: Methods for Ex Vivo Expansion 119

the authors were followed and were in accor- 16. Nakahara M, Okumura N, Kay EP, et  al. Corneal
dance with the ethical standards of the responsi- endothelial expansion promoted by human bone
marrow mesenchymal stem cell-derived conditioned
ble committee on human experimentation medium. PLoS One. 2013;8(7):e69009.
(institutional and national) and with the Helsinki 17. Okumura N, Ueno M, Koizumi N, et al. Enhancement
Declaration of 1975, as revised in 2000. Informed on primate corneal endothelial cell survival in vitro
consent was obtained from all patients for being by a ROCK inhibitor. Invest Ophthalmol Vis Sci.
2009;50(8):3680–7.
included in the study. All institutional and 18. Peh GS, Chng Z, Ang HP, et  al. Propagation of
national guidelines for the care and use of labora- human corneal endothelial cells: a novel dual media
tory animals were followed. approach. Cell Transplant. 2015;24(2):287–304.
19. Peh GSL, Ang HP, Lwin CN, et al. Regulatory com-
pliant tissue-engineered human corneal endothelial
grafts restore corneal function of rabbits with bul-
References lous keratopathy. Sci Rep. 2017;7(1):14149.
20. Tan TE, Peh GS, George BL, et  al. A cost-­
1. Bonanno JA.  Molecular mechanisms underly- minimization analysis of tissue-engineered con-
ing the corneal endothelial pump. Exp Eye Res. structs for corneal endothelial transplantation. PLoS
2012;95(1):2–7. One. 2014;9(6):e100563.
2. Maurice DM. The location of the fluid pump in the 21. Joyce NC, Harris DL, Mello DM.  Mechanisms of
cornea. J Physiol. 1972;221(1):43–54. mitotic inhibition in corneal endothelium: contact
3. Bourne WM. Biology of the corneal endothelium in inhibition and TGF-beta2. Invest Ophthalmol Vis
health and disease. Eye (Lond). 2003;17(8):912–8. Sci. 2002;43(7):2152–9.
4. Engelmann K, Bednarz J, Valtink M.  Prospects 22. Joyce NC, Meklir B, Joyce SJ, et  al. Cell cycle
for endothelial transplantation. Exp Eye Res. protein expression and proliferative status in
2004;78(3):573–8. human corneal cells. Invest Ophthalmol Vis Sci.
5. Nuyts RM, Boot N, van Best JA, et  al. Long term 1996;37(4):645–55.
changes in human corneal endothelium follow- 23. Yokoi T, Seko Y, Yokoi T, et  al. Establishment of
ing toxic endothelial cell destruction: a specular functioning human corneal endothelial cell line with
microscopic and fluorophotometric study. Br J high growth potential. PLoS One. 2012;7(1):e29677.
Ophthalmol. 1996;80(1):15–20. 24. Peh GS, Toh KP, Wu FY, et al. Cultivation of human
6. O'Neal MR, Polse KA. Decreased endothelial pump corneal endothelial cells isolated from paired donor
function with aging. Invest Ophthalmol Vis Sci. corneas. PLoS One. 2011;6(12):e28310.
1986;27(4):457–63. 25. Zhu C, Joyce NC.  Proliferative response of cor-
7. Peh GS, Beuerman RW, Colman A, et al. Human cor- neal endothelial cells from young and older donors.
neal endothelial cell expansion for corneal endothe- Invest Ophthalmol Vis Sci. 2004;45(6):1743–51.
lium transplantation: an overview. Transplantation. 26. Lin F, Wang N, Zhang TC. The role of endothelial-­
2011;91(8):811–9. mesenchymal transition in development and patho-
8. Wilson SE, Bourne WM. Fuchs' dystrophy. Cornea. logical process. IUBMB Life. 2012;64(9):717–23.
1988;7(1):2–18. 27. Zhu YT, Chen HC, Chen SY, et  al. Nuclear p120
9. Joyce NC. Proliferative capacity of the corneal endo- catenin unlocks mitotic block of contact-inhibited
thelium. Prog Retin Eye Res. 2003;22(3):359–89. human corneal endothelial monolayers without dis-
10. Tan DT, Dart JK, Holland EJ, et  al. Corneal trans- rupting adherent junctions. J Cell Sci. 2012;125(Pt
plantation. Lancet. 2012;379(9827):1749–61. 15):3636–48.
11. Borderie VM, Boelle PY, Touzeau O, et al. Predicted 28. Joyce NC, Harris DL, Zieske JD.  Mitotic inhibi-
long-term outcome of corneal transplantation. tion of corneal endothelium in neonatal rats. Invest
Ophthalmology. 2009;116(12):2354–60. Ophthalmol Vis Sci. 1998;39(13):2572–83.
12. Eye Bank Association of America. 2016; 2015 Eye 29. Ding V, Chin A, Peh G, et  al. Generation of novel
Banking Statistical Report. monoclonal antibodies for the enrichment and
13. Gain P, Jullienne R, He Z, et  al. Global survey of characterization of human corneal endothelial cells
corneal transplantation and eye banking. JAMA (hCENC) necessary for the treatment of corneal
Ophthalmol. 2016;134(2):167–73. endothelial blindness. MAbs. 2014;6(6):1439–52.
14. Baum JL, Niedra R, Davis C, et al. Mass culture of 30. Senoo T, Joyce NC.  Cell cycle kinetics in corneal
human corneal endothelial cells. Arch Ophthalmol. endothelium from old and young donors. Invest
1979;97(6):1136–40. Ophthalmol Vis Sci. 2000;41(3):660–7.
15. Choi JS, Kim EY, Kim MJ, et  al. Factors affect- 31. Zaniolo K, Bostan C, Rochette Drouin O, et  al.
ing successful isolation of human corneal endo- Culture of human corneal endothelial cells isolated
thelial cells for clinical use. Cell Transplant. from corneas with Fuchs endothelial corneal dystro-
2014;23(7):845–54. phy. Exp Eye Res. 2012;94(1):22–31.
120 S. Wahlig et al.

32. Matthaei M, Meng H, Meeker AK, et al. Endothelial yields functional hexagonal monolayers. PLoS One.
Cdkn1a (p21) overexpression and accelerated 2012;7(12):e51427.
senescence in a mouse model of Fuchs endothe- 48. McCabe KL, Kunzevitzky NJ, Chiswell BP, et  al.
lial corneal dystrophy. Invest Ophthalmol Vis Sci. Efficient generation of human embryonic stem cell-­
2012;53(10):6718–27. derived corneal endothelial cells by directed differ-
33. He J, Kakazu AH, Bazan NG, et al. Aspirin-triggered entiation. PLoS One. 2015;10(12):e0145266.
lipoxin A4 (15-epi-LXA4) increases the endothelial 49. Zhang K, Pang K, Wu X.  Isolation and transplan-
viability of human corneas storage in Optisol-GS. J tation of corneal endothelial cell-like cells derived
Ocul Pharmacol Ther. 2011;27(3):235–41. from in-vitro-differentiated human embryonic stem
34. Schroeter J, Ruggeri A, Thieme H, et al. Impact of cells. Stem Cells Dev. 2014;23(12):1340–54.
temporary hyperthermia on corneal endothelial cell 50. Joyce NC, Harris DL, Markov V, et  al. Potential
survival during organ culture preservation. Graefes of human umbilical cord blood mesenchymal stem
Arch Clin Exp Ophthalmol. 2015;253(5):753–8. cells to heal damaged corneal endothelium. Mol Vis.
35. Spelsberg H, Reinhard T, Sengler U, et  al. Organ-­ 2012;18:547–64.
cultured corneal grafts from septic donors: a retro- 51. Zhao JJ, Afshari NA. Generation of human corneal
spective study. Eye (Lond). 2002;16(5):622–7. endothelial cells via in  vitro ocular lineage restric-
36. Cotsarelis G, Cheng SZ, Dong G, et al. Existence of tion of pluripotent stem cells. Invest Ophthalmol Vis
slow-cycling limbal epithelial basal cells that can be Sci. 2016;57(15):6878–84.
preferentially stimulated to proliferate: implications 52. Trounson A, Thakar RG, Lomax G, et  al. Clinical
on epithelial stem cells. Cell. 1989;57(2):201–9. trials for stem cell therapies. BMC Med. 2011;9:52.
37. Pellegrini G, Traverso CE, Franzi AT, et  al. Long-­ 53. Chambers SM, Fasano CA, Papapetrou EP, et  al.
term restoration of damaged corneal surfaces with Highly efficient neural conversion of human ES and
autologous cultivated corneal epithelium. Lancet. iPS cells by dual inhibition of SMAD signaling. Nat
1997;349(9057):990–3. Biotechnol. 2009;27(3):275–80.
38. Raviola G. Schwalbe line's cells: a new cell type in 54. Song Q, Yuan S, An Q, et  al. Directed differentia-
the trabecular meshwork of Macaca mulatta. Invest tion of human embryonic stem cells to corneal endo-
Ophthalmol Vis Sci. 1982;22(1):45–56. thelial cell-like cells: a transcriptomic analysis. Exp
39. Acott TS, Samples JR, Bradley JM, et al. Trabecular Eye Res. 2016;151:107–14.
repopulation by anterior trabecular meshwork 55. Wu J, Izpisua Belmonte JC.  Dynamic pluripotent
cells after laser trabeculoplasty. Am J Ophthalmol. stem cell states and their applications. Cell Stem
1989;107(1):1–6. Cell. 2015;17(5):509–25.
40. He Z, Campolmi N, Gain P, et al. Revisited micro- 56. Gutierrez-Aranda I, Ramos-Mejia V, Bueno C, et al.
anatomy of the corneal endothelial periphery: Human induced pluripotent stem cells develop tera-
new evidence for continuous centripetal migra- toma more efficiently and faster than human embry-
tion of endothelial cells in humans. Stem Cells. onic stem cells regardless the site of injection. Stem
2012;30(11):2523–34. Cells. 2010;28(9):1568–70.
41. McGowan SL, Edelhauser HF, Pfister RR, et  al. 57. Lister R, Pelizzola M, Kida YS, et  al. Hotspots
Stem cell markers in the human posterior limbus and of aberrant epigenomic reprogramming in
corneal endothelium of unwounded and wounded human induced pluripotent stem cells. Nature.
corneas. Mol Vis. 2007;13:1984–2000. 2011;471(7336):68–73.
42. Whikehart DR, Parikh CH, Vaughn AV, et  al. 58. Gamm DM, Phillips MJ, Singh R. Modeling retinal
Evidence suggesting the existence of stem cells degenerative diseases with human iPS-derived cells:
for the human corneal endothelium. Mol Vis. current status and future implications. Expert Rev
2005;11:816–24. Ophthalmol. 2013;8(3):213–6.
43. Lwigale PY.  Corneal development: different cells 59. Yue BY, Sugar J, Gilboy JE, et  al. Growth of
from a common progenitor. Prog Mol Biol Transl human corneal endothelial cells in culture. Invest
Sci. 2015;134:43–59. Ophthalmol Vis Sci. 1989;30(2):248–53.
44. Hara S, Hayashi R, Soma T, et  al. Identification 60. Fabricant RN, Alpar AJ, Centifanto YM, et  al.
and potential application of human corneal Epidermal growth factor receptors on corneal endo-
endothelial progenitor cells. Stem Cells Dev. thelium. Arch Ophthalmol. 1981;99(2):305–8.
2014;23(18):2190–201. 61. Tripathi RC, Tripathi BJ. Human trabecular endothe-
45. Lovatt M, Yam GH, Peh GS, et al. Directed differ- lium, corneal endothelium, keratocytes, and scleral
entiation of periocular mesenchyme from human fibroblasts in primary cell culture. A comparative
embryonic stem cells, differentiation. 2018;99:62–9. study of growth characteristics, morphology, and
46. Katikireddy KR, Schmedt T, Price MO, et  al. phagocytic activity by light and scanning electron
Existence of neural crest-derived progenitor cells microscopy. Exp Eye Res. 1982;35(6):611–24.
in Normal and Fuchs endothelial dystrophy corneal 62. Lie JT, Birbal R, Ham L, et  al. Donor tissue
endothelium. Am J Pathol. 2016;186(10):2736–50. preparation for Descemet membrane endo-
47. Schmedt T, Chen Y, Nguyen TT, et  al. Telomerase thelial keratoplasty. J Cataract Refract Surg.
immortalization of human corneal endothelial cells 2008;34(9):1578–83.
8  Corneal Endothelial Cells: Methods for Ex Vivo Expansion 121

63. Melles GR, Lander F, Rietveld FJ.  Transplantation 79. Peh GS, Adnan K, George BL, et al. The effects of
of Descemet's membrane carrying viable endo- rho-associated kinase inhibitor Y-27632 on primary
thelium through a small scleral incision. Cornea. human corneal endothelial cells propagated using a
2002;21(4):415–8. dual media approach. Sci Rep. 2015;5:9167.
64. Engelmann K, Bohnke M, Friedl P. Isolation and long-­ 80. De Becker A, Van Riet I. Mesenchymal stromal cell
term cultivation of human corneal endothelial cells. therapy in hematology: from laboratory to clinic and
Invest Ophthalmol Vis Sci. 1988;29(11):1656–62. Back again. Stem Cells Dev. 2015;24(15):1713–29.
65. Engelmann K, Friedl P. Optimization of culture con- 81. Vianna LM, Kallay L, Toyono T, et al. Use of human
ditions for human corneal endothelial cells. In Vitro serum for human corneal endothelial cell culture. Br
Cell Dev Biol. 1989;25(11):1065–72. J Ophthalmol. 2015;99(2):267–71.
66. Engelmann K, Friedl P.  Growth of human cor- 82. Chou ML, Burnouf T, Wang TJ.  Ex vivo expan-
neal endothelial cells in a serum-reduced medium. sion of bovine corneal endothelial cells in xeno-free
Cornea. 1995;14(1):62–70. medium supplemented with platelet releasate. PLoS
67. Proulx S, Bourget JM, Gagnon N, et al. Optimization One. 2014;9(6):e99145.
of culture conditions for porcine corneal endothelial 83. Rao BM, Zandstra PW.  Culture development for
cells. Mol Vis. 2007;13:524–33. human embryonic stem cell propagation: molecu-
68. Niu G, Choi JS, Wang Z, et  al. Heparin- lar aspects and challenges. Curr Opin Biotechnol.
modified gelatin scaffolds for human corneal 2005;16(5):568–76.
endothelial cell transplantation. Biomaterials. 84. Gstraunthaler G.  Alternatives to the use of fetal
2014;35(13):4005–14. bovine serum: serum-free cell culture. ALTEX.
69. Shima N, Kimoto M, Yamaguchi M, et al. Increased 2003;20(4):275–81.
proliferation and replicative lifespan of isolated 85. Miyata K, Drake J, Osakabe Y, et al. Effect of donor
human corneal endothelial cells with L-ascorbic age on morphologic variation of cultured human cor-
acid 2-phosphate. Invest Ophthalmol Vis Sci. neal endothelial cells. Cornea. 2001;20(1):59–63.
2011;52(12):8711–7. 86. Yamaguchi M, Ebihara N, Shima N, et al. Adhesion,
70. Chen CH, Chen VN, Chen SC. Effect of chondroi- migration, and proliferation of cultured human
tin sulfate on the endothelium in corneal storage. corneal endothelial cells by laminin-5. Invest
Cornea. 1996;15(1):35–40. Ophthalmol Vis Sci. 2011;52(2):679–84.
71. Kimoto M, Shima N, Yamaguchi M, et  al. Role of 87. Palchesko RN, Lathrop KL, Funderburgh JL, et al.
hepatocyte growth factor in promoting the growth In vitro expansion of corneal endothelial cells on
of human corneal endothelial cells stimulated by biomimetic substrates. Sci Rep. 2015;5:7955.
L-ascorbic acid 2-phosphate. Invest Ophthalmol Vis 88. Okumura N, Kakutani K, Numata R, et  al.
Sci. 2012;53(12):7583–9. Laminin-511 and -521 enable efficient in  vitro
72. Roy O, Leclerc VB, Bourget JM, et al. Understanding expansion of human corneal endothelial cells. Invest
the process of corneal endothelial morphologi- Ophthalmol Vis Sci. 2015;56(5):2933–42.
cal change in  vitro. Invest Ophthalmol Vis Sci. 89. Kabosova A, Azar DT, Bannikov GA, et  al.
2015;56(2):1228–37. Compositional differences between infant and
73. Okumura N, Koizumi N, Ueno M, et  al. ROCK adult human corneal basement membranes. Invest
inhibitor converts corneal endothelial cells into a Ophthalmol Vis Sci. 2007;48(11):4989–99.
phenotype capable of regenerating in vivo endothe- 90. Numata R, Okumura N, Nakahara M, et  al.
lial tissue. Am J Pathol. 2012;181(1):268–77. Cultivation of corneal endothelial cells on a pericel-
74. Joko T, Shiraishi A, Akune Y, et  al. Involvement lular matrix prepared from human decidua-derived
of P38MAPK in human corneal endothelial cell mesenchymal cells. PLoS One. 2014;9(2):e88169.
migration induced by TGF-beta(2). Exp Eye Res. 91. Muhammad R, Peh GS, Adnan K, et al. Micro- and
2013;108:23–32. nano-topography to enhance proliferation and sus-
75. Kim TY, Kim WI, Smith RE, et al. Role of p27(Kip1) tain functional markers of donor-derived primary
in cAMP- and TGF-beta2-mediated antiproliferation human corneal endothelial cells. Acta Biomater.
in rabbit corneal endothelial cells. Invest Ophthalmol 2015;19:138–48.
Vis Sci. 2001;42(13):3142–9. 92. Bartakova A, Alvarez-Delfin K, Weisman AD, et al.
76. Okumura N, Kay EP, Nakahara M, et al. Inhibition Novel identity and functional markers for human
of TGF-beta signaling enables human corneal endo- corneal endothelial cells. Invest Ophthalmol Vis Sci.
thelial cell expansion in vitro for use in regenerative 2016;57(6):2749–62.
medicine. PLoS One. 2013;8(2):e58000. 93. Chng Z, Peh GS, Herath WB, et al. High throughput
77. Saika S.  TGFbeta pathobiology in the eye. Lab gene expression analysis identifies reliable expres-
Investig. 2006;86(2):106–15. sion markers of human corneal endothelial cells.
78. Cheong YK, Ngoh ZX, Peh GS, et al. Identification PLoS One. 2013;8(7):e67546.
of cell surface markers glypican-4 and CD200 that 94. Okumura N, Hirano H, Numata R, et  al. Cell sur-
differentiate human corneal endothelium from face markers of functional phenotypic corneal
stromal fibroblasts. Invest Ophthalmol Vis Sci. endothelial cells. Invest Ophthalmol Vis Sci.
2013;54(7):4538–47. 2014;55(11):7610–8.
122 S. Wahlig et al.

95. Dorfmueller S, Tan HC, Ngoh ZX, et al. Isolation of is tissue protective in a model of bleomycin-induced
a recombinant antibody specific for a surface marker lung injury. J Immunol. 2006;176(7):4449–58.
of the corneal endothelium by phage display. Sci 101. Masedunskas A, King JA, Tan F, et  al. Activated
Rep. 2016;6:21661. leukocyte cell adhesion molecule is a compo-
96. He Z, Forest F, Gain P, et al. 3D map of the human nent of the endothelial junction involved in tran-
corneal endothelial cell. Sci Rep. 2016;6:29047. sendothelial monocyte migration. FEBS Lett.
97. Yoshihara M, Ohmiya H, Hara S, et  al. Discovery 2006;580(11):2637–45.
of molecular markers to discriminate corneal 102. Lee JG, Heur M.  Interleukin-1beta-induced Wnt5a
endothelial cells in the human body. PLoS One. enhances human corneal endothelial cell migration
2015;10(3):e0117581. through regulation of Cdc42 and RhoA.  Mol Cell
98. Hamuro J, Toda M, Asada K, et al. Cell homogeneity Biol. 2014;34(18):3535–45.
indispensable for regenerative medicine by cultured 103. Sakane H, Yamamoto H, Matsumoto S, et  al.
human corneal endothelial cells. Invest Ophthalmol Localization of glypican-4  in different mem-
Vis Sci. 2016;57(11):4749–61. brane microdomains is involved in the regula-
99. Quintanilla RH Jr, Asprer JS, Vaz C, et al. CD44 is a tion of Wnt signaling. J Cell Sci. 2012;125(Pt 2):
negative cell surface marker for pluripotent stem cell 449–60.
identification during human fibroblast reprogram- 104. Frausto RF, Le DJ, Aldave AJ. Transcriptomic anal-
ming. PLoS One. 2014;9(1):e85419. ysis of cultured corneal endothelial cells as a valida-
100. Volmer JB, Thompson LF, Blackburn MR. Ecto-5′- tion for their use in cell replacement therapy. Cell
nucleotidase (CD73)-mediated adenosine production Transplant. 2016;25(6):1159–76.
Corneal Regeneration: Use
of Extracorneal Stem Cells
9
Christian Claude Lachaud, Abdelkrim Hmadcha,
and Bernat Soria

9.1 Introduction able [2, 3]. For instance, the transplantation of a


donor’s corneal tissue is at present the only thera-
A major issue for regenerative therapies of dam- peutic option to regain vision for most corneal
aged or diseased tissues is identifying suitable disorders causing vision loss and blindness.
SCs able to recreate a functional tissue. Ideally, Unfortunately, the availability of donated ocular
this task should be performed with the use of SCs tissues is however globally much lower than the
of autologous and endogenous origin, thereby demand, a frustrating situation which has forced
avoiding immune rejection risk and ensuring the research of alternative approaches, such as
optimal regenerative outcomes [1]. This approach tissue engineering of artificial corneal tissues
is however currently unviable for most corneal with scaffolds and SCs.
disorders, such as corneal dystrophies, where SCs are unspecialized cells with the unique
endogenous stem and/or progenitor cells and ability to undergo asymmetric cell division. By
their differentiated cellular progenies are pre- means of this mechanism, SCs can maintain their
cisely affected or lacking or because effective self-renewal potential while generating differen-
technologies based on their use are still not avail- tiating progenies [4, 5]. Embryonic SCs (ESCs)
are pluripotent SCs giving rise to the three germ
layers of the developing embryo. Upon an ade-
quate stepwise differentiation protocol, embryo-­
C. C. Lachaud · A. Hmadcha
Department of Cell Regeneration and Advanced derived ESC lines can virtually give rise to any
Therapies, Andalusian Center of Molecular Biology differentiated cell type of the adult body, includ-
and Regenerative Medicine-CABIMER, Junta de ing corneal cells [6]. Somatic SCs residing in
Andalucía-University of Pablo Olavide-University of fetal and adult tissues have by contrast a more
Seville-CSIC, Seville, Andalusia, Spain
e-mail: christian.lachaud@cabimer.es restrictive differentiation potential in vivo, which
is committed to produce cellular components of
B. Soria (*)
Department of Cell Regeneration and Advanced the tissues where they are housed, making them
Therapies, Andalusian Center of Molecular Biology cells with progenitor characteristics [7].
and Regenerative Medicine-CABIMER, Junta de Cumulative evidence has however indicated how
Andalucía-University of Pablo Olavide-University of adult SCs have a broader spectrum of differenti-
Seville-CSIC, Seville, Andalusia, Spain
ating capacities than initially thought [8]. As
Centro de Investigación Biomédica en Red de such, and under new in  vitro chemical environ-
Diabetes y Enfermedades Metabólicas Asociadas
(CIBERDEM), Madrid, Spain ments, several types of adult SCs such as
e-mail: bernat.soria@cabimer.es ­mesenchymal SCs (MSCs) are able to acquire

© Springer Nature Switzerland AG 2019 123


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_9
124 C. C. Lachaud et al.

differentiated characteristics of neurons or hepa- adipose tissue and gingiva) and (iiii) multipotent
tocytes, among others, two cell types which are SCs of neural crest origin (dental pulp and
derived from the embryonic ectoderm and endo- follicle).
derm germ layers, respectively [8, 9]. Such abil- Table 9.1 shows extracorneal stem cells used
ity to switch their phenotype across distinct cell for corneal epithelial regeneration.
lineages is known as transdifferentiation poten-
tial, a property which has attracted growing inter-
est for using adult SCs in the development of 9.2.1 Pluripotent Stem Cells
novel therapies for corneal regeneration [10, 11].
To date, human embryonic and human-­ 9.2.1.1 Embryonic Stem Cells
induced pluripotent SCs, as well as MSCs, are In 2001, Yu et al. addressed for the first time the
the types of SCs which have been more broadly ability of ESC to differentiate into corneal epithe-
applied for experimental corneal regeneration. In lial cells (CEC) by using a Transwell™ coculture
slight contrast, fetal and adult stem/progenitor system incorporating rabbit primary
cells have been in general more restrictively CEC.  Authors indicated how soluble factors
applied to the regeneration of a single corneal secreted by the cocultured rabbit CEC could
cellular layer (epithelium, stroma or endothe- induce mouse ESC to differentiate into epithelial-­
lium), principally on the basis of their phenotypic like cells with developed microvilli and tight
similarities with the type of corneal cells to junctions and express the CEC markers cytokera-
regenerate. As such, epithelial SCs from different tins 3 y 12 (CK3/12) [17]. Thereafter, Wang et al.
extracorneal sources such as the skin or hair fol- reported that D3 mouse ESC-GFP cells pre-­
licle have been specifically applied for corneal induced with retinoic acid (RA) could form a cor-
epithelium regeneration [12–14], whereas endo- neal epithelium-like tissue expressing the CEC
thelial progenitor cells of haematopoietic origin markers CK3 and p63 after their seeding onto
have been tested for repairing the corneal endo- deepithelialized corneoscleral slices [18]. By rec-
thelium [15, 16]. reating a corneal niche microenvironment, human
We herein recover the different types of extra- ESC (hESC) cultured onto collagen IV-coated
corneal SCs which have been experimentally tissue culture dishes containing conditioned
and/or clinically applied for regenerating the cor- media by limbal fibroblasts could acquire epithe-
neal epithelium, stroma and endothelium layers lioid morphologies and express CK12 and p63
and detail their applications. [19]. Interestingly, a study reported how murine
ESC transduced with pax6, a transcription factor
critical for eye development, could efficiently
9.2 Corneal Epithelium generate monolayered corneal epithelial-like
Regeneration cells expressing CK12, E-cadherin and CD44
which could adhere and remain alive onto the
Different types of extracorneal SCs have been surface of injured mouse corneas [20]. The same
proposed for corneal epithelium regeneration, laboratory further reported almost similar results
either because of their ability to differentiate into with non-human primate ESC cells cultured onto
cells with corneal epithelium cells characteristics collagen IV-coated surfaces. In this case, the
or indirectly due to their secretion of factors authors reported their moderated capacity to gen-
inducing in situ corneal epithelium cells regen- erate a corneal epithelium-like tissue (positive for
eration. They can be basically classified into four pax6, CK3/12 and p63) onto injured mouse cor-
categories: (i) pluripotent SCs (embryonic or neas [21].
induced by artificial genetic reprogrammation), The importance of the corneal epithelium
(ii) epithelial stem/progenitor cells (oral mucus, niche environment to direct the differentiation of
hair follicle, epidermis and amnion), (iii) MSCs ESC towards corneal epithelial-like cells was
(umbilical cord, placenta, amnion, bone marrow, further reinforced by other studies [22–24].
­
9  Corneal Regeneration: Use of Extracorneal Stem Cells 125

Table 9.1  Extracorneal stem cells used for corneal epithelial regeneration
SC type Main procedures Principal results References
ESC Transwell coculture with rabbit CEC Epithelial-like, exp° CK3/12 [17]
Preinduction with RA, seeding on Epithelium-like, exp° CK3 and p63 [18]
corneoscleral slices
Culture on collagen IV coating, CM by Epithelioid, exp° CK3 and p63 [19]
limbal fibroblasts
Transduction with pax6 Epithelial-like, exp° CK12, E-cadherin, CD44 [20]
Application on mouse injured corneal Integration on mouse injured corneal surface
surface
Culture on collagen IV coating Limited capacity to form an epithelium [21]
Graft on mouse injured corneal surface in vivo
Exp° CK3/12, pax6, p63
Spontaneously differentiated hESC seeded Expansion, formation multilayers [22]
onto Epithelial-like, exp° CK3, pax6
Bowman’s membrane of deepithelialized
human cornea
Use limbal epithelial SC (LESC) CM Generation LESC-like cells. Multilayers onto [23]
Seeding on APCM, implantation into rabbit APCM, restore transparency in vivo
LSCD model
Culture in KSFM with 7% CO2. Airlifting CEC-like, formation multilayered sheet [24]
culture
Diff° into CEC-like cells and CEnC-like Multilayered epithelium-like onto APCM [24]
cells Restore corneal transparency in rabbit
Generation corneal equivalent onto APCM
iPSC iPSC from human corneal limbal epithelium Diff° into CEC-like CK12, pax6 [25]
cells
Use of stromal inducing activity of PA6
fibroblasts
Coculture with corneal limbal stromal cells Diff° into CEC-like CK12 [26]
Media containing bFGF, EGF and NGF
Culture on collagen IV coating Diff° into CEC-like CK3/12, p63 [27]
Media with inhibitor TGF-β (SB-5051249
and Wnt (IWP-2) + bFGF
Partial reprogrammation into LESC-like Diff° into limbal epithelial cell-like CK3/12 [28]
cells
OMEC Culture onto denuded AM. Graft onto Restore epithelium and corneal transparency [29]
injured surface rabbit corneas
COMET in humans with LSCD Restore epithelium and corneal transparency [30]
COMET in human with Stevens-Johnson Restore epithelium and corneal transparency [31]
syndrome
Culture OMEC into SFM onto collagen IV Generation OMEC sheets [32]
coating
HFSC Culture in limbal fibroblasts CM, collagen Diff° into CEC-like CK12, exp° α6-integrin [13]
IV coating and CK15
Epithelial cell sheets onto fibrin gels
Transplantation in mouse injured surface Successful re-epithelialization [33]
corneas
SESC Culture onto denuded AM. Graft in goat Restore epithelium and corneal transparency [12]
injured cornea (LSCD)
HAEC Culture onto collagen corneal shields Adhesion onto third part of corneas [34]
Transplantation onto deepithelialized rabbit
corneas
Culture with CM by immortalized hCEC Diff° into CEC-like CK3/12 [35]
Airlifting culture onto deepithelialized Multilayered epithelium-like, exp° CK3/12 [36]
rabbit corneas
(continued)
126 C. C. Lachaud et al.

Table 9.1 (continued)
SC typeMain procedures Principal results References
CLEC Organotypic culture umbilical cord lining Formation stratified epithelium in vitro [37]
Cultivation onto denuded AM Formation stratified sheet of CK3/12 cells [38]
BM-MSC Cultivation onto denuded AM Graft adhesion and survival. Lower [39]
Transplantation injured corneal surface (rat) inflammation and neovascularization. No diff°
into CK3+ cells
Coculture system with rat CSSC. Graft in Acquire CK12 exp°. Graft reduces corneal [40]
rat LSCD model opacity and neovascularization
IV and IP injection in rat with injured Limited corneal homing of MSC. MSC [41]
corneal surface reduces corneal opacity and inflammation via
TSG-6
IV injection in mice with injured surface Corneal homing. Induces corneal epithelial [42, 43]
cornea regeneration
Culture onto denuded AM with Corneal healing, reduces corneal opacity [43]
KGF-2 + AS
Grafting onto injured corneal surface
Diff° and culture onto denuded AM with Increased CEC markers ABCG2 and p63 [44]
limbal epithelial CM. Grafting into LSCD Graft improve corneal regeneration
model
Growth on nanofiber scaffolds. Grafting Improves corneal healing. Reduces [45]
onto injured corneal surface (rabbit) inflammation, thickness, neovascularization
Seeding and culture onto nanofiber Graft restores antioxidant enzymes [46]
scaffolds
Grafting onto injured corneal surface
(rabbit)
Subconjunctival injection in a model of Diabetic corneal epithelial wound healing via [47]
corneal epithelial wound healing in type 1 TSG-6
diabetic mice
ADSC Adherent subculture Exp° p63, ABCG2, upregulated CK12 [48]
collagen IV coating, use of CM by hCEC Epithelioid, exp° CK12 [49]
Seeding onto PLA-based nanofibers Adhesion graft onto injured rabbit corneas [45]
scaffold
Topical application onto rabbit injured Adhesion onto cornea [50]
surface cornea Reduces neovascularization and corneal
opacity
HWJSC Culture and seeding onto fibrin-agarose Generation multilayered sheet onto scaffolds [51]
scaffolds Exp° CK3712, CX43, PKG, ZO-1
IDPSC Transplantation onto rabbit injured surface Reconstruction epithelium, exp° LESC and [52]
corneas CEC markers
hIDPSC sheets. Graft onto rabbit injured Reconstruction epithelium, exp° LESC and [53]
surface corneas CEC markers
Abbreviations: Diff differentiation, Exp expression, ADSC adipose-derived SCs, AM amniotic membrane, ALDH1A1
aldehyde dehydrogenase 1 family, member A1, ALDH3A1 aldehyde-3-dehydrogenase 1 family, AS autologous serum,
bFGF basic fibroblast growth factor, BM-MSC bone marrow mesenchymal SCs, CEC corneal epithelial cells, CLEC
cord lining epithelial cells, CM conditioned medium, COMET cultivated oral mucosa epithelial transplantation, DPSC
dental pulp SCs, ESC embryonic SCs, HAEC human amniotic epithelial cells, HFSC hair follicle SCs, HSHS platelet-­
poor horse serum, HWJSC human Wharton’s jelly SCs, IGF-1 insulin-like growth factor 1, iPSC immature dental SCs,
LESC Limbal epithelial SCs, LSCD limbal stem cell deficiency, KS keratan sulphate, OMEC oral mucosa epithelial
cells, PEA poly-(ethyl acrylate), PLGA polylactic-co-glycolic acid, RA retinoic acid, SESC skin epidermal SCs, SFM
serum-free media, TGF-β1 transforming growth factor beta 1, UC-MSC umbilical cord mesenchymal SCs

Indeed, spontaneously differentiated hESC 3–4 stratified layers of pax6+/CK3+ CEC-like


seeded onto the Bowman’s membrane of deepi- cells [22]. Additionally, the use of a culture media
thelialized human cornea buttons were shown to conditioned by limbal epithelial SCs (LESC) was
successfully expand and establish a differentiated shown to be enough by itself to direct hESC
9  Corneal Regeneration: Use of Extracorneal Stem Cells 127

towards LESC-like cells [23]. Their subsequent thus suggesting that their specific epigenomic
seeding onto an acellular porcine cornea matrix status might be responsible of their higher pro-
(APCM) could generate a stratified epithelium pensity to differentiate along the CEC lineage
with basal cells retaining LESC markers. [25].
Transplantation of this neo-corneal tissue graft By replication of the limbal epithelial stem
into a rabbit model of LSCD could restore the cell niche, Yu et  al. reported the generation of
ocular surface damage and prevent corneal neo- CK12+ CEC-like cells from mouse iPSC cocul-
vascularization [23]. Zhang et al. recently gener- tured with limbal stromal cells in a media supple-
ated hESC-derived corneal epithelial progenitor mented with a mix of bFGF, EGF and NGF [26].
cells (CEPC) through culture into a mixture of Mikhailova et al. could efficiently direct human
DMEM/F12 and keratinocyte serum-free medium iPSC into p63+ and CK3/12+ CEC-like cells,
(KSFM) (1:1) under distinct carbon dioxide through sequential culture onto collagen
(CO2) concentrations. The concentration of 7% IV-coated surface, followed by their suspension
of CO2 yielded highest numbers of ABCG2+ and culture in a media containing two small mole-
p63+ CEPC. Purified hESC-derived CEPC could cules (TGF-β inhibitor SB-505124 and Wnt
efficiently generate multilayered epithelioid cells inhibitor IWP-2) and bFGF and a final matura-
by airlifting culture [24]. In the same year, and tion step into the corneal epithelial media CnT-30
using the same protocols as previously published [27]. Of special interest, Cieslar-Pobuda et  al.
[23, 54], Zhang et al. tested the functionality of generated CK3/12+ limbal epithelial cells from
their hESC-derived CEPC to engineer a full-­ induced LESC-like cells by partially reprogram-
thickness corneal equivalent by coculturing them mation of HDFs with three limbal-specific tran-
together with hESC-derived corneal endothelial-­ scription factors: TCF4, CEBPD and ΔNp63α.
like cells onto an APCM. Of note, CEPC gener- Such reprogrammation shortcut avoiding a com-
ated onto the APCM a multilayered corneal plete return to the pluripotent step is suggested to
epithelium equivalent with cells expressing the circumvent the potential teratoma formation
corneal epithelium markers ABCG2+ and CK3+ associated with the use of iPSC-derived corneal
cells. Of particular interest, the recellularized epithelial-like cells [28].
APCM corneal substitute increased progressively
their transparency over the 8 weeks of follow-up
after transplantation in rabbit eyes, thus indicat- 9.2.2 Epithelial Stem Cells
ing the functionality of the hESC-derived neo-­
epithelium and endothelium [55]. 9.2.2.1 Oral Mucus
The oral mucus is a non-keratinized stratified
9.2.1.2 Induced Pluripotent Stem Cells squamous epithelium displaying morphostruc-
Hayashi et  al. reported the first generation of tural and phenotypical similarities with the cor-
iPSC from both human adult dermal fibroblast neal epithelium. As such, Nakamura et  al.
(HDF) and human adult corneal limbal epithelial published in 2003 a pioneer study reporting the
cells (HLEC) by using the retrovirus-mediated cultivation of oral mucosal tissue biopsies onto a
transfection of Yamanaka’s four factors [56], denuded amniotic membrane (AM) as carrier
namely, Oct3/4, Sox2, c-Myc and Klf4 [25]. [29]. The re-epithelialized AM could success-
Colonies of pax6+/CK12+ CEC-like cells could fully graft onto injured corneas of rabbits pro-
be obtained by subjecting both HDF- and HLEC-­ duced by lamellar keratectomy and generate a
derived iPSC to a stromal cell-derived inducing corneal neo-epithelium substitute re-establishing
activity differentiation method based on the sec- corneal transparency [29]. Soon after, in 2004,
retome activity of mouse stromal cells (PA6 Nishida et al. generated autologous cultured oral
cells). Significantly better corneal epithelial dif- mucosa epithelial cell (OMEC) sheets from
ferentiation was however achieved with the patients with bilateral LSCD, by culturing OMEC
HLEC-derived iPSC, which still retained a small into temperature-responsive cell-culture inserts
proportion of differentially methylated regions, placed into larger culture plates containing 3T3
128 C. C. Lachaud et al.

feeder cells. Transplantation of OMEC sheets and CK15 and superficial squamous epithelioid
onto their corneas’ ocular surface has shown to cells expressing dermal (CK10) and corneal
successfully restore their corneal epithelium and (CK12) differentiated markers [13]. A more
cornea transparency [30]. This landmark clinical definitive evidence of the re-epithelialization
study was further followed by the report of the potential of HFSC for the corneal epithelium was
successful restoration of corneal functionality in further reported in a study where a fibrin gel car-
patients with Stevens-Johnson syndrome and rier seeded with mouse HFSC was transplanted
chemical eye injury after a sequential transplan- onto the injured ocular surface of mouse corneas
tation of cultivated oral mucosal epithelial trans- from which were removed their limbal and cor-
plantation (COMET) and penetrating keratoplasty neal epithelium [33]. Altogether, these studies
(PKP) [31]. Of special interest, a recent study therefore identify HFSC as strong candidate cells
reported the generation of tissue-engineered oral for being applied to human LSCD.
mucus epithelial cell sheets in a xenogeneic-free
culture system for potential application into ocu- 9.2.2.3 Epidermis
lar surface surgery. The best protocol consisted in The skin epidermis is a complex multilayered
generating epithelial cell sheets onto collagen stratified epithelium with a strong regenerative
IV-coated culture inserts in a chemically defined potential provided from a pool of skin epidermal
media previously conditioned by inactivated SCs (SESC) residing in the basal layer. Howard
human foreskin fibroblasts and containing 10 ng/ Green’s laboratory established the first successful
ml recombinant EGF [32]. long-term in vitro expansion of human epidermal
keratinocytes obtained from organotypic culture
9.2.2.2 Hair Follicle of epidermis biopsies onto inactivated 3T3 embry-
In 1990, Costarelis et al. identified a population onic fibroblasts [61]. This landmark study laid the
of slow-cycling SCs in the bulge area of the hair basis for the successful grafting of in vitro gener-
follicle [57]. These follicular bulge epithelial SCs ated skin substitutes onto massive full-­thickness
were further shown to contribute to the epidermis burn wounds in humans [62]. Of interest, the skin
regeneration during wound healing [58–60]. and corneal epithelium share common lineage
Later on, Blazejewska et al. reported for the first origin, the basal layer of the embryonic skin, and
time that hair follicle SCs (HFSC) could transdif- their respective SCs, namely, epidermal and lim-
ferentiate in vitro into corneal epithelial-like cells bal epithelial SCs, are both two types of keratino-
upon exposure to limbal environment [13]. Of cyte SCs [10]. In line of their phenotypic
particular interest, the authors reported formation similarities, Yang et  al. demonstrated that SESC
of holoclones (stem cell-derived) only from cells isolated from goat skin and cultured onto denuded
of the bulge area. HFSC obtained either by enzy- human AM could restore a transparent corneal
matic digestion or explant culture of the bulge epithelium after transplantation onto the damaged
area were expanded onto 3T3 feeder cells. corneal surface of goats with total LSCD [12, 63].
Subcultivated HFSC were found to form regular
epithelioid cell layers onto collagen IV- and 9.2.2.4 Amniotic Membrane
laminin-­5-coated surfaces, but not onto laminin 1 The amniotic membrane encloses the amniotic
or fibronectin. Additionally, HFSC could be effi- cavity housing the foetus. Its inner side in contact
ciently differentiated into CK12+ CEC-like cells with the amniotic fluid is a continuous layer of
when cultured in a media conditioned by limbal flattened cells of ectodermal origin, the amniotic
fibroblasts, but not by peripheral corneal or 3T3 epithelium. Different studies have suggested that
fibroblasts. HFSC-derived CEC-like cell sheets human amniotic epithelial cells (HAEC) repre-
generated onto fibrin gels precoated with lam- sent putative extracorneal SCs for corneal surface
inin-­5 displayed morphostructural hallmarks of regeneration. In this way, He et  al. reported for
the native corneal epithelium, with basal cells the first time in 1999 the feasibility of achieving
expressing the epithelial SC markers α6-integrin corneal epithelium repair with HAEC. This was
9  Corneal Regeneration: Use of Extracorneal Stem Cells 129

done by seeding and culturing HAEC onto the connective tissues, principally at the perivascular
concave surface of collagen corneal shields. The niche [66], and as such, they can be harvested
cellularized shields were finally transplanted from virtually all vascularized tissues and organs
onto the deepithelialized surface of rabbit cor- [67, 68]. Reports in the late 1980s described the
neas. HAEC were found to adhere and be retained ability of mesenchymal cells to acquire simple
onto the third part of the corneas transplanted epithelial markers (CK18 and CK19) upon
[34]. Of interest, Yao et al. further reported that in vitro culture into serum-containing media [69].
HAEC cultured into a conditioned media secreted Of interest, the process of mesenchymal-to-­
by spontaneously immortalized hCEC [64] could epithelial transition (MET) was shown to be nec-
differentiate into CK3/12+ CEC-like cells [35]. essary for the reprogrammation of adult
Fatimah et  al. also reported how subcultured fibroblasts into iPSC [70]. Whether adult MSC
HAEC strongly express CK18 and β1 integrin possesses such transdifferentiative ability and
and how their expression of CK19 and CK3 can convert into corneal epithelial cells has
increased upon subculture in a normal media attracted great interest for corneal surface regen-
containing 10 ng/ml EGF, while in turn they lost erative therapies. To date, MSCs of distinct origin
their expression of the stem cell marker p63. have been tested, including the bone marrow, adi-
Zhou et al. recently reported how the seeding and pose tissue and fetal tissues.
airlifting culture of HAEC onto the Bowman’s
membrane of deepithelialized rabbit corneal 9.2.3.1 Bone Marrow MSC
stroma could generate 4–5 stratified layers of Bone marrow-derived MSCs were the first type
CK3/12+ cells mimicking the native corneal epi- of MSC investigated for corneal regeneration.
thelium. Transplantation of the HAEC-corneal Ma et  al. in 2006 reported that a denuded AM
stroma onto deepithelialized rabbit cornea could seeded with human BM-MSC could induce the
restore corneal transparency and a normal cor- re-establishment of a corneal epithelium-like tis-
neal epithelium within 4 weeks [36]. sue onto injured rat corneas. BM-MSC survived
onto the allograft and could inhibit neovascular-
9.2.2.5 Umbilical Cord Lining ization and inflammation, although they failed to
Epithelium transdifferentiate into CK3+ corneal epithelial-­
The umbilical cord lining is a multilayered epi- like cells [39]. Further studies also reported the
thelium [65]. An initial report described how ability of BM-MSC to acquire some of the CEC
umbilical cord lining epithelial cells (CLEC) markers (CK3/CK12, beta1-integrin, C/
were able to form a stratified epithelium upon EBPdelta, ABCG2 and p63) upon directed differ-
organotypic culture [37]. Reza et  al. cultivated entiation culture and their ability to induce the
human CLEC onto a denuded AM carrier. They regeneration of the corneal epithelium in animal
report their initial expression of SC markers models of ocular surface epithelial injury [40–47,
(HES1, ABCG2, BMI1, CK15) and their ability 71–73]. A detailed analysis of the studies pub-
to adhere, grow and form a stratified sheet of lished so far clearly demonstrated how BM-MSC
CK3/12+ CEC-like cells. Transplantation of the can accelerate the healing and regeneration of the
neo-epithelial tissue onto rabbit eyes with limbal injured corneal epithelium. However, the diver-
stem cell deficiency restored a normal clear cor- gent outcomes published so far question whether
neal surface, showing the promise of this cell such regenerative mechanism is directly medi-
type for ocular surface repair [38]. ated through their transdifferentiation into CEC
or is instead rather indirectly mediated by their
secretion of anti-inflammatory and regenerative
9.2.3 Mesenchymal Stem Cells factors sustaining endogenous corneal epithe-
lium regeneration [11].
Mesenchymal SCs (MSCs) are multipotent In support of their indirect contribution are
mesodermal progenitors residing in vascularized several reports indicating how BM-MSC stimu-
130 C. C. Lachaud et al.

late corneal epithelium regeneration through how topically transplanted human ADSC could
secretion of transforming growth factor beta 1 adhere and migrate onto injured corneas and
(TGF-β1) [73] and tumour necrosis factor-­ reduce inflammation and fluorescein positive epi-
inducible gene 6 protein (TSG-6), being this later thelial defects. ADSC could partially revert the
able to induce pro-inflammatory macrophages course of ocular surface neovascularization and
(M1) to adopt an anti-inflammatory phenotype opacity in the partial LSCD model [50].
(M2) [41, 47]. Additionally, BM-MSC transplan-
tation onto injured rabbit corneas after alkali 9.2.3.3 Fetal Tissue MSC: Placenta,
burns has also shown to reduce the expression of Amniotic Membrane
pro-­inflammatory and angiogenic markers and Umbilical Cord
(iNOS, MMP9, alpha-SMA, TGF-β1 and VEGF) The accessibility of discarded fetal tissues such
in subepithelial corneal stromal cells [46]. as fetal placenta, umbilical cord or amniotic
membrane, from which can be harvested clini-
9.2.3.2 Adipose Tissue MSC cally relevant numbers of MSC, has attracted
Several studies addressed the potential of human great interest for regenerative therapies of cor-
subcutaneous adipose tissue-derived MSC (usu- neal disorders. In this way, Garzón et al. isolated
ally termed h-ADSC) to acquire CEC character- human Wharton’s jelly SCs, which are MSC in
istics in  vitro [48, 49] and/or to induce corneal nature [75]. They report how their seeding and
epithelium regeneration in animal models of ocu- culture at high density onto a fibrin-agarose scaf-
lar surface injury [45, 46, 50, 74]. Martinez-­ fold could led to the formation of a sheet of 4–5
Conesa et  al. reported how early cultured layers of cells expressing a set of markers
h-ADSC express the SC markers p63 and ABCG2 [CK3/12, connexin 43 (CX43), plakoglobin
and upregulate their expression of CK12 upon (PKG) and zonula occludens 1 (ZO1)] consistent
time culture, suggesting they possess CEC differ- with CEC [51]. Of special interest, Nam et  al.
entiation capacity [48]. In this way, h-ADSC cul- isolated MSC from human placenta and umbili-
tured onto a collagen IV-coated surface and into a cal cord and expanded them initially in a mini-
conditioned media (CM) by human CEC were mum essential medium-alpha +10% fetal bovine
found to undergo a phenotypical switch into epi- serum. Serum-free conditioned media by P-MSC
thelioid cells expressing CK12. Such transdiffer- or UC-MSC were then produced and found to
entiation however didn’t occur when using limbal exert clonal growth expansion of human limbal
fibroblast CM [49]. In their work, Holan et  al. epithelial cells through secretion of the matrix
established cultures of rabbit limbal epithelial molecule protein transforming growth factor-­
SCs (LESC) and ADSC, which were then seeded beta-­induced protein (TGF-βIp) [76].
for separately onto poly(L-lactic) acid (PLA)-
based nanofibers laminar scaffolds. Of interest,
they report how both LSC and ADSC scaffolds 9.2.4 Immature Dental Pulp Stem
could adhere onto the injured surface of rabbit Cells
corneas and promote corneal epithelium regen-
eration and restore corneal transparency with Vascularized tissues of the oral cavity have been
similar efficiency [45]. The same laboratory fur- identified as a rich source of SCs with MSC char-
ther described how the ADSC scaffolds trans- acteristics. Oral sources for isolating MSC
planted onto injured rabbit corneas led to a include the gingiva, periodontal ligament, devel-
restoration of antioxidant enzymes in the regen- oping tooth (papilla and follicle) and exfoliated
erated epithelium, as well as to a downregulation deciduous teeth (pulp) [77, 78]. Few studies have
of pro-inflammatory and angiogenic markers addressed the potential of oral MSC for corneal
(iNOS, MMP9, alpha-SMA, TGF-β1 and VEGF) epithelium regeneration. Monteiro et al. isolated
[46]. Galindo et al. developed models of partial human immature dental pulp SCs (hIDPSC, also
and total LSCD in rabbits. The authors report termed SHED) from exfoliated teeth [52]. The
9  Corneal Regeneration: Use of Extracorneal Stem Cells 131

hIDPSC were shown to express MSC and ESC available for severe keratoconus forms where
markers. Of interest, undifferentiated hIDPSC corneal opacity has developed. SC-based thera-
also expressed the limbal SC markers p63 and pies are suggested promising, as they may offer
ABCG2. The healing potential of hIDPSC was the possibility of producing healthy keratocytes
evaluated in a rabbit model of total limbal SC to be reincorporated in the diseased corneal
deficiency (TLSCD) induced by sequential stroma or for biomedical engineering applica-
chemical burn and keratectomy. Transplanted tions such as the manufacture of corneal stromal
hIDPSC onto the deepithelialized corneas could equivalents. In this way, we below recover the
reconstruct the corneal epithelium by direct cel- use of different types of extraocular SCs with
lular incorporation, and hIDPSC expressed lim- keratocyte differentiation potential proposed in
bal SCs and corneal epithelial cells markers [52]. the last decade.
In 2010, the same laboratory generated undiffer- Table 9.2 shows extracorneal stem cells used
entiated hIDPSC compact sheets onto a thermo-­ for corneal stromal regeneration.
responsive polymer, poly-N-isopropyl acrylamide
(PNI-PAAm) culture surface. Transplantation of
the hIDPSC sheet onto the same rabbit model of 9.3.1 E
 mbryonic Stem Cells
TLSCD as previously described [52] led to the and Induced Pluripotent Stem
reconstruction of the epithelium, with cells cor- Cells
responding to hIDPSC and with higher expres-
sion of the CEC marker CK3 in suprabasal cells, The derivation of phenotypically and function-
whereas basal cells showed highest expression of ally differentiated keratocytes from human
the LESC markers ABCG2 and p63 [53]. In fur- embryonic SCs (hESC) has been reported quite
ther support of this concept, Tsai et  al. also recently [81–83]. Chan et  al. in 2013 generated
reported the ability of hIDPSC to acquire expres- and purified hESC-derived neural crest progeni-
sion of the CEC marker CK3/12 after 7 days of tors (NCP) expressing the cell surface nerve
coculture in a transwell system with immortal- growth factor receptor (NGFR), by coculturing
ized human corneal epithelial (HCE-T) [79]. them onto the PA6 embryonic fibroblasts, a cell
line with reported neuronal-inducing activity
[81]. Culture of NGFR+ NCP as compact 3D pel-
9.3 Corneal Stromal lets in serum-free media containing ascorbate has
Regeneration shown to upregulate their expression of the kera-
tocytes genes AQP1, B3GNT7, PTDGS and
The corneal stroma harbours specialized fibro- ALDH3A1 and keratocyte phenotype as demon-
blasts, the keratocytes, which are critical to main- strated by their 10,000-fold upregulation of kera-
tain a specific arrangement of extracellular matrix tocan (KERA), a proteoglycan principally found
(ECM) molecules, necessary to confer transpar- in the corneal stroma [100]. Additionally, culture
ency. Keratocytes are neural crest-derived mes- medium from pellets was also found to contain
enchymal cells [80]. Although they are quiescent high molecular weight keratocan modified with
in the healthy cornea, corneal trauma can how- keratan sulphate, a unique molecular component
ever partially induce their reactivation into myo- of corneal stroma [81]. Later on, the same labora-
fibroblasts and their secretion of de novo ECM tory modified slightly their previous protocol, in
fibrotic components leading to fibrosis/scarring a two-step procedure omitting the monolayered
and loss of transparency. Additionally, corneal culture of purified hESC-derived NGFR+, which
stromal diseases also originate when keratocytes were instead directly put into 3D pellets culture
become dysfunctional and/or are progressively in a defined serum-free keratocyte differentiation
lost through apoptosis, such as in keratoconus, medium (KDM) consisting of advanced DMEM,
one of the most common form of corneal dystro- with 10 ng/ml bFGF and 0.1 mM ascorbic acid-­
phy. Keratoplasty is currently the only treatment 2-­phosphate to induce a keratocyte phenotype as
132 C. C. Lachaud et al.

Table 9.2  Extracorneal stem cells used for corneal stromal regeneration
SC typeSummary procedures Principal results References
ESC Cocultured with PA6 fibroblasts. 3D pellet culture Exp° AQP1, B3GNT7, PTDGS, [81]
in SFM + ascorbate ALDH3A1, keratocan, KS
3D pellet culture. SFM + RA + ascorbic Exp° keratocan, KS [82]
acid-2-phosphate
iPSC 3D pellet culture in DMEM + bFGF + ascorbic Exp° keratocan, ABCB5, vimentin and [83]
acid-2-phosphate HNK1
DPSC 3D pellet culture in DMEM + ascorbate-2-­ Exp° type I collagen, keratocan, KS [84]
phosphate, bFGF and TGF-β1. Intrastromal Restore transparency
injection in mouse corneas
UC-MSC Standard adherent culture. Intrastromal injection Exp° ALDH3A1, lumican and [85]
in lumican−/− null mice keratocan
Standard adherent culture. Intrastromal injection Reduced stromal GAGs and corneal [86]
in mucopolysaccharidosis VII mice haze
BM-MSC Intrastromal injection in keratocan null mice Diff° into keratocan+ keratocytes [87]
Culture with keratocyte CM Diff° into keratocytes-like, keratocan+, [88]
lumican+ and ALDH1A1+
Generation GFP+-BM-MSC. Intrastromal Increased stromal keratocyte markers [89]
injection in injured rat corneal stroma (keratocan, ALDH and CD34). Diff°
into GFP+/keratocan+ cells
Culture with mouse corneas extract + IGF-1 Diff° into keratocan+, lumican+ [71]
keratocytes-like
Topical corneal application of CM by BM-MSC Enhanced stromal keratocyte survival [90]
in injured surface cornea
IV injection in injured mouse cornea Reduced corneal swelling, [91]
inflammation and opacity. Reduced
TNF-α, IL-1β and IL-6
ADSC Intrastromal injection in rabbit cornea Rearranged into multilayered [92]
keratocytes. Exp° ALDH3A1+ and
keratocan+
Culture into fibrin gels or 3D pellet culture into Diff° in vitro into keratocan+, [93]
SFM + ascorbate ALDH3A1+
Culture onto PLGA scaffold. Intrastromal graft Good intrastromal implantation. [94]
ADSC-PLGA construct Restore corneal transparency
Cocultured with human keratocytes Diff° in vitro into keratocan+, [95]
ALDH3A1+
Seeding on PEA-based porous membranes. Good intrastromal implantation inside [96]
Creation of corneal stromal equivalent the stroma of rabbit corneas
Seeding onto acellular human corneal stroma Good intrastromal graft integration. [97]
sheet. Graft into rabbit cornea Diff° of h-ADSC into human
keratocan+
Phase-1 study intrastromal implantation of Good graft implantation at 6 months [98]
h-ADSC seeded human cornea matrix for Moderate amelioration visual
advanced keratoconus parameters
Culture in DMEM + bFGF, RA, Diff° in vitro into keratocan+, [99]
ascorbate-2-phosphate ALDH3A1+
Abbreviations: Diff differentiation, Exp expression, ADSC adipose-derived SCs, ALDH1A1 aldehyde dehydroge-
nase 1 family, member A1, ALDH3A1 aldehyde-3-dehydrogenase 1 family, bFGF basic fibroblast growth factor,
BM-MSC bone marrow mesenchymal SCs, CM conditioned medium, DPSC dental pulp SCs, ESC embryonic
SCs, GAGs glycosaminoglycans, HSHS platelet-poor horse serum, IGF-1 insulin-like growth factor 1, iPSC
immature dental SCs, KS keratan sulphate, PEA poly-(ethyl acrylate), PLGA polylactic-co-glycolic acid, RA reti-
noic acid, SFM serum-free media, TGF-β1 transforming growth factor beta 1, UC-MSC umbilical cord mesen-
chymal SCs
9  Corneal Regeneration: Use of Extracorneal Stem Cells 133

evidenced by their increased secretion of kerato- stromal BM-MSC-derived GFP+/keratocan+ cells


can and keratan sulphate proteoglycans after were found homed at the site of injury [89]. In
3  weeks of differentiating culture [82]. Naylor other study, purified mouse BM-MSC were cul-
et  al. reported the derivation of AP2a+ and tured with mouse corneas extract (MCE) obtained
p27NTR+ neural crest cells (NCCs) from human-­ through repeated freezing-thawing cycles and
induced pluripotent SCs (hiPSC). 3D pellet cul- with or without 20 ng/ml insulin-like growth fac-
ture of the hiPSC-derived NCCs into a keratocyte tor 1 (IGF-1). Combination of MCE and IGF-1
differentiation (KD) media could efficiently has shown to be the best condition for differenti-
direct their differentiation into keratocytes-like ating BM-MSC into K12+, keratocan+ and lumi-
cells expressing keratocan, ABCB5, vimentin can+ keratocyte-like cells [71]. To study the
and HNK1. Additionally, hiPSC-derived NCCs anti-inflammatory role of MSC for corneal
seeded onto donor’s corneal tissue could migrate stroma injury, Jiang et  al. performed a mouse
inside the corneal stroma and developed a model of corneal injury by epithelial scraping
keratocytes-­like morphology and expressed kera- after ethanol injury. They showed that topical
tocyte markers [83]. corneal surface application of a conditioned
media by BM-MSC could significantly enhance
stromal keratocyte survival [90]. By using the
9.3.2 M
 esenchymal Stem Cells same mouse corneal injury model, Yun et  al.
(MSCs) demonstrated that an intravenous transplantation
of 1.106 human BM-MSC was able to reduce cor-
A special interest has been focussed onto adult neal swelling and inflammation through reduced
MSC as extracorneal SCs for obtaining kerato- leukocytic infiltration. BM-MSC treatment
cytes, which are indeed cells with MSC-like fea- resulted in decreased corneal opacity and reduc-
tures [101]. The sources of MSC used to generate tion of the inflammatory markers TNF-α, IL-1β
keratocyte-like cells include the bone marrow, and IL-6 [91].
adipose tissue, umbilical cord and dental pulp.
9.3.2.2 Adipose Tissue MSC
9.3.2.1 Bone Marrow MSC MSC isolated from processed lipoaspirate (PLA)
The first evidence of the keratocyte differentia- cells of the human subcutaneous adipose tissue
tion capacity of MSC of bone marrow origin was (termed ADSC or ASC) are a type of extracorneal
reported by Liu et al. [87]. They showed how a MSC that has received specific attention for
corneal intrastromal transplantation of ex  vivo establishing regenerating therapies of corneal
Dil-labelled mouse BM-MSC in keratocan null stromal disorders. Arnalich-Montiel et  al.
(Kera−/−) mice led to their correct stromal inte- reported for the first time that human PLA cells
gration and differentiation into Dil+/keratocan+ could satisfactorily incorporate for up to 10
keratocytes [87]. Park et al. reported how human weeks in the corneal stroma of immune-­
BM-MSC cultured in a keratocyte-conditioned competent rabbits [92]. In a second set of experi-
media (KCM) could efficiently differentiate into ments, they further showed that PLA cells
keratocyte-like cells, as evidenced by their loss of injected in rabbit corneas with partial stromal
α-SMA expression and up-expression of kerato- ablation could home and rearrange into the typi-
can, lumican and aldehyde dehydrogenase 1 fam- cal multilayered keratocytes organization and
ily, member A1 (ALDH1A1) [88]. Demirayak express ALDH and keratocan [92]. In their study,
et al. created a penetrating injury in rat corneas. Ma et  al. report their creation of a thin disc of
Rat GFP+ BM-MSC were further injected into polylactic-co-glycolic acid (PLGA) as a scaffold
the anterior chamber, and their homing to the that was further cellularized with autologous rab-
injury site was monitored. At 8  weeks post-­ bit ASC.  The rASC-PGLA stromal equivalent
transplantation, rats displayed increased kerato- constructs were further grafted in a rabbit corneal
cyte markers (keratocan, ALDH and CD34), and stromal pocket. Implanted rASC-PGLA con-
134 C. C. Lachaud et al.

structs became progressively transparent and cytes in a transwell system and using a kerato-
with histological features similar to the native cytes media consisting of DMEM media
corneal stroma [94]. Alió del Barrio et al. reported containing 10 ng/ml FGF2 0.1 mM ascorbate-­2-­
the creation of corneal stromal equivalents by phosphate 1% heparin-stripped platelet-poor
in  vitro cellularization of poly-(ethyl acrylate) horse serum (HSHS) [95]. Cultured h-ADSC into
(PEA)-based macroporous membranes with DMEM containing 10  ng/ml FGF2 0.1  mM
human ADSC.  The authors reported how PEA ascorbate-2-phosphate and retinoic acid (RA)
scaffolds containing 10% hydroxyethyl acrylate were shown to upregulate keratocyte markers. A
(PEA-HEA10) display optimal outcomes in concentration of 1 μM RA produced the highest
terms of h-ADSC colonization rate and lack of upregulation of keratocan and ALDH3A1 [99]
extrusion at 3 months post implantation inside
the stroma of rabbit corneas. [96]. In the same 9.3.2.3 Umbilical Cord MSC
year, the same laboratory successfully recolo- Liu et al. performed an intrastromal implantation
nized in vitro decellularized sheets of human cor- of Dio- and/or Dil-labelled human umbilical cord
neal stroma with h-ADSC, which were further MSC (h-UMSC) into lumican−/− null mice.
intrastromally implanted in rabbit corneas. Transplanted h-UMSC in lumican−/− mice prolif-
Stromal equivalents demonstrated an excellent erated transiently inside the stroma and expressed
biointegration, transparency and survival, and lumican. Corneal transcript expression for
differentiation of h-ADSC into human kerato- ALDH3A1, lumican and keratocan was also
can+ cells was patent at 3 months post implanta- increased in response to h-UMSC transplantation
tion. The combined use of h-ADSC and donor´s [85]. Later on, an intrastromal transplantation of
corneal stroma may allow tissue engineering of h-UMSC was also performed into mucopolysac-
bioequivalents for corneal stromal disorders [97]. charidosis VII mice which develop corneal cloud-
Alió del Barrio et  al. recently published results ing due to glycosaminoglycans accumulation. Of
from a human Phase-1 study addressing the effi- note, h-UMSC implantation reduced corneal
cacy of the implantation of a human corneal stro- haze, restored keratocyte morphology and
mal lamina cellularized or not with autologous decreased stromal content of chondroitin and
h-ADSC for the treatment of advanced keratoco- heparan sulphate [86].
nus [98]. Intrastromal implantation of decellular-
ized and recellularized laminas were similarly
well accepted after 6 months, and both approaches 9.3.3 O
 ral Stem Cells: Dental Pulp
provided a similar moderated recuperation of Stem Cells
visual parameters and keratometric values, thus
suggesting h-ADSC didn´t exerted significant In their work, Syed-Picard et al. isolated human
improvements over that of the stromal implant adult dental pulp cells (hDPCs) from extracted
alone [98]. third molars [84]. Collected pulp tissue was sub-
In addition to these preclinical studies, other jected to collagenase I treatment to liberate
studies also addressed the in  vitro ability of DPCs, which were then expanded on plastic
human subcutaneous adipose tissue MSC [93, adherent plates into a DMEM media containing
95, 99] to differentiate into keratocytes. Human 20% FBS. Subcultured hDPCs between passages
ADSC-derived keratocyte-like cells (keratocan+/ 2 and 4 were subjected to pellet culture for
ALDH3A1+) could be obtained through different 3  weeks into a DMEM basal media containing
methods. In one study, they were obtained by cul- ascorbate-2-phosphate, FGF2 and TGF-β1 [84].
turing h-ADSC into fibrin gels or as pellet cul- After 2 weeks of culture in keratocyte differentia-
tures into serum-free or reduced-serum media tion medium, hDPCs displayed increased type I
containing ascorbate [93]. Alternatively Zhang collagen and keratocan expression and secreted
et  al. generated keratocyte-like cells from keratan sulphate-containing proteoglycans.
h-ADSC cocultured with human primary kerato- Intrastromal injection of Dio-labelled hDPC in
9  Corneal Regeneration: Use of Extracorneal Stem Cells 135

healthy mouse corneas revealed good implanta- chymal progenitors (POMPs). POMPs cultured
tion outcomes over time, without alterations of into a lens epithelial cell-conditioned medium
corneal transparency or thickness. Of special could generate N-cadherin+ CEnC-like cells
interest, 2 weeks after implantation, Dio+ hDPCs expressing the transcription factors FoxC1 and
were found to have produced specific human col- Pitx2. Seeding of hESC-derived CEnC-like cells
lagen type I and keratocan [84]. onto a porcine corneal stromal lamella was fur-
ther performed to create a corneal endothelium
equivalent. Implantation of the construct into
9.4 Corneal Endothelium eyes of rabbit with corneal endothelium dysfunc-
Regeneration tion has shown to gradually increase corneal
transparency [54]. Soon after, McCabe et al. also
The search for extracorneal SCs able to differ- reported the successful directed differentiation of
entiate into functional CEnC or of adult cells hESC into CEnC-like cells by promoting their
displaying similar functions (surrogate) has initial differentiation into neural crest progenitors
focussed significant attention in the last decade. (NCPs) in a culture media containing the Smad
In that way, different laboratories have experi- inhibitors Noggin and SB431542. NCPs were
mentally evaluated the suitably of obtaining further differentiated towards CEnC-like cells in
CEnC by in vitro ocular lineage restriction of a corneal endothelium culture media containing
their precursor SCs, such are ESC or postnatal B27 supplement, 10  ng/ml platelet-derived
neural crest progenitors. The use of iPSC to growth factor BB (PDGF-BB) and 10  ng/ml
generate CEnC has also been object of interest. dickkopf WNT signalling pathway inhibitor 2
On the other hand, other laboratories also eval- (Dkk-2). hESC-derived CEnC-like cells have
uated the suitability of adult stem or progenitor shown to express the corneal endothelium mark-
cells from distinct lineages, such as MSC, vas- ers COL8A1, COL8A2 and Na+/K+ATPase α1
cular endothelial progenitor cell (VEPC) and (ATPA1) [102]. In their work, Chen et al. show
skin-derived precursors (SKPs). Finally, other that treatment with all-trans retinoic acid (RA) of
possibilities would be to use differentiated embryoid body (EB) generated from mouse-­
cells from tissues which are morphostructur- induced pluripotent SCs (iPSC) could induce
ally and functionality closely similar to the their differentiation into neural crest cells [104].
corneal endothelium such are some simple epi- Replating of the differentiated EB onto gelatin-­
theliums such as the vascular endothelium or coated plates into a conditioned medium pro-
mesothelium. duced by primary lens epithelial cells has shown
We below detail in more extent the use of the to produce outgrowth monolayered cells express-
different types of extracorneal SCs that have been ing the CEnC markers ATPA1, ZO-1 and AQP1
proposed for corneal endothelium regeneration. [104]. Following a similar two-step differentia-
Table 9.3 shows extracorneal stem cells used tion protocol, Song et al. generated hESC-derived
for corneal endothelium regeneration. EB, which were further induced to NCPs by
using a culture media containing Noggin and
SB431542. Differentiating EBs were further cul-
9.4.1 Embryonic and Induced tured on adherence onto chondroitin sulphate and
Pluripotent Stem Cells laminin-­coated plates in a medium supplemented
with PDGF-BB, Dkk-2 and TGF-β or alterna-
Zhang et al. addressed for the first time in 2014 tively in a CM produced by bovine primary
the corneal endothelial differentiation capacity of CEnC, this later producing the best differentia-
human ESC using a two-step induction proce- tion into CEnC-like cells expressing COL8A1,
dure. A transwell-based coculture system incor- ATPA1, ZO-1, AQP1 and S100A4 [103]. Zhao
porating human corneal stromal cells was first and Afshari reported the successful generation of
used to direct hESC towards periocular mesen- CEnC-like cells through in  vitro ocular lineage
136 C. C. Lachaud et al.

Table 9.3  Extracorneal stem cells used for corneal endothelium regeneration
SC type Summary procedures Principal Results References
ESC Use of lens epithelial cells CM Polygonal, CDH2+, Vimentin+. Functional [54]
in vivo
SF media + B27, PDGF-BB, Dkk-2 Polygonal-shaped, COL8A1+, COL8A2+, [102]
ATPA1+
Chondroitin sulphate and laminin coating Polygonal, COL8A1+, ATPA1+, ZO-1+, AQP1+ [103]
Use of CM by bovine primary CEnC and S100A4+
Three-step diff° protocol Functional full-thickness cornea substitute [55]
(TECS)
iPSC EB formation. Treatment with RA Polygonal ATPA1+, ZO-1+, AQP1+ [104]
Gelatin coating, lens epithelial cells CM
Three-step diff° protocol. Suppression Polygonal ZO-1+, ATPA1+, CDH2+ [105]
TGF-β and ROCK signalling
NCSC Explant culture neural tube. Laminin and Polygonal, ATPA1+, ZO-1+ [106]
chondroitin-6 sulphate coating Generation of CE-like monolayer onto bovine
Use of rat primary CEnC CM APCM
SKP Spheres into SFM containing B27, EGF Polygonal, Atp1a1+, Cdh2+ and Pitx2+. Seeded [107]
and bFGF. Media with RA, BIO and on
Y-27632 Type I atelocollagen, functional in vivo
Chondroitin sulphate and laminin coating Polygonal, express genes (atp1a1, ZO-1, [108]
Coculture with hCEnC line B4G12 cdh2, ca2, col4a2, col8a2) and proteins
(ZO-1, ATP1A1)
BM-EPC Transwell coculture with human CEnC Polygonal, AQP1+, NSE+, functional in cat [15]
corneas
UCB-­ Collagen I coating into EGM-2 media Integrate into rabbit injured CE using [16]
EPC magnetic field
UCB-­ Use of human lens epithelial cells CM Not polygonal, intercellular expression ZO-1, [109]
MSC CDH2
BM-MSC BM-MSC-CM improves CEnC growth Improves polygonal shape of human CEnC [110]
AT-MC Culture in MRPM. Seeding onto human Polygonal. ZO-1, CDH2, ATP1A1, COL4A2, [111]
lens capsules (HLCs) COL8A2, SLC4A4. Adhere onto HLCs
APCM acellular porcine cornea matrix, AT-MC adipose tissue mesothelial cells, BM-EPC bone marrow endothelial
progenitor cells, bFGF basic fibroblast growth factor, BM-MSC bone marrow mesenchymal SCs, CDH2 N-cadherin,
CEnC corneal endothelial cells, CM conditioned medium, EGF epidermal growth factor, ESC embryonic SCs, IPSC
immature dental SCs, MRPM mesothelial retaining phenotype media, NCSCs neural crest SCs, PDGF-BB platelet
growth factor BB, RA retinoic acid, SFM serum-free media, SKP skin-derived precursor, UCB-EPC umbilical cord
blood endothelial progenitor cells, UCB-MSC umbilical cord blood mesenchymal SCs

restriction of human iPSC, using a three-step [105]. Zhang et  al.’s laboratory recently pub-
induction protocol [105]. The iPSC were first lished a relevant study demonstrating successful
directed towards eye field SCs (EFSC) in a tissue engineering of a full-thickness cornea sub-
serum-free media containing bFGF, N2 and B27 stitute (TECS) through sequential cellularization
growth supplements and small molecule inhibi- of an acellular porcine cornea matrix (APCM)
tors (SB431542, LDN193189 and IWP-2). The with limbal epithelial-like cells and CEnC-like
iPSC-derived EFSC were then directed towards cells generated from hESC, using for these later a
HNK-1+/p75NTR+ ocular neural crest SCs similar approach as previously published in 2014
(NCSCs) onto Matrigel-coated surface in a [54]. Of special interest, TECS implanted within
serum-free media containing N2, B27, L-ascorbic rabbit eyes increased their transparency after few
acid 2-phosphate and CHIR99021 (GSK-3 weeks and developed significant lower neovascu-
Inhibitor). NCSCs were finally directed towards larization than implanted APCM scaffolds with-
ZO-1+, ATPA1+ and N-cadherin+ CEnC-like cells out cellularization, that instead became
by suppressing TGF-β and ROCK signalling increasingly cloudy [55].
9  Corneal Regeneration: Use of Extracorneal Stem Cells 137

9.4.2 Neural Crest Stem Cells transparency by comparison to the carrier with-
out cells [107]. In 2017, Shen et al. published an
Ju et  al. isolated P75 and HNK-1 dual positive excellent preclinical study based on the use of
neural crest SCs (NCSCs) from an explant cul- human adult SKPs isolated from eyelid operation
ture of the neural tube of rat embryos. [108]. SKPs were isolated from the dermal layer
Differentiation of NSCs into ATPA1+ and ZO-1+ after enzymatic digestion into Liberase DH solu-
polygonal CEnC-like cells was achieved by cul- tion. SKPs were induced to proliferate in a
turing NSCs onto plastic culture plate coated serum-free DMEM/F12-based media supple-
with laminin and chondroitin-6 sulphate in an mented with 2% B27, bFGF and EGF and sub-
inductive media containing a CM by rat primary cultured for 2–4 passages. Their differentiation
CEnC.  Seeding and culture of the NSC-derived was performed for 8 days into a transwell cocul-
CEnC-like cells onto an APCM could generate a ture system with SKPs cultured onto a chondroi-
corneal endothelium-like monolayer [106]. tin sulphate- and laminin-coated surface and
immortalized B4G12 hCEnC cells in the upper
insert. By this method, hSKPs progressively
9.4.3 Skin-Derived Precursors underwent a fibroblastic-to-polygonal morpho-
logic shift, which was associated with increased
Skin-derived precursors (SKPs) reside in the CEnC genes (Atp1a1, ZO-1, N-cadherin, CA2,
adult dermal layer and were shown to represent a Col4a2 and Col8a2) and proteins (ZO-1,
pool of primitive cells sharing hallmarks with ATP1A1). Functional evaluation of hSKP-­
embryonic neural crest SCs, the population from derived CEnC-like cells was demonstrated after
which are derived CEnC [112]. In light of this implantation as a cell suspension in the anterior
concept, Inagaki et al. reported a work addressing chamber of rabbit and monkey eyes, from which
the ability of SKPs isolated from neonatal mouse were scrapped their corneal endothelium to
facial skin to differentiate along the corneal expose a decellularized Descemet’s membrane.
endothelial cells lineage [107]. For this issue, the Eyes treated with CEnC-like cells significantly
dermal skin layer was separated from the epider- maintain good corneal thickness and clarity, by
mis after dispase II enzymatic digestion and was comparison to non-treated operated eyes that
then fully disaggregated using collagenase. SKPs developed severe stromal oedema and corneal
were propagated as floating spheres in low attach- opacity [108].
ment plate and in a serum-free media containing
B27 supplements, EGF and bFGF. SKP spheres
expressed the neural crest markers p75NTR and 9.4.4 Vascular Endothelial
N-cadherin. Adherent culture of SKPs onto 0.1% Progenitors: Bone Marrow
gelatin-coated plate in a corneal endothelium-­ and Cord Blood
inducing medium (CEIM) containing 5% FBS,
insulin, all-trans retinoic acid, the GSK 3-beta Shao et al. reported the isolation of human fetal
inhibitor BIO, TGF-β2 and the ROCK inhibitor bone marrow-derived endothelial progenitor cells
Y-27632 was able to direct their differentiation (BEPC) [15]. Using a coculture transwell system,
into polygonal CEnC-like cells expressing the corneal endothelial differentiation of BEPCs was
corneal endothelium genes Atp1a1 (Na+/ induced in a non-contact fashion with human pri-
K+ATPase α1), Cdh2 (N-cadherin) and Pitx2. mary CEnC and their conditioned medium. Upon
Culture of the SKP-derived CEnC onto type I stimulation, the induced BEPC-derived CEnC-­
atelocollagen sheets was performed to generate a like cells adopted polygonal morphologies and
corneal endothelium (CE) equivalent. expressed AQP1, neuron-specific enolase (NSE)
Transplantation of the CE equivalent into a rabbit and tight junctions. BEPC-derived CEnC-like
model of bullous keratopathy has shown to sub- cells seeded and grown onto an APCM could
stantially reduce corneal thickness and maintain form a corneal endothelium equivalent with the
138 C. C. Lachaud et al.

capacity to restore the clarity in cat’s corneas CEnC.  For this purpose, hBM-MSC were first
after stripping of their Descemet’s membrane and expanded in a DMEM containing 10% FBS, and
corneal endothelium [15]. In a next study, Shao confluent hBM-MSC were then used to condition
et  al. isolated EPCs from the mononuclear cell a basal growth medium designed for culturing
(MNC) fraction of human umbilical cord blood. CEnC cultures (OptiMEM-I containing 8% FBS,
Selective outgrowth of umbilical cord EPCs was EGF, ascorbic acid and chondroitin sulphate).
performed through culture of adherent MNCs Application of this conditioned medium to
onto collagen I-coated plates into the EGM-2 hCEnC culture has shown to modify their mor-
endothelial culture media. EPCs positive for von phology to a more and regular polygonal mor-
Willebrand factor (vWF), Ac-LDL, CD133 and phology similar to that displayed in  vivo by
CD34 were obtained with this method. CEnC. The hBM-MSC-­conditioned medium also
Trypsinized EPCs were further labelled with stimulates CEnC in vitro proliferation to a higher
CD34 immunomagnetic nanoparticles and degree than the conditioned medium obtained
infused into the anterior chamber of rabbit eyes from 3T3 mouse fibroblasts [110].
after surgical removal of the Descemet’s mem-
brane and corneal endothelium. Their apposition
and adhesion onto the denuded anterior corneal 9.4.6 Adipose Tissue Mesothelial
surface were achieved by placing during several Cells
hours a magnet on top of the eyes. EPC trans-
plantation could progressively reduce corneal Mesothelial cells are squamous epithelial-like
oedema and opacity [16]. cells of mesodermal origin [113]. They are found
lining the surface of visceral organs house into
coelomic cavities and also the walls of these cavi-
9.4.5 Mesenchymal Stem Cells ties. Although mesothelial cells are not demon-
strated SCs, cumulating evidence is however
9.4.5.1 Umbilical Cord Blood MSC demonstrating their in  vitro capacity to acquire
The first experimental evaluation of the capacity hallmarks of several differentiated mesodermal
of mesenchymal SCs (MSCs) to differentiate into cell lineage such as vascular smooth muscle cells,
CEnC-like cells was published by Joyce et al. in adipocytes and osteocytes, and thus the existence
2012. MSCs were isolated from human umbilical of mesothelial progenitors has been proposed
cord blood (UCB-MSC) and established under [114–116]. Because the mesothelium and corneal
adherent culture. The authors reported significant endothelium are two types of simple epithelium,
morphological modification of UCB-MSC cul- it is therefore possible that mesothelial cells are
tured in a culture media supplemented with 20% endowed with some similar functions of
FBS previously conditioned by human lens epi- CEnC.  Our laboratory reported that mesothelial
thelial cells. Although the differentiated UCB-­ cells represent a putative cellular surrogate to
MSC expressed ZO-1 and N-cadherin at regenerate the corneal endothelium [111]. In this
intercellular contact, they however failed to adopt way, we reported that freshly isolated murine adi-
the typical polygonal morphology of pose tissue mesothelial cells (ATMCs) express
CEnC.  UCB-MSC could successfully adhere different genes expressed in CEnC in similar
in vitro onto areas of damaged corneal endothe- (N-cadherin) or inclusively higher levels (col4a2,
lium of human corneas but however didn’t adopt slc4a4, car-2 and atp1a1). Only col8a2, a highly
polygonal morphologies similar to CEnC [109]. specific CEnC marker, was expressed in lower
extent in the ATMCs. Subcultured ATMCs were
9.4.5.2 Bone Marrow MSC shown to fully adhere and form a continuous
Nakamura et al. evaluated whether the secretome monolayer onto decellularized human lens cap-
of human bone marrow-MSC (hBM-MSC) is sules [111]. Further studies are currently under
able to enhance the proliferation of human development to assess whether human ATMCs
9  Corneal Regeneration: Use of Extracorneal Stem Cells 139

have similar phenotype with CEnC and could endothelium. Whether these cells are plastic
serve for tissue engineering of a functional cor- enough to adapt to the anterior chamber microen-
neal endothelium equivalent. vironment and acquire functions of native cor-
neal endothelial cells is a critical issue to better
determine. Altogether, the use of extracorneal
9.5 Concluding Remarks SCs of autologous origin for tissue engineering
of corneal tissue equivalents is expected to pro-
The approach of using extracorneal SCs for vide great achievements in the next decades and
regenerating the cornea is an incipient field which might converge towards the manufacture of a
is however experimenting a very fast develop- full-thickness corneal autograft.
ment, principally for tissue engineering applica-
tions of corneal tissue equivalents. A global Acknowledgement  Authors are supported by the nonprofit
examination of the different types of SCs of Fundación Progreso y Salud, Consejería de Salud, Junta de
Andalucía; FEDER cofounded grants from Instituto de
extracorneal origin proposed so far for corneal Salud Carlos III and the Ministry of Economy, Industry and
regeneration leads to the conclusion that using Competitiveness (Red TerCel: RD12/0019/0028 and
SCs belonging to a similar or closely related lin- RD16/00259; CIBERDEM: CB07/08/0006; PI14/01015,
eage to that of the corneal cell type to replace PI16/00259, PI17/02104 and CD16/00118); and Junta de
Andalucía (PAI-BIO311, CTS-576, CTS 11-727, PI-0109-
appears to be the most promising cell replace- 2014, PI0007/2016 and PI0272/2017). CIBERDEM is an
ment strategy to use in the near term. Given that initiative of the Instituto de Salud Carlos III.
much of the progresses accomplished until now
in this matter are globally related to the regenera- Competing Interests  The authors declare no
tion of the corneal epithelium, increased efforts conflict of interest.
are thus required to accelerate strategies for
regenerating the corneal stroma and endothelium. Informed Consent  No human studies were car-
Tissue engineering of a corneal stromal equiva- ried out by the authors for this article.
lent appears actually the most challenging goal to
achieve in corneal regeneration, principally Animal Studies  No animal studies were carried
because it will require to identify or develop ade- out by the authors for this article.
quate scaffolds mimicking more closely the
extremely complex biochemical and morpho-
structural properties of the native stromal extra- References
cellular matrix and also to better define which is
the best type of SCs able to replace the lost or 1. Olson JL, Atala A, Yoo JJ.  Tissue engineering: cur-
diseased keratocytes. In that way, autologous rent strategies and future directions. Chonnam Med J.
2011;47(1):1–13. Epub 2011/11/24.
adult MSCs seem a very promising cellular phe- 2. Lin ZN, Chen J, Cui HP.  Characteristics of corneal
notype able to provide neo-keratocytes. As stated dystrophies: a review from clinical, histological and
above, regenerating the corneal endothelium with genetic perspectives. Int J Ophthalmol. 2016;9(6):904–
SCs of extracorneal origin is also a great priority. 13. Epub 2016/07/02.
3. Klintworth GK. Corneal dystrophies. Orphanet J Rare
The simple epithelium-like structure of the cor- Dis. 2009;4:7. Epub 2009/02/25.
neal endothelium makes of its tissue engineering 4. Ito K.  Metabolism and the control of cell fate deci-
with laminar scaffolds and SCs a goal less com- sions and stem cell renewal. Annu Rev Cell Dev Biol.
plex to achieve by comparison to the corneal 2016;32:399–409. Epub 2016/08/03.
5. Januschke J, Nathke I.  Stem cell decisions: a twist
stroma. Using extracorneal SCs or progenitors of fate or a niche market? Semin Cell Dev Biol.
from monolayered cells tissues, such as vascular 2014;34:116–23. Epub 2014/03/13.
endothelium progenitors or squamous epithelial 6. Hmadcha A, Dominguez-Bendala J, Wakeman J,
cells from the amniotic membrane or adult sim- Arredouani M, Soria B.  The immune boundaries for
stem cell based therapies: problems and prospective
ple epitheliums, is a promising approach to recre- solutions. J Cell Mol Med. 2009;13(8A):1464–75.
ate a tissue surrogate of the native corneal Epub 2009/07/09.
140 C. C. Lachaud et al.

7. Tweedell KS. The adaptability of somatic stem cells: in  vitro replication of the corneal epithelial stem
a review. J Stem Cells Regen Med. 2017;13(1):3–13. cell niche. Stem Cells. 2007;25(5):1145–55. Epub
Epub 2017/07/08. 2007/01/27.
8. Hombach-Klonisch S, Panigrahi S, Rashedi I, 20. Ueno H, Kurokawa MS, Kayama M, Homma R,
Seifert A, Alberti E, Pocar P, et al. Adult stem cells Kumagai Y, Masuda C, et  al. Experimental trans-
and their trans-differentiation potential--perspec- plantation of corneal epithelium-like cells induced
tives and therapeutic applications. J Mol Med (Berl). by Pax6 gene transfection of mouse embryonic
2008;86(12):1301–14. Epub 2008/07/17. stem cells. Cornea. 2007;26(10):1220–7. Epub
9. Jadalannagari S, Aljitawi OS. Ectodermal differenti- 2007/11/29.
ation of Wharton's jelly mesenchymal stem cells for 21. Kumagai Y, Kurokawa MS, Ueno H, Kayama
tissue engineering and regenerative medicine appli- M, Tsubota K, Nakatsuji N, et  al. Induction of
cations. Tissue Eng Part B Rev. 2015;21(3):314–22. corneal epithelium-like cells from cynomolgus
Epub 2014/12/18. monkey embryonic stem cells and their experi-
10. Pellegrini G, Rama P, Mavilio F, De Luca mental transplantation to damaged cornea. Cornea.
M. Epithelial stem cells in corneal regeneration and 2010;29(4):432–8. Epub 2010/02/19.
epidermal gene therapy. J Pathol. 2009;217(2):217– 22. Hanson C, Hardarson T, Ellerstrom C, Nordberg
28. Epub 2008/10/16 M, Caisander G, Rao M, et  al. Transplantation
11. Harkin DG, Foyn L, Bray LJ, Sutherland AJ, Li of human embryonic stem cells onto a partially
FJ, Cronin BG.  Concise reviews: can mesenchy- wounded human cornea in vitro. Acta Ophthalmol.
mal stromal cells differentiate into corneal cells? 2013;91(2):127–30. Epub 2012/01/28.
A systematic review of published data. Stem Cells. 23. Zhu J, Zhang K, Sun Y, Gao X, Li Y, Chen Z, et al.
2015;33(3):785–91. Epub 2014/11/18. Reconstruction of functional ocular surface by acel-
12. Yang X, Moldovan NI, Zhao Q, Mi S, Zhou Z, Chen lular porcine cornea matrix scaffold and limbal stem
D, et al. Reconstruction of damaged cornea by autol- cells derived from human embryonic stem cells.
ogous transplantation of epidermal adult stem cells. Tissue Eng Part A. 2013;19(21–22):2412–25. Epub
Mol Vis. 2008;14:1064–70. Epub 2008/06/17. 2013/05/17.
13. Blazejewska EA, Schlotzer-Schrehardt U, Zenkel 24. Zhang C, Du L, Pang K, Wu X.  Differentiation of
M, Bachmann B, Chankiewitz E, Jacobi C, et  al. human embryonic stem cells into corneal epithelial
Corneal limbal microenvironment can induce trans- progenitor cells under defined conditions. PLoS
differentiation of hair follicle stem cells into corneal One. 2017;12(8):e0183303. Epub 2017/08/17.
epithelial-like cells. Stem Cells. 2009;27(3):642–52. 25. Hayashi R, Ishikawa Y, Ito M, Kageyama T,
Epub 2008/12/17. Takashiba K, Fujioka T, et al. Generation of corneal
14. Inatomi T, Nakamura T, Koizumi N, Sotozono C, epithelial cells from induced pluripotent stem cells
Yokoi N, Kinoshita S. Midterm results on ocular sur- derived from human dermal fibroblast and corneal
face reconstruction using cultivated autologous oral limbal epithelium. PLoS One. 2012;7(9):e45435.
mucosal epithelial transplantation. Am J Ophthalmol. Epub 2012/10/03.
2006;141(2):267–75. Epub 2006/02/07. 26. Yu D, Chen M, Sun X, Ge J.  Differentiation of
15. Shao C, Fu Y, Lu W, Fan X. Bone marrow-derived mouse induced pluripotent stem cells into corneal
endothelial progenitor cells: a promising therapeu- epithelial-like cells. Cell Biol Int. 2013;37(1):87–94.
tic alternative for corneal endothelial dysfunction. Epub 2013/01/23.
Cells Tissues Organs. 2011;193(4):253–63. Epub 27. Mikhailova A, Ilmarinen T, Uusitalo H, Skottman
2010/10/22. H.  Small-molecule induction promotes cor-
16. Shao C, Chen J, Chen P, Zhu M, Yao Q, Gu P, et al. neal epithelial cell differentiation from human
Targeted transplantation of human umbilical cord induced pluripotent stem cells. Stem Cell Rep.
blood endothelial progenitor cells with immunomag- 2014;2(2):219–31. Epub 2014/02/15.
netic nanoparticles to repair corneal endothelium 28. Cieslar-Pobuda A, Rafat M, Knoflach V, Skonieczna
defect. Stem Cells Dev. 2015;24(6):756–67. Epub M, Hudecki A, Malecki A, et  al. Human induced
2014/10/16. pluripotent stem cell differentiation and direct
17. Yu L, Ge J, Wang Z, Huang B, Yu K, Long C, et al. transdifferentiation into corneal epithelial-like
The preliminary experimental study of induced cells. Oncotarget. 2016;7(27):42314–29. Epub
differentiation of embryonic stem cells into cor- 2016/06/09.
neal epithelial cells. Yan ke xue bao = Eye science. 29. Nakamura T, Endo K, Cooper LJ, Fullwood NJ,
2001;17(3):138–43. Epub 2003/02/06. Tanifuji N, Tsuzuki M, et al. The successful culture
18. Wang Z, Ge J, Huang B, Gao Q, Liu B, Wang L, et al. and autologous transplantation of rabbit oral muco-
Differentiation of embryonic stem cells into corneal sal epithelial cells on amniotic membrane. Invest
epithelium. Sci China C Life Sci. 2005;48(5):471– Ophthalmol Vis Sci. 2003;44(1):106–16. Epub
80. Epub 2005/12/01. 2002/12/31.
19. Ahmad S, Stewart R, Yung S, Kolli S, Armstrong L, 30. Nishida K, Yamato M, Hayashida Y, Watanabe K,
Stojkovic M, et al. Differentiation of human embry- Yamamoto K, Adachi E, et  al. Corneal reconstruc-
onic stem cells into corneal epithelial-like cells by tion with tissue-engineered cell sheets composed of
9  Corneal Regeneration: Use of Extracorneal Stem Cells 141

autologous oral mucosal epithelium. N Engl J Med. 42. Lan Y, Kodati S, Lee HS, Omoto M, Jin Y, Chauhan
2004;351(12):1187–96. Epub 2004/09/17. SK.  Kinetics and function of mesenchymal stem
31. Inatomi T, Nakamura T, Kojyo M, Koizumi N, cells in corneal injury. Invest Ophthalmol Vis Sci.
Sotozono C, Kinoshita S. Ocular surface reconstruc- 2012;53(7):3638–44. Epub 2012/05/09.
tion with combination of cultivated autologous oral 43. Pinarli FA, Okten G, Beden U, Fisgin T, Kefeli M,
mucosal epithelial transplantation and penetrating Kara N, et  al. Keratinocyte growth factor-2 and
keratoplasty. Am J Ophthalmol. 2006;142(5):757– autologous serum potentiate the regenerative effect
64. Epub 2006/09/23. of mesenchymal stem cells in cornea damage in
32. Ilmarinen T, Laine J, Juuti-Uusitalo K, Numminen J, rats. Int J Ophthalmol. 2014;7(2):211–9. Epub
Seppanen-Suuronen R, Uusitalo H, et al. Towards a 2014/05/03.
defined, serum- and feeder-free culture of stratified 44. Rohaina CM, Then KY, Ng AM, Wan Abdul Halim
human oral mucosal epithelium for ocular surface WH, Zahidin AZ, Saim A, et  al. Reconstruction
reconstruction. Acta Ophthalmol. 2013;91(8):744– of limbal stem cell deficient corneal surface
50. Epub 2012/09/12. with induced human bone marrow mesenchymal
33. Meyer-Blazejewska EA, Call MK, Yamanaka O, Liu stem cells on amniotic membrane. Transl Res.
H, Schlotzer-Schrehardt U, Kruse FE, et  al. From 2014;163(3):200–10. Epub 2013/11/30.
hair to cornea: toward the therapeutic use of hair 45. Holan V, Trosan P, Cejka C, Javorkova E, Zajicova
follicle-derived stem cells in the treatment of limbal A, Hermankova B, et al. A comparative study of the
stem cell deficiency. Stem Cells. 2011;29(1):57–66. therapeutic potential of mesenchymal stem cells and
Epub 2010/10/20. Limbal epithelial stem cells for ocular surface recon-
34. He YG, Alizadeh H, Kinoshita K, McCulley struction. Stem Cells Transl Med. 2015;4(9):1052–
JP. Experimental transplantation of cultured human 63. Epub 2015/07/18.
limbal and amniotic epithelial cells onto the cor- 46. Cejka C, Holan V, Trosan P, Zajicova A, Javorkova
neal surface. Cornea. 1999;18(5):570–9. Epub E, Cejkova J. The favorable effect of mesenchymal
1999/09/16. stem cell treatment on the antioxidant protective
35. Yao M, Chen J, Yang XX, Zhang XL, Ji QS, Zhou mechanism in the corneal epithelium and renewal of
Q, et al. Differentiation of human amniotic epithelial corneal optical properties changed after alkali burns.
cells into corneal epithelial-like cells in vitro. Int J Oxidative Med Cell Longev. 2016;2016:5843809.
Ophthalmol. 2013;6(5):564–72. Epub 2013/11/07. Epub 2016/04/09.
36. Zhou Q, Liu XY, Ruan YX, Wang L, Jiang MM, 47. Di G, Du X, Qi X, Zhao X, Duan H, Li S, et  al.
Wu J, et  al. Construction of corneal epithelium Mesenchymal stem cells promote diabetic corneal
with human amniotic epithelial cells and repair epithelial wound healing through TSG-6-dependent
of limbal deficiency in rabbit models. Hum Cell. stem cell activation and macrophage switch. Invest
2015;28(1):22–36. Epub 2014/08/20. Ophthalmol Vis Sci. 2017;58(10):4344–54. Epub
37. Huang L, Wong YP, Gu H, Cai YJ, Ho Y, Wang CC, 2017/08/16.
et  al. Stem cell-like properties of human umbilical 48. Martinez-Conesa EM, Espel E, Reina M, Casaroli-­
cord lining epithelial cells and the potential for epi- Marano RP.  Characterization of ocular surface
dermal reconstitution. Cytotherapy. 2011;13(2):145– ­epithelial and progenitor cell markers in human adi-
55. Epub 2010/08/26. pose stromal cells derived from lipoaspirates. Invest
38. Reza HM, Ng BY, Gimeno FL, Phan TT, Ang Ophthalmol Vis Sci. 2012;53(1):513–20. Epub
LP. Umbilical cord lining stem cells as a novel and 2011/12/27.
promising source for ocular surface regeneration. 49. Nieto-Miguel T, Galindo S, Reinoso R, Corell A,
Stem Cell Rev. 2011;7(4):935–47. Epub 2011/03/25. Martino M, Perez-Simon JA, et  al. In vitro simu-
39. Ma Y, Xu Y, Xiao Z, Yang W, Zhang C, Song E, et al. lation of corneal epithelium microenvironment
Reconstruction of chemically burned rat corneal sur- induces a corneal epithelial-like cell phenotype from
face by bone marrow-derived human mesenchymal human adipose tissue mesenchymal stem cells. Curr
stem cells. Stem Cells. 2006;24(2):315–21. Epub Eye Res. 2013;38(9):933–44. Epub 2013/06/19.
2005/08/20. 50. Galindo S, Herreras JM, Lopez-Paniagua M, Rey E,
40. Jiang TS, Cai L, Ji WY, Hui YN, Wang YS, Hu D, de la Mata A, Plata-Cordero M, et  al. Therapeutic
et al. Reconstruction of the corneal epithelium with effect of human adipose tissue-derived mesenchy-
induced marrow mesenchymal stem cells in rats. mal stem cells in experimental corneal failure due
Mol Vis. 2010;16:1304–16. Epub 2010/07/29. to Limbal stem cell niche damage. Stem Cells.
41. Roddy GW, Oh JY, Lee RH, Bartosh TJ, Ylostalo 2017;35(10):2160–74. Epub 2017/08/02.
J, Coble K, et  al. Action at a distance: systemi- 51. Garzon I, Martin-Piedra MA, Alfonso-Rodriguez
cally administered adult stem/progenitor cells C, Gonzalez-Andrades M, Carriel V, Martinez-­
(MSCs) reduce inflammatory damage to the cornea Gomez C, et al. Generation of a biomimetic human
without engraftment and primarily by secretion of artificial cornea model using Wharton's jelly mes-
TNF-alpha stimulated gene/protein 6. Stem Cells. enchymal stem cells. Invest Ophthalmol Vis Sci.
2011;29(10):1572–9. Epub 2011/08/13. 2014;55(7):4073–83. Epub 2014/06/08.
142 C. C. Lachaud et al.

52. Monteiro BG, Serafim RC, Melo GB, Silva MC, 65. Regauer S, Franke WW, Virtanen I.  Intermediate
Lizier NF, Maranduba CM, et  al. Human imma- filament cytoskeleton of amnion epithelium and cul-
ture dental pulp stem cells share key character- tured amnion epithelial cells: expression of epider-
istic features with limbal stem cells. Cell Prolif. mal cytokeratins in cells of a simple epithelium. J
2009;42(5):587–94. Epub 2009/07/21. Cell Biol. 1985;100(4):997–1009. Epub 1985/04/01.
53. Gomes JA, Geraldes Monteiro B, Melo GB, Smith 66. Crisan M, Yap S, Casteilla L, Chen CW, Corselli M,
RL, Cavenaghi Pereira da Silva M, Lizier NF, et al. Park TS, et al. A perivascular origin for mesenchy-
Corneal reconstruction with tissue-engineered mal stem cells in multiple human organs. Cell Stem
cell sheets composed of human immature den- Cell. 2008;3(3):301–13. Epub 2008/09/13
tal pulp stem cells. Invest Ophthalmol Vis Sci. 67. Oh M, Nor JE.  The perivascular niche and self-­
2010;51(3):1408–14. Epub 2009/11/07. renewal of stem cells. Front Physiol. 2015;6:367.
54. Zhang K, Pang K, Wu X.  Isolation and transplan- Epub 2015/12/24.
tation of corneal endothelial cell-like cells derived 68. da Silva ML, Chagastelles PC, Nardi
from in-vitro-differentiated human embryonic stem NB. Mesenchymal stem cells reside in virtually all
cells. Stem Cells Dev. 2014;23(12):1340–54. Epub post-natal organs and tissues. J Cell Sci. 2006;119(Pt
2014/02/07. 11):2204–13. Epub 2006/05/11.
55. Zhang C, Du L, Sun P, Shen L, Zhu J, Pang K, et al. 69. von Koskull H, Virtanen I.  Induction of cytokera-
Construction of tissue-engineered full-­ thickness tin expression in human mesenchymal cells. J Cell
cornea substitute using limbal epithelial cell-like Physiol. 1987;133(2):321–9. Epub 1987/11/01.
and corneal endothelial cell-like cells derived 70. Li R, Liang J, Ni S, Zhou T, Qing X, Li H, et  al.
from human embryonic stem cells. Biomaterials. A mesenchymal-to-epithelial transition initiates and
2017;124:180–94. Epub 2017/02/16. is required for the nuclear reprogramming of mouse
56. Takahashi K, Yamanaka S. Induction of pluripotent fibroblasts. Cell Stem Cell. 2010;7(1):51–63. Epub
stem cells from mouse embryonic and adult fibroblast 2010/07/14.
cultures by defined factors. Cell. 2006;126(4):663– 71. Trosan P, Javorkova E, Zajicova A, Hajkova M,
76. Epub 2006/08/15. Hermankova B, Kossl J, et  al. The supportive role
57. Cotsarelis G, Sun TT, Lavker RM.  Label-retaining of insulin-like growth factor-I in the differentiation
cells reside in the bulge area of pilosebaceous unit: of murine mesenchymal stem cells into corneal-like
implications for follicular stem cells, hair cycle, cells. Stem Cells Dev. 2016;25(11):874–81. Epub
and skin carcinogenesis. Cell. 1990;61(7):1329–37. 2016/04/07.
Epub 1990/06/29. 72. Sanchez-Abarca LI, Hernandez-Galilea E, Lorenzo
58. Taylor G, Lehrer MS, Jensen PJ, Sun TT, Lavker R, Herrero C, Velasco A, Carrancio S, et al. Human
RM.  Involvement of follicular stem cells in form- bone marrow stromal cells differentiate into corneal
ing not only the follicle but also the epidermis. Cell. tissue and prevent ocular graft-versus-host disease in
2000;102(4):451–61. Epub 2000/08/31. mice. Cell Transplant. 2015;24(12):2423–33. Epub
59. Ito M, Liu Y, Yang Z, Nguyen J, Liang F, Morris RJ, 2015/02/20.
et  al. Stem cells in the hair follicle bulge contrib- 73. Wen L, Zhu M, Madigan MC, You J, King NJ,
ute to wound repair but not to homeostasis of the Billson FA, et  al. Immunomodulatory effects of
epidermis. Nat Med. 2005;11(12):1351–4. Epub bone marrow-derived mesenchymal stem cells on
2005/11/17. pro-inflammatory cytokine-stimulated human cor-
60. Ohyama M, Terunuma A, Tock CL, Radonovich neal epithelial cells. PLoS One. 2014;9(7):e101841.
MF, Pise-Masison CA, Hopping SB, et  al. Epub 2014/07/09.
Characterization and isolation of stem cell-enriched 74. Zeppieri M, Salvetat ML, Beltrami A, Cesselli D,
human hair follicle bulge cells. J Clin Invest. Russo R, Alcalde I, et al. Adipose derived stem cells
2006;116(1):249–60. Epub 2006/01/06. for corneal wound healing after laser induced cor-
61. Rheinwald JG, Green H. Serial cultivation of strains neal lesions in mice. J Clin Med. 2017;6(12). Epub
of human epidermal keratinocytes: the formation 2017/12/06.
of keratinizing colonies from single cells. Cell. 75. Davies JE, Walker JT, Keating A.  Concise review:
1975;6(3):331–43. Epub 1975/11/01. Wharton's jelly: the rich, but enigmatic, source of
62. Grafting of burns with cultured epithelium pre- mesenchymal stromal cells. Stem Cells Transl Med.
pared from autologous epidermal cells. Lancet. 2017;6(7):1620–30. Epub 2017/05/11.
1981;1(8211):75–8. Epub 1981/01/10. 76. Nam SM, Maeng YS, Kim EK, Seo KY, Lew H. Ex
63. Yang X, Qu L, Wang X, Zhao M, Li W, Hua J, vivo expansion of human Limbal epithelial cells
et  al. Plasticity of epidermal adult stem cells using human placenta-derived and umbilical cord-­
derived from adult goat ear skin. Mol Reprod Dev. derived mesenchymal stem cells. Stem Cells Int.
2007;74(3):386–96. Epub 2006/09/26. 2017;2017:4206187. Epub 2017/09/13.
64. Liu J, Song G, Wang Z, Huang B, Gao Q, Liu B, 77. Xiao L, Nasu M.  From regenerative dentistry to
et al. Establishment of a corneal epithelial cell line regenerative medicine: progress, challenges, and
spontaneously derived from human limbal cells. Exp potential applications of oral stem cells. Stem Cells
Eye Res. 2007;84(3):599–609. Epub 2007/01/16. Cloning. 2014;7:89–99. Epub 2014/12/17
9  Corneal Regeneration: Use of Extracorneal Stem Cells 143

78. Gorski B. Gingiva as a new and the most accessible on the activation of keratocytes. Br J Ophthalmol.
source of mesenchymal stem cells from the oral 2017;101(11):1583–90. Epub 2017/08/28.
cavity to be used in regenerative therapies. Postepy 91. Yun YI, Park SY, Lee HJ, Ko JH, Kim MK, Wee
Hig Med Dosw (Online). 2016;70(0):858–71. Epub WR, et  al. Comparison of the anti-inflammatory
2016/09/07. effects of induced pluripotent stem cell-derived and
79. Tsai CL, Chuang PC, Kuo HK, Chen YH, Su WH, Wu bone marrow-derived mesenchymal stromal cells
PC. Differentiation of stem cells from human exfoli- in a murine model of corneal injury. Cytotherapy.
ated deciduous teeth toward a phenotype of corneal 2017;19(1):28–35. Epub 2016/11/15.
epithelium in  vitro. Cornea. 2015;34(11):1471–7. 92. Arnalich-Montiel F, Pastor S, Blazquez-Martinez
Epub 2015/07/15. A, Fernandez-Delgado J, Nistal M, Alio JL,
80. Lwigale PY, Cressy PA, Bronner-Fraser M. Corneal et  al. Adipose-derived stem cells are a source for
keratocytes retain neural crest progenitor cell cell therapy of the corneal stroma. Stem Cells.
properties. Dev Biol. 2005;288(1):284–93. Epub 2008;26(2):570–9. Epub 2007/12/11.
2005/11/03. 93. Du Y, Roh DS, Funderburgh ML, Mann MM,
81. Chan AA, Hertsenberg AJ, Funderburgh ML, Mann Marra KG, Rubin JP, et  al. Adipose-derived stem
MM, Du Y, Davoli KA, et  al. Differentiation of cells differentiate to keratocytes in  vitro. Mol Vis.
human embryonic stem cells into cells with corneal 2010;16:2680–9. Epub 2010/12/24.
keratocyte phenotype. PLoS One. 2013;8(2):e56831. 94. Ma XY, Bao HJ, Cui L, Zou J.  The graft of
Epub 2013/02/26. autologous adipose-derived stem cells in the cor-
82. Hertsenberg AJ, Funderburgh JL.  Generation of neal stromal after mechanic damage. PLoS One.
corneal Keratocytes from human embryonic stem 2013;8(10):e76103. Epub 2013/10/08.
cells. Methods Mol Biol. 2016;1341:285–94. Epub 95. Zhang S, Espandar L, Imhof KM, Bunnell
2015/06/01. BA. Differentiation of human adipose-derived stem
83. Naylor RW, McGhee CN, Cowan CA, Davidson cells along the keratocyte lineage In vitro. J Clin Exp
AJ, Holm TM, Sherwin T.  Derivation of corneal Ophthalmol. 2013;4(270). Epub 2013/08/13.
Keratocyte-like cells from human induced pluripo- 96. Alio del Barrio JL, Chiesa M, Gallego Ferrer G,
tent stem cells. PLoS One. 2016;11(10):e0165464. Garagorri N, Briz N, Fernandez-Delgado J, et  al.
Epub 2016/10/30. Biointegration of corneal macroporous membranes
84. Syed-Picard FN, Du Y, Lathrop KL, Mann MM, based on poly(ethyl acrylate) copolymers in an
Funderburgh ML, Funderburgh JL.  Dental pulp experimental animal model. J Biomed Mater Res A.
stem cells: a new cellular resource for corneal 2015;103(3):1106–18. Epub 2014/06/10.
stromal regeneration. Stem Cells Transl Med. 97. Alio del Barrio JL, Chiesa M, Garagorri N, Garcia-­
2015;4(3):276–85. Epub 2015/02/26. Urquia N, Fernandez-Delgado J, Bataille L, et  al.
85. Liu H, Zhang J, Liu CY, Wang IJ, Sieber M, Chang Acellular human corneal matrix sheets seeded
J, et al. Cell therapy of congenital corneal diseases with human adipose-derived mesenchymal stem
with umbilical mesenchymal stem cells: lumican cells integrate functionally in an experimental ani-
null mice. PLoS One. 2010;5(5):e10707. Epub mal model. Exp Eye Res. 2015;132:91–100. Epub
2010/05/27. 2015/01/28.
86. Coulson-Thomas VJ, Caterson B, Kao 98. Alio Del Barrio JL, El Zarif M, Azaar A, Makdissy
WW.  Transplantation of human umbilical mes- N, Khalil C, Harb W, et al. Corneal stroma enhance-
enchymal stem cells cures the corneal defects of ment with Decellularized stromal laminas with or
mucopolysaccharidosis VII mice. Stem Cells. without stem cell Recellularization for advanced
2013;31(10):2116–26. Epub 2013/07/31. keratoconus. Am J Ophthalmol. 2017.; Epub
87. Liu H, Zhang J, Liu CY, Hayashi Y, Kao WW. Bone 2017/11/07.
marrow mesenchymal stem cells can differenti- 99. Lynch AP, Ahearne M.  Retinoic acid enhances
ate and assume corneal keratocyte phenotype. the differentiation of adipose-derived stem cells
J Cell Mol Med. 2012;16(5):1114–24. Epub to Keratocytes in  vitro. Transl Vis Sci Technol.
2011/09/03. 2017;6(1):6. Epub 2017/02/01.
88. Park SH, Kim KW, Chun YS, Kim JC. Human mes-
100. Corpuz LM, Funderburgh JL, Funderburgh ML,
enchymal stem cells differentiate into keratocyte-­ Bottomley GS, Prakash S, Conrad GW.  Molecular
like cells in keratocyte-conditioned medium. Exp cloning and tissue distribution of keratocan. Bovine
Eye Res. 2012;101:16–26. Epub 2012/06/12. corneal keratan sulfate proteoglycan 37A.  J Biol
89. Demirayak B, Yuksel N, Celik OS, Subasi C, Chem. 1996;271(16):9759–63. Epub 1996/04/19.
Duruksu G, Unal ZS, et al. Effect of bone marrow
101. Choong PF, Mok PL, Cheong SK, Then
and adipose tissue-derived mesenchymal stem cells KY.  Mesenchymal stromal cell-like characteristics
on the natural course of corneal scarring after pene- of corneal keratocytes. Cytotherapy. 2007;9(3):252–
trating injury. Exp Eye Res. 2016;151:227–35. Epub 8. Epub 2007/04/28.
2016/08/29.
102. McCabe KL, Kunzevitzky NJ, Chiswell BP, Xia
90. Jiang Z, Liu G, Meng F, Wang W, Hao P, Xiang Y, X, Goldberg JL, Lanza R.  Efficient generation of
et  al. Paracrine effects of mesenchymal stem cells human embryonic stem cell-derived corneal endo-
144 C. C. Lachaud et al.

thelial cells by directed differentiation. PLoS One. chymal stem cells to heal damaged corneal endothe-
2015;10(12):e0145266. Epub 2015/12/23. lium. Mol Vis. 2012;18:547–64. Epub 2012/03/16.
103. Song Q, Yuan S, An Q, Chen Y, Mao FF, Liu Y, et al. 110. Nakahara M, Okumura N, Kay EP, Hagiya M,
Directed differentiation of human embryonic stem Imagawa K, Hosoda Y, et  al. Corneal endothelial
cells to corneal endothelial cell-like cells: a tran- expansion promoted by human bone marrow mes-
scriptomic analysis. Exp Eye Res. 2016;151:107– enchymal stem cell-derived conditioned medium.
14. Epub 2016/08/16. PLoS One. 2013;8(7):e69009. Epub 2013/07/31.
104. Chen P, Chen JZ, Shao CY, Li CY, Zhang YD, Lu 111. Lachaud CC, Soria F, Escacena N, Quesada-­
WJ, et al. Treatment with retinoic acid and lens epi- Hernandez E, Hmadcha A, Alio J, et al. Mesothelial
thelial cell-conditioned medium in  vitro directed cells: a cellular surrogate for tissue engineering of
the differentiation of pluripotent stem cells towards corneal endothelium. Invest Ophthalmol Vis Sci.
corneal endothelial cell-like cells. Exp Ther Med. 2014;55(9):5967–78. Epub 2014/08/21.
2015;9(2):351–60. Epub 2015/01/13. 112. Tuft SJ, Coster DJ.  The corneal endothelium. Eye
105. Zhao JJ, Afshari NA. Generation of human corneal (Lond). 1990;4(Pt 3):389–424. Epub 1990/01/01.
endothelial cells via in  vitro ocular lineage restric- 113. Mutsaers SE, Wilkosz S.  Structure and function of
tion of pluripotent stem cells. Invest Ophthalmol Vis mesothelial cells. Cancer Treat Res. 2007;134:1–19.
Sci. 2016;57(15):6878–84. Epub 2016/12/22. Epub 2007/07/18.
106. Ju C, Zhang K, Wu X. Derivation of corneal endothe- 114. Lachaud CC, Lopez-Beas J, Soria B, Hmadcha
lial cell-like cells from rat neural crest cells in vitro. A.  EGF-induced adipose tissue mesothelial cells
PLoS One. 2012;7(7):e42378. Epub 2012/08/04. undergo functional vascular smooth muscle dif-
107. Inagaki E, Hatou S, Higa K, Yoshida S, Shibata S, ferentiation. Cell Death Dis. 2014;5:e1304. Epub
Okano H, et al. Skin-derived precursors as a source 2014/06/27.
of progenitors for corneal endothelial regeneration. 115. Lansley SM, Searles RG, Hoi A, Thomas C, Moneta
Stem Cells Transl Med. 2017;6(3):788–98. Epub H, Herrick SE, et al. Mesothelial cell differentiation
2017/02/12. into osteoblast- and adipocyte-like cells. J Cell Mol
108. Shen L, Sun P, Zhang C, Yang L, Du L, Wu Med. 2011;15(10):2095–105. Epub 2010/11/13.
X. Therapy of corneal endothelial dysfunction with 116. Herrick SE, Mutsaers SE.  Mesothelial progeni-
corneal endothelial cell-like cells derived from skin-­ tor cells and their potential in tissue engineering.
derived precursors. Sci Rep. 2017;7(1):13400. Epub Int J Biochem Cell Biol. 2004;36(4):621–42. Epub
2017/10/19. 2004/03/11.
109. Joyce NC, Harris DL, Markov V, Zhang Z, Saitta
B. Potential of human umbilical cord blood mesen-
One Cell, Two Phenotypes:
Capturing Pluripotency
10
for Corneal Regeneration

Trevor Sherwin, Carol Ann Greene, Colin R. Green,


and Kushant R. Kapadia

10.1 Introduction slowly replaced by the keratocytes in later devel-


opment with type I collagen [6, 7], the most
A major goal for corneal restoration and regen- abundant collagen of the adult human cornea [8].
eration over the last decade has been the produc- The spacing of collagen fibrils and thickness of
tion of bioengineered corneas in  vitro for collagen lamellae and their arrangement within
transplant, with many research laboratories the stroma are critical to the passage of light and
investigating a variety of scaffold materials aimed thus the transparency of the cornea. Thus, post-­
at preserving the physiological and optical prop- partum, the extracellular matrix of the corneal
erties of the cornea [1]. However, advances in stroma is thought to turn over very slowly in
cell reprogramming [2, 3] and in gene therapy [4, order not to disturb the delicate balance of consti-
5] have made the possibility of in  vivo corneal tution and spacing of the lamellae and fibrils. The
engineering a distinct possibility. Here we keratocyte thus has long been thought of as a
describe the methods by which we have been largely quiescent cell residing in the stroma ready
able to induce in vivo keratocytes to produce pro- to jump into action if needed, such as following
teins normally associated with other cell pheno- trauma to the cornea [9].
types, which may prove to be simple and effective
examples of in vivo corneal engineering.
10.3 Keratocyte Activation
(Fig. 10.1)
10.2 Homoeostasis in the Stroma
Quiescent keratocytes close to a traumatic
The corneal stroma and associated keratocytes wound soon become activated and acquire the
arise from the neural crest during embryonic cellular properties necessary to deal with the
development. The stroma of the early cornea is wound [10]. As the cells directly traumatized by
largely composed of type II collagen which is the wound and those in close association will die
soon after the damage, there is a need to derive
T. Sherwin (*) · C. A. Greene · C. R. Green new cells and for them to be able to migrate into
K. R. Kapadia the wound site. Thus, upon activation following
Department of Ophthalmology, New Zealand trauma, the previously quiescent keratocyte
National Eye Centre, Faculty of Medical and Health acquires the fibroblastic phenotype able to
Sciences, The University of Auckland,
Auckland, New Zealand undergo rapid cell division and the mechanistic
e-mail: t.sherwin@auckland.ac.nz molecules that promote migration. Upon arrival

© Springer Nature Switzerland AG 2019 145


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_10
146 T. Sherwin et al.

a b

Fig. 10.1  Cultured human keratocytes display typical accentuates the difference in cell size between these two
stellate morphology of compact cell body with long very differentiated keratocyte lineages (b). The myofibroblast
fine dendritic extensions that interconnect between kera- can be molecularly distinguished from the surrounding
tocytes to form a syncytium (a). Upon wounding stimuli, keratocytes by using smooth muscle actin staining to
keratocytes differentiate into fibroblasts with the more reveal the stress fibres (red) that underlie the contractile
spindle-shaped cell bodies revealed by tubulin staining. A nature of the myofibroblast. Cell nuclei are stained with
large myofibroblast lies amongst the fibroblasts and DAPI (blue). Scale bar = 50 μ

at the wound edge, these fibroblastic cells The fate of the myofibroblast is unclear fol-
remodel their actin cytoskeletons and become lowing wound closure. Evidence suggests that
the giant myofibroblast with prominent stress they disappear; however, whether this is through
fibres that are capable of straddling and contract- targeted apoptosis or by dedifferentiation still
ing the wound edges. The fibroblasts also down- remains to be elucidated [12].
regulate the recognized proteins of the
differentiated keratocyte and start expressing the
enzymatic proteinases required to remodel the 10.4 Keratocyte Differentiation
damaged collagen fibres and lamellae [11, 12].
They also lay down type III collagen which can Given the keratocyte activation following trauma,
be responsible for scarring and, as a conse- it is strange to think of the keratocyte as a quies-
quence, visual defects [13, 14]. cent cell when it is capable of such responses [9].
10  One Cell, Two Phenotypes: Capturing Pluripotency for Corneal Regeneration 147

Furthermore, it follows that we should actually transcription factors by genetic manipulation [19]
think of the keratocyte as a cell of greater potency but also subsequently by using small molecule
and differentiation potential than previously chemical stimuli [20]. Thus, it is not beyond the
thought [15]. The archaic view of the terminally realms of belief to imagine that keratocyte cells
differentiated cell being the dead end of a one-­way may well be inducible into other phenotypes
street, with the cell neither being able to transdif- should the right stimulus present itself.
ferentiate into another cell type or to dedifferenti-
ate into an earlier stage, has been shown to be
misguided [16]. For instance, the now well- 10.5 Untapped Potential
accepted epithelial to mesenchymal transition pro- of Keratocytes (Figs. 10.2
cesses [17] that occur in corneal epithelial and 10.3)
development, fibrosis and pathogenesis show that
transdifferentiation occurs in the cornea [18]. The Serendipitously, we were able to show that kera-
potential of cells to dedifferentiate into much ear- tocytes possessed greater differentiation potential
lier and much more potent stages has also been than previously thought [21]. In a series of exper-
demonstrated elegantly by the production of iments where we isolated keratocytes and induced
induced pluripotent stem cells by the induction of them to form spheres (one parameter indicative

a b

c d

Fig. 10.2  Neuronal reprogramming of corneal kerato- with corneal fibroblasts revealed that they too are amena-
cytes and fibroblasts. Keratocytes within human corneal ble to neuronal reprogramming, expressing nestin (c) and
slices cultured in neuronal reprogramming medium map-2 (d) in cell culture. These results suggest that the
expressed the early neuronal marker, nestin (a), within corneal fibroblast may not be a terminally differentiated
3  days in culture, and by day 8, the cells expressed the cell type as once thought
mature neuronal marker, map-2 (b). Similar experiments
148 T. Sherwin et al.

Fig. 10.3 Chondrogenic
a b
reprogramming of
keratocytes. Keratocytes
cultured for 3 weeks in
chondrogenic
differentiation medium
containing TGFβ3 and
dexamethasone formed
spheres (b) which
labelled for collagen
type II. Keratocytes in
the control medium
failed to form spheres
(a). Human corneal
slices (d) cultured for
2 weeks in the same
chondrogenic
reprogramming medium
(with control medium-­ c d
treated cornea shown in
(c)) and rat corneas
treated in vivo (f) (with
control medium-treated
cornea shown in (e)) for
at least 2 weeks also
labelled positively for
collagen type II

e f

of stem or progenitor potential) by growing them Initially we planned to use this observation to
in neuronal-specific media, we were able to show return to cultured slices of corneal tissue and
the production of cells with neuronal-specific induce the neuronal phenotype with the aim of
protein markers [21] (Fig. 10.2). showing definitively that the cells capable of
10  One Cell, Two Phenotypes: Capturing Pluripotency for Corneal Regeneration 149

p­ roducing neuronal markers would be located in We identified that the TGFβ family were pow-
the stroma underlying the limbus. erful promoters of chondrogenic differentiation
However, we were surprised to find that upon and also stimulators of collagen production [23].
culture of corneal slices in our neuronal-specific Noting that TGFβ1 and TGFβ2 are associated
media, all of the keratocytes were capable of with matrix production aligned with scarring, we
expressing the neuronal proteins. Furthermore, if tested TGFβ3 as it acts without instigating fibro-
we continued culturing the corneal slices under sis or scar tissue formation [24, 25]. We tested
our specific conditions, the reprogrammed kerato- TGFβ3 coupled with the synthetic steroid dexa-
cytes expressed early, intermediate and mature methasone, a combination which has been shown
neuronal markers in the same timeframe and to be effective in promoting chondrogenesis [26],
sequence as occurs during the differentiation of and found that this combination of factors pro-
immature neuronal cells to mature neuronal cells. duced strongly positive and consistently repro-
Concommitant with this change in protein expres- ducible results for matrix production.
sion, we also saw a change in morphology from
keratocyte to neural phenotype. We were able to
reproduce this effect in both rodent and human tis- 10.7 Potential Therapeutic
sues, including in vivo treatments in rodents with a for the Treatment
simple eye drop delivery of the reprogramming of Keratoconus
factors, but also show that the molecules that
induced the neuronal switch were species specific After successfully inducing the production of the
[21]. These findings enabled us to propose that the developmental type II collagen in isolated kerato-
keratocyte cell retains a memory of its neural crest cytes (Fig. 10.2), we investigated the potential (if
origins that can be tapped into for therapeutic pur- translatable to keratocytes within the stroma) for
poses. We further tested this hypothesis to deter- retarding not only the loss of stroma in kerato-
mine if other cells that descend from the neural conic corneas but also the possibility of restoring
crest also retain some memory of their lineage, some lost matrix and increasing the thickness of
and we were able to show that neural crest-derived the corneal stroma in this ectatic disease.
cartilaginous cells were also capable of switching Furthermore, we hypothesized that if we could
on neural protein expression [22]. stimulate the production of type II collagen
within a keratoconic cornea, then this would
present an opportunity to also reshape the cornea
10.6 F
 urther Untapped Potential at the same time as strengthening it to restore
of Keratocytes both integrity and optical functionality. This has
the potential to become the first regenerative
This serendipitous finding of a neural crest mem- treatment for keratoconus. To this aim we started
ory in keratocytes led us to hypothesize that we to ask a series of experimental questions [27].
could turn on other molecules that are expressed
by neural crest cells during the developing cor-
nea. We aligned this hypothesis with the observa- 10.7.1 Can We Induce the Production
tion that type II collagen is expressed in the of the Developmental Type II
developing cornea but is later replaced by type I Collagen in Human Tissue?
collagen [6–8]. Knowing also that certain carti-
laginous tissues are derived from the neural crest Following initial promising studies on isolated
and that type II collagen is a cartilage-specific keratocytes, we used our combination of TGFβ3
collagen, we applied molecules used to induce and dexamethasone treatment to see if we could
cartilage cell differentiation to our keratocytes to induce collagen II production in keratocytes in
assess whether the keratocytes would be capable situ by incubating slices of human corneal tissue
of once again turning on type II collagen. in an organotypic system. Production of collagen
150 T. Sherwin et al.

II was not evident in either the control or the keratoconic tissue. Such production of a corneal
treated corneas after 1  week of incubation. developmental collagen may return the cornea to
However, slices incubated in the induction a less dystrophic stage and restore the integrity of
medium for 3 weeks showed the de novo appear- the collagen lamellae. Such a treatment may rep-
ance of collagen II, while there was no corre- resent the first restorative therapeutic treatment
sponding labelling of collagen II apparent in the for keratoconus. However, the possibility
control samples. It was important to note that the remained that the keratocytes within keratoconic
deposition of type II collagen that we observed tissue might already be significantly affected by
was uniform and evenly distributed along the the disease process and could no longer be
existing collagen lamellae. No clumping of pro- induced to produce the developmental collagen II.
tein deposition was seen, and no disruption of the We examined this by using human kerato-
lamellar spacing was observed (Fig. 10.3). conic buttons obtained from corneal transplant
Furthermore, no labelling for collagen III surgery. Corneal buttons were divided into half
(associated with scar formation) or αSMA (asso- and placed into either collagen II induction
ciated with fibrosis) was observed in either the medium or into control medium. For compari-
control or treated corneas, supporting the use of son, normal tissue was also induced. Again,
TGFβ3 as a non-scar-producing factor. after 3 weeks, collagen II production had been
switched on in the treated corneas despite the
presence of keratoconus. The number of kerato-
10.7.2 Can We Induce Collagen II cytes was much lower than in comparison to
Production in Keratoconic normal tissue, so it was not surprising that the
Tissue? labelling for collagen II was lower than that
observed in normal corneas under induction
Having observed the induction of collagen II in conditions. Nonetheless, significant deposition
situ in human corneal slices under organotypic occurred with the collagen II in keratoconic cor-
conditions, we investigated the use of this tech- neas evenly distributed along lamellae and with-
nology to potentially restore matrix production in out clumping (Fig. 10.4).

a b

Fig. 10.4  Collagen type II induction in keratoconic cor- clear and strong labelling for collagen type II (green)
neal tissue. Slices of keratoconic cornea were cultured for within the stromal matrix of corneal tissue cultured in the
3 weeks in the control medium (a) and chondrogenic chondrogenic medium
medium (b) and labelled for collagen type II. There was
10  One Cell, Two Phenotypes: Capturing Pluripotency for Corneal Regeneration 151

10.7.3 Does Collagen II Production Elasticity


0.001
Alter the Biomechanical Untreated cornea
Properties of Dystrophic 0.0009
Treated cornea
Corneas? 0.0008
0.0007

ER (Gpa)
We were fortunate to obtain a pair of keratoglo- 0.0006
bus corneas that were initially donated for trans- 0.0005
plant but were unable to be used for this purpose 0.0004
due to the presence of keratoglobus. Treatment of
0.0003
these corneas for extended periods in our induc-
0.0002
tion media was followed by the measurement of
0.0001
biomechanical properties in our collaborator’s
laboratory. Subsequent tests of both the elasticity 0
Rat cornea (after Keratoglobus
modulus and the hardness of the corneas follow- in vivo cornea (after ex
ing type II collagen induction showed a signifi- treatment) vivo treatment)
cant increase in both parameters. This showed
that the deposition of type II collagen was Hardness
0.0005
­associated with a recovery of corneal elasticity, 0.00045
Untreated cornea
that is, the ability to restore shape after deforma- Treated cornea
0.0004
tion and, in corneal stiffness, the ability to resist
0.00035
deformation (Fig. 10.5).
H (Gpa)

0.0003
0.00025

10.7.4 Can We Induce the Production 0.0002

of the Developmental Type II 0.00015


Collagen in an In Vivo Model? 0.0001
0.00005
Having shown that the use of TGFΒ3 and dexa- 0
Rat cornea (after Keratoglobus
methasone was capable of inducing the production in vivo cornea (after ex
of collagen II in both normal human and kerato- treatment) vivo treatment)
conic corneas, we realized that to advance the
technology we needed to transfer the treatment Fig. 10.5  Results from ex  vivo corneal biomechanical
testing via nanoindentation. Elastic modulus (Er) and
into animal models that could be tested in vivo. hardness (H) measures of treated and control rat corneas
Experimentation using the collagen II induc- treated in  vivo for 3  weeks reveal a marked increase in
tion media on rat tissue in vitro showed exactly both parameters in the treated corneas. Similarly, in a pair
the same results as those observed for human tis- of ex  vivo treated and untreated keratoglobus corneas,
there is a clear increase in elastic modulus (Er) and hard-
sue in vitro. ness (H) in the treated cornea
Thus, after obtaining ethical approval for
in vivo experimentation, we were able to treat rat
corneas once daily with our TGFβ3 and dexa- c­ ornea but in close association with the existing
methasone in an eye drop formulation (a tenfold lamellae and without observed clumping.
concentration of the factors was used to compen- Slit lamp examination of the cornea revealed
sate for washout by the tear film and the reduced normal stromal appearance without any hint of
time of contact). stromal opacity, and fundoscopy revealed easy
Production of collagen II in the in vivo rat cor- imaging of the optic nerve and retina to illustrate
neas followed the same pattern as seen in human the clarity of the tissue in front of the fundus,
corneas ex vivo when examined by immunohis- including the cornea.
tochemistry, with collagen II being deposited The rat corneas were also investigated for any
throughout the depth and the breadth of the changes in biomechanical properties. Once again,
152 T. Sherwin et al.

a 3-week treatment with the collagen II induction a 3-week treatment period, while two further
factors produced a significantly higher elasticity cohorts of sheep were treated for 3 weeks or
modulus and corneal stiffness when compared to 6  weeks and then grazed out for a further
control corneas. Interestingly, in vivo rat corneas 6  months prior to collection. Similarly to the
treated for only 1 week with the induction factors human and rat studies, the sheep corneas showed
did not show increased biomechanical properties. deposition of collagen II in only the treated group
This aligns nicely with our immunohistochemis- with the collagen deposition being uniformly
try data which shows that collagen II deposition even along the pre-existing lamellae.
is not apparent until week 2. To assess changes in the biomechanical prop-
erties of treated sheep corneas, sheep eyes were
collected from controls, from 3-week-treated ani-
10.7.5 Can We Switch Off Collagen II mals and from 3- or 6-week-treated animals left
Production? for by 6 months with no treatment prior to collec-
tion. The intraocular pressure was raised to
We also processed tissue from treated and con- 20  mmHg in all eyes to normalize the pressure
trol corneas from the in vivo rat experiments for reading; then the use of a Tono-Pen was engaged
PCR analysis. We looked at the levels of colla- to determine changes in corneal resistance. The
gen II mRNA after induction in treated and con- Tono-Pen normally records intraocular pressure,
trol corneas for a 1-week and a 3-week treatment but the readings are confounded in part by corneal
period. Following treatment, we withdrew the biomechanics. In our study, the intraocular pres-
factors from the culture medium and cultured for sure was set as a constant, and changes in read-
a further 4 weeks. For comparison (as an internal ings therefore solely reflect changes in corneal
control), we also examined the mRNA levels of biomechanics. In all 11 treated eyes tested, the
collagen I. resistance was greater than non-treated eyes by an
Interestingly, after 1  week of treatment, the average of 23%. Interestingly the increased resis-
mRNA for collagen II was maximally induced tance of treated sheep corneas was still present
which coincides with the protein deposition mea- after a 6-month lay-off period, indicating that the
surable by week 2 after treatment. Upon withdrawal treatment has a lasting effect of at least 6 months.
of treatment for 4 weeks, the levels of collagen II
mRNA were returning to baseline indicating that
the induction of collagen II production is controlla- 10.7.7 Is It Toxic?
ble. Interestingly in our treated samples, collagen I
production is also upregulated at week one and may Full systemic toxicity testing has been analysed
be a helpful byproduct of the treatment. for the sheep, and no evidence of toxicity has
been found in any of the body fluids or tissue.
Analysis of the corneal and retinal tissue post-
10.7.6 Is the Technology Transferable treatment also showed no sign of inflammation of
to a Large Animal Model? any tissues of the eye following treatment [28].

Next, we asked if the technology was transferable


to a large animal model as we wanted to know if 10.7.8 Can We Reshape the Cornea?
we could induce collagen II production in an
in vivo cornea of similar size and anatomy to that We have turned our attention to whether our col-
of a human. We also wanted to examine the lagen II protocol could be used to reshape the
­longevity of the treatment effect. We commenced cornea. We envisage the treatment being applied
sheep models where we treated one eye for 3 in combination with a reshaping contact lens or
weeks with our TGFβ3 and dexamethasone for- stromal inserts which could hold the cornea in a
mulation and used the other eye as the control. new conformation while the collagen II deposi-
The sheep tissue was then collected directly after tion occurs. With the increased corneal stiffness
10  One Cell, Two Phenotypes: Capturing Pluripotency for Corneal Regeneration 153

and increased elasticity of the corneas following Compliance with Ethical Requirements Carol Greene,
Colin Green and Trevor Sherwin are coinventors listed on
our treatment, we hypothesize that we may be patent PCT/NZ2016/050033, application date: 05 Mar
able to hold the newly reshaped cornea for an 2015. Kushant Kapadia has no conflict of interest to declare.
extended period after treatment. This would, All procedures followed were in accordance with the ethi-
therefore, not only see this therapeutic as being cal standards of the responsible committee on human
experimentation (institutional and national) and with the
the only restorative treatment for keratoconus in Helsinki Declaration of 1975, as revised in 2000. Informed
terms of regenerating the corneal stroma but may consent was obtained from all patients for being included in
also provide a mechanism to restore some visual the study. All institutional and national guidelines for the
acuity or function. care and use of laboratory animals were followed.
To this aim we have performed some prelimi-
nary experiments in the sheep model where we
have inserted compression sutures in the cornea References
to induce a classical bowtie astigmatism, then
1. Chen Z, You J, Liu X, Cooper S, Hodge C, Sutton
treated with our TGFβ3 and dexamethasone for- G, Crook JM, Wallace GG. Biomaterials for corneal
mulation. The hypothesis being that the induced bioengineering. Biomed Mater. Accepted Manuscript
astigmatism would remain for an extended period online 12 October 2017.
after the removal of the compression suture. Pilot 2. Sasamoto Y, Hayashi R, Park SJ, Saito-Adachi
M, Suzuki Y, Kawasaki S, Quantock AJ, Nakai K,
experiments were highly encouraging in that Tsujikawa M, Nishida K.  PAX6 isoforms, along
even after only 1 week of compression suture and with reprogramming factors, differentially regu-
treatment a proportion of the induced astigma- late the induction of cornea-specific genes. Sci Rep.
tism (reshaping) was retained compared to the 2016;6:20807.
3. Amici AW, Onikoyi FO, Bonfanti P.  Lineage poten-
control eye which showed no retention of the tial, plasticity and environmental reprogramming of
induced shape change following a compression epithelial stem/progenitor cells. Biochem Soc Trans.
suture without treatment. 2014;42(3):637–44.
4. Solinis MA, del Pozo-Rodriguez A, Apaolaza PS,
Rodriguez-Gascon A.  Treatment of ocular disorders
by gene therapy. Eur J Pharm Biopharm. 2015;95(Pt
10.8 Conclusion: Restorative B):331–42.
Treatment for Keratoconus? 5. Farjadnia M, Naderan M, Mohammadpour M. Gene ther-
apy in keratoconus. Oman J Ophthalmol. 2015;8(1):3–8.
6. Hendrix MJ, Hay ED, von der Mark K, Linsenmayer
We have identified a combination drug therapy TF.  Immunohistochemical localization of collagen
that is capable of turning on the corneal develop- types I and II in the developing chick cornea and tibia
ment collagen II expression in living corneas by electron microscopy. Invest Ophthalmol Vis Sci.
from three species rat, sheep and human. 1982;22(3):359–75.
7. Toole BP. Transitions in extracellular macromolecules
Surprisingly, both components are used at a very during avian ocular development. Prog Clin Biol Res.
low concentration, 2000-fold less than for current 1982;82:17–34.
clinical ocular applications of dexamethasone 8. Marshall GE, Konstas AG, Lee WR. Immunogold fine
and around 0.5% of normal circulating blood lev- structural localization of extracellular matrix compo-
nents in aged human cornea. I.  Types I-IV collagen
els for TGFβ3. The collagen II is deposited in a and laminin. Graefes Arch Clin Exp Ophthalmol.
manner in which the transparency of the cornea is 1991;229(2):157–63.
maintained, while the cornea becomes stiffer but 9. Snyder MC, Bergmanson JP, Doughty MJ.
also more elastic. The potential of this drug com- Keratocytes: no more the quiet cells. J Am Optom
Assoc. 1998;69(3):180–7.
bination therapy is to provide a unique restorative 10. West-Mays JA, Dwivedi DJ. The keratocyte: corneal
drug therapy for keratoconus. However, as the stromal cell with variable repair phenotypes. Int J
drug therapy works in normal as well as Biochem Cell Biol. 2006;38(10):1625–31.
­dystrophic corneas, it could be used to stiffen 11. Torricelli AA, Santhanam A, Wu J, Singh V, Wilson
SE. The corneal fibrosis response to epithelial-stromal
induced reshaping of normal corneas and thus is injury. Exp Eye Res. 2016;142:110–8.
a potential therapeutic for myopia, a concept cur- 12. Ljubimov AV, Saghizadeh M.  Progress in corneal

rently being explored. wound healing. Prog Retin Eye Res. 2015;49:17–45.
154 T. Sherwin et al.

13. Cintron C, Hong BS, Covington HI, Macarak


Green CR.  Cells from the adult corneal stroma can
EJ.  Heterogeneity of collagens in rabbit cor- be reprogrammed to a neuron-like cell using exog-
nea: type III collagen. Invest Ophthalmol Vis Sci. enous growth factors. Exp Cell Res. 2014;322(1):
1988;29(5):767–75. 122–32.
14. Manificat H, Rovère MR, Galiacy SD, Malecaze F, 22. Greene CA, Green CR, Sherwin T. Transdifferentiation
Hulmes DJ, Moali C, Damour O.  Development of of chondrocytes into neuron-like cells induced by
ex vivo organ culture models to mimic human corneal neuronal lineage specifying growth factors. Cell Biol
scarring. Mol Vis. 2012;18:2896–908. Epub 2012 Int. 2015;39(2):185–91.
Dec 1. 23.
Quintana L, zur Nieden NI, Semino CE.
15. Kaltschmidt B; Kaltschmidt C, Widera D. Adult cra- Morphogenetic and regulatory mechanisms during
niofacial stem cells: sources and relation to the neural developmental chondrogenesis: new paradigms for
crest. Stem Cell Rev. 2012;8(3):658–71. cartilage tissue engineering. Tissue Eng Part B Rev.
16. Merrell AJ, Stanger BZ. Adult cell plasticity in vivo: 2009;15(1):29–41.
de-differentiation and transdifferentiation are back in 24. Shah M, Foreman DM, Ferguson MW. Neutralisation
style. Nat Rev Mol Cell Biol. 2016;17(7):413–25. of TGF-beta 1 and TGF-beta 2 or exogenous addition
17. Hay ED, Zuk A.  Transformations between epi-
of TGF-beta 3 to cutaneous rat wounds reduces scar-
thelium and mesenchyme: normal, pathological, ring. J Cell Sci. 1995;108(3):985–1002.
and experimentally induced. Am J Kidney Dis. 25.
Karamichos D, Hutcheon AEK, Zieske JD.
1995;26(4):678–90. Transforming growth factor-β3 regulates assembly of
18. Suzuki K, Yokoyama C, Higashi Y, Daikoku T,
a non-fibrotic matrix in a 3D corneal model. J Tissue
Mizoguchi S, Saika S, Wakayama YG.  Symposium: Eng Regen Med. 2011;5(8):e228.
epithelial-mesenchymal interaction regulates tissue 26. Wa Q, Gao M, Dai X, Yu T, Zhou Z, Xu D, Zou X.
formation and characteristics: insights for corneal Induction of chondrogenic differentiation of mouse
development. Ocul Surf. 2012;10(4):217–20. embryonic mesenchymal stem cells through an in vitro
19. Takahashi K, Yamanaka S.  Induction of pluripotent pellet model. Cell Biol Int. 2015;39(6):657–65.
stem cells from mouse embryonic and adult fibro- 27. Greene CA, Green CR, Dickinson ME, Johnson

blast cultures by defined factors. Cell. 2006;126: V, Sherwin T.  Keratocytes are induced to produce
663–76. collagen type II: a new strategy for in  vivo cor-
20. Lin T, Wu S.  Reprogramming with small molecules neal matrix regeneration. Exp Cell Res. 2016;347:
instead of exogenous transcription factors. Stem Cells 241–9.
Int. 2015;2015:794632. 28. Kapadia K.  Safety of a novel in  vivo corneal tissue
21. Greene CA, Chang C-Y, Fraser CJ, Nelidova DE, regeneration therapy for Keratoconus. Master of
Chen JA, Lim A, Brebner A, McGhee J, Sherwin T, Science in Biomedical Science Thesis; 2017.
Corneal Stem Cell-Based Therapies
11
Yuzuru Sasamoto, Yoshinori Oie,
and Kohji Nishida

11.1 C
 urrent Stem Cell-Based focus on the current and future stem cell-based
Therapy therapies for corneal epithelium and also refer to
future stem cell-based therapy for corneal stroma
Currently, stem cell-based therapy is the way to and endothelium.
restore a cornea with stem cell deficiency. Stem
cells are well investigated in the corneal epithe-
lium (we’re going to refer to stem cells in corneal 11.2 C
 urrent Stem Cell-Based
stroma and endothelium later). Corneal epithelial Therapy: Corneal Epithelium
stem cells reside in the basal layer of limbus and
are thus called limbal stem cells (LSCs) [1]. They There are various causes of LSCD, from con-
give rise to transient amplifying cells (TACs) and genital diseases (e.g., aniridia and multiple
provide terminally differentiated epithelial cells endocrine deficiency) to acquired diseases (e.g.,
by proliferation and migration toward the apical Stevens-­Johnson syndrome [SJS], ocular cicatri-
central cornea (X, Y, Z hypothesis) [2, 3]. They cial pemphigoid [OCP], chemical or thermal
play an important role in the maintenance and burns, and contact lens overwear). LSCD
wound healing of the corneal epithelium. Severe patients suffer from severe visual impairment
loss of LSCs leads the defect of corneal epithelial with conjunctival epithelial ingrowth, neovascu-
regeneration, and this situation is called limbal larization of corneal stroma, and corneal opacifi-
stem cell deficiency (LSCD). Other than artificial cation [4, 5]. Partial LSC abnormality is
cell-free devices such as Boston type I kerato- presumed to be one of the causes of pterygium, a
prostheses, transplantation of LSCs or their sub- fibrovascular tissue which grows over the limbus
stitute containing tissue is the only way to onto the cornea [6]. Patients with pterygium suf-
overcome the deficiency of stem cells. We mainly fer from impaired vision due to high astigma-
tism, altered tear film, binocular diplopia,
irritation, etc. Partial LSCD is also brought by a
surgical excision of limbus in patients with ocu-
Y. Sasamoto lar surface neoplasia.
Division of Genetics, Department of Medicine, The effort to treat LSCD with LSCs starts
Brigham and Women’s Hospital, Boston, MA, USA after the establishment of the concept of LSCs in
Y. Oie · K. Nishida (*) 1986 [1]. After the development of conjunctival
Department of Ophthalmology, Osaka University limbal autograft (CLAU), keratolimbal allograft
Graduate School of Medicine, Suita, Osaka, Japan (KLAL), living-related conjunctival limbal
e-mail: knishida@ophthal.med.osaka-u.ac.jp

© Springer Nature Switzerland AG 2019 155


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_11
156 Y. Sasamoto et al.

allograft (Lr-CLAL), cultivated limbal epithe- 11.2.1.2 Procedure


lial transplantation (CLET), and simple limbal
epithelial transplantation (SLET) were developed LSCD (Fig. 11.1)
one after the other, and cultivated oral mucosal Abnormal corneal epithelium and fibrovascular
epithelial transplantation (COMET) was devel- pannus are stripped by blunt dissection using for-
oped as a transplantation using alternative source ceps, spatula, and/or scarifier and polish the
of LSCs. exposed stromal surface to improve clarity. Two
grafts from limbus and adjacent conjunctiva are
excised from the contralateral eye. The grafts are
11.2.1 Conjunctival Limbal Autograft transferred to their corresponding sites in the
(CLAU) recipient eye and sutured at the both corneal and
conjunctival edge [8]. Simultaneous or subse-
Conjunctival autograft was performed for the quent penetrating keratoplasty (PK) or deep
treatment of LSCD in the early 1980s [7]. After lamellar keratoplasty should be considered when
the establishment of the concept of LSCs in 1986 the stromal scarring hampers improvement of
by Schermer et al. [1], Kenyon et al. introduced visual acuity.
CLAU, which contains limbal stem cells in the Combination of CLAU with AM transplanta-
graft [8]. This was the first stem cell-based ther- tion (AMT) may result in a higher success rate
apy for LSCD. because AM provide a basement membrane for
epithelial migration and healing and reduce
11.2.1.1 Indications inflammation, scarring, and vascularization [15].
Indication of CLAU includes corneal vascu- AM, with the basement membrane side up, is
larization, scarring, and long-term ocular sur- grafted onto the defect of the donor site and onto
face failures including epithelial defects, the recipient corneal and limbal sclera before
keratinization, symblepharon, inflammation transplantation of grafts.
and ocular pain, and reduced visual acuity in Traditionally, CLAU has involved the removal
patients with LSCD [8]. As CLAU uses the of two 60° large free grafts from the healthy fel-
contralateral uninjured limbal tissue as a graft, low eye [16]. There is a report suggesting only
its indication is limited to the unilateral
LSCD.  CLAU is also adopted as one of the
surgical treatments of pterygium because
destruction of LSCs resulting from chronic
exposure to ultraviolet radiation has been
thought to play a major role (besides CLAU,
bare sclera technique, free conjunctival auto-
graft, and amniotic membrane (AM) are the
other surgical options) [9–12]. Visually sig-
nificant induced astigmatism, threat of
involvement of the visual axis, severe symp-
toms of irritation and cosmesis are the surgical
indications of pterigium. CLAU is also used to
promote epithelialization and prevent fibro-
vascular proliferation and conjunctivalization
Fig. 11.1  Limbal stem cell deficiency. Limbal stem cell
of the cornea after excision of neoplasia when deficiency caused by Stevens-Johnson syndrome. Corneal
the ocular surface neoplasia extensively surface is totally covered by vascularized conjunctival
involves the limbal area [13, 14]. epithelium
11  Corneal Stem Cell-Based Therapies 157

resolution of chronic external ocular inflamma-


tion. After 6-month follow-up or more, visual
acuity was improved in 17 cases (81%), and sta-
ble epithelial adhesion without recurrent erosion
or persistent epithelial defect was recorded in 20
cases (95.2%). There was no immune rejection as
expected. They did not encounter infection, lim-
bal graft failure, and sloughing of the grafts.
Meallet MA et al. reported that CLAU com-
bined with AMT improved visual acuity, corneal
clarity, and successful ocular surface reconstruc-
tion in all five patients (mean follow-up period:
22  months) [15]. In the other study of CLAU
Fig. 11.2  Pterygium. Thickened vascularized pterygium
invaded into pupillary zone of patient’s cornea at 9 o’clock combined with AMT, the overall cumulative graft
position survival was obtained in eight of ten eyes (80%)
(mean follow-up period: 33  months) [28].
However, comparative study showed no differ-
one 60° graft together with AMT and fibrin glue ence in visual acuity and ocular surface recon-
successfully reconstructed a corneal surface [17], struction between CLAU with AMT and without
but it’s still controversial [18, 19]. AMT [29].
Recurrence rate of pterygium treated with
Pterygium (Fig. 11.2) CLAU was reported to be 2.14% (5 recurrences
The head of pterygium is scraped from the cor- in 234 procedures) after 2-year follow-up.
neal surface followed by excision of the body of Recurrences were observed in 4 of 58 (6.9%)
pterygium from the surrounding conjunctiva and cases with recurrent pterygium and 1 of 178
removal of episcleral scarring. A graft excised (0.57%) cases with primary pterygium in this
from the limbus and conjunctiva of the contralat- study [30]. In the study of 10-year follow-up, 2 of
eral eye is sutured onto the recipient bed [20, 21]. 29 (6.9%) cases had recurrent pterygium after
Fibrin glue can be used instead of suture [22–24]. CLAU [31].
Applying mitomycin C (MMC) and/or AMT on Systematic review shows that CLAU signifi-
the limbus and bare sclera is optional [11, cantly reduces the risk of pterygium recurrence
25–27]. compared with pterygium excision with bare
sclera technique (odds ratios [OR]  =  0.08,
Ocular Surface Neoplasia p  <  0.01), bulbar conjunctival autograft
After the excision of neoplastic region in the cor- (OR = 0.10, p < 0.01), and intraoperative MMC
nea and conjunctiva combined with the intraop- (OR = 0.22, p < 0.01) [32]. There was no signifi-
erative use of 0.02% MMC, CLAU obtained from cant difference between CLAU and AMT
the contralateral eye is transplanted on the exci- (OR = 0.66, p = 0.39). Another systematic review
sion site and sutured onto the sclera. Graft should shows that CLAU significantly reduced the risk
be enough large to cover the bare area [14]. of pterygium recurrence compared with con-
junctival autograft (relative risk [RR]  =  0.09,
11.2.1.3 Outcome p = 0.02) [33].
First series of CLAU was reported by Kanyon As for combination therapy, development of
et al. in 1989 [8]. They performed CLAU in 21 recurrence is less in CLAU with AMT group than
patients with unilateral LSCD.  Most patients MMC with AMT group (0% vs 20%) for the treat-
experienced prompt reduction and permanent ment of recurrent pterygium [34]. There is no signifi-
158 Y. Sasamoto et al.

cant difference in recurrence rates between CLAU 11.2.2.3 Outcome


with MMC and conjunctival autograft with MMC Holland et al. reported that regeneration of normal
in primary and recurrent pterygium [11]. Also, ocular surface obtained in 23 of 31 patients
there is no significant difference in recurrence (74.2%) and visual acuity improvement was
rates between CLAU with MMC and AMT with obtained in 27 of 31 patients (87.1%) and there
MMC in recurrent pterygium [35]. were no adverse reactions in patients with bilateral
It is still controversial whether the application LSCD caused by aniridia [37]. Of note, 20 eyes
of CLAU in addition to the excision of ocular (64.5%) underwent subsequent PK. A systematic
neoplasia is necessary to prevent recurrence [14]. review shows that 20 of 29 eyes (69%) had best-
corrected visual acuity (BCVA) ≥20/200 at last
follow-up (median follow-up period: 42 months)
11.2.2 Keratolimbal Allograft (KLAL) and 32 of 36 eyes (80%) underwent simultaneous
or subsequent PK [40]. Solomon A. et al. reported
As CLAU requires relatively large biopsy of lim- long-term outcome of KLAL and AMT with or
bus, this procedure frequently brings LSC deple- without PK, but simultaneous KLAL and PK may
tion in the healthy donor eye. In addition, CLAU be associated with a less favorable outcome [39].
cannot apply to the patients with bilateral In eyes with unilateral LSCD, Shimazaki et  al.
LSCD. KLAL was developed to overcome these reported that KLAL resulted in a worse clinical
situations [36]. KLAL uses cadaveric donor lim- outcome than CLAU [41].
bal tissue, so that there is no concern about the
LSC depletion in the contralateral eye and this
procedure is applicable to bilateral LSCD 11.2.3 Living-Related Conjunctival
patients. However, systemic immunosuppression Limbal Allograft (Lr-CLAL)
is required to avoid graft rejection.
Lr-CLAL was developed to reduce the risk of
11.2.2.1 Indications rejection by KLAL [42, 43]. The limbal and
KLAL is applicable to both unilateral and bilat- conjunctival tissues are harvested from a
eral LSCD. patient’s living relative. Systemic immunosup-
pression is still required at least in HLA-
11.2.2.2 Procedure (Fig. 11.3) incompatible cases.
Corneoscleral rims obtained from two cadaveric
donor eyes are sectioned into two equal halves, 11.2.3.1 Indications
and they are thinned by removing the posterior Indications of Lr-CLAL are similar to that of
two thirds of corneoscleral tissue. On the recipi- KLAL (both unilateral and bilateral LSCD), but
ent eye, a 360° limbal peritomy is performed. Lr-CLAL is applicable to the patients who has a
After the removal of abnormal fibrovascular pan- willing living relative. A good HLA match is
nus and epithelium, three of the four dissected preferable (Ophthalmology. 2001 Daya SM,
crescents are transplanted on the recipient eye Ophthalmology 1999, Srinivas KRaoDO).
with the donor corneal edge overlying the recipi- For the eyes with severe LSCD, combination
ent limbus and are secured with sutures [37]. of KLAL and Lr-CLAL can be considered
Fibrin glue can be used instead of suture [38]. (Cincinnati Procedure) [47].
Simultaneous AMT is optional [39]. Other than
topical immunosuppression using corticosteroid 11.2.3.2 Procedure
and cyclosporine A, oral immunosuppressants, Two grafts from limbus and adjacent conjunc-
including prednisone, cyclosporine A, and aza- tiva are excised from the donor eye [44].
thioprine, should be considered [37]. Transplantation onto the recipient cornea is
Simultaneous or subsequent keratoplasty can be same as CLAU.  Simultaneous AMT, PK, and
considered when the stromal scarring hampers the use of fibrin glue instead of suture are
improvement of visual acuity. optional [24, 45].
11  Corneal Stem Cell-Based Therapies 159

a e

b f

c g

d h

Fig. 11.3  KLAL procedure. (a) Preoperative status; (b) surface. (d–f) Limbal graft was created using cadaver
peritomy was performed to separate abnormal fibrovascu- sclerocorneal tissue using trephine, golf blade, and scis-
lar tissue inside the cornea from conjunctiva on the sclera. sors. (g, h) Limbal tissue from donor was sutured to host
(c) Fibrovascular pannus was removed from the corneal limbus
160 Y. Sasamoto et al.

11.2.3.3 Outcome [49]. They utilized cells harvested from the small
Daya et al. reported the outcome of ten eyes of biopsy specimen of contralateral limbus.
Lr-CLAL (mean follow-up period: 26.2 months) Allogenic CLET was developed for the eyes with
[44]. Systemic immunosuppression was adminis- bilateral LSCD [50]. This technique is advanta-
tered in all the patients. Restoration of corneal geous in regard to the risk of donor cornea dam-
epithelium and reduction of vascularization were age when compared to KLAL and Lr-CLAL.
obtained in eight eyes (80%). Visual improve-
ment was achieved in seven eyes (70%). Allograft 11.2.4.1 Indications
rejection occurred in two eyes (25%), which had Autologous CLET is applicable to the patients
two class I HLA mismatches and underlying with unilateral LSCD, and allogeneic CLET is
diagnosis of SJS. Gomes et al. reported the out- applicable to the patients with bilateral LSCD.
come of Lr-CLAL with AMT in ten LSCD eyes
caused by SJS (mean follow-up period: 11.2.4.2 Procedure
1.67 months) [45]. Nine patients (90%) were not One of the standard procedures was reported by
HLA-compatible with their respective living-­ Rama et  al. [51]. A 1–2  mm2 limbal biopsy is
related donors and required systemic immuno- taken from contralateral eye, and primary limbal
suppression. Ocular surface reconstruction was keratinocytes are cultivated on fibrin and feeder
obtained only in two eyes (20%), infection layer of lethally irradiated 3T3-J2 cells for
occurred in four eyes (40%), and visual acuity 14–16 days. After a 360° limbal peritomy and the
improvement was achieved in four eyes (40%). removal of fibrovascular corneal pannus in the
When compared to KLAL, Lr-CLAL had a recipient eye, the fibrin-cultured epithelial sheet
higher gain in vision (p = 0.029), decrease in con- is transplanted on the prepared corneal wound
junctivalization (p  =  0.009), and increase in bed. The conjunctiva is then sutured over the
Schirmer’s values (p = 0.009) in eyes with ocular peripheral fibrin sheet. Xeno-free explant culture
burns [46]. The other group showed that CLAU system on human AM was developed to reduce
achieved better long-term corneal surface results the risk of transmission of zoonotic infection
compared to KLAL and Lr-CLAL [33]. Although from mouse 3T3 cells and fetal bovine serum in
they showed no significant difference between the medium [52].
KLAL and Lr-CLAL, they indicate that the out- Use of systemic immunosuppression is con-
come of Lr-CLAL may be improved by HLA troversial after allogenic CLET. When the central
matching and systemic immunosuppression. cornea scarring exists, simultaneous or subse-
Biber et  al. reported clinical outcome of 24 quent keratoplasty is optional [53].
eyes which performed combination of KLAL and
Lr-CLAL (mean follow-up period: 43.4 months) 11.2.4.3 Outcome
[47]. Keratoplasty was undergone in 79.2% eyes Long-term CLET study of patients with unilateral
after ocular surface reconstruction. Stable or LSCD caused by ocular surface burns was reported
improved ocular surface was achieved in 87.5%, by Rama et  al. in 2010 (mean follow-up period:
and improved visual acuity was obtained in 75%. 2.91  ±  1.99  years) [51]. They cultivated limbal
keratocyte on fibrin and 3T3-J2 feeder cells. Of the
107 eyes treated, 73 eyes (68.2%) were classified
11.2.4 Cultivated Limbal Epithelial as success, and 60 eyes (56.1%) had improved
Transplantation (CLET) visual acuity. Sangwan et  al. reported long-term
outcome of CLET using a xeno-free explant cul-
CLAU potentially damages the donor cornea due ture system and human AM as a substrate for the
to a large excision [48]. To overcome this disad- treatment of LSCD caused by ocular surface burns
vantage, Pellegrini et  al. developed autologous (mean follow-up period: 3 ± 1.6 years) [52]. Of the
cultivated limbal epithelial transplantation 200 eyes treated, completely epithelialized, avas-
(CLET) for the treatment of unilateral LSCD cular, and clinically stable corneal surface was
11  Corneal Stem Cell-Based Therapies 161

obtained in 142 (71%) eyes, and visual acuity was 11.2.5.1 Indications
improved in 121 eyes (60.5%). Patients with unilateral LSCD.
Basu et al. reported the outcome of the combi-
nation of CLET and PK in 47 patients (mean 11.2.5.2 Procedure [60, 61]
follow-up was 4.2 ± 1.9 years) [54]. Visual acuity A 1-clock hour limbal tissue is excised from the
of 20/40 or better was attained by 71.4% of the contralateral donor eye. After the removal of
eyes. However, as simultaneous PK lead higher vascular pannus covering the cornea, human
recurrence rate of LSCD (58.3%) than subse- AM graft is placed over the bared ocular sur-
quent PK (14.3%), they suggest that simultane- face and secured with fibrin glue. The donor tis-
ous PK should be avoided. Allogeneic CLET, sue is then cut into small pieces, placed on the
followed by PK when needed, can restore the human AM, and fixed with fibrin glue.
ocular surface and improve vision in patients Keratoplasty can be combined with SLET to
with bilateral LSCD [55]. Qi et al. administered improve vision.
systemic steroid and topical steroid and cyclo-
sporine A after allogenic CLET and rejection 11.2.5.3 Outcome
were seen in 10 eyes of 42 eyes (23.8%) during Basu et  al. reported the outcome of SLET in
the follow-up of 17.8 ± 3.8 months [56]. 125 patients with unilateral LSCD caused by
A couple of systematic reviews show the chemical or thermal burns (median follow-up
effectiveness of CLET.  Holland reported long-­ period: 1.5 years) [61]. Successful outcome (a
term outcome of 720 eyes treated with either completely epithelialized, clinically stable,
autologous (641 eyes) or allogenic CLET (79 and avascular corneal surface) was observed in
eyes) [57]. Overall success rate was 72%, and 95 eyes (76%), and two-line improvement in
improved visual acuity was obtained in 63%. visual acuity was seen in 94 eyes (75.2%).
Haagdorens et  al. reported clinical outcomes of Randomized study shows that a stable epitheli-
1029 cases of autograft and 135 cases of allografts alized corneal surface was obtained in all
[58]. Overall success rate was 70%, and improved patients and there was no significant difference
visual acuity was obtained in 55%. Zhao et  al. in the vision improvement between CLAU and
reported clinical outcomes of 572 cases of CLET SLET [62].
with human AM as the substrate [59]. The suc-
cess rate was 67%, and improved visual acuity
was achieved in 62%. 11.2.6 Cultivated Oral Mucosal
Epithelial Transplantation
(COMET)
11.2.5 Simple Limbal Epithelial
Transplantation (SLET) KLAL, Lr-CLAL, and allogenic CLET are per-
formed for the treatment of bilateral LSCD, but
CLET was developed to overcome the disadvan- those techniques have disadvantages arisen
tage of CLAU, which may damage the donor cor- from their allogenic manner, that is, rejection,
nea due to a large excision. However, CLET adverse events associated with immunosuppres-
requires high-cost clinical-grade laboratory sup- sion, and disease transmission. Autologous
port for the graft cultivation and two steps of sur- COMET was developed to eliminate these con-
gical procedures (limbal excision and cerns [63]. Cultivated oral mucosal epithelial
transplantation). To overcome these situations, cells contain epithelial stem cells which work as
Sangwan et  al. developed SLET by combining an alternative source of LSCs on the ocular
the benefits of CLAU and CLET [60]. This tech- surface.
nique utilizes small portion of autologous donor
limbus for transplantation and does not require 11.2.6.1 Indications
clinical-grade laboratory. Patients with bilateral LSCD.
162 Y. Sasamoto et al.

11.2.6.2 Procedure [63, 64] (Fig. 11.4) sal epithelial sheets are placed directly onto the
3-by-3-mm specimen of oral mucosal tissue is corneal surface after excision of conjunctival and
excised from the interior buccal mucosal epithe- subconjunctival scar tissue. Although xenogenic
lium. Dissociated oral mucosal epithelial cells are feeder cells have a potential risk to transmit virus
seeded onto a human AM or fibrin-coated culture infection, autologous feeder layer can be a resolu-
plate or a temperature-responsive dish, under tion [65]. Transportation technique for cultivated
which 3T3 cells are cultured as feeder layers and cell sheet has been developed for multicenter clini-
incubated for 14  days to obtain epithelial cell cal trial [66]. To improve corneal stromal transpar-
sheet. Three to five layers of cultivated oral muco- ency, secondary keratoplasty can be considered.

a b

c d

e f

Fig. 11.4  COMET procedure. (a) Preoperative status; host ocular surface. (e) Continuous suture using 10-0
(b) fibrovascular pannus was removed from cornea. (c) nylon was added to limbus. (f) Ocular surface was covered
Cultivated oral mucosal epithelial cell sheet was harvested by cell sheet and host conjunctiva
with temperature reduction. (d) Cell sheet was placed on
11  Corneal Stem Cell-Based Therapies 163

11.2.6.3 Outcome subconjunctival hemorrhage, corneal thinning (or


Satake et al. reported a long-term outcome of 40 melting, perforation), inflammation, and infec-
bilateral LSCD eyes treated by COMET (mean tious keratitis. In the donor eyes of CLAU and
follow-up period: 25.5 months) [64]. Stable ocular Lr-CLAL, biopsy-related epithelial abnormali-
surface was achieved in 57.5% eyes, and improved ties, filamentary keratitis, and subconjunctival
visual acuity was obtained in 59% eyes. A second- hemorrhage have been reported [43, 48, 68].
ary keratoplasty was performed to improve visual However, LSCs repopulate the donor area within
acuity in seven eyes. Another long-term clinical 1 year, and removal of 2 pieces of 2 clock hours
report shows stable ocular surface was achieved in each of donor limbus does not bring any adverse
15 of 20 eyes (75%) and improved visual acuity effect in the donor cornea [48, 69].
was obtained in 14 eyes (70%) (mean follow-up In KLAL, CLAL, and allogenic CLET, there
period: 31.9 months) [67]. are risks of transplant rejection and systemic
immunosuppression-related adverse events such
as infection, anemia, hyperglycemia, and renal
11.2.7 Complications (Fig. 11.5) and liver function abnormalities [70–72]. In those
allogenic techniques, there are some reports of
Postoperative complications include recurrence disease transmissions, such as tumor (melanoma
or persistence of epithelial defects, continued and breast carcinoma) and cytomegalovirus dis-
conjunctivalization (recurrence of pterygium), ease [73–75].

Fig. 11.5  Complications. (a) Fungal keratitis caused by Candida parapsilosis following limbal transplantation. (b)
Rejection following allogenic limbal transplantation. Rejection line was observed (arrow head)
164 Y. Sasamoto et al.

Development of superficial corneal neovascu- iPSCs were developed by introducing


larization under the epithelial sheet is the major Yamanaka four factors into the differentiated cells
concern in COMET [63, 76]. such as fibroblasts and blood cells and were
3T3 cells and fetal bovine serum for cell cul- shown that they have the similar characteristics as
ture can increase the risk of transmission of zoo- ESCs [80]. This technique can overcome the ethi-
notic infection in CLET and COMET [77]. cal problem. Hayashi et  al. reported a way to
induce corneal stem and progenitor cells from
iPSCs by mimicking whole-eye development [81,
11.3 F
 uture Stem Cell-Based 82]. They successfully generated a corneal epithe-
Therapy: Corneal Epithelium lial cell sheet, and its transplantation recovered
the LSCD model eye in rabbit (Fig.  11.6).
As each current method has its pros and cons, no However, generation of iPSCs and induction of
gold standard exists so far. Especially for the corneal epithelial cells take long time and have
treatment of bilateral LSCD, the necessity of sys- high cost. Creation of HLA-typed iPSC banks can
temic immunosuppression is a big concern. be one of the solutions [83]. It is calculated that 50
COMET can solve this barrier, but the recon- iPSC lines of the highest-ranked homozygous
structed ocular surface does not allow a full, HLA types can provide a zero HLA mismatch for
long-term improvement of visual acuity and is 79% of potential UK recipients. One strategy is to
prone to problems associated with induce corneal epithelial stem and progenitor
neovascularization. cells from top-ranked HLA-typed iPSCs and store
Therefore, the researchers are seeking the them in advance, and then we can thaw and cul-
other sources of stem cells to overcome the ture them when needed. There still remain some
situation. issues to be solved about tumorigenicity, immu-
The cell sources being investigated can be cat- nogenicity, and mutagenicity of iPSCs [84, 85].
egorized into four groups: (i) corneal epithelial MSCs are multipotent adult stem cells which
cells induced from pluripotent stem cells, (ii) cor- reside in multiple tissues, including the umbilical
neal epithelial cells induced from the other dif- cord, bone marrow, and adipose tissue. MSCs not
ferentiated cell source (direct reprogramming), only can overcome the ethical issues of ESCs but
(iii) corneal epithelial alternatives from the other also can be obtained much easier than ESCs and
surface ectoderm-derived cells, and (iv) purified iPSCs. MSCs also have less immunogenicity
LSCs. even when being used as allogenic transplanta-
tion and may have less mutagenicity [86, 87].
MSCs were shown to differentiate into corneal
11.3.1 Corneal Epithelial Cells epithelial-like cells in LSC niche environment
Induced from Pluripotent in  vitro [88–90] and in  vivo [91]. Dental pulp
Stem Cells stem cells, which share the characteristics of
MSCs and ESCs, can generate a cell sheet and
Researchers are trying to generate corneal epithe- reconstruct the corneal epithelium in a rabbit
lial cells or LSCs from three different sources of model of LSCD [92]. There still remains a con-
pluripotent stem cell, that is, embryonic stem cern of tumorigenicity [93].
cells (ESCs), induced pluripotent stem cells
(iPSCs), and mesenchymal stem cells (MSCs).
Corneal epithelial-like cells were successfully 11.3.2 Corneal Epithelial Cells
induced from ESCs in limbal fibroblast-­ Induced from the Other
conditioned medium [78]. However, there are Differentiated Cell Source
some obstacles such as immunogenicity, uncon- (Direct Reprogramming)
trolled growth which results in increased tumori-
genicity, and ethical issues arisen from its Direct reprogramming, the way to convert fully
embryonic origin [79]. differentiated mature cells into the other cell
11  Corneal Stem Cell-Based Therapies 165

b Before 7d 14d
iCEC
Control (sham)

Fig. 11.6 Ocular surface reconstruction using iPSC-­ rabbit LSCD model on days 0, 7, and 14 post-surgery.
derived corneal epithelial cell sheet in rabbit model. (a) A iPSC-derived corneal epithelial cell sheet successfully
corneal epithelial cell sheet derived from human iPSCs. recovered a healthy corneal barrier function. iCEC iPSC-­
(b) Barrier function assay using fluorescein staining for derived corneal epithelial cells. (Reproduced from
iPSC-derived corneal epithelial cell sheet transplanted Hayashi et al. [82])
166 Y. Sasamoto et al.

types while bypassing a pluripotent state, can 11.4 F


 uture Stem Cell-Based
shorten the induction time of corneal epithelial Therapy: Corneal Stroma
cells from the other types of cells and reduce the
risk of tumorigenicity [94]. Ouyang et al. induced Corneal stroma consists of collagens, proteogly-
LSC-like cells by transduction of PAX6 in skin cans, glycoproteins, and a small amount of stro-
epithelial stem cells in  vitro [95]. Corneal mal cells known as keratocytes. PK and deep
epithelial-­like cells can also be induced from anterior lamellar keratoplasty have been per-
fibroblasts and oral mucosal epithelial cells by formed for the treatment of corneal stromal dis-
overexpression of gene sets including PAX6 [96, eases [106].
97]. Although gene delivery has a potential of Corneal stromal stem cells are a population of
insertional mutagenesis, transgene-free methods neural crest-derived MSC which resides in the
such as Sendai virus, direct mRNA, or protein limbal stroma, subjacent to the epithelial base-
delivery can solve this issue [98]. Interestingly, ment membrane [107]. Corneal stromal stem
hair follicle epithelial stem cells and limbal fibro- cells injected into the stroma can restore the
blasts have a potential to differentiate into cor- transparency and thickness of lumican-defective
neal epithelial-like cells without any gene murine corneas [108].
modification [99, 100]. Keratocyte-like cells have been induced from
ESCs [109, 110], iPSCs [111], MSCs [112], and
human dental pulp stem cells [113] in  vitro.
11.3.3 Corneal Epithelial Alternatives MSCs are shown to have an ability to treat cor-
from the Other Surface neal stromal diseases in animal models. MSCs
Ectoderm-Derived Cells injected into the mouse corneal stroma can sur-
vive for more than 3  months, and these cells
Other than oral mucosa, there are some other sur- assumed a keratocyte phenotype [114]. Also,
face ectoderm-derived cells that can be alterna- MSCs injected into the anterior chamber can
tives of corneal epithelium. Conjunctival migrate to the stroma and differentiate into func-
epithelial cell cultivated ex  vivo (EVCAU) is tional keratocytes [115].
already performed as a clinical trial for the treat-
ment of bilateral LSCD [101]. Clinical improve-
ment was observed in 83.3%, and visual acuity 11.5 F
 uture Stem Cell-Based
was improved in 75% cases. Nasal mucosal epi- Therapy: Corneal
thelial cells also have a potential to reconstruct Endothelium
functional ocular surface epithelial cells [102].
Loss or dysfunction of corneal endothelium leads
to bullous keratopathy (BK). Major causes of BK
11.3.4 Purified LSCs are cataract surgery, peripheral iridotomy, and
Fuchs endothelial corneal dystrophy (FECD)
Ksander et al. reported that cell membrane pro- [116]. As there were no efficient ways to prolifer-
tein ABCB5 is a marker of LSCs [103]. They ate corneal endothelial cells, BK has been treated
showed that LSCs can be purified by anti-ABCB5 by corneal transplantation such as PK, Descemet
antibody. Transplantation of purified ABCB5-­ stripping automated endothelial keratoplasty
positive LSCs can restore the corneal epithelium (DSAEK), and Descemet membrane endothelial
in a LSCD model mouse. As ABCB5-positive keratoplasty (DMEK) [117–119]. Recently,
dermal stem cells are reported to induce allograft ROCK inhibitor was shown to enhance prolifera-
tolerance to HLA-mismatched transplants, tion, promote cell adhesion, and suppress apopto-
ABCB5-positive LSCs may be a good candidate sis of corneal endothelial cells [120, 121]. Pilot
for allogenic transplantation in patients with data from a clinical trial which injected cultured
bilateral LSCD [104, 105]. corneal endothelium in combination with a
11  Corneal Stem Cell-Based Therapies 167

ROCK inhibitor into the anterior chamber has has been developed [133]. Drastic development
shown its safety and effectiveness [121]. of technologies is expediting the research of cor-
Autologous or allogenic cultured corneal endo- neal stem cell therapy.
thelial cell transplantation can be applied to the
treatment of BK in the future. Compliance with Ethical Requirements Yuzuru
The existence of corneal endothelial stem Sasamoto and Yoshinori Oie declare no conflict of inter-
est. Prof. Kohji Nishida received a Grant for Translational
cells had been discussed for a long time [122, Research Network Program supported by the Ministry of
123]. Recently, multipotent neural crest-derived Education, Culture, Sports, Science and Technology of
progenitor (NCDP) cells were isolated from cor- Japan. All procedures followed were in accordance with
neal endothelium, and corneal endothelial cells the ethical standards of the responsible committee on
human experimentation (institutional and national) and
were induced from NCDPs [124]. As normal with the Helsinki Declaration of 1975, as revised in 2000.
NCDPs were detected not only in normal but also Informed consent was obtained from all patients for being
in Fuchs endothelial corneal dystrophy (FECD) included in the study. All institutional and national guide-
endothelial tissue, this cell population has prom- lines for the care and use of laboratory animals were
followed.
ise for potential autologous cell therapies.
Human corneal endothelium-like cells have
been induced from ESCs [125–127], iPSCs
[128], MSCs [129], adipose-derived stem cells References
[130], and corneal stroma stem cells [131].
1. Schermer A, Galvin S, Sun TT.  Differentiation-­
related expression of a major 64K corneal kera-
tin in  vivo and in culture suggests limbal location
11.6 Summary and Conclusions of corneal epithelial stem cells. J Cell Biol.
1986;103(1):49–62.
Clinical studies have shown certain degrees of 2. Thoft RA, Friend J. The X, Y, Z hypothesis of cor-
success in several stem cell-based therapies espe- neal epithelial maintenance. Invest Ophthalmol Vis
Sci. 1983;24(10):1442–3.
cially in the treatment of LSCD.  However, we 3. Yoon JJ, Ismail S, Sherwin T.  Limbal stem cells:
still need to establish more evidence to decide central concepts of corneal epithelial homeostasis.
which procedure is preferable for a specific con- World J Stem Cells. 2014;6(4):391–403.
dition. In addition to the therapies currently per- 4. Chen JJ, Tseng SC. Corneal epithelial wound heal-
ing in partial limbal deficiency. Invest Ophthalmol
formed, an increasing number of novel techniques Vis Sci. 1990;31(7):1301–14.
using various stem cells are being investigated. 5. Tseng SC.  Concept and application of limbal stem
ESCs, iPSCs, and MSCs are increasing their cells. Eye (Lond). 1989;3(Pt 2):141–57.
presence for the induction of not only corneal 6. Kwok LS, Coroneo MT. A model for pterygium for-
mation. Cornea. 1994;13(3):219–24.
epithelial cells but also keratocytes and corneal 7. Thoft RA.  Conjunctival transplantation. Arch
endothelial cells. These techniques can be one of Ophthalmol. 1977;95(8):1425–7.
the promising future therapies, but there are many 8. Kenyon KR, Tseng SC. Limbal autograft transplan-
obstacles to be solved, e.g., ethical issues, tumor- tation for ocular surface disorders. Ophthalmology.
1989;96(5):709–22; discussion 22–3.
igenicity, immunogenicity, and mutagenicity. 9. Koch JM, Mellin KB, Waubke TN. The pterygium,
Besides possible future therapies shown autologous conjunctiva-limbus transplantation as
above, building an entire functional cornea, treatment. Ophthalmologe. 1992;89(2):143–6.
which has organized epithelium, stroma, and 10. Prabhasawat P, Barton K, Burkett G, Tseng
SC.  Comparison of conjunctival autografts,
endothelium, may be the next breakthrough. Very amniotic membrane grafts, and primary clo-
recently, a corneal organoid which contains these sure for pterygium excision. Ophthalmology.
three components has been induced from human 1997;104(6):974–85.
iPSCs [132], and an artificial cornea which con- 11. Kheirkhah A, Hashemi H, Adelpour M, Nikdel M,
Rajabi MB, Behrouz MJ.  Randomized trial of pte-
sists of limbal epithelial cell-like cells and cor- rygium surgery with mitomycin C application using
neal endothelial cell-like cells derived from conjunctival autograft versus conjunctival-limbal
human ESCs and acellular porcine cornea matrix autograft. Ophthalmology. 2012;119(2):227–32.
168 Y. Sasamoto et al.

12. Chui J, Coroneo MT, Tat LT, Crouch R, Wakefield D, associated with symblepharon. Br J Ophthalmol.
Di Girolamo N. Ophthalmic pterygium: a stem cell 1998;82(3):235–40.
disorder with premalignant features. Am J Pathol. 27. Yao YF, Qiu WY, Zhang YM, Tseng SC. Mitomycin
2011;178(2):817–27. C, amniotic membrane transplantation and lim-
13. Copeland RA Jr, Char DH. Limbal autograft recon- bal conjunctival autograft for treating multire-
struction after conjunctival squamous cell carci- current pterygia with symblepharon and motility
noma. Am J Ophthalmol. 1990;110(4):412–5. restriction. Graefes Arch Clin Exp Ophthalmol.
14. Siganos CS, Kozobolis VP, Christodoulakis EV. The 2006;244(2):232–6.
intraoperative use of mitomycin-C in excision of 28. Santos MS, Gomes JA, Hofling-Lima AL, Rizzo
ocular surface neoplasia with or without limbal auto- LV, Romano AC, Belfort R Jr. Survival analysis of
graft transplantation. Cornea. 2002;21(1):12–6. conjunctival limbal grafts and amniotic membrane
15. Meallet MA, Espana EM, Grueterich M, Ti SE, transplantation in eyes with total limbal stem cell
Goto E, Tseng SC.  Amniotic membrane trans- deficiency. Am J Ophthalmol. 2005;140(2):223–30.
plantation with conjunctival limbal autograft for 29. Barreiro TP, Santos MS, Vieira AC, de Nadai Barros
total limbal stem cell deficiency. Ophthalmology. J, Hazarbassanov RM, Gomes JA. Comparative study
2003;110(8):1585–92. of conjunctival limbal transplantation not associated
16. Dua HS, Azuara-Blanco A.  Autologous lim- with the use of amniotic membrane transplantation
bal transplantation in patients with unilateral for treatment of total limbal deficiency secondary to
corneal stem cell deficiency. Br J Ophthalmol. chemical injury. Cornea. 2014;33(7):716–20.
2000;84(3):273–8. 30. Masters JS, Harris DJ Jr. Low recurrence rate of
17. Kheirkhah A, Raju VK, Tseng SC.  Minimal con- pterygium after excision with conjunctival limbal
junctival limbal autograft for total limbal stem cell autograft: a retrospective study with long-term fol-
deficiency. Cornea. 2008;27(6):730–3. low-­up. Cornea. 2015;34(12):1569–72.
18. Baradaran-Rafii A, Akbari M, Shirzadeh E, Shams 31. Young AL, Ho M, Jhanji V, Cheng LL.  Ten-year
M.  Single block conjunctival limbal autograft results of a randomized controlled trial compar-
for unilateral total limbal stem cell deficiency. J ing 0.02% mitomycin C and limbal conjunctival
Ophthalmic Vis Res. 2015;10(1):90–2. autograft in pterygium surgery. Ophthalmology.
19. Marinho D, Hofling-Lima AL, Kwitko S, Tseng 2013;120(12):2390–5.
SC.  Does amniotic membrane transplantation 32. Zheng K, Cai J, Jhanji V, Chen H.  Comparison of
improve the outcome of autologous limbal trans- pterygium recurrence rates after limbal conjunctival
plantation? Cornea. 2003;22(4):338–42. autograft transplantation and other techniques: meta-­
20. Kenyon KR, Wagoner MD, Hettinger analysis. Cornea. 2012;31(12):1422–7.
ME.  Conjunctival autograft transplantation for 33. Ontario HQ.  Limbal stem cell transplantation: an
advanced and recurrent pterygium. Ophthalmology. evidence-based analysis. Ont Health Technol Assess
1985;92(11):1461–70. Ser. 2008;8(7):1–58.
21. Al Fayez MF.  Limbal-conjunctival vs conjunctival 34. Fallah MR, Golabdar MR, Amozadeh J, Zare MA,
autograft transplant for recurrent pterygia: a prospec- Moghimi S, Fakhraee G.  Transplantation of con-
tive randomized controlled trial. JAMA Ophthalmol. junctival limbal autograft and amniotic membrane vs
2013;131(1):11–6. mitomycin C and amniotic membrane in treatment of
22. Ozdamar Y, Mutevelli S, Han U, Ileri D, Onal B, recurrent pterygium. Eye (Lond). 2008;22(3):420–4.
Ilhan O, et  al. A comparative study of tissue glue 35. Chen R, Huang G, Liu S, Ma W, Yin X, Zhou
and vicryl suture for closing limbal-conjunctival S.  Limbal conjunctival versus amniotic mem-
autografts and histologic evaluation after pterygium brane in the intraoperative application of mito-
excision. Cornea. 2008;27(5):552–8. mycin C for recurrent pterygium: a randomized
23. Jiang J, Gong J, Li W, Hong C.  Comparison of controlled trial. Graefes Arch Clin Exp Ophthalmol.
intra-operative 0.02% mitomycin C and sutureless 2017;255(2):375–85.
limbal conjunctival autograft fixation in pterygium 36. Tsai RJ, Tseng SC.  Human allograft limbal trans-
surgery: five-year follow-up. Acta Ophthalmol. plantation for corneal surface reconstruction.
2015;93(7):e568–72. Cornea. 1994;13(5):389–400.
24. Welder JD, Pandya HK, Nassiri N, Djalilian 37. Holland EJ, Djalilian AR, Schwartz GS. Management
AR.  Conjunctival limbal autograft and allograft of aniridic keratopathy with keratolimbal allograft:
transplantation using fibrin glue. Ophthalmic Surg a limbal stem cell transplantation technique.
Lasers Imaging. 2012;43(4):323–7. Ophthalmology. 2003;110(1):125–30.
25. Fakhry MA.  The use of mitomycin C with autolo- 38. Sonmez B, Beden U.  Fibrin glue-assisted suture-
gous limbal-conjunctival autograft transplanta- less limbal stem cell transplantation surgery for the
tion for management of recurrent pterygium. Clin treatment of severe ocular chemical injury. Cornea.
Ophthalmol. 2011;5:123–7. 2011;30(3):296–300.
26. Shimazaki J, Shinozaki N, Tsubota 39. Solomon A, Ellies P, Anderson DF, Touhami A,
K. Transplantation of amniotic membrane and lim- Grueterich M, Espana EM, et al. Long-term outcome
bal autograft for patients with recurrent pterygium of keratolimbal allograft with or without penetrating
11  Corneal Stem Cell-Based Therapies 169

keratoplasty for total limbal stem cell deficiency. penetrating keratoplasty after autologous cultivated
Ophthalmology. 2002;109(6):1159–66. limbal epithelial transplantation for ocular surface
40. Shanbhag SS, Saeed HN, Paschalis EI, Chodosh J. burns. Am J Ophthalmol. 2011;152(6):917–24 e1.
Keratolimbal allograft for limbal stem cell deficiency 55. Basu S, Fernandez MM, Das S, Gaddipati S,
after severe corneal chemical injury: a systematic Vemuganti GK, Sangwan VS.  Clinical outcomes
review. Br J Ophthalmol. 2018;102(8):1114–21. of xeno-free allogeneic cultivated limbal epithelial
41. Shimazaki J, Shimmura S, Tsubota K. Donor source transplantation for bilateral limbal stem cell defi-
affects the outcome of ocular surface reconstruc- ciency. Br J Ophthalmol. 2012;96(12):1504–9.
tion in chemical or thermal burns of the cornea. 56. Qi X, Xie L, Cheng J, Zhai H, Zhou Q. Characteristics
Ophthalmology. 2004;111(1):38–44. of immune rejection after allogeneic cultivated
42. Kenyon KR, Rapoza PA. Limbal allograft transplan- limbal epithelial transplantation. Ophthalmology.
tation for ocular surface disorders. Ophthalmology. 2013;120(5):931–6.
1995;102(suppl):101–2. 57. Holland EJ.  Management of limbal stem cell defi-
43. Tan DT, Ficker LA, Buckley RJ. Limbal transplanta- ciency: a historical perspective, past, present, and
tion. Ophthalmology. 1996;103(1):29–36. future. Cornea. 2015;34(Suppl 10):S9–15.
44. Daya SM, Ilari FA. Living related conjunctival lim- 58. Haagdorens M, Van Acker SI, Van Gerwen V, Ni
bal allograft for the treatment of stem cell deficiency. Dhubhghaill S, Koppen C, Tassignon MJ, et  al.
Ophthalmology. 2001;108(1):126–33; discussion Limbal stem cell deficiency: current treatment
33–4. options and emerging therapies. Stem Cells Int.
45. Gomes JA, Santos MS, Ventura AS, Donato WB, 2016;2016:9798374.
Cunha MC, Hofling-Lima AL.  Amniotic mem- 59. Zhao Y, Ma L. Systematic review and meta-analysis
brane with living related corneal limbal/conjunc- on transplantation of ex vivo cultivated limbal epi-
tival allograft for ocular surface reconstruction thelial stem cell on amniotic membrane in limbal
in Stevens-Johnson syndrome. Arch Ophthalmol. stem cell deficiency. Cornea. 2015;34(5):592–600.
2003;121(10):1369–74. 60. Sangwan VS, Basu S, MacNeil S, Balasubramanian
46. Titiyal JS, Sharma N, Agarwal AK, Prakash G, D. Simple limbal epithelial transplantation (SLET):
Tandon R, Vajpayee R. Live related versus cadaveric a novel surgical technique for the treatment of uni-
limbal allograft in limbal stem cell deficiency. Ocul lateral limbal stem cell deficiency. Br J Ophthalmol.
Immunol Inflamm. 2015;23(3):232–9. 2012;96(7):931–4.
47. Biber JM, Skeens HM, Neff KD, Holland EJ.  The 61. Basu S, Sureka SP, Shanbhag SS, Kethiri AR,
Cincinnati procedure: technique and outcomes Singh V, Sangwan VS.  Simple limbal epithelial
of combined living-related conjunctival limbal transplantation: long-term clinical outcomes in 125
allografts and keratolimbal allografts in severe ocu- cases of unilateral chronic ocular surface burns.
lar surface failure. Cornea. 2011;30(7):765–71. Ophthalmology. 2016;123(5):1000–10.
48. Miri A, Said DG, Dua HS. Donor site complications 62. Arora R, Dokania P, Manudhane A, Goyal
in autolimbal and living-related allolimbal trans- JL.  Preliminary results from the comparison of
plantation. Ophthalmology. 2011;118(7):1265–71. simple limbal epithelial transplantation with con-
49. Pellegrini G, Traverso CE, Franzi AT, Zingirian M, junctival limbal autologous transplantation in severe
Cancedda R, De Luca M. Long-term restoration of unilateral chronic ocular burns. Indian J Ophthalmol.
damaged corneal surfaces with autologous cultivated 2017;65(1):35–40.
corneal epithelium. Lancet. 1997;349(9057):990–3. 63. Nishida K, Yamato M, Hayashida Y, Watanabe K,
50. Schwab IR, Reyes M, Isseroff RR. Successful trans- Yamamoto K, Adachi E, et  al. Corneal reconstruc-
plantation of bioengineered tissue replacements tion with tissue-engineered cell sheets composed of
in patients with ocular surface disease. Cornea. autologous oral mucosal epithelium. N Engl J Med.
2000;19(4):421–6. 2004;351(12):1187–96.
51. Rama P, Matuska S, Paganoni G, Spinelli A, De 64. Satake Y, Higa K, Tsubota K, Shimazaki J. Long-term
Luca M, Pellegrini G.  Limbal stem-cell therapy outcome of cultivated oral mucosal epithelial sheet
and long-term corneal regeneration. N Engl J Med. transplantation in treatment of total limbal stem cell
2010;363(2):147–55. deficiency. Ophthalmology. 2011;118(8):1524–30.
52. Sangwan VS, Basu S, Vemuganti GK, Sejpal K, 65. Oie Y, Hayashi R, Takagi R, Yamato M, Takayanagi
Subramaniam SV, Bandyopadhyay S, et al. Clinical H, Tano Y, et  al. A novel method of culturing
outcomes of xeno-free autologous cultivated lim- human oral mucosal epithelial cell sheet using
bal epithelial transplantation: a 10-year study. Br J post-mitotic human dermal fibroblast feeder cells
Ophthalmol. 2011;95(11):1525–9. and modified keratinocyte culture medium for
53. Basu S, Ali H, Sangwan VS.  Clinical outcomes ocular surface reconstruction. Br J Ophthalmol.
of repeat autologous cultivated limbal epithelial 2010;94(9):1244–50.
transplantation for ocular surface burns. Am J 66. Oie Y, Nozaki T, Takayanagi H, Hara S, Hayashi R,
Ophthalmol. 2012;153(4):643–50, 50 e1–2. Takeda S, et al. Development of a cell sheet transpor-
54. Basu S, Mohamed A, Chaurasia S, Sejpal K, tation technique for regenerative medicine. Tissue
Vemuganti GK, Sangwan VS. Clinical outcomes of Eng Part C Methods. 2014;20(5):373–82.
170 Y. Sasamoto et al.

67. Prabhasawat P, Ekpo P, Uiprasertkul M, tion of multiple ocular-like cell lineages and fabrica-
Chotikavanich S, Tesavibul N, Pornpanich K, et al. tion of functional corneal epithelial cell sheets from
Long-term result of autologous cultivated oral muco- human iPS cells. Nat Protoc. 2017;12(4):683–96.
sal epithelial transplantation for severe ocular sur- 82. Hayashi R, Ishikawa Y, Sasamoto Y, Katori R,
face disease. Cell Tissue Bank. 2016;17(3):491–503. Nomura N, Ichikawa T, et  al. Co-ordinated
68. Jenkins C, Tuft S, Liu C, Buckley R. Limbal trans- ocular development from human iPS cells
plantation in the management of chronic contact-­ and recovery of corneal function. Nature.
lens-­associated epitheliopathy. Eye (Lond). 1993;7 2016;531(7594):376–80.
.(Pt 5:629–33. 83. Taylor CJ, Peacock S, Chaudhry AN, Bradley JA,
69. Busin M, Breda C, Bertolin M, Bovone C, Ponzin Bolton EM.  Generating an iPSC bank for HLA-­
D, Ferrari S, et  al. Corneal epithelial stem cells matched tissue transplantation based on known
repopulate the donor area within 1 year from lim- donor and recipient HLA types. Cell Stem Cell.
bus removal for limbal autograft. Ophthalmology. 2012;11(2):147–52.
2016;123(12):2481–8. 84. Miura K, Okada Y, Aoi T, Okada A, Takahashi K,
70. Satake Y, Dogru M, Yamaguchi T, Tomida D, Okita K, et  al. Variation in the safety of induced
Hirayama M, Shimazaki J.  Immunological rejec- pluripotent stem cell lines. Nat Biotechnol.
tion following allogeneic cultivated limbal epi- 2009;27(8):743–5.
thelial transplantation. JAMA Ophthalmol. 85. Zhao T, Zhang ZN, Rong Z, Xu Y. Immunogenicity
2013;131(7):920–2. of induced pluripotent stem cells. Nature.
71. Tsubota K.  Ocular surface management in cor- 2011;474(7350):212–5.
neal transplantation, a review. Jpn J Ophthalmol. 86. Zhang J, Huang X, Wang H, Liu X, Zhang T, Wang
1999;43(6):502–8. Y, et al. The challenges and promises of allogeneic
72. Krakauer M, Welder JD, Pandya HK, Nassiri N, mesenchymal stem cells for use as a cell-based ther-
Djalilian AR. Adverse effects of systemic immuno- apy. Stem Cell Res Ther. 2015;6(1):234.
suppression in keratolimbal allograft. J Ophthalmol. 87. Bellagamba BC, Abreu BR, Grivicich I, Markarian
2012;2012:576712. CF, Chem E, Camassola M, et al. Human mesenchy-
73. Sepsakos L, Cheung AY, Nerad JA, Mogilishetty mal stem cells are resistant to cytotoxic and geno-
G, Holland EJ.  Donor-derived conjunctival-limbal toxic effects of cisplatin in  vitro. Genet Mol Biol.
melanoma after a keratolimbal allograft. Cornea. 2016;39(1):129–34.
2017;36(11):1415–8. 88. Gu S, Xing C, Han J, Tso MO, Hong J. Differentiation
74. Miller AK, Young JW, Wilson DJ, Dunlap J, of rabbit bone marrow mesenchymal stem cells into
Chamberlain W.  Transmission of donor-derived corneal epithelial cells in vivo and ex vivo. Mol Vis.
breast carcinoma as a recurrent mass in a keratolim- 2009;15:99–107.
bal allograft. Cornea. 2017;36(6):736–9. 89. Jiang TS, Cai L, Ji WY, Hui YN, Wang YS, Hu D,
75. Cheung AY, Govil A, Friedstrom SR, Holland et al. Reconstruction of the corneal epithelium with
EJ. Probable donor-derived cytomegalovirus disease induced marrow mesenchymal stem cells in rats.
after keratolimbal allograft transplantation. Cornea. Mol Vis. 2010;16:1304–16.
2017;36(8):1006–8. 90. Rohaina CM, Then KY, Ng AM, Wan Abdul Halim
76. Nakamura T, Inatomi T, Sotozono C, Amemiya T, WH, Zahidin AZ, Saim A, et  al. Reconstruction
Kanamura N, Kinoshita S.  Transplantation of cul- of limbal stem cell deficient corneal surface
tivated autologous oral mucosal epithelial cells in with induced human bone marrow mesenchymal
patients with severe ocular surface disorders. Br J stem cells on amniotic membrane. Transl Res.
Ophthal. 2004;88(10):1280–4. 2014;163(3):200–10.
77. Murphy FA. The public health risk of animal organ 91. Reinshagen H, Auw-Haedrich C, Sorg RV,
and tissue transplantation into humans. Science. Boehringer D, Eberwein P, Schwartzkopff J,
1996;273(5276):746–7. et  al. Corneal surface reconstruction using adult
78. Ahmad S, Stewart R, Yung S, Kolli S, Armstrong L, mesenchymal stem cells in experimental limbal
Stojkovic M, et al. Differentiation of human embry- stem cell deficiency in rabbits. Acta Ophthalmol.
onic stem cells into corneal epithelial-like cells by 2011;89(8):741–8.
in vitro replication of the corneal epithelial stem cell 92. Gomes JA, Geraldes Monteiro B, Melo GB, Smith
niche. Stem Cells. 2007;25(5):1145–55. RL, Cavenaghi Pereira da Silva M, Lizier NF, et al.
79. Grinnemo KH, Sylven C, Hovatta O, Dellgren G, Corneal reconstruction with tissue-engineered
Corbascio M. Immunogenicity of human embryonic cell sheets composed of human immature den-
stem cells. Cell Tissue Res. 2008;331(1):67–78. tal pulp stem cells. Invest Ophthalmol Vis Sci.
80. Takahashi K, Tanabe K, Ohnuki M, Narita M, 2010;51(3):1408–14.
Ichisaka T, Tomoda K, et al. Induction of pluripotent 93. Barkholt L, Flory E, Jekerle V, Lucas-Samuel S,
stem cells from adult human fibroblasts by defined Ahnert P, Bisset L, et al. Risk of tumorigenicity in
factors. Cell. 2007;131(5):861–72. mesenchymal stromal cell-based therapies–bridging
81. Hayashi R, Ishikawa Y, Katori R, Sasamoto Y, scientific observations and regulatory viewpoints.
Taniwaki Y, Takayanagi H, et al. Coordinated genera- Cytotherapy. 2013;15(7):753–9.
11  Corneal Stem Cell-Based Therapies 171

94. Kelaini S, Cochrane A, Margariti A.  Direct repro- peripheral neurons, and corneal Keratocytes.
gramming of adult cells: avoiding the pluripotent Biotechnol J. 2017;12(12). https://doi.org/10.1002/
state. Stem Cells Cloning. 2014;7:19–29. biot.201700067. Epub 2017 Aug 22.
95. Ouyang H, Xue Y, Lin Y, Zhang X, Xi L, Patel 110. Hertsenberg AJ, Funderburgh JL. Generation of cor-
S, et  al. WNT7A and PAX6 define corneal epi- neal keratocytes from human embryonic stem cells.
thelium homeostasis and pathogenesis. Nature. Methods Mol Biol. 2016;1341:285–94.
2014;511(7509):358–61. 111. Naylor RW, McGhee CN, Cowan CA, Davidson
96. Kitazawa K, Hikichi T, Nakamura T, Mitsunaga K, AJ, Holm TM, Sherwin T.  Derivation of corneal
Tanaka A, Nakamura M, et  al. OVOL2 maintains keratocyte-­like cells from human induced pluripo-
the transcriptional program of human corneal epi- tent stem cells. PLoS One. 2016;11(10):e0165464.
thelium by suppressing epithelial-to-mesenchymal 112. Park SH, Kim KW, Chun YS, Kim JC. Human mes-
transition. Cell Rep. 2016;15(6):1359–68. enchymal stem cells differentiate into keratocyte-­
97. Sasamoto Y, Hayashi R, Park SJ, Saito-Adachi M, like cells in keratocyte-conditioned medium. Exp
Suzuki Y, Kawasaki S, et al. PAX6 isoforms, along Eye Res. 2012;101:16–26.
with reprogramming factors, differentially regulate 113. Syed-Picard FN, Du Y, Lathrop KL, Mann MM,
the induction of cornea-specific genes. Sci Rep. Funderburgh ML, Funderburgh JL.  Dental pulp
2016;6:20807. stem cells: a new cellular resource for corneal
98. Bayart E, Cohen-Haguenauer O.  Technological stromal regeneration. Stem Cells Transl Med.
overview of iPS induction from human adult somatic 2015;4(3):276–85.
cells. Curr Gene Ther. 2013;13(2):73–92. 114. Liu H, Zhang J, Liu CY, Wang IJ, Sieber M, Chang
99. Blazejewska EA, Schlotzer-Schrehardt U, Zenkel J, et al. Cell therapy of congenital corneal diseases
M, Bachmann B, Chankiewitz E, Jacobi C, et  al. with umbilical mesenchymal stem cells: lumican
Corneal limbal microenvironment can induce trans- null mice. PLoS One. 2010;5(5):e10707.
differentiation of hair follicle stem cells into corneal 115. Demirayak B, Yuksel N, Celik OS, Subasi C,
epithelial-like cells. Stem Cells. 2009;27(3):642–52. Duruksu G, Unal ZS, et al. Effect of bone marrow
100. Katikireddy KR, Dana R, Jurkunas and adipose tissue-derived mesenchymal stem cells
UV.  Differentiation potential of limbal fibroblasts on the natural course of corneal scarring after pen-
and bone marrow mesenchymal stem cells to corneal etrating injury. Exp Eye Res. 2016;151:227–35.
epithelial cells. Stem Cells. 2014;32(3):717–29. 116. Shimazaki J, Amano S, Uno T, Maeda N, Yokoi
101. Ricardo JR, Cristovam PC, Filho PA, Farias CC, N, Japan Bullous Keratopathy Study G.  National
de Araujo AL, Loureiro RR, et  al. Transplantation survey on bullous keratopathy in Japan. Cornea.
of conjunctival epithelial cells cultivated ex  vivo 2007;26(3):274–8.
in patients with total limbal stem cell deficiency. 117. Gorovoy MS.  Descemet-stripping automated endo-
Cornea. 2013;32(3):221–8. thelial keratoplasty. Cornea. 2006;25(8):886–9.
102. Kobayashi M, Nakamura T, Yasuda M, Hata Y, Okura 118. Price MO, Giebel AW, Fairchild KM, Price FW Jr.
S, Iwamoto M, et al. Ocular surface reconstruction Descemet’s membrane endothelial keratoplasty: pro-
with a tissue-engineered nasal mucosal epithelial spective multicenter study of visual and refractive
cell sheet for the treatment of severe ocular surface outcomes and endothelial survival. Ophthalmology.
diseases. Stem Cells Transl Med. 2015;4(1):99–109. 2009;116(12):2361–8.
103. Ksander BR, Kolovou PE, Wilson BJ, Saab KR, Guo 119. Price MO, Gorovoy M, Benetz BA, Price FW Jr,
Q, Ma J, et  al. ABCB5 is a limbal stem cell gene Menegay HJ, Debanne SM, et al. Descemet’s stripping
required for corneal development and repair. Nature. automated endothelial keratoplasty outcomes com-
2014;511(7509):353–7. pared with penetrating keratoplasty from the Cornea
104. Schatton T, Yang J, Kleffel S, Uehara M, Barthel SR, Donor Study. Ophthalmology. 2010;117(3):438–44.
Schlapbach C, et al. ABCB5 identifies immunoregu- 120. Okumura N, Ueno M, Koizumi N, Sakamoto Y,
latory dermal cells. Cell Rep. 2015;12(10):1564–74. Hirata K, Hamuro J, et  al. Enhancement on pri-
105. Frank MH, Frank NY.  Restoring the cornea from mate corneal endothelial cell survival in  vitro by
limbal stem cells. Regen Med. 2015;10(1):1–4. a ROCK inhibitor. Invest Ophthalmol Vis Sci.
106. Tan DT, Dart JK, Holland EJ, Kinoshita S. Corneal 2009;50(8):3680–7.
transplantation. Lancet. 2012;379(9827):1749–61. 121. Okumura N, Kinoshita S, Koizumi N.  Application
107. Pinnamaneni N, Funderburgh JL.  Concise review: of rho kinase inhibitors for the treatment of
stem cells in the corneal stroma. Stem Cells. corneal endothelial diseases. J Ophthalmol.
2012;30(6):1059–63. 2017;2017:2646904.
108. Du Y, Carlson EC, Funderburgh ML, Birk DE, 122. Hara S, Hayashi R, Soma T, Kageyama T, Duncan
Pearlman E, Guo N, et al. Stem cell therapy restores T, Tsujikawa M, et  al. Identification and potential
transparency to defective murine corneas. Stem application of human corneal endothelial progenitor
Cells. 2009;27(7):1635–42. cells. Stem Cells Dev. 2014;23(18):2190–201.
109. Zhu Q, Li M, Yan C, Lu Q, Wei S, Gao R, et al. 123. He Z, Campolmi N, Gain P, Ha Thi BM, Dumollard
Directed differentiation of human embryonic JM, Duband S, et al. Revisited microanatomy of the
stem cells to neural crest stem cells, functional corneal endothelial periphery: new evidence for con-
172 Y. Sasamoto et al.

tinuous centripetal migration of endothelial cells in corneal endothelial cell-like cells. Exp Ther Med.
humans. Stem Cells. 2012;30(11):2523–34. 2015;9(2):351–60.
124. Katikireddy KR, Schmedt T, Price MO, Price FW, 129. Joyce NC, Harris DL, Markov V, Zhang Z, Saitta
Jurkunas UV.  Existence of neural crest-derived B. Potential of human umbilical cord blood mesen-
progenitor cells in normal and fuchs endothe- chymal stem cells to heal damaged corneal endothe-
lial dystrophy corneal endothelium. Am J Pathol. lium. Mol Vis. 2012;18:547–64.
2016;186(10):2736–50. 130. Dai Y, Guo Y, Wang C, Liu Q, Yang Y, Li S, et al.
125. Zhao JJ, Afshari NA. Generation of human corneal Non-genetic direct reprogramming and biomimetic
endothelial cells via in  vitro ocular lineage restric- platforms in a preliminary study for adipose-derived
tion of pluripotent stem cells. Invest Ophthalmol Vis stem cells into corneal endothelia-like cells. PLoS
Sci. 2016;57(15):6878–84. One. 2014;9(10):e109856.
126. McCabe KL, Kunzevitzky NJ, Chiswell BP, Xia 131. Hatou S, Yoshida S, Higa K, Miyashita H, Inagaki
X, Goldberg JL, Lanza R.  Efficient generation of E, Okano H, et  al. Functional corneal endothelium
human embryonic stem cell-derived corneal endo- derived from corneal stroma stem cells of neural
thelial cells by directed differentiation. PLoS One. crest origin by retinoic acid and Wnt/beta-catenin
2015;10(12):e0145266. signaling. Stem Cells Dev. 2013;22(5):828–39.
1 27. Zhang K, Pang K, Wu X.  Isolation and trans- 132. Foster JW, Wahlin K, Adams SM, Birk DE, Zack DJ,
plantation of corneal endothelial cell-like Chakravarti S. Cornea organoids from human induced
cells derived from in-vitro-differentiated pluripotent stem cells. Sci Rep. 2017;7:41286.
human embryonic stem cells. Stem Cells Dev. 133. Zhang C, Du L, Sun P, Shen L, Zhu J, Pang K, et al.
2014;23(12):1340–54. Construction of tissue-engineered full-­ thickness
128. Chen P, Chen JZ, Shao CY, Li CY, Zhang YD, Lu cornea substitute using limbal epithelial cell-like
WJ, et al. Treatment with retinoic acid and lens epi- and corneal endothelial cell-like cells derived
thelial cell-conditioned medium in  vitro directed from human embryonic stem cells. Biomaterials.
the differentiation of pluripotent stem cells towards 2017;124:180–94.
Part III
Regenerative Surgery and Therapy of the
Ocular Surface Epithelium
Ocular Surface Epithelium:
Applied Anatomy
12
Harminder Singh Dua and Dalia G. Said

Key Points mitotic activity and are transient ampli-


• Anatomically the ocular surface con- fying cells. The adhesion complexes of
sists of the conjunctiva, limbal and cor- the basal epithelium are stronger  and
neal epithelium covered by the tear film. manifold at the periphery than centre;
Functionally, the blink reflex and the hence the central epithelium abrades
lacrimal drainage apparatus are also easily.
important components. • Turnover of corneal epithelium is main-
• The tear film affords an optical polish to tained by stem cells that reside in pali-
the corneal surface and also plays an sades of Vogt and limbal epithelial
important role in ocular surface defence. crypts.
• The corneal epithelium is made of five • In physiological homeostasis, the basal
layers of cells. The basal cells have epithelial cells can maintain the epithe-
lial cell mass, but the limbal stem cells
play a key role in response to injury and
disease.
• Damage or destruction of the limbus
leads to limbal stem cell deficiency that
H. S. Dua (*)
Academic Section of Ophthalmology, Division of
can alter the nature of the corneal epi-
Clinical Neuroscience, University of Nottingham, thelium and affect vision.
Nottingham, UK • The conjunctiva is an integral part of the
Department of Ophthalmology, Queens Medical mucosal immune system. Conjunctival
Centre, University Hospitals NHS Trust, epithelium is of variable thickness and
Nottingham, UK is endowed with goblet cells that secrete
e-mail: harminder.dua@nottingham.ac.uk
the major component of the  mucins of
D. G. Said the tear film.
Academic Section of Ophthalmology, Division of
Clinical Neuroscience, University of Nottingham,
• The cornea is the most sensitive struc-
Nottingham, UK ture of the human body and is supplied
Department of Ophthalmology, Queens Medical
by a dense network of sensory and tro-
Centre, University Hospitals NHS Trust, phic nerves from the fifth cranial nerve.
Nottingham, UK It also has an autonomic nerve supply,
Research Institute of Ophthalmology (RIO), which is less well defined in humans.
Cairo, Egypt

© Springer Nature Switzerland AG 2019 175


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_12
176 H. S. Dua and D. G. Said

The term ‘ocular surface’ (OS) has anatomical The corneal epithelium is a non-keratinized
and functional connotations. Anatomically it uniform stratified squamous epithelium derived
includes the mucosal lining of the lid margins from the surface ectoderm at 5–6 weeks of gesta-
and conjunctiva and the limbal and corneal epi- tion. It extends from limbus to limbus and is con-
thelium. Functionally the OS includes the eye tinuous with the limbal epithelium. It is formed
lids and associated blink reflex, the tear film of 4–6 layers and ranges in thickness from 40 to
together with the various glands that secrete tear 50 microns. Like all ‘covering epithelia’, its
components and the lacrimal drainage system. functions include selective diffusion of gases
The transparent cornea is the key component of (oxygen from the atmosphere), nutrients and
the OS, permitting the regular passage and focusing waste products of cell metabolism and at the
of light on its way to the retina. The OS functions as same time preventing excessive loss of fluid;
a unit to maintain the health and optical qualities of secretion of a variety of substances as part of nor-
the cornea. The OS also interfaces with the envi- mal physiology and in response to insult, such as
ronment and hence is endowed with an array of mucins and antimicrobial peptides; physical pro-
innate and adaptive immune mechanisms, coupled tection against abrasive forces; and a barrier
with an intricate sensory neuronal input providing function between the cornea and the outside
both trophic and protective defence responses, world. All layers of the epithelium are extremely
emphasizing the importance of vision, in evolution- well organized in the interest of maintaining
ary terms, for the survival of the species. transparency of the cornea. The absence of blood
The cornea extends from limbus (corneo-­ vessels and relative sparsity of intracellular
scleral junction) to limbus with a horizontal diam- organelles also contribute to corneal epithelial
eter of 11.5–12 mm and approximately 1 mm less transparency. The corneal epithelium is made of
vertically. It appears to be oval as the sclera three types of cells that demonstrate polarity
encroaches on the superior and inferior limbus; between the external and internal environment
however, if viewed from its inner aspect, the cor- [4] (Fig. 12.1).
nea appears circular with an average diameter of
11.7  mm. It is prolate in shape with a steeper Superficial Flat Cells  The superficial 2–3 lay-
radius of curvature centrally than peripherally. ers of epithelium are made of polygonal flat cells
Similarly, its thickness increases from about (50 × 5 microns) with flat nuclei (Fig. 12.1a, b).
0.53 mm centrally to approximately 0.71 mm at The anterior surface of the anterior most layer of
the periphery [1, 2]. Despite the highly organized cells shows well-developed ‘apical microvilli
and regular structure of the corneal epithelium, it “finger like projections” and micro plicae “ridge
is not a perfect optical surface. The optical polish like projections”’, which increase the surface
is provided by the tear film, which is crucial for area of the cell membrane and are covered by a
both vision and corneal health. charged glycocalyx layer secreted by the epithe-
The cornea is made of five layers, namely, lial plasma membrane. This strengthens the adhe-
from anterior to posterior, the epithelium, sion of the mucinous layer of the tear film to the
Bowman’s layer, stroma, Descemet’s membrane epithelial cell membrane. Deficiency in this layer
(DM) and endothelium. There is increasing evi- leads to tear film instability due to poor mucin
dence to support the presence of a distinct layer, adhesion that can be detected by a rapid tear film
the pre-Descemet’s layer (Dua’s layer), just ante- breakup time [5]. Wing cells (suprabasal cells):
rior to the DM [3]. Beneath the superficial layers, there are 2–3 lay-
ers of wing cells which have a central oval or
polyhedral cell body from which wing-like lat-
12.1 Corneal Epithelium eral processes extend on either side (20 microns
in diameter), convex anterior surface and a con-
Ocular surface epithelium can be considered in cave posterior surface. Both superficial cells and
three distinct regions: the corneal epithelium, the wing cells are postmitotic. Basal cells: The deep-
limbal epithelium and the conjunctival epithelium. est layer of the corneal epithelium, ‘the basal
12  Ocular Surface Epithelium: Applied Anatomy 177

a b c

d e f

Fig. 12.1 (a) Transmission electron micrograph (TEM) of separated from the underlying stroma showing the detached
the stratified corneal epithelium showing columnar cells in hemidesmosomes. (e) SEM of the multilayered corneal epi-
the basal layer; wing cells in the middle layers and flat thelium and the basement membrane from which the cells
squamous cells in the superficial layer. (b) Scanning elec- have been stripped off. (f) TEM of two basal epithelial cells
tron micrograph (SEM) of the anterior surface of the super- from a case of recurrent corneal erosion syndrome. The
ficial squamous cells showing the polyhedral outline. The boundary between the two cells (arrowhead) is clearly vis-
cell outlines are clearly visible. The dark ‘holes’ seen at the ible. The hemidesmosomes are abnormal (arrow) and rest
junction of cells are fixation artefacts. (c) TEM of corneal on collagenous debris (star) between the basal cells and
epithelial cells illustrating hemidesmosomal attachments Bowman’s membrane. (Adapted from author’s own publi-
(arrows) with the underlying substratum (amniotic mem- cation Browning et al. [16], Fig. 12.1f is from author’s own
brane). (d) TEM of basal epithelial cells that have been publication Dua et al. [15])

layer’, comprises of a single layer of columnar epithelium does not stain. However, if this barrier
cells (20 microns tall) that sit on a basement is breached or poor, fluorescein dye can outline
membrane, secreted by the basal cells. These cells in intact epithelial sheets as is seen when
cells are mitotically active (transient amplifying there is active migration of epithelium during
cells) (see below) [4, 6]. wound healing or when conjunctivalization of the
cornea occurs [7, 8].
Cell Attachments  Like all epithelia, the corneal
epithelial cells exhibit specialized cell-cell and
cell-basement membrane attachments that inte- Below the tight junctions and in all underlying
grate individual cells into a cohesive sheet of epi- cell layers, the space between adjacent cells is
thelium. The cell membranes of adjacent filled with a protein called E-cadherin, a glyco-
superficial cells close to the apex are tightly protein, which ‘cements’ the cells together to
attached or fused together along the circumfer- form anchoring junctions or zonula adherens.
ence to form tight junctions or zonula occludens. Zonula adherens in turn link with intracellular
This forms an effective seal that restricts passage cytoskeletal actin filaments. Desmosomes or
of small molecules, water and electrolytes. The macula adherens are another well-defined type of
main protein of this junction is occludin. Occludin cell-cell adhesion structure made of electron-­
is closely associated with another cell protein dense localized (spots) plaques on adjacent cell
called ZO-1, which in turn connects with the walls that are linked by transmembrane cadherin
intracellular cytoskeletal proteins. Occludin and proteins called desmogleins and desmocollins.
ZO-1 are targeted by invading pathogens that These, like in zonula adherens, link with intracel-
must breach this barrier to gain access to the epi- lular cytoskeletal filaments. These attachments
thelium. Fluorescein dye normally does not pen- give some flexibility to the cells, allowing indi-
etrate this barrier, and hence the normal vidual cells to change shape, within limits, whilst
178 H. S. Dua and D. G. Said

retaining solidarity with the epithelial sheet. microtubules. The latter are especially seen in
Communicating channels, called gap junctions, mitotic cells where they provide directional guid-
also exist between cells. These are not adherens-­ ance to the separating chromosomes.
type junctions. Specialized molecules called con-
nexins form a microscopic channel called a Basement Membrane  The basement mem-
connexon. Connexons from adjacent cells line up brane (BM), approximately 0.11–0.33 microns
to form a continuous channel between the two thick, is secreted by the basal epithelial cells
cells, called the gap junction. These channels and is made of several proteins mainly type IV
allow molecules less than a thousand kilo dal- collagen, laminin and nidogens (glycopro-
tons, but not proteins, to pass through [4, 9]. teins),  fibrin and fibronectin [13, 14].
Specialized adhesion complexes exist between Structurally, by transmission electron micros-
the basal cells and underlying basement mem- copy, it demonstrates a pale anterior part, the
brane and Bowman’s layer. About 25 percent of lamina lucida, and a dense posterior part, the
the cytoplasmic surface of the basal cell mem- lamina densa. It serves to provide a substrate for
brane of the central basal epithelial cells is cell adhesion, like a foundation for the epithelial
occupied by electron-dense plaques called
­ construct above it, facilitating cell migration
hemidesmosomes, into which cytoskeletal fila- and importantly providing a basis for cell polar-
ments (keratin) insert (Fig. 12.1c, d). Posterior to ity and maintenance of cell stratification [6]. In
and corresponding to the hemidesmosome is a diabetes, where the BM is thickened and dupli-
much narrower electron-dense line which gives cated, the anchoring filaments are not long
origin to fine fibrils, the anchoring fibrils, made enough to penetrate anterior Bowman’s layer
predominantly of collagen type VII.  Anchoring causing the epithelium to be loosely attached
fibrils traverse the lamina lucida and lamina and readily come off, making diabetics suscep-
densa of the basement membrane and insert into tible to corneal abrasions. A similar mechanism
anchoring plaques located in the anterior one to has been reported in recurrent corneal erosion
two microns of the Bowman’s layer. The major syndrome, where there is an accumulation of
protein of the anchoring plaques is laminin. Here, subepithelial collagenous debris that weakens
the type VII collagen of the anchoring fibrils and the attachment of the anchoring filaments [15]
plaques is intricately interwoven with the type I (Fig.  12.1e, f). When the BM is damaged, as
collagen of the Bowman’s layer. The key trans- after laser surgery, it takes a week or longer for
membrane protein of the adhesion complexes is new BM to be formed, but fibronectin provides
integrin. In addition to these specific attachments, a temporary matrix for the cells to migrate on.
the cell membranes of all suprabasal cells and of Attachment of epithelium after laser keratec-
the apical surface of basal cells show undula- tomy is strong making the removal of epithe-
tions, which interdigitate with adjacent cells lium for retreatment more difficult [6, 16].
making them ‘fit’ with each other like the pieces
of a jigsaw puzzle. In epidermolysis bullosa there The corneal epithelium is self-renewing.
is a genetic defect of type VII collagen causing Superficial corneal epithelial cells are continu-
the epithelium to lose attachment to the substrate, ously undergoing apoptosis,  shedding and are
all over the body [10]. replaced by cell migration from the deeper layers
The intracytoplasmic structural element of all of the epithelium with a complete turnover of the
epithelial cells, known as the cytoskeleton, is epithelium occurring in 5–7  days [17, 18].
made of intermediate filament proteins, the cyto- However, following corneal transplantation, epi-
keratins. Cytokeratin (CK) 3 (basic)/CK12 thelial rejection can occur for up to a year, sug-
(acidic) pair are specific to corneal epithelium gesting that donor cells are not ‘turned around’
[11, 12]. Cytoskeletal filaments attach to the dif- for at least a year in humans. It is likely that the
ferent adhesion structures described above. Other basal epithelium survives for much longer [19].
cytoskeletal filaments are actin filaments and Like all self-renewing systems, the drive for
12  Ocular Surface Epithelium: Applied Anatomy 179

regeneration and replacement of cells is provided also termed transdifferentiation, is the basis of the
by stem cells located at the limbus. clinical application of SC derived from one source
to regenerate and rejuvenate another failing organ.
Corneal epithelial stem cells are unipotent or com-
12.2 Limbus and Limbal mitted progenitors that in their physiological role
Epithelium only differentiate to corneal epithelial phenotype,
though in experimental conditions can be made to
The palisades of Vogt with their inter-palisade rete transdifferentiate into hair follicles. Stem cells
ridges, the limbal epithelial crypts, the specialized usually reside in a defined SC niche, which repre-
connective tissue architecture and the different sents the anatomical and physiological microenvi-
epithelial immunophenotypes, supported by a ronment of the SC. The inter-palisade rete ridges
highly organized vascular and nervous supply, of the palisades of Vogt and the limbal epithelial
provide the anatomical basis and the physiological crypts constitute the limbal SC niche [20–24].
microenvironment for the limbus to sustain the
limbal epithelial stem cells. Stem cells (SC) are Palisades of Vogt  The palisades of Vogt are radi-
progenitor cells that are r­esponsible for cellular ally oriented fibrovascular ridges that are most
replacement and tissue regeneration. They are the prominent along the upper and lower limbus and
ultimate source cells from which arise almost all readily visible in pigmented individuals due to the
other cells that constitute a given organ served by increased melanin content of the epithelial cells
the SC.  Stem cells are slow cycling and live as (Fig. 12.2a). The ‘valley’ between adjacent pali-
long as the organ or organism they serve. They sades is packed with epithelial cells that constitute
constitute a very small proportion of the total cell the inter-palisade (epithelial) rete ridge.
mass of the organ. Their major attributes are Impression cytology studies show that the limbal
potency and plasticity. Potency is the ability of a epithelial cells are smaller, with a higher cell den-
single SC to differentiate along different lineages sity and a greater nucleus to cytoplasm ratio com-
of cell types, for example, the fertilized ovum pared to central corneal epithelial cells [25, 26]
(totipotent SC), the mother of all SC, can produce (Fig. 12.2b, c). The connective tissue of the pali-
a whole baby and part of the placenta. With subse- sade contains arteries and veins, radially oriented
quent cell divisions, some of the potency is lost as narrow hairpin loops, nerves and lymphatics.
though the cells retain qualities of SC. There thus The arterial supply is from the anterior ciliary
exists a hierarchy of SC  – totipotent, pluripotent arteries derived from arteries of the rectus mus-
and multipotent. Plasticity is the ability of SC cles. The epithelium covering the palisade is two
serving one organ to start serving a different organ to three cells thick. The epithelial cells in inter-
when transplanted into that organ. This attribute, palisade rete ridges are 10–15 cells thick [27, 28].

a b c

Fig. 12.2 (a) The limbal palisades of Vogt are clearly and peripheral corneal epithelium. The limbal epithelial
seen as radially oriented slender lines in a pigmented indi- cells are small, densely packed and show a greater
vidual. They blend with the conjunctival epithelium nucleus-to-cytoplasm ratio compared to the peripheral
peripherally and transition into dot-like circular palisade and (c) central epithelial cells. (From author’s own publi-
structures centrally. (b) Impression cytology of the limbal cation Dua et al. [52])
180 H. S. Dua and D. G. Said

Basal cells of the rete ridges are (CK) 5/14+, palisade rete ridge and extend into the surround-
CK19+ and CK3/12+ [4, 12]. These cells also ing substantia propria peripherally, centrally
show staining for vimentin, epidermal growth fac- towards the cornea or circumferentially in a
tor receptors and alpha-enolase and are connexin clockwise or anticlockwise direction [26, 38]
43 (gap junction protein) negative [29–32]. An (Fig. 12.3). LEC are classed as minor and major,
ATP binding cassette transporter protein, ABCG2, with the major ones ranging from 40 to over 200
is considered to be a marker of a subpopulation of microns. They are widest at their origin and taper
cells that contain SC. ABCG2-­positive cells have to a narrow-rounded end at their termination. The
been located in the basal cells of the limbus [33, number of LEC range from 3 to 13 with an aver-
34]. These cells also show a specific isoform of age of 9 per eye. LEC are packed with epithelial
P63 (ΔNp63α), a transcriptional factor believed phenotype of basal, suprabasal and central cells.
to correlate with SC [35]. Other putative stem cell Basal cells are small with prominent nucleoli and
markers like nestin, cytokeratin 15 and SOX-9 dense marginated chromatin, which is different
transcription factor have also been identified in from the more superficial cells (Fig.  12.4a, b).
relation to limbal stem cells [36, 37]. The cell membrane of the basal cells in contact
with the basement membrane have complex
finger-­like processes that match corresponding
Limbal Epithelial Crypts  Limbal epithelial invaginations in the basement membrane and
crypts (LEC) are distinct solid cords of epithelial express multiple hemidesmosomes, allowing for
cells that arise from the posterior end of the inter-­ a firm attachment between these putative SC and

Fig. 12.3  Serial sections of a limbal epithelial crypt (LEC) from start to termination. The LEC is seen to arise from the
posterior end of the inter-palisade rete peg and extend into the underlying stroma, tapering to a point at its termination
12  Ocular Surface Epithelium: Applied Anatomy 181

a b c

Fig. 12.4 (a) Transmission electron micrograph (TEM) cell membrane are seen extending into corresponding
of the basal cells of the limbal epithelial crypt (LEC) invaginations of the basement membrane with multiple
showing prominent nucleoli and marginated chromatin. hemidesmosome attachments (arrows). (Adapted from the
(b) Central cells of the LEC. (c) Showing attachments of author’s own publication Shanmuganathan et al. [38])
the basal cells of the LEC. Cytoplasmic extensions of the

a b c

Fig. 12.5  In vivo confocal microscopic (IVCM) features plane is moved to the stroma underlying the cells. (c) A
of (a) the limbal palisades showing basal cells of the rete cluster of hyper-reflective cells is seen in deep stroma, rep-
pegs (arrow heads) and stroma of the palisades (arrows). (b) resenting the limbal epithelial crypt cells. (Adapted from
The reflectivity of these structures reverses when the IVCM the authors’ own publication Miri et al. [39])

the substrate (Fig. 12.4c). The peripheral corneal of the rete pegs appears hyper-reflective (bright),
basal cell attachments too are stronger than the and the stroma of the palisades, in the corre-
central epithelial attachments. Hence following sponding plane, appears hypo-reflective (dark)
injury, a peripheral rim of corneal epithelium (Fig.  12.5b). Deeper in the stroma, at places,
(and the limbal epithelium) remains attached clusters of bright cells are seen, probably repre-
even when a large area of the central cornea is senting limbal epithelial crypts (Fig.  12.5c). In
denuded. Cells from this rim migrate centripe- vivo demonstration of crypts with optical coher-
tally as convex sheets that close the defect and ence tomography is possible and helps in assess-
restore epithelial integrity. Lines of contact ment of limbal stem cell deficiency [41, 42].
between these sheets stain with fluorescein and
appear as the pseudo-dendrites seen during heal- Immunophenotypically, the LEC cells show
ing of corneal epithelial defects [7]. In vivo con- SC characteristics. Cytokeratin 3, which is a dif-
focal microscopy (IVCM) is a good clinical tool ferentiation marker, is present in central corneal
to examine the eye for assessment of both normal epithelium but absent from limbal and LEC
anatomy and evaluation of limbal stem cell defi- cells. CK19, which is a predominant cytokeratin
ciency [39, 40]. The stroma of the palisades of conjunctival cells, is seen in the LEC. Vimentin,
appears as a hyper-reflective band and the basal an intermediate filament that is largely expressed
limbal epithelial cells in the rete pegs appear in stromal cells and less differentiated cells, is
dark, outlined with a bright border (Fig. 12.5a). present in clusters of limbal basal cells and in
In the plane deeper to the basal cells. The stroma basal and suprabasal LEC cells. P63 and ABCG2,
182 H. S. Dua and D. G. Said

putative SC markers, are expressed maximally in suggested to have a role in corneal growth,
the LEC (Fig.  12.6a). Desmoglein 3 is a cell development and maintaining corneal stem cell
membrane protein that has a role in the forma- “stemness” by anchoring SC [43, 44]. Ki67
tion of desmosomes and in cell-cell adhesion. (MIB-1), a marker of cellular proliferation, and
Decreased expression of desmoglein 3, as seen CD34, a haematopoietic stem cell marker, are
in the LEC, is associated with increased colony not expressed in the LEC cells nor in the limbus
formation efficacy. Tenascin C, an extracellular adjacent to LEC bearing rete pegs. SC are known
matrix protein, is expressed prominently by the to lack gap junctions (connexin (Cx) 43). This is
BM of LEC.  It modulates cell adhesion and is believed to confer a survival advantage to the SC

b c d

Fig. 12.6 (a) A limbal epithelial crypt (LEC) stained for overlying epithelium are both positive for Cx43. (b) All
the transporter protein ABCG2 showing strong staining of cells of the central corneal epithelium are positive for
all cells in the LEC and along the basal cells of the rete Cx43. The rete peg basal cells some distance from the
peg. The staining becomes less intense towards the con- LEC are negative for Cx43, whilst the adjacent peripheral
junctival side of the LEC. (Adapted from author’s own corneal epithelial cells are positive (d). (Adapted from the
publication Dua et  al. [26]). (b–d) Immunoperoxidase author’s own publication Shanmuganathan et al. [38])
staining of LEC for connexin 43 (Cx43). The LEC and
12  Ocular Surface Epithelium: Applied Anatomy 183

a b c

Fig. 12.7  In vivo confocal microscopic (IVCM) features transition zone of the conjunctivalized area (*) demonstrate
of limbal stem cell deficiency. (a) The superficial layer prominent cell nuclei with no cell borders. The adjoining
shows hyper-reflective cells with ill-defined margins repre- area (+) of normal corneal epithelial cells shows bright,
senting conjunctivalization of the cornea (*). The adjacent well-defined borders. Arrowheads mark the junction. (c)
normal corneal epithelium (+) shows dark cells with well- Cystic spaces (arrowhead) and goblet cells, singly or in
defined margins. Arrowheads mark the junction of the two rosettes (arrow), are seen in conjunctivalized area. (Adapted
epithelial phenotypes. (b) The subsuperficial cells of the from the author’s own publication Dua et al. [19])

in that it isolates them from their progeny and a crucial role in healing and regeneration of
factors related to infection and stress in an adja- epithelium after injury but have a minimal role
cent cell cannot pass into the healthy cell. in physiological homeostasis of the corneal epi-
However, LEC cells express abundant Cx43. thelium, which can be served by the basal cells
This could be related to the range of non-gap [19, 49].
junction functions attributed to Cx43, which Clinically, the limbal stems cells can be dam-
includes a role in suppression of tumour growth aged by a variety of causes ranging from con-
and in differentiation and migration. Thus, the genital defects to chronic inflammation and
finding of Cx43+ cells in the LEC would be con- chemical burns. The corneal epithelium is
sistent with it being a putative niche for stem replaced by conjunctival epithelium (in mild
cells, and the Cx43+ cells could be niche cells cases) or by an extensive fibrovascular pannus
that support the division of stem cells [45–47] (in severe cases) [50, 51]. Conjunctival epithe-
(Fig. 12.6b). lium on the cornea on IVCM demonstrates
The corneal epithelial cell mass is main- hyper-reflective metaplastic cells with ill-defined
tained by the SC of the limbus/LEC, the prog- margins and goblet cells distributed singly or in
eny of which step out of the niche as transient rosettes (Fig. 12.7a–c). The fact that these eyes
cells (TC) and transient amplifying cells (TAC, can be successfully treated by limbal transplan-
basal cells), which migrate centripetally on the tation or ex vivo cultured limbal SC is a strong
corneal BM.  Postmitotic cells migrate anteri- indication of the importance of the limbus and
orly as the wing cells and the superficial cells. direct clinical confirmation of the limbal loca-
This is embodied in the X, Y, Z hypothesis of tion of SC [52–54].
corneal epithelial maintenance, where X = pro-
liferation of basal epithelial cells, Y = contribu-
tion to the cell mass by centripetal movement of 12.3 Conjunctiva
peripheral cells and Z = the epithelial cell loss
from the surface [48]. Healthy central islands The conjunctiva is the highly specialized muco-
of corneal epithelium survive despite total loss sal lining of the ocular surface that extends from
of limbal epithelium indicating that the central the lid margin, posterior to the grey line. It lines
epithelial cell mass can be sustained by the the posterior surface of the lids as the palpebral
basal cells (TAC). It is postulated that SC play or tarsal conjunctiva folds over to cover the eye
184 H. S. Dua and D. G. Said

a b

Fig. 12.8 (a) Haematoxylin- and eosin-stained section of staining of blood vessels in a section of inflamed conjunc-
the normal conjunctiva showing a multilayered epithelium tiva. The endothelial cells stain strongly for vascular adhe-
with goblet cells sitting on a basement membrane with sion protein-1. (Adapted from the author’s own publication
underlying substantia packed with resident immune cells, Haynes et al. [55])
vessels and connective tissue cells. (b) Immunoperoxidase

ball as the bulbar conjunctiva. The folds between which are non-specific responses of the conjunc-
the palpebral and bulbar conjunctiva are termed tiva to chronic inflammation.
the superior, inferior, medial and lateral fornices.
Palpebral conjunctiva is firmly attached to the Conjunctival Epithelium  The conjunctival epi-
underlying tarsal plates but the forniceal and bul- thelium is non-keratinized and squamous in
bar conjunctiva is loosely attached to the under- nature, two to three cells thick in all parts and
lying substantia propria, except close to the closely integrated with goblet cells, which are
limbus, where for about 2–3 mm the bulbar con- spread throughout the conjunctiva but especially
junctiva is firmly attached. Hence in grasping the concentrated in the nasal palpebral and fornicial
globe with forceps, it should be applied immedi- regions. At the lid margin, it is thicker and can be
ately next to the limbus where grasping the con- 10–12 cells thick. Conjunctival cells have abun-
junctiva provides a grasp on the globe as well. dant cytoplasmic organelles, unlike corneal
The substantia propria, the loose connective tis- ­epithelial cells. Along the basolateral surface of
sue underlying the conjunctival epithelium, is the basal cells of the conjunctiva are intraepithe-
highly vascularized and contains diffuse and lial CD8+ lymphocytes that are a feature of most
localized collections of lymphoid tissue made of mucosal epithelia [58] (Fig. 12.10). Goblet cells
lymphocytes, plasma cells, mast cells and neutro- are columnar, glandular cells found among most
phils (Fig. 12.8a). In inflamed states, the vessels mucosal epithelial cells including the conjunctiva.
of the substantia propria express high endothelial They are part of the mucosal barrier in innate
venules and lymphocyte homing receptor pro- defence and their secretion serves as an important
tein, vascular adhesion protein-1 for attachment part of the tear film to afford lubrication to the
and egress of circulating white cells [55, 56] ocular surface that interfaces with the environ-
(Fig. 12.8b). Unlike the corneal epithelial base- ment. They have a distinct morphology with
ment membrane, the conjunctival basement mucin vesicles and granules filling the apical part
membrane shows fenestrations through which of the cell, and the nucleus and other cell organ-
immune cells can migrate from the substantia elles packed at the bottom. When the mucin is dis-
propria to the surface during inflammation [57] charged, the cell assumes the shape of a ‘goblet’.
(Fig.  12.9a, b). Tufts of vascular proliferation Secretion of the mucin occurs by holocrine (the
covered by thickened epithelium constitute a cell membrane bursts to release the secretion
papillary reaction and collections of lymphoid effectively implying destruction of the cell) or
cells constitute a follicular reaction, both of apocrine (the apex of the cell pinches off to release
12  Ocular Surface Epithelium: Applied Anatomy 185

a b

Fig. 12.9 (a) Scanning electron micrograph (SEM) of junctiva a few hours after maintaining in tissue culture
basement membrane of the conjunctival epithelium show- medium showing mononuclear cells that have migrated
ing multiple fenestrations. (b) SEM of the denuded con- through the fenestrations

­ entioned above, is an integral part of the tear


m
film. Extensive injury to the conjunctiva like
chemical burns or chronic inflammation results in
loss of goblet cells as well with consequent mucin
deficiency and rapid tear film breakup time. The
ducts of the lacrimal and Meibomian glands open
into the conjunctiva. Extensive scarring of the
conjunctiva can lead to obstruction of the gland
orifices, depriving the OS of lacrimal secretions
leading to dry eyes [4, 9].

Tear Film and Conjunctiva  The tear film is a


dynamic complex structure served by numerous
cells and glands and continuously replenished
and evenly distributed across the OS by the
Fig. 12.10  Section of normal human conjunctiva stained blinking action of the eyelids. The hydrophobic
with the alkaline phosphatase anti-alkaline phosphatase surface of the corneal epithelium is rendered
technique for the human mucosal lymphocyte antigen-1 hydrophilic by the binding of mucin to the epi-
(HML-1). HML-1 positive cells are seen associated with
the basal epithelial cells as red dots. These cells are seen in
thelial cells. This allows the aqueous component
mucosal surfaces elsewhere in the body as well. (Adapted to form a relatively thick layer on the mucin bed,
from the author’s own publication Dua et al. [58]) upon which the Meibomian oil (meibum) spreads
as a thin layer, delaying both evaporation and the
the secretory granules or mucin-filled vesicles and breakup of the tear film. The tears contain a
some cytoplasm, but the cell survives). The con- number of proteins, growth factors, vitamins,
junctiva serves two major functions of protection electrolytes and a few cells that collectively con-
and lubrication. For OS defence the epithelial tribute to the nourishment, defence and health of
cells are capable of directly phagocytosing bacte- the ocular surface [59]. These include immuno-
ria and secrete substances such as antimicrobial globulins, lysozyme, lactoferrin and a number of
peptides that are a­ntibacterial and antiviral. antimicrobial peptides. Adaptive cell-mediated
Lubrication is provided by mucin, which, as and humoral responses by the resident and circu-
186 H. S. Dua and D. G. Said

lating lymphocytes contribute to the protection a distinct whorl pattern [65, 66] (Fig. 12.11). From
provided by the conjunctival mucosal immune the subbasal plexus, terminal nerve endings enter
system [60, 61]. the epithelium and terminate both inter and intra-
cellularly [67].
The cornea receives its nutrients from the Autonomic innervation consists essentially of
aqueous humour, the tear film and the limbal ves- sympathetic nerves from the superior cervical
sels. Oxygen from the atmosphere dissolved in ganglion (SCG). Fibres destined for the eye form
the tear film reaches the cornea; hence contact the sympathetic root of the ciliary ganglion.
lenses can potentially interfere with this and Some fibres directly merge with the long ciliary
cause adverse consequences. Equally noxious nerves and others pass through the ciliary gan-
agents such as carbon monoxide from smoke can glion, without synapse, to emerge in the short
also dissolve in tears and reach the cornea. ciliary nerves. Though there are few sympathetic
Deficiency of tears causes dry eyes with trouble- fibres in the cornea in humans [68, 69], they are
some symptoms and blurring of vision that an important component of the peri-limbal nerve
affects the quality of life of the patient. Surgical plexus.
interventions to restore sight loss such as corneal OS innervation serves sensory, trophic and
and limbal transplantation generally do not sur- vasomotor functions. The limbal vascular arcade
vive for long in a dry environment [53]. responds to corneal nerve stimuli by vasodilation
(circumcorneal congestion) which facilitates dia-
pedesis of white blood cells (triple response)
12.4 Corneal Innervation which invade the cornea to combat noxious and
infective agents. In situations where the innerva-
The cornea is the most sensitive structure in the tion is compromised, infective organisms can
human body, being a hundred times more sensi- flourish without the host response in a unique
tive than the conjunctiva [62] and over 400 times pattern of infection called infectious crystalline
more than the skin [63]. keratopathy [70]. Loss of sensory innervation of
Corneal innervation is predominantly sensory the cornea leads to neurotrophic keratopathy
from the ophthalmic division of the trigeminal wherein the OS suffers from tear film instability,
nerve via the long posterior ciliary nerves. These epithelial irregularity, persistent epithelial defects
nerves arborize to form the peri-limbal plexus, and stromal melting. Damage to corneal innerva-
which has both sensory and autonomic nerves and tion peripheral to and involving the trigeminal
is predominantly vasomotor in function [64]. ganglion results in loss of sensory and trophic
From the peri-limbal plexus, demyelinated radial function with risk of epithelial loss and stromal
nerves enter the mid-thickness of the corneal lysis. Damage central to the ganglion results in
stroma, about 11  in each quadrant. These travel sensory loss, whilst trophic functions are pre-
centrally and superficially to form a sub-­Bowman’s served [71, 72]. The prognosis in such cases is
plexus. Smaller twigs penetrate Bowman’s layer likely to be good.
and end in the subbasal plane of the epithelium, in
distinct terminal bulbs [65]. From the bulb-like Declaration of Interest None of the authors have any
structures finer neurites emerge as naked endings conflict of interest related to the subject matter and con-
(without the Schwann cells) to form the subbasal tent of the chapter. HS Dua is the consultant for Dompe,
Santen, Thea and Shire. He has shares in NuVision
plexus. The neurites are generally oriented such
BioTherapeutics and GlaxoSmithKline. No human or ani-
that they converge to an area between the upper mal studies were carried out by the authors for this
two thirds and the lower one third where they form chapter.
12  Ocular Surface Epithelium: Applied Anatomy 187

a b

c d

e f

Fig. 12.11  Photomicrographs of whole mount of normal Bowman’s membrane and each branch ends in a terminal
human cornea stained by the acetylcholinesterase tech- blub (arrowheads) which give rise to slender neurites. (e)
nique. (a) Subbasal nerve plexus with characteristic A single sub-Bowman’s nerve (arrow) gives rise to multi-
branching (arrows) and union/reunion (arrowheads). (b) ple branches (arrowheads), each of which ends in a termi-
A stromal nerve (arrowhead) penetrates Bowman’s mem- nal bulb on penetrating Bowman’s membrane. (f) A higher
brane (arrow) and ends in a terminal bulb from which a magnification of (e) showing the terminal bulbs (arrow-
leash of neurites of the subbasal plexus emerge. (c) A heads) and the leash of neurites emerging from the bulbs.
thick stromal nerve (arrow) emerges anterior to the (Adapted from the author’s own publication Al-Aqaba
Bowman’s membrane and ends in a blub (arrowhead). (d) et al. [65])
A stromal nerve bifurcates immediately after penetrating
188 H. S. Dua and D. G. Said

References 17. Hanna C, O’Brien JE. Cell production and migration


in the epithelial layer of the cornea. Arch Ophthalmol.
1960;64:536–9.
1. Warwick R. Eugene Wolff’s anatomy of the eye and
18. Hanna C, Bicknell DS, O’Brien JE. Cell turnover in the
orbit. 7th ed. Philadelphia: Saunders; 1976. p. 3.
adult human eye. Arch Ophthalmol. 1961;65:695–8.
2. Rufer F, Schroder A, Erb C.  White-to-white corneal
19. Dua HS, Miri A, Alomar T, Yeung AM, Said

diameter; normal values in healthy humans obtained
DG. The role of limbal stem cells in corneal epithe-
with the Orbscan II topography system. Cornea.
lial maintenance: testing the dogma. Ophthalmology.
2005;24:259–61.
2009;116:856–63.
3. Dua HS, Faraj LA, Said DG, Gray T, Lowe
20. Dua HS.  Stem cells of the ocular surface: scientific
J. Human corneal anatomy redefined: a novel
principles and clinical applications. Br J Ophthalmol.
­pre-­Descemet’s layer (Dua’s layer). Ophthalmology.
1995;79:968.
2013;120:1778–85.
21. Alison MR, Poulsom R, Forbes S, Wright NA.  An
4. Gipson IK.  Anatomy of the conjunctiva, cornea and
introduction to stem cells. J Pathol. 2002;197:419–23.
limbus. In: Smolin G, Thoft RA, editors. The cornea,
22. Fuchs E, Segre JA.  Stem cells: a new lease on life.
scientific foundations and clinical practice. New York:
Cell. 2000;100:143–55.
Little Brown and Company; 1994.
23. Janes SM, Lowell S, Hutter C. Epidermal stem cells. J
5. Farjo A, McDermott M, Soong HK. Corneal anatomy,
Pathol. 2002;197:479–91.
physiology, and wound healing. In: Yanoff M, Duker
24. Sehic A, Utheim ØA, Ommundsen K, Utheim

JS, editors. Ophthalmology. 3rd ed. St. Louis: Mosby;
TP.  Pre-clinical cell-based therapy for limbal stem
2008. p. 203–8.
cell deficiency. J Funct Biomater. 2015;6:863–88.
6. Nishida T.  Cornea. In: Krachmer JH, Mannis MJ,
25. Dua HS, Azuara-Blanco A.  Limbal stem cells

Holland EJ, editors. Fundamentals of cornea and
of the corneal epithelium. Surv Ophthalmol.
external disease. St Louis: Mosby; 1997.
2000;44:415–25.
7. Dua HS, Forrester JV.  Clinical patterns of cor-
26. Dua HS, Shanmuganathan VA, Powell-Richards AO,
neal epithelial wound healing. Am J Ophthalmol.
Tighe PJ, Joseph A. Limbal epithelial crypts: a novel
1987;104:481–9.
anatomical structure and a putative limbal stem cell
8. Dua HS. The conjunctiva in corneal epithelial wound
niche. Br J Ophthalmol. 2005;89:529–32.
healing. Br J Ophthalmol. 1998;82:1407–11.
27. Goldberg MF, Bron A.  Limbal palisades of Vogt.

9. Scott RA, Lauweryns B, Snead DM, Haynes RJ,
Trans Am Ophthalmol Soc. 1982;80:155–71.
Mahida Y, Dua HS.  E-cadherin distribution and
28. Van Buskirk EM.  Anatomy of the limbus. Eye.

epithelial basement membrane characteristics of
­
1989;3:101–8.
the normal human conjunctiva and cornea. Eye.
29. Zieske JD, Wasson M. Regional variation in distribu-
1997;11:607–12.
tion of EGF receptor in developing and adult corneal
10.
Gipson IK, Grill SM, Spurr SJ, Brennan
epithelium. J Cell Sci. 1993;106:145–52.
SJ. Hemidesmosome formation in vitro. J Cell Biol.
30. Chung EH, DeGregorio PG, Wasson M, et  al.

1983;97:849–57.
Epithelial regeneration after limbus-to-limbus
11. Chaloin-Dufau C, Pavitt I, Delorme P, Dhouailly D.
debridement. Expression of alpha-enolase in stem and
Identification of keratins 3 and 12 in corneal epithelium
transient amplifying cells. Invest Ophthalmol Vis Sci.
of vertebrates. Epithelial Cell Biol. 1993;2:120–5.
1995;36:1336–43.
12. Schermer A, Galvin S, Sun TT. Differentiation related
31. Zieske JD. Perpetuation of stem cells in the eye. Eye.
expression of a major 64k corneal keratin in vivo and
1994;8:163–9.
in culture suggests limbal location of corneal epithe-
32. Matic M, Petrov IN, Chen S, Wang C, Dimitrijevich
lial stem cells. J Cell Biol. 1986;102:49–62.
SD, Wolosin JM.  Stem cells of the corneal epithe-
13. Taylor HR, Kimsey RA. Corneal epithelial basement
lium lack connexins and metabolite transfer capacity.
membrane changes in diabetes. Invest Ophthalmol Vis
Differentiation. 1997;61:251–60.
Sci. 1981;20:548–53.
33. Chen Z, De Pavia CS, Luo L, Kretzer FL, Pflugfelder
14. Torricelli AAM, Singh V, Santhiago MR, Wilson

SC, Li DQ.  Characterization of putative stem cell
SE. The corneal epithelial basement membrane: struc-
phenotype in human limbal epithelia. Stem Cells.
ture, function, and disease. Invest Ophthalmol Vis Sci.
2004;22:355–66.
2013;54:6390–400.
34. Watanabe H, Okano T, Tano Y.  Human limbal epi-
15. Dua HS, Lagnado R, Raj D, Singh R, Mantry S, Gray
thelium contains side population cells expressing the
T, Lowe J. Alcohol delamination of the corneal epi-
ATP-binding cassette transporter ABCG2. FEBS Lett.
thelium: an alternative in the management of recurrent
2004;565:6–10.
corneal erosions. Ophthalmology. 2006;113:404–11.
35. Kawasaki S, Tanioka H, Yamasaki K, Connon CJ,
16. Browning AC, Shah S, Dua HS, Maharajan SV,

Kinoshita S. Expression and tissue distribution of p63
Gray T, Bragheeth MA. Alcohol debridement of the
isoforms in human ocular surface epithelia. Exp Eye
corneal epithelium in PRK and LASEK: an elec-
Res. 2006;82:293–9.
tron microscopic study. Invest Ophthalmol Vis Sci.
36. Sartaj R, Zhang C, Wan P, Pasha Z, Guaiquil V, Liu A,
2003;44:510–3.
Liu J, Luo Y, Fuchs E, Rosenblatt MI. Characterization
12  Ocular Surface Epithelium: Applied Anatomy 189

of slow cycling corneal limbal epithelial cells identi- 51. Dua HS, Joseph A, Shanmuganathan VA, Jones

fies putative stem cell markers. Sci Rep. 2017;7:3793. RE. Stem cell differentiation and the effects of defi-
37. Ghoubay-Benallaoua D, de Sousa C, Martos R, Latour ciency. Eye. 2003;17:877–85.
G, Schanne-Klein MC, Dupin E, Borderie V.  Easy 52. Dua HS.  Transplantation of limbal stem cells.

xeno-free and feeder-free method for isolating and Essentials in ophthalmology. Series Editors
growing limbal stromal and epithelial stem cells of Krieglstein GK and Weinreb RN. Cornea and external
the human cornea. PLoS One. 2017;12:e0188398. eye disease. Section editor Thomas Rienhard. Berlin/
38.
Shanmuganathan VA, Foster T, Kulkarni BB, Heidelberg: Springer; 2005. p. 34–56.
Hopkinson A, Gray T, Powe DG, Lowe J, Dua 53. Dua HS, Said DG. The ocular surface functional anat-
HS. Morphological characteristics of the limbal epi- omy, medical and surgical management. In: Guell JL,
thelial crypt. Br J Ophthalmol. 2007;91:514–9. editor. ESASO course series, Cornea. Basel: Krager;
39. Miri A, Al-Aqaba M, Otri AM, Fares U, Said
2015. p. 1–25.
DG, Faraj  LA, Dua HS.  In vivo confocal micro- 54. Rama P, Ferrari G, Pellegrini G.  Cultivated limbal
scopic features of normal limbus. Br J Ophthalmol. epithelial transplantation. Curr Opin Ophthalmol.
2012;96:530–6. 2017;28:387–9.
40. Miri A, Alomar T, Nubile M, Al-Aqaba M, Lanzini M, 55. Haynes RJ, Tighe PJ, Scott RA, Singh Dua H. Human
Fares U, Said DG, Lowe J, Dua HS. In vivo confocal conjunctiva contains high endothelial venules that
microscopic findings in patients with limbal stem cell express lymphocyte homing receptors. Exp Eye Res.
deficiency. Br J Ophthalmol. 2012;96:523–9. 1999;69:397–403.
41. Haagdorens M, Behaegel J, Rozema J, Van Gerwen 56. Knop N, Knop E.  Conjunctiva-associated lymphoid
V, Michiels S, Ní Dhubhghaill S, Tassignon MJ, tissue in the human eye. Invest Ophthalmol Vis Sci.
Zakaria N.  A method for quantifying limbal stem 2000;41:1270–9.
cell niches using OCT imaging. Br J Ophthalmol. 57. Chan JH, Amankwah R, Robins RA, Gray T, Dua
2017;101:1250–5. HS. Kinetics of immune cell migration at the human
42. Grieve K, Ghoubay D, Georgeon C, Thouvenin O, ocular surface. Br J Ophthalmol. 2008;92:970–5.
Bouheraoua N, Paques M, Borderie VM.  Three-­ 58. Dua HS, Gomes JA, Jindal VK, Appa SN, Schwarting
dimensional structure of the mammalian limbal stem R, Eagle RC Jr, Donoso LA, Laibson PR.  Mucosa
cell niche. Exp Eye Res. 2015;140:75–8. specific lymphocytes in the human conjunctiva, cor-
43. Yeung AM, Schlötzer-Schrehardt U, Kulkarni B,
neoscleral limbus and lacrimal gland. Curr Eye Res.
Tint NL, Hopkinson A, Dua HS.  Limbal epithelial 1994;13:87–93.
crypt: a model for corneal epithelial maintenance and 59. Paulsen F.  Functional anatomy and immunologi-

novel limbal regional variations. Arch Ophthalmol. cal interactions of ocular surface and adnexa. Dev
2008;126:665–9. Ophthalmol. 2008;41:21–35.
44. Kulkarni BB, Tighe PJ, Mohammed I, Yeung AM, 60. Knop E, Knop N.  Anatomy and immunology

Powe DG, Hopkinson A, Shanmuganathan VA, Dua of the ocular surface. Chem Immunol Allergy.
HS. Comparative transcriptional profiling of the lim- 2007;92:36–49.
bal epithelial crypt demonstrates its putative stem cell 61. Mohammed I, Said DG, Dua HS. Human antimicro-
niche characteristics. BMC Genomics. 2010;29:526. bial peptides in ocular surface defense. Prog Retin
45. Goodenough DA, Paul DL. Beyond the gap: functions Eye Res. 2017;61:1–22.
of unpaired connexon channels. Nat Rev Mol Cell 62. Wells JR, Michelson MA.  Diagnosing and treat-

Biol. 2003;4:285–94. ing neurotrophic keratopathy, EyeNet Magazine.
46. Omori Y, Yamasaki H. Mutated connexin43 proteins American Academy of Ophthalmology. 2008. https://
inhibit rat glioma cell growth suppression mediated www.aao.org/eyenet/article/diagnosing-treating-neu-
by wild-type connexin43 in a dominant-negative man- rotrophic-keratopathy. Accessed 02 Apr 2018.
ner. Int J Cancer. 1998;9:446–53. 63. Bonini S, Rama P, Olzi D, Lambiase A. Neurotrophic
47. Huang RP, Fan Y, Hossain MZ, Peng A, Zeng ZL, keratitis. Eye. 2003;17:989–95.
Boynton AL.  Reversion of the neoplastic phenotype 64. Marfurt CF, Cox J, Deek S, Dvorscak L.  Anatomy
of human glioblastoma cells by connexin 43 (cx43). of the human corneal innervation. Exp Eye Res.
Cancer Res. 1998;58:5089–96. 2010;90:478–92.
48. Thoft RA, Friend J. XYZ thoft hypothesis the X, Y, Z 65. Al-Aqaba MA, Fares U, Suleman H, Lowe J, Dua
hypothesis of corneal epithelial maintenance. Invest HS.  Architecture and distribution of human corneal
Ophthalmol Vis Sci. 1983;24:1442–3. nerves. Br J Ophthalmol. 2010;94:784–9.
49. Majo F, Rochat A, Nicolas M, Jaoudé GA, Barrandon 66. Patel DV, McGhee CN. In vivo confocal microscopy
Y.  Oligopotent stem cells are distributed throughout of human corneal nerves in health, in ocular and
the mammalian ocular surface. Nature. 2008;456: systemic disease, and following corneal surgery: a
250–4. review. Br J Ophthalmol. 2009;93:853–60.
50. Dua HS, Forrester JV.  The corneoscleral limbus
67. Stepp MA, Tadvalkar G, Hakh R, Pal-Ghosh S. Corneal
in human corneal epithelial wound healing. Am J epithelial cells function as surrogate Schwann cells
Ophthalmol. 1990;110:646–56. for their sensory nerves. Glia. 2017;65:851–63.
190 H. S. Dua and D. G. Said

68. Toivanen M, Tervo T, Partanen M, Vannas A,


architecture determines pattern of microbial spread.
Hervonen A. Histochemical demonstration of adren- Invest Ophthalmol Vis Sci. 2001;42:1243–6.
ergic nerves in the stroma of human cornea. Invest 71.
Dhillon VK, Elalfy MS, Al-Aqaba M, Dua
Ophthalmol Vis Sci. 1987;28:398–400. HS.  Anaesthetic corneas with intact sub-basal nerve
69. Sugiura S, Yamaga C.  Studies on the adrenergic
plexus. Br J Ophthalmol. 2014;983:417–8.
nerve of the cornea. Nippon Ganka Gakkai Zasshi. 72. Dhillon VK, Elalfy MS, Al-Aqaba M, Gupta A, Basu
1968;72:872–9. S, Dua HS.  Corneal hypoesthesia with normal sub-­
70. Butler TK, Dua HS, Edwards R, Lowe JS.  In vitro basal nerve density following surgery for trigeminal
model of infectious crystalline keratopathy: tissue neuralgia. Acta Ophthalmol. 2016;94:e6–e10.
Classical Techniques for Limbal
Transplantation
13
Rafael I. Barraquer and Juan Alvarez de Toledo

13.1 Historical Background In 1977, Thoft proposed “conjunctival trans-


plantation” for unilateral chemical burns [6].
Limbal transplantation for the management of This technique did not result in a functional cor-
severe ocular surface disease was proposed as neal surface, as it probably did not include cor-
early as 1964 by José I. Barraquer, during a discus- neal epithelial stem cells (CESC) in the donor
sion at the first World Cornea Congress [1]. He tissue. He further described a “kerato-­
described the use of “epithelial conjunctivo-­ epithelioplasty” procedure in 1984 [7], later
corneal limbus taken from the other eye” for the modified by Turgeon et  al. in 1990 [8]. This
treatment of superficial burns of a single affected involved the application of several thin disks or
eye [2]. In 1966, Strampelli and Restivo Manfredi lenticels from cadaveric peripheral cornea. These
published a cases of highly vascularized and grafts probably included only a small amount of
opaque corneas improved after a transplant of a donor CESC.  However, they represent the first
complete ring from the limbus of the fellow eye attempt at using allografts for ocular surface
[3]. The technique was described in more detail, reconstruction, which would allow treating bilat-
together with a second case, at the 2nd International eral diseases.
Corneo-Plastic Conference in London in 1967 [4]. Relevant research about epithelial regenera-
Joaquin Barraquer also performed some cases in tion and homeostasis can be traced back to the
Barcelona (Fig. 13.1) and later published his tech- 1940s. Studies showed increased frequency of
nique (Figs. 13.2 and 13.3) [5]. mitoses in the basal layer of peripheral cornea
using mitotic figure counts and radiated thymi-
dine [9–11]. However, the crucial role of the
R. I. Barraquer (*) “limbal stem cells” – or, more accurately, CESC –
Centro de Oftalmología Barraquer, Anterior Segment was understood only during the 1970s and 1980s,
Department, Barcelona, Spain following the works of Davenger and Evenson
Universitat Internacional de Catalunya, [12], Schermer [13], Kinoshita [14], and Potten
Barcelona, Spain and Loeffler [15].
Institut Universitari Barraquer, Universitat Autónoma In 1989, Kenyon and Tseng published a
de Barcelona, Barcelona, Spain series of limbal transplantations acknowledg-
J. Álvarez de Toledo ing the CESC theory [16]. They employed two
Centro de Oftalmología Barraquer, Anterior Segment arcuate segments of conjunctival and periph-
Department, Barcelona, Spain
eral corneal tissue from the fellow eye to treat
Institut Universitari Barraquer, Universitat Autónoma unilateral limbal stem cell deficiency (LSCD).
de Barcelona, Barcelona, Spain

© Springer Nature Switzerland AG 2019 191


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_13
192 R. I. Barraquer and J. Álvarez de Toledo

a b

c d

Fig. 13.1 (a) Right eye (RE) of a female patient who had affected by a chemical burn before autologous limbal trans-
a chemical burn in her left eye. This RE was amblyopic due plantation. (c) Immediate postoperative result of the CLAU
to childhood unilateral aphakia. The image shows the status 15 days after surgery. (d) One year after a penetrating kera-
of the ocular surface 2 months after removal of a 360° ring toplasty performed to rehabilitate the visual function. (Case
of limbal conjunctiva. (b) Left eye (LE) of the patient courtesy of Prof. Joaquin Barraquer, performed in 1981)

Their technique continues being the standard was modified by Kenyon and Rapoza to include
treatment for most unilateral severe ocular sur- limbus and conjunctiva from the living ­relative,
face disease, especially where ex  vivo expan- being the first description of a living-­related
sion techniques are not available. In 1994, Tsai conjunctival-limbal autograft (lr-CLAL) pro-
and Tseng first described a proper limbal cedure [19].
allograft technique using cadaveric donor tis- As an increasing number of techniques were
sue [17]. One year later, Kwitko et al. reported being published with different names, in 1996
the use of conjunctiva from a living-­related Holland and Schwartz proposed a classification
donor – siblings or other living relatives – for based on the anatomical location of the donor
treating bilateral LSCD [18]. Their technique ­tissue and whether the source is or not the same
13  Classical Techniques for Limbal Transplantation 193

a b

c d

Fig. 13.2 (a) Schematic design of a ring-shaped CLAU and (c) a peripheral keratectomy marked with a non-­
(Emilio Iglesias MD, PhD [5], 1981). A ring of limbal tis- perforating superficial 9 mm. trephination are performed
sue including 5 mm. of limbal conjunctiva and peripheral in the recipient’s eye. (d) Dissection of the peripheral lim-
corneal epithelium is obtained from the donor and bal area is performed with a crescent blade
anchored in the recipient’s eye. (b) A complete peritomy

a b

c d

Fig. 13.3 (a) Complete removal of the fibrovascular tissue is then anchored with sutures in the conjunctival and cor-
that covers the limbal area and cornea. (b) Recipient’s eye neal circumferences. (d) Limbal area starts repopulating the
is prepared with a bare ocular surface. (c) The ring autograft epithelial layer from the stem cells present in this area
194 R. I. Barraquer and J. Álvarez de Toledo

individual [20]. This should allow a better under- ogous and allograft conjunctival transplantation,
standing of each technique and its outcomes and kerato-epithelioplasty, homotransplantation of
improvement in the communication between sur- limbal cells, limbal transplantation, and autolo-
geons. A committee of the Cornea Society recently gous and allograft limbal transplantation. As this
reviewed and modified this classification and pro- was prone to cause confusion, the Cornea Society
posed that all corneal surgeons use the same established in 2008 a steering committee to pro-
nomenclature [21]. pose a more rational nomenclature. The ensuing
In this chapter, we will review the classical classification – and set of abbreviations – is based
surgical techniques for ocular surface rehabilita- on the anatomical origin of the transplanted tis-
tion, excluding ex  vivo expansion techniques sue (conjunctival, limbal, other mucosal), the
which will be described elsewhere. relation between donor and recipient (autograft,
cadaveric, living relative or non-relative
allografts), and whether the tissue was cultivated
13.2 Classification of Surgical ex vivo (Table 13.1). This classification and com-
Procedures mon nomenclature contribute to clarify the com-
munication, allowing a more accurate comparison
The techniques for ocular surface reconstruction of the outcomes of each procedure. This classifi-
were named as they emerged along several cation can be reviewed to eventually incorporate
decades, using an array of terms including autol- novel techniques.

Table 13.1  Classification of surgical and tissue-engineered procedures for ocular surface rehabilitation
Procedure Abbreviation Donor Transplanted tissue
Conjunctival transplantation
 Conjunctival autograft CAU Fellow eye Conjunctiva
 Cadaveric conjunctival allograft c-CAL Cadaveric Conjunctiva
 Living-related conjunctival allograft lr-CAL Living relative Conjunctiva
 Living-nonrelated conjunctival allograft lnr-CAL Living non-relative Conjunctiva
Limbal transplantation
 Conjunctival-limbal autograft CLAU Fellow eye Limbus/conjunctiva
 Cadaveric conjunctival-limbal allograft c-CLAL Cadaveric Limbus/conjunctiva
 Living-related conjunctival-limbal allograft lr-CLAL Living relative Limbus/conjunctiva
 Living-nonrelated conjunctival-limbal allograft lnr-ClAL Living non-relative Limbus/conjunctiva
 Keratolimbal autograft KLAU Fellow eye Limbus/cornea
 Keratolimbal allograft KLAL Cadaveric Limbus/cornea
Other mucosal transplantation
 Oral mucosa autograft OMAU Recipient Oral mucosa
 Nasal mucosa autograft NMAU Recipient Nasal mucosa
 Intestinal mucosa autograft IMAU Recipient Intestinal mucosa
 Peritoneal mucosa autograft PMAU Recipient Peritoneum
Ex vivo cultivated conjunctival transplantation
 Ex vivo cultivated conjunctival autograft EVCAU Recipients eye(s) Conjunctiva
 Ex vivo cultivated cadaveric conjunctival allograft EVc-CAL Cadaveric Conjunctiva
 Ex vivo cultivated living-related conjunctival allograft EVlr-CAL Living relative Conjunctiva
 Ex vivo cultivated living-nonrelated conjunctival EVlnr-CAL Living non-relative Conjunctiva
allograft
Ex vivo limbal transplantation
 Ex vivo cultivated limbal autograft EVLAU Recipients eye(s) Limbus/cornea
 Ex vivo cultivated cadaveric limbal allograft EVc-LAL Cadaveric Limbus/cornea
 Ex vivo cultivated living-related limbal allograft EVlr-LAL Living relative Limbus/cornea
 Ex vivo cultivated living-nonrelated limbal allograft EVlnr-LAL Living non-relative Limbus/cornea
Other ex vivo cultivated mucosal transplantation
 Ex vivo cultivated oral mucosa autograft EVOMAU Recipient Oral mucosa
13  Classical Techniques for Limbal Transplantation 195

The first feature to be considered is the ana- regarding the potential consequences for the
tomical type of transplanted tissue for ocular sur- donor eye – the patient’s fellow eye or from a liv-
face reconstruction. These include most ing relative.
frequently conjunctiva and/or limbus. However, The relation between donor and recipient is a
other tissues such as oral [22], nasal [23], perito- key parameter for the success and long-term
neal [24], or intestinal (rectal) mucosa have been prognosis in ocular surface transplantation, as
used in some cases. Oral mucosa is probably the can be represented by the degree of histocompat-
extraocular mucosal tissue most frequently used ibility. In this respect, the best source of tissue is
for ocular surface reconstruction, especially for the autologous origin, as no immune reaction will
recurrent pterygia, symblepharon, and recon- be present. This is, nevertheless, not an option in
struction of conjunctival fornices, as well as in cases with bilateral disease. The second best
biological keratoprosthesis procedures. source is a compatible living donor: a parent,
Conjunctival tissue is recognized as an impor- brother, or sister with at least half of the major
tant contributor to the ocular surface homeosta- histocompatibility antigens identical [26]. In this
sis. A healthy conjunctiva is absolutely necessary scenario, HLA types I and II should be deter-
in cases of LSCD, due to its crucial role in the mined for all the relatives and potential family
production of mucins  – from the goblet cells  – donors. In cases of non-related living or cadav-
and of cytokines. Conjunctival tissue can be har- eric donors, some degree of tissue matching
vested from the same eye, from the patient’s should be attempted to improve the results. Ex
fellow eye, or from a donor. It can be obtained vivo expansion techniques might offer the possi-
from either the bulbar or fornix conjunctiva. The bility of modifying the immunogenicity of the
latter has been reported as a greater source of cultivated cells [27] to decrease the immune
conjunctival stem cells [25]. However, bulbar rejection to the grafted cells [28].
conjunctiva is more commonly used due to an
easier harvesting.
Limbal tissue has been shown to be the locus 13.3 Conjunctival Transplants
of CESC. These appear to be present around the
limbus in a nonuniform distribution, as a number Conjunctival tissue has been widely used as
of discrete niches related to Vogt’s palisades. repairing tissue for many ocular surface diseases
When the tissue is obtained from limbal conjunc- such as pterygium [29], symblepharon [30], post-­
tiva alone (conjunctival-limbal graft), it will glaucoma surgery complications [31] like bleb
include few if any CESC.  On the other hand, a perforations and valve or tube extrusions, tumor
keratolimbal tissue segment including the periph- excision [32], and other conjunctival defects. In
eral and superficial cornea will contain most of most cases conjunctiva is used as an autograft
the CESC present in the harvested limbal sector. (CLAU), from the same eye or from the patient’s
The first type is mainly used for the reconstruc- fellow eye. In cases of bilateral disease, conjunc-
tion of a conjunctival defect adjacent to the cor- tiva from a living-related donor or a cadaveric
nea, as in pterygium or limbal tumor surgery, or donor can be employed. Nevertheless, conjuncti-
in cases with localized LSCD in which there is a val allograft represents nowadays only a small
wide, healthy, and functional limbus in the percentage of conjunctival transplants.
remaining sectors of the limbus. In conjunctival-­ At the donor site, conjunctiva must be care-
limbal grafts, the CESC from the donor site are fully dissected and separated from Tenon’s cap-
preserved. The keratolimbal tissue is the pre- sule. Inclusion of this layer will induce retraction
ferred tissue in cases with severe or complete and fibrosis of the graft in the postoperative
LSCD, as it not only repairs the limbal area but period. It is advised to mark the perimeter of the
contributes to the repopulation of the corneal sur- graft with a marker pen before harvesting
face with new epithelial corneal cells. As CESC (Fig. 13.4). This will allow a better size match-
are removed from the harvested site in keratolim- ing to the recipient bed (Fig. 13.5). The graft can
bal grafts, this should be taken into account be secured in place with biological (fibrin) glue
196 R. I. Barraquer and J. Álvarez de Toledo

Fig. 13.6  Conjunctival graft is placed with epithelium


facing up and adjusted in position with its limbal area 1 to
2 mm. away from the recipient’s corneal limits to avoid
postoperative dellen formation when the grafts naturally
swell until revascularization occurs. Peripheral borders
are also ironed, extended, and contacting with the recipi-
Fig. 13.4  Marking of the conjunctival autograft with a ent’s conjunctiva
pen to delimitate the area of the conjunctiva to be
dissected

Fig. 13.5  Conjunctival autograft is placed over the cor- Fig. 13.7  Donor area is covered again with the conjunc-
nea with stromal face up, and fibrin glue bioadhesive is tival tissue attached with fibrin glue to avoid scarring and
applied onto the bare sclera and on the autograft. Care symblepharon formation that may rarely occur in some
must be taken in positioning the limbal area of the donor races
graft in the limbal area of the recipient when it is flipped
and adjusted
Conjunctival allograft is used only in cases in
which an autograft tissue cannot be obtained, and
(Fig.  13.6), several 10-0 nylon interrupted amniotic membrane is not suitable or indicated.
stitches, or a continuous running suture. The A combined procedure using CLAU and lr-­
donor site can be closed with the same tech- CLAL has been recently described in two cases
niques (Fig. 13.7). However, a purely conjuncti- of unilateral severe chemical/thermal burn. The
val (non-­Tenon) donor site of moderate size in rationale for this approach is the combined use of
an otherwise healthy upper conjunctival sector – conjunctiva from the patient’s other eye to
which will be protected by the upper eyelid  – account for the conjunctival and goblet cell
may be left to spontaneous reepithelialization. ­deficiency and an HLA-compatible conjunctival
The conjunctival autograft technique is mostly allograft from a living-related donor with a good
employed in pterygium surgery but can be used ABO and HLA matching to cover the larger
for repairing a perforated filtering bleb or the defect. This combination reduces the antigen
extrusion of a glaucoma drainage device load, and no immune episodes were documented
(Fig. 13.8). during the follow-up period  – although both
13  Classical Techniques for Limbal Transplantation 197

a b

c d

Fig. 13.8 (a) Extrusion of a Baerveldt implant through fellow eye and a thin layer of a collagen implant
the conjunctiva in a patient affected of secondary glau- (Tutopatch®). (c) Dehiscence of the conjunctival wound
coma due to anterior chamber epithelial ingrowth. (b) and recurrence of the valve extrusion. (d) Definitive repair
Reconstruction of the extrusion with conjunctiva from the with a new Tutopatch® implant and oral mucosal graft

patients were under immunosuppression with trolled using a long-term immunosuppression


oral tacrolimus and mycophenolate mofetil. protocol under a strict monitoring and supervi-
Using a conjunctival or limbal allograft sion by a specialist [36]. Multiple drug protocols
involves a number of risks including immune include combinations of oral steroids, double
rejection, systemic toxicity due to the long-term immunosuppression with tacrolimus and myco-
immunosuppressive drugs, the possibility of phenolate mofetil, valganciclovir, and TMP/
inducing a LSCD in the donor eye [33], trans- SMX for prophylaxis of cytomegalovirus and
mission of a donor infection [34], even the Pneumocystis carinii. These are usually contin-
recently described transmission of a conjunctival ued for 3 years after the surgery if ocular condi-
neoplasia [35]. Immune rejection can be con- tions are stable.
198 R. I. Barraquer and J. Álvarez de Toledo

LSCD has never been described in a healthy tion of the peripheral superficial corneal stroma
donor eye when at least half of the limbal circum- is then performed, dissecting approximately 1.0–
ference had been respected. However, the potential 1.5 mm of the peripheral cornea. This donor tis-
donor eye must be carefully inspected at the slit sue will include most of the CESC niches present
lamp to rule out any subclinical LSCD. The pres- at the local Vogt’s palisades. A KLAU can be
ence of any kind of malignancy or infectious trans- obtained from the same eye (Fig. 13.11), when a
missible disease disqualifies a potential allograft translocation of the limbus is used for treating a
donor, because the required immunosuppression localized LSCD, or from the fellow eye in cases
would enhance risk of transmission and could favor with severe unilateral LSCD (Fig. 13.12).
the growth of dormant metastatic cells. However, CLAU or KLAU can be combined with a
in another reported case of transmission of a donor- lamellar (Fig.  13.13) or penetrating keratoplasty
derived breast carcinoma [37], the activation of the (Fig.  13.14), to improve ocular surface stability
malignant cells in the recipient coincided with the and avoid neovascularization of the graft in cases
reduction of the oral immunosuppression. of LSCD associated with corneal opacities. After
dissecting all the superficial corneal neovascular-
ization and limbal conjunctiva, eliminating all the
13.4 Limbal Autograft Techniques perilimbal fibrotic tissue, a lamellar or penetrating
keratoplasty is performed following the standard
Conjunctival-limbal and keratolimbal autografts technique. Once the suture is completed, a 90°
(respectively, CLAU and KLAU) are the ocular (3-h) autograft is obtained from the fellow eye and
surface rehabilitation procedures with the best anchored with monofilament sutures (10-0 nylon
prognosis [38]. The use of autologous tissue or 11-0 polyester), usually at the upper limbus.
avoids the risk of immune rejection, as well as A combined conjunctival autograft and kera-
the traumatic maneuvers involved in harvesting a tolimbal allograft (CLAU + KLAL), sometimes
postmortem tissue and preserving the tissue in referred as “the modified Cincinnati procedure”
culture media  – as is the case in a majority of [39], uses two fragments of recipient’s ­conjunctiva
allograft techniques, except for the relatively rare obtained from the fellow eye and two sectors of a
living-related sources of tissue. cadaveric donor keratolimbal ring. The conjunc-
The donor tissue in CLAU comprises con- tival grafts are placed superiorly and inferiorly,
junctiva with some peripheral limbal epithelial while the keratolimbal allograft sectors are
cells (Fig. 13.9a, b), while in KLAU, it includes placed nasally and temporally.
the limbal conjunctiva and peripheral superficial Another way of transplanting CESC from a
corneal  – stroma and epithelium  – including donor fellow eye or from an allograft has been
Vogt’s palisades (Fig. 13.9c, d). proposed as “single epithelial limbal transplanta-
In CLAU, donor conjunctiva is dissected care- tion” (SLET) [40]. In this technique, a human
fully separating Tenon’s capsule to its limbal amniotic membrane (hAM) is grafted on the bare
insertion, and a superficial sheet of limbal ocular surface with fibrin glue. Small fragments
(peripheral corneal) epithelium is cut with scis- of limbal tissue from the fellow eye or a fresh
sors, without including Bowman’s membrane. cadaveric corneoscleral rim are placed on the
This technique is mainly used for pterygium sur- cornea covered with the hAM graft in a circular
gery (Fig.  13.10). However, there is little hard fashion avoiding the visual axis. A layer of fibrin
evidence on whether the limbal component of glue is applied to fixate the small pieces of grafted
these grafts represents any significant benefit tissue, and finally a bandage soft contact lens is
compared to a standard conjunctival graft. fitted over the cornea. These multiple fragments
In KLAU, a portion of the peripheral superfi- of CESC-containing tissue will originate a new
cial corneal stroma and epithelium is included in epithelial multilayer, stimulated by the known
the donor tissue. After dissecting the conjunctiva beneficial effects of hAM has on cell growth.
centripetally, a 150–200 μm groove is performed In cases of autografts, the patient’s main con-
with a blade at the limbal sclera; lamellar dissec- cern relates to the possible consequences of
13  Classical Techniques for Limbal Transplantation 199

a b

c d

Fig. 13.9 (a) Delimitation of the area of a CLAU. Only included. (c, d) KLAU dissection includes peripheral
an epithelial part of the limbal area is removed. (b) superficial corneal stroma, epithelium, and superficial
Dissection of the conjunctiva is done splitting the Tenon’s limbal sclera to ensure the inclusion of all the niches of
capsule, which will induce retraction of the graft if the limbal epithelial stem cells

removing a limbal area from the healthy eye. In Rare and relatively minor complications have
the CLAU procedure, the preferred area for har- been reported after KLAU [41], including infec-
vesting the donor tissue is the superior or supero-­ tions of the donor site, filamentary keratitis, neg-
temporal conjunctival quadrant  – of either the ative fluorescein staining, and subconjunctival
fellow or the same eye  – due to the protection hemorrhage. Astigmatism can be induced if the
offered by the upper lid. Scarring and fibrosis of corneal stroma is removed too deep or too cen-
the donor area, even symblepharon in the upper trally. It has been documented that the CESC
fornix, can occur when the donor area is not repopulate completely the donor area within
properly repaired, especially in certain ethnic 1  year after the procedure [42]. In only one
groups predisposed to scarring, and in cases instance, a case of Mooren’s ulcer has been
when postoperative steroid treatment poorly reported after a CLAU procedure for pterygium
complied by the patient. surgery [43].
200 R. I. Barraquer and J. Álvarez de Toledo

a b

c d

Fig. 13.10 (a) Recurrent pterygium after four surgeries the lateroversion. (c) Result after careful removal of the
with multiple corneal adherences. (b) Symblepharon of symblepharon, CLAU, and superficial keratectomy. (d)
the infero-nasal fornix with involvement of the inferior Slit lamp appearance of the cornea with moderate thin-
lacrimal punctum, retraction of the eyelid, and diplopia in ning of the inferior stroma after the keratectomy

a b

c d

Fig. 13.11 (a) Keratolimbal autograft (KLAU) in a Complete epithelial layer with no fluorescein staining
patient with a unilateral chemical burn in his RE. One day 6  days after the surgery. (d) Stable ocular surface and
after surgery with two autografts positioned in the supe- improvement in corneal transparency 9  years after
rior and inferior limbus. (b) Centripetal reepithelialization KLAU. Patient’s BCVA reached 0.9 with RGP CL
from the limbal autografts (fluorescein staining). (c)
13  Classical Techniques for Limbal Transplantation 201

a b

c d

Fig. 13.12 (a) Limbal stem cell deficiency (LSCD) in the inferior quadrant of the same eye was used because of
the only eye of a patient in which more than 20 intravitreal the absence of the other eye (lost in his infancy) and
anti-VEGF have been injected through the upper nasal grafted in the supero-nasal area. (c) A complete healthy
sclera. Toxic effect of the antiseptic or anesthetic drugs epithelium covers the cornea 2 months after the KLAU.
applied during the injections had been involved in the (d) Corneal superficial stroma is still with a tenuous opac-
etiopathogenesis of this case of LSCD. (b) KLAU from ity but BCVA and subjective symptomatology improved

a b

c d

Fig. 13.13 (a) Corneal opacity and neovascularization limbal conjunctiva with a CLAU from the fellow eye, a
after previous failed pterygium surgery. (b) Lamellar ker- lamellar corneal graft was sutured in the corneal bed. (d)
atectomy was performed manually until reaching a trans- Final result 6 months after the procedure with a stable
parent corneal stromal plane. (c) After reconstructing the ocular surface and a transparent corneal graft
202 R. I. Barraquer and J. Álvarez de Toledo

a b

c d

Fig. 13.14 (a) Total superficial corneal neovasculariza- bus. (c) KLAU fixed in position with good revasculariza-
tion after a previous conjunctival flap performed to treat a tion. (d) Transparent corneal graft and stable ocular
bacterial keratitis with risk of perforation. (b) Combined surface 1 year after the surgical procedure
penetrating keratoplasty with KLAU in the superior lim-

KLAU are generally very successful provided SLET combined with keratoplasty, and postopera-
some guidelines are strictly followed. The placement tive loss of the transplants, which highlights the
of a healthy keratolimbal tissue from the fellow eye importance of performing the procedure in quiet
replenishes the pool of CESC of the recipient eye, eyes without inflammation and with previously
promotes increase in transparency of the corneal repaired eyelid or conjunctival malposition.
stroma improving vision, and enhances ocular com-
fort. However, KLAU will not work if the ocular sur-
face of the recipient’s eye is severely inflamed, and 13.5 Limbal Allograft Techniques
there is a tear film deficiency or eyelid malposition.
All these factors should be medically or surgically In cases with total bilateral LSCD, allogenic
corrected to ensure the survival of the CESC and a CESC transplantation is the only way to repopu-
correct regrowth of a corneal epithelial layer. Corneal late the affected ocular surface with these pro-
stroma should also be uninflamed; intrastromal neo- genitor cells of the corneal epithelium. The first
vascularization should be treated or decreased before descriptions of successful limbal allografts were
the KLAU is performed. Ocular surface reconstruc- reported in the mid-1990s. Turgeon et al. reported
tion procedures must always be rationally staged in on 13 patients in which kerato-epithelioplasty was
order to achieve an optimal final result. performed to stabilize the ocular surface affected
A large series of 125 eyes treated with autolo- with persistent epithelial defects [8]. Tsai and
gous SLET found an overall success of 76% after Tseng reported a series of 16 eyes with several
1.5 years of follow-up, with progressive conjuncti- causes of LSCD (chemical burns, Stevens-Johnson
valization in 18.4% of treated eyes [44]. Main fac- syndrome, congenital sclerocornea, Terrien’s
tors of failure were acid injury, severe symblepharon, degeneration, and chronic conjunctivitis) in which
13  Classical Techniques for Limbal Transplantation 203

a limbal ring-shaped allograft was grafted as a cularization and scarring of the ocular surface
source of CESC [17]. (Fig. 13.16).
Limbal (keratolimbal) allograft or KLAL can
be performed in different modalities: (1) sectorial
KLAL, which corresponds to the first description
of kerato-epithelioplasty by Thoft; (2) ring-­shaped
KLAL; (3) a combined ring-shaped KLAL with
lamellar or penetrating keratoplasty; and (4) mul-
tiple fragments of limbal allogenic tissue covered
by hAM and fibrin glue (SLET allograft).
Sectorial KLAL (Fig. 13.15) consists in graft-
ing two arcuate segments of limbal tissue from a
donor (living related or cadaveric) after perform-
ing a 360° peritomy and removal of the pannus
that usually covers the cornea in these cases.
When the recipient Bowman’s membrane is
absent, a hAM can be transplanted under the lim-
bal allografts, covering the corneal stroma to pro-
mote the epithelial repopulation, decrease the Fig. 13.15 Schematic representation of a sectorial
stromal inflammation, and inhibit the neovascu- KLAL. After removal of the pannus and fibrovascular tis-
sue in the recipient’s cornea and limbal area, two
larization. Sectorial KLAL is indicated in cases
3-o’clock-hour-wide limbal allografts are placed in the
with total bilateral LSCD without extensive vas- superior and inferior limbus

a b

c d

Fig. 13.16 (a) Bilateral LSCD in a patient due to chronic New epithelial centripetal growth from the two limbal
contact lens abuse. Persistent epithelial defects and super- allografts. (d) Complete reepithelialization and recovery
ficial stromal scarring. (b) Sectorial KLAL with two lim- of the corneal transparency observed 1 year after the
bal grafts placed in the vertical meridian of the limbus. (c) reconstructive surgery
204 R. I. Barraquer and J. Álvarez de Toledo

Ring-shaped KLAL is the most widely used


technique to treat bilateral LSCD (Fig. 13.17). A
ring of keratolimbal tissue is obtained from a
fresh cadaver eye (Fig. 13.18), with the inner and
outer diameters appropriately marked with dif-
ferent trephines. Ring width and thickness should
be enough to include the corneal-scleral limbal
area including the CESC niches. As in the secto-
rial technique, a complete dissection of the super-
ficial corneal and limbal pannus in the recipient’s
eye is mandatory, hAM can be applied to pro-
mote reepithelialization, and a stable tear film
and good eyelid function is required to achieve a
long-term success (Fig. 13.19).
Ring-shaped KLAL may also be combined Fig. 13.17  Schematic representation of a ring-shaped
KLAL. After removal of the pannus and fibrovascular tis-
with simultaneous lamellar or penetrating kerato-
sue in the recipient’s cornea and limbal area, a ring-shaped
plasty (Fig.  13.20). This multiple procedure is limbal allograft including limbal conjunctiva, superficial
indicated when the central cornea is in poor con- sclera, and cornea is placed in the limbal area and fixed
dition or rapid visual recovery is demanded by the with sutures

a b

c d

Fig. 13.18 (a) Harvesting of a ring-shaped KLAL from a (c) The outer scleral diameter is also marked and tre-
fresh cadaveric eye. Conjunctiva is cut 3–5 mm. from the phined with a 13–14 mm. trephine. (d) A careful lamellar
limbus and reflected over the corneal surface. (b) The dissection is performed to obtain a ring of limbal tissue
inner diameter of the ring is marked with an 8–9 mm. tre- containing all the pool of the donor’s limbal stem cells
phine which penetrates 150–200 μ in the corneal stroma.
13  Classical Techniques for Limbal Transplantation 205

a b

c d

Fig. 13.19 (a) Bilateral chemical burn with superficial layer with epithelial healing lines that are visible with
pannus and central corneal leukoma. (b) Ring-shaped fluorescein in the central cornea. (d) Final result 1 year
KLAL 4 days after surgery. Only the outer border of the after surgery with stable epithelium and clear cornea.
tissue was sutured with 8-0 Vicryl sutures, and no suture Patient is under oral cyclosporine A
was used to fix the inner border. (c) Complete epithelial

patient (Fig. 13.21). After removing all the limbal


fibrotic tissue and superficial neovascularization,
a penetrating or deep lamellar corneal graft is
anchored with eight 10-0 nylon single sutures; the
ring-shaped allograft is then placed and sutured
with several 10-0 nylon or 8-0 Vicryl sutures to
the sclera in the outer borders of the graft. After
that, eight more nylon sutures are placed includ-
ing the central donor graft, the recipient’s periph-
eral cornea, and the inner border of the limbal
allograft. The previous eight sutures that were
used to secure the central corneal graft at the
Fig. 13.20 Schematic representation of a ring-shaped beginning are then removed, and eight new nylon
KLAL combined with central lamellar or penetrating ker- sutures that attach all the three tissues are placed
atoplasty. After removal of the pannus and fibrovascular
again. Finally, conjunctiva of the donor and the
tissue in the recipient’s cornea and limbal area, a central
lamellar or penetrating keratoplasty is performed and recipient is secured with 8-0 or 9-0 Vicryl.
fixed with 8-0 nylon sutures. Then, a ring-shaped limbal A single large-diameter lamellar or penetrat-
allograft (A) including limbal conjunctiva, superficial ing keratoplasty, trephined eccentrically in the
sclera, and cornea is placed in the limbal area and fixed
donor (Fig.  13.22), will include a sector of the
with 8-0 Vicryl sutures in the scleral side. The central graft
(B) is sutured with eight more sutures, and the previously donor’s limbal area with its stem cells. This pro-
placed eight sutures are removed and changed for another cedure, named limbo-keratoplasty by its authors
eight sutures that include the corneal graft, the recipient’s [45] in 1994, is technically less demanding than
peripheral cornea, and the rink KLAL
206 R. I. Barraquer and J. Álvarez de Toledo

Fig. 13.21  Combined ring-shaped KLAL with penetrat- Typical postoperative intra-tissular hemorrhage that
ing keratoplasty in a case of Stevens-Johnson syndrome. occurs before re-connection of the blood microcirculation
Limits of the donor’s conjunctiva are highlighted (arrows). has been completed

a b

c d

Fig. 13.22 (a) Limbo-keratoplasty of 9 mm of diameter with fluorescein in the superior and central cornea; epithe-
performed in the single eye of a congenital aniridia-­ lial defects and conjunctival cell staining pattern in the
affected patient. (b) Around 40% of the donor limbus is inferior cornea, demonstrating that LSCD is present in the
included in the graft when trephined eccentrically. (c) inferior limbus. Graft failed 3 years after surgery due to
Epithelial irregularities in the inferior cornea seen under recurrence of LSCD
retro-illumination. (d) Normal corneal epithelium pattern

the previously described but has more chance of procedure in non-severe special indications like
failure due to the smaller proportion (up to 40%) gelatinous corneal dystrophy [46].
of the limbal zone included in the graft. With the Success with allogeneic SLET  – with donor
recent advances in immunosuppression, better tissue  – has also been reported. The minimal
long-term results have been reported with this amount of tissue needed to assure a correct
13  Classical Techniques for Limbal Transplantation 207

growth of epithelial cells has been established by a strict clinical or laboratory follow-up or medica-
in vitro studies in about 0.3 mm2 of a live limbal tions, and significant health issues like diabetes,
fragment including a CESC niche [47]. uncontrolled hypertension, renal insufficiency,
Depending on the donor source (cadaver or live severe heart diseases, or other organ failures. Age
tissue), the growth potential is different, being over 70 is also a contraindication, and patients
necessary a larger amount of cadaveric tissue among 60–70 years will be selected for immuno-
(0.5 mm2) to obtain a similar proliferative rate as suppression depending on their general status.
with the live tissue. This limbal allograft tech- Discontinuation of the oral immunosuppression
nique represents an in  vivo or in situ stem cell remains a controversial issue. Acute limbal
and epithelium expansion which bypasses the allograft rejection has been described recently [49]
need for the sophisticated and expensive cultur- in a series of six patients more than 3 years after a
ing technology of ex vivo cell expansion, particu- KLAL procedure, suggesting that donor cells are
larly in countries where these are not available. still present, thus at risk of acute rejection. Long-
Oral immunosuppression is a key factor in the term or indefinite i­mmunosuppression should be
long-term success rate of limbal allograft tech- considered despite a good midterm result.
niques. The limbal tissue antigen load is much Advanced topical treatments to promote and
larger than that of a standard penetrating protect epithelial cell growth constitute a signifi-
­keratoplasty, due to the presence of different cell cant aspect in the postoperative period.
types including Langerhans cells. Limbal Autologous serum or growth factor-enriched
allografts are at high risk of immune rejection plasma drops, amniotic membrane extracts, topi-
also due to the vascularity of the limbal area, cal immunosupressives [50], and epithelializa-
which negates the immune privilege of the central tion promoters like carboxylmethyl glucose
cornea. Most ophthalmologists are not familiar poly-sulfate should be rationally used to enhance
with the required immunosuppression protocols the epithelial healing and maintain epithelial
and their general side effects. Despite the fact that layer homeostasis.
these are not serious in the majority of cases [48], Regarding the long-term outcomes, a recent
specific knowledge and experience are required report found KLAL to achieve a true ocular sur-
for a proper management. As these protocols have face stability in 72.7% of cases with a mean fol-
to be frequently adjusted to the postoperative sta- low-­ up of 9.1  years, when the appropriate
tus of each patient, the cooperation of specialist in selection criteria and proper immunosuppression
immunosuppression  – usually an internist  – are applied, and the procedure is repeated as
becomes a requirement in their monitoring. needed [51]. Including this last option obviously
A recently reported protocol involves two oral conditions the meaning of that long-term “suc-
immunosuppressants (mycophenolate mofetil cess,” as the duration of each individual trans-
and tacrolimus) combined with 1  mg/kg oral plant is actually shorter. In a deep review of the
prednisone, all of which should be started 1 week evidence-based published results of limbal trans-
before surgery. Tacrolimus levels are adjusted to plantations [52], there were statistically signifi-
8–10 ng/ml the first month postoperative and to cantly better results with the use of autologous
5–8 ng/ml at 6 months postoperative. Oral pred- tissue, but no statistically significant differences
nisone is slowly tapered and discontinued after between the types of limbal allograft technique
3 months. One year after the procedure, mono- (KLAL and lr-CLAL, p = 0.328). When lr-CLAL
therapy can be considered, and 3 years after sur- is performed under a close HLA matching (0-1
gery, oral medication can be stopped if the HLA mismatches), the graft survival is increased
ocular surface is stable [36] (but see below). [26]. Other authors also reported better results
Other protocols consist in combinations of aza- when using living-related donor tissue [53].
thioprine and cyclosporine A with prednisone. Patients with Stevens-Johnson syndrome and
Absolute contraindications for oral immuno- patients with concurrent hAM transplantation
suppression include patients with a history of pre- had poorer prognosis and long-term surface
vious malignancy 5 years before, nonadherence to improvements [52].
208 R. I. Barraquer and J. Álvarez de Toledo

Donor age seems not to be a critical factor for that can be obtained, and its histological and ana-
a successful clinical result. Limbal stem cell cul- tomical characteristics have made its use as a rou-
tures obtained from older donors show a mini- tine procedure in diseases like pterygium for
mum required of >3% of p63+ cells considered decades. Aside from its ophthalmological indica-
as the minimum value to predict a favorable out- tions, oral mucosa is the new gold standard tissue
come [54]. Routine hypothermic storage in liquid used in urethral reconstructive surgery [56].
media at 4  °C is generally used to preserve the Recently, interest in this tissue has re-emerged as
donor limbal tissue, but novel methods of preser- a source of cultivated epithelial cells to be utilized
vation like hypothermic airlifted conditions [55] in ex vivo cell expansion techniques [57, 58].
have demonstrated better maintenance of epithe- Oral mucosa has been used in many procedures
lial structure, cell phenotype, and higher viability to repair symblepharon and primary or recurrent
of the stem cell pool. pterygium, as a component of biological kerato-
prostheses, fornix reconstruction, and other recon-
structive techniques of the ocular surface and
13.6 Other Mucosal oculoplastic procedures. The larger experience with
Transplantation the oral mucosal tissue relates to pterygium surgery.
With an appropriate dissection of a thin graft with
Oral mucosa is the human mucosal tissue more an electro-keratome (such as Castroviejo’s), excel-
frequently used to repair ocular surface defects. lent cosmetic and anatomical long-term results can
Its ease of harvesting, the large amount of tissue be obtained (Fig.  13.23). In this technique, a thin

a b

c d

Fig. 13.23 (a) Castroviejo’s electro-keratome with the advanced through the surface of the mucosa. (c) An excel-
motor in the handle, the head in which the oscillating lent and thin graft is then removed with scissors from the
blade is mounted, and the gauge to obtain the desired lip. (d) Graft is placed in the inferior conjunctiva of a
thickness of the graft. (b) After injecting saline solution patient affected of ocular pemphigoid to reconstruct the
under the submucosal area of the inferior lip to create inferior fornix
intra-tissular tension, the electro-keratome is firmly
13  Classical Techniques for Limbal Transplantation 209

layer (0.2–0.5 mm thick) of oral mucosa of up to 50 References


by 20 mm is obtained from a clamped and serum-
hydrated inner lower lip and placed on the conjunc- 1. Barraquer J.  Panel three discussion. In: King JH,
McTigue JW, editors. The world cornea congress
tival or fornix site to be reconstructed. This method I. Washington, DC: Butterworths; 1965. p. 354.
of mechanized harvesting the graft favors the 2. Holland EJ.  Management of limbal stem cell defi-
absence of submucosal tissue and fat lobules, avoid- ciency: a historical perspective, past, present, and
ing tissue retraction. The cosmetic appearance is future. Human Cell. 2015;34(Suppl:10):S9–15.
3. Strampelli B, Restivo Manfredi ML.  Total kera-
improved compared to a full-thickness graft tectomy in leukomatous eye associated with auto-
obtained manually with scissors. graft of a keratoconjunctival ring removed from
the contralateral eye. Ann Ottalmol Clin Ocul.
1966;92:778–86.
4. Strampelli B. Ring autokeratoplasty. In: Rycroft PV,
13.7 Conclusions editor. Corneoplastic surgery. Oxford: Pergamon
Press; 1969. p. 253–75.
Classic ocular surface reconstruction techniques 5. Barraquer J, Rutllán J. Microsurgery of the cornea: an
are still the surgical procedures of choice in many atlas and text book. Barcelona: Ed: Ediciones Scriba;
1984. p. 160–2.
ocular surface conditions involving conjunctival 6. Thoft RA.  Conjunctival transplantation. Arch
defects or LSCD. An adequate knowledge of the Ophthalmol. 1977;95:1425–7.
anatomy and pathophysiology of the condition, 7. Thoft RA.  Keratoepithelioplasty. Am J Ophthalmol.
together with familiarity of all surgical options and 1984;97:1–6.
8. Turgeon PW, Nauhein RC, Roat MI, et al. Indications
armamentarium, is a requisite for the selection of for keratoepitelioplasty. Arch Ophthalmol.
the best technique and for success. Autograft tech- 1990;108:33–6.
niques are preferable whenever possible, and an 9. Buschke W, Friedenwald JS, Fleischmann
adequate management of the tissues in both the W. Studies on the mitotic activity of the corneal epi-
thelium; methods; the effects of colchicine, ether,
donor and recipient areas is the clues for excellent cocaine and ephedrin. Bull Johns Hopkins Hosp.
results avoiding complications. When allografts are 1943;73:143–67.
required, it is paramount to establish a close col- 10. Hanna C, O’Brien JE. Cell production and migration
laboration with the specialist in systemic immuno- in the epithelial layer of the cornea. Arch Ophthalmol.
1960;64:536–9.
suppression, as the application of adequate 11. Mann I. A study of epithelial regeneration in the liv-
treatment protocols appears to substantially ing eye. Br J Ophthalmol. 1944;28:26–40.
improve the long-term results. Clear information of 12. Davanger M, Evensen A.  Role of the pericorneal

the advantages and possible complications of the papillary structure in renewal of corneal epithelium.
Nature. 1971;229:560–1.
options should be given to the patients and to 13. Schermer S, Galvin S, Sun TT. Differentiation-related
potential living-­related donors, also highlighting expression of a major 64K corneal keratin in vivo and
the benefits of being the source of tissue for their in culture suggests limbal location of corneal epithe-
affected relatives. Some techniques that may have lial stem cells. J Cell Biol. 1986;103:49–62.
14. Kinoshita S, Friend J, Thoft RA. Biphasic cell prolif-
been considered old-fashioned at a certain point, eration in transdifferentiation of conjunctival to cor-
like oral mucosal transplantation, can be very use- neal epithelium in rabbits. Invest Ophthalmol Vis Sci.
ful to solve complex situations. As long as the tech- 1983;24:1008–14.
nological, availability, and cost issues of ex  vivo 15. Potten CS, Loeffler M.  Epidermal cell proliferation.
I.  Changes with time in the proportion of isolated,
cell expansion and regeneration limit their practical paried, and clustered labeled cells in sheets of murine
application, the classical techniques remain a very epidermis. Virchows Arch B Cell Pathol Incl Mol
valid option offering excellent results when indi- Pathol. 1987;53:286–300.
cated and performed in an appropriate manner. 16. Kenyon KR, Tseng SCG. Limbal autograft transplan-
tation for ocular surface disorders. Ophthalmology.
1989;96:709–23.
Compliance with Ethical Requirements Rafael 17. Tsai RJF, Tseng SCG. Human allograft limbal trans-
I. Barraquer and Juan Alvarez de Toledo declare that they plantation for corneal surface reconstruction. Cornea.
have no conflict of interest. No human or animal studies 1994;13:389–400.
were carried out by the authors for this chapter.
210 R. I. Barraquer and J. Álvarez de Toledo

18. Kwitko S, Raminho D, Barcaro S, et al. Allograft con- 34. Cheung AY, Govil A, Friedstrom SR, Holland

junctival transplantation for bilateral ocular surface EJ.  Probable donor-derived cytomegalovirus disease
disorders. Ophthalmology. 1995;102:1020–5. after keratolimbal allograft transplantation. Cornea.
19. Kenyon KR, Rapoza PA. Limbal allograft transplan- 2017;36(8):1006–8.
tation for ocular surface disorders. Ophthalmology. 35. Sepsakos L, Cheung AY, Nerad JA, Mogilishetty

1995;102(suppl):101–2. G, Holland EJ.  Donor-derived conjunctival-limbal
20. Holland EJ, Schwartz OS.  The evolution of epithe- melanoma after a keratolimbal allograft. Cornea.
lial transplantation for severe ocular surface dis- 2017;36(11):1415–8.
ease and a proposed classification system. Cornea. 36. Holland EJ, Mogilishetty G, Skeens HM, et  al.

1996;15:549–56. Systemic immunosuppression in ocular surface stem
21. Daya SM, Chan CC, Holland EJ.  Cornea society
cell transplantation: results of a 10-year experience.
nomenclature for ocular surface rehabilitative proce- Cornea. 2012;31:655–61.
dures. Cornea. 2011;30:1115–9. 37. Miller AK, Young JW, Wilson DJ, Dunlap J,

22. Forbes J, Collin R, Dart J.  Split thickness buccal
Chamberlain W.  Transmission of donor-derived
mucous membrane grafts and beta irradiation in the breast carcinoma as a recurrent mass in a keratolimbal
treatment of recurrent pterygium. Br J Ophthalmol. allograft. Cornea. 2017;36(6):736–9.
1998;82(12):1420–3. 38. Daya SM.  Conjunctival-limbal autograft. Curr Opin
23. Naumann GO, Lang GK, Rummelt V, Wigand
Ophthalmol. 2017;28(4):370–6.
ME.  Autologous nasal mucosa transplantation in 39. Chan CC, Biber JM, Holland EJ. The modified Cincinnati
severe bilateral conjunctival mucus deficiency syn- procedure: combined conjunctival limbal autografts and
drome. Ophthalmology. 1990;97(8):1011–7. keratolimbal allografts for severe unilateral ocular sur-
24. llen JH. The use of peritoneum as a substitute for con- face failure. Cornea. 2012;31(11):1264–72.
junctiva in plastic surgery; a preliminary report. Am J 40. Sangwan VS, Basu S, MacNeil S, Balasubramanian
Ophthalmol. 1953;36:1249–52. D.  Simple limbal epithelial transplantation (SLET):
25. Harun MH, Sepian SN, Chua KH, Ropilah AR,
a novel surgical technique for the treatment of uni-
Abd Ghafar N, Che-Hamzah J, Idrus RBH, Annuar lateral limbal stem cell deficiency. Br J Ophthalmol.
FH.  Human forniceal region is the stem cell-rich 2012;96(7):931–4.
zone of the conjunctival epithelium. Hum Cell. 41. Miri A, Said DG, Dua HS. Donor site complications
2013;26(1):35–40. in autolimbal and living-related allolimbal transplan-
26. Daya SM, Ilari FA.  Living related conjunctiva lim- tation. Ophthalmology. 2011;118(7):1265–71.
bal allograft for the treatment of stem cell deficiency. 42. Busin M, Breda C, Bertolin M, Bovone C, Ponzin
Ophthalmology. 2001;108:126–33; discussion D, Ferrari S, Barbaro V, Elbadawy HM. Corneal epi-
133–134. thelial stem cells repopulate the donor area within
2 7.
Shaharuddin B, Ahmad S, Meeson A, Ali 1 year from limbus removal for limbal autograft.
S.  Concise review: immunological properties Ophthalmology. 2016;123(12):2481–8.
of ocular surface and importance of limbal stem 43. Kim EC, Jun AS, Kim MS, Jee D.  Mooren ulcer
cells for transplantation. Stem Cells Transl Med. occurring at donor site after contralateral conjunc-
2013;2(8):614–24. tivolimbal autograft for recurrent pterygium. Cornea.
28. Qi X, Xie L, Cheng J, Zhai H, Zhou Q. Characteristics 2012;31(11):1357–8.
of immune rejection after allogeneic cultivated lim- 44. Basu S, Sureka SP, Shanbhag SS, Kethiri AR, Singh
bal epithelial transplantation. Ophthalmology. V, Sangwan VS. Simple limbal epithelial transplanta-
2013;120(5):931–6. tion: long-term clinical outcomes in 125 cases of uni-
29. Lee JS, Ha SW, Yu S, Lee GJ, Park YJ. Efficacy and lateral chronic ocular surface burns. Ophthalmology.
safety of a large conjunctival autograft for recurrent 2016;123(5):1000–10.
pterygium. Korean J Ophthalmol. 2017;31(6):469–78. 45. Sundmacher R, Reinhard T, Althaus C. Homologous
30. Kheirkhah A, Blanco G, Casas V, Hayashida Y, Raju central limbo-keratoplasty in limbus stem cell dam-
VK, Tseng SC.  Surgical strategies for fornix recon- age. Retrospective study of 3 years’ experience.
struction based on symblepharon severity. Am J Ophthalmologe. 1997;94(12):897–901.
Ophthalmol. 2008;146(2):266–75. 46. Lang SJ, Böhringer D, Reinhard T. Penetrating lim-
31. Thakur S, Ichhpujani P, Kumar S.  Grafts in glau- bokeratoplasty for gelatinous corneal dystrophy. Klin
coma surgery: a review of the literature. Asia Pac J Monatsbl Augenheilkd. 2017. [Epub ahead of print].
Ophthalmol (Phila). 2017;6(5):469–76. 47. Kethiri AR, Basu S, Shukla S, Sangwan VS, Singh
32. Choi YJ, Kim IH, Choi JH, Lee MJ, Kim N, Choung V.  Optimizing the role of limbal explant size
HK, Khwarg SI.  Early results of surgical manage- and source in determining the outcomes of lim-
ment of conjunctival dermolipoma: partial excision bal transplantation: an in  vitro study. PLoS One.
and free conjunctival autograft. Br J Ophthalmol. 2017;12(9):e0185623.
2015;99(8):1031–6. 48. Krakauer M, Welder JD, Pandya HK, Nassiri N,

33. Cheung AY, Sarnicola E, Holland EJ. Long-term ocu- Djalilian AR.  Adverse effects of systemic immuno-
lar surface stability in conjunctival limbal autograft suppression in keratolimbal allograft. J Ophthalmol.
donor eyes. Cornea. 2017;36(9):1031–5. 2012;2012:576712.
13  Classical Techniques for Limbal Transplantation 211

49. Eslani M, Haq Z, Movahedan A, Moss A, Baradaran-­ 54. Nieto-Nicolau N, Martínez-Conesa EM, Casaroli-­

Rafii A, Mogilishetty G, Holland EJ, Djalilian Marano RP. Limbal stem cells from aged donors are a
AR.  Late acute rejection after allograft limbal stem suitable source for clinical application. Stem Cells Int.
cell transplantation: evidence for long-term donor sur- 2016;2016:3032128.
vival. Cornea. 2017;36(1):26–31. 55. Li C, Dong N, Wu H, Dong F, Xu Y, Du H, He H, Liu
50. Pfau B, Kruse FE, Rohrschneider K, Zorn M, Fiehn Z, Li W. A novel method for preservation of human
W, Burk RO, Völcker HE. Comparison between local corneal limbal tissue. Invest Ophthalmol Vis Sci.
and systemic administration of cyclosporin A on the 2013;54(6):4041–7.
effective level in conjunctiva, aqueous humor and 56. Barbagli G, Balò S, Montorsi F, Sansalone S,

serum. Ophthalmologe. 1995;92(6):833–9. Lazzeri M.  History and evolution of the use of oral
51. Movahedan A, Cheung AY, Eslani M, Mogilishetty mucosa for urethral reconstruction. Asian J Urol.
G, Govil A, Holland EJ. Long-term outcomes of ocu- 2017;4(2):96–101.
lar surface stem cell allograft transplantation. Am J 57.
Prabhasawat P, Ekpo P, Uiprasertkul M,
Ophthalmol. 2017;184:97–107. Chotikavanich S, Tesavibul N, Pornpanich K,
52. Health Quality Ontario. Limbal stem cell transplanta- Luemsamran P.  Long-­ term result of autologous
tion: an evidence-based analysis. Ont Health Technol cultivated oral mucosal epithelial transplantation
Assess Ser. 2008;8(7):1–58. for severe ocular surface disease. Cell Tissue Bank.
53. Titiyal JS, Sharma N, Agarwal AK, Prakash G,
2016;17(3):491–503.
Tandon R, Vajpayee R. Live related versus cadaveric 58. Kinoshita S, Koizumi N, Nakamura T. Transplantable
limbal allograft in limbal stem cell deficiency. Ocul cultivated mucosal epithelial sheet for ocular surface
Immunol Inflamm. 2015;23(3):232–9. reconstruction. Exp Eye Res. 2004;78(3):483–91.
Simple Limbal Epithelial
Transplantation: An Update
14
Nandini Venkateswaran and Guillermo Amescua

14.1 Causes and Manifestations cal and/or thermal burns, long-term contact lens
of Limbal Stem Cell use, inflammatory eye diseases such as Stevens-
Deficiency Johnson syndrome, ocular cicatricial pemphigoid,
atopic conjunctivitis, microbial keratitis, neuro-
The corneal epithelium plays an integral role in pro- trophic keratitis, history of multiple surgical inter-
tecting the cornea, maintaining corneal optical prop- ventions, radiation or systemic chemotherapy
erties as well as corneal transparency. The corneal exposure, or the use of topical antimetabolites such
epithelium undergoes constant shedding and regen- as 5-fluorouracil or mitomycin C [1].
eration, and the limbal stem cells that are critical for The diagnosis of LSCD is largely clinical.
the renewal of healthy corneal epithelium are located Ophthalmologists arrive at the diagnosis based
at the corneal limbus. Within the limbal region, cor- on a patient’s clinical history and ophthalmic
neal stem cells are located specifically in niches examination. Patients often present with symp-
within radial fibrovascular ridges called the pali- toms of pain, decreased vision, and photophobia.
sades of Vogt. Limbal stem cell deficiency (LSCD) On examination, there is loss of normal limbal
is a condition defined by the delay or failure of nor- anatomy, a “whorled-like” corneal epithelium
mal corneal epithelialization. The absence of normal with fluorescein staining, as well as corneal con-
corneal epithelium can lead to conjunctivalization, junctivalization, persistent epithelial defects, and
superficial corneal neovascularization, and poor corneal scarring [1, 2] (Fig. 14.1). Corneal impres-
healing of the corneal surface, ultimately causing
corneal scarring and secondary blindness [1–3].
LSCD can occur due to a myriad of primary and
secondary causes. The process is driven largely by
destruction of the microenvironment of the stem
cell niche [2]. Primary causes include conditions
such as aniridia, multiple endocrine deficiency, epi-
dermal dysplasias, congenital erythrokeratoder-
mia, and dyskeratosis congenita. Secondary causes,
which are typically more common, include chemi-

N. Venkateswaran · G. Amescua (*)


Bascom Palmer Eye Institute, University of Miami, Fig. 14.1  Clinical photograph of a patient who devel-
Department of Ophthalmology, Miami, FL, USA oped 160° limbal stem cell deficiency after sustaining a
e-mail: Gamescua@med.miami.edu chemical burn

© Springer Nature Switzerland AG 2019 213


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_14
214 N. Venkateswaran and G. Amescua

sion cytology can reveal goblet cells and confo- 14.3 S


 imple Limbal Epithelial
cal microscopy can show loss of the palisades of Transplantation
Vogt [4].
Simple limbal epithelial transplantation (SLET)
is a novel, surgical approach heralded by
14.2 Treatments for Limbal Stem Sangwan in 2010 that permits the in vivo expan-
Cell Deficiency sion of harvested limbal stem cells [7]. In SLET
surgery, a 2 × 2 mm donor limbal graft from the
The treatment of LSCD is primarily geared unaffected eye is carefully harvested, typically
toward restoring the normal, phenotypic corneal from the superior quadrant. Once the limbal graft
epithelium and rehabilitating the ocular surface. is harvested, it is left in balance salt solution. In
Prior to any surgical intervention, the ocular sur- the recipient eye, a 360° conjunctival peritomy is
face must be optimized and augmented to made and the fibrovascular pannus secondary to
improve surgical outcomes. Active, comorbid LSCD is mechanically removed. A fresh amni-
autoimmune diseases must be managed with the otic membrane transplant is secured onto the cor-
proper medications such as immunomodulatory nea with fibrin glue. The donor limbal graft is cut
agents with the goal of having a quiescent ocular into eight to ten small pieces, and these pieces are
surface. Exposure or neurotrophic keratopathy evenly distributed and placed epithelial side up
should be treated with botox-induced ptosis and/ on the amniotic membrane transplant and secured
or tarsorraphies. Mechanical irritation secondary with fibrin glue. A bandage contact lens is then
to lid or lash abnormalities should be co-­ placed over both the basal amniotic membrane
managed with oculoplastics, and surface protec- and overlying limbal transplants. Patients are
tion can be achieved with scleral lenses or subsequently followed to monitor surface healing
mucous membrane grafts. Aqueous tear defi- and assess for LSCD recurrence. Patients are
ciency should be addressed with punctal occlu- treated with topical antibiotics and a course of
sion, lubricating drops, or autologous serum tapered topical steroid drops. The bandage con-
tears [5]. tact lens is typically removed within 1 week of
For surgical approaches, main consider- the postoperative period [7]. Recently, Amescua
ations include the extent of limbal involvement et al. in 2014 posed a modification to the above
(sectoral vs total) and laterality of disease (uni- SLET surgery procedure, implementing the use
lateral or bilateral) [2]. It is important to note of a double-layer cryopreserved amniotic mem-
that corneal transplants almost universally fail brane, which is readily available in the United
in cases of LSCD [6]. Mechanical debridement States. The limbal transplants with this modified
as well as amniotic membrane grafting can ini- technique are sandwiched within two layers of
tially be pursued for partial LCSD.  However, amniotic membrane, as opposed to only being
full recovery is often limited by the degree of laid on the layer of amniotic membrane that is
remaining corneal stem cell reserves, and for secured to the corneal surface. This double layer
more extensive cases, limbal stem cell trans- of amniotic membrane aims to protect and repli-
plantation must be pursued [1]. Various tech- cate a fetal environment for the limbal stem cells
niques have been described and differ based [6]. Epithelialization after SLET can occur as
upon the anatomical source of transplanted tis- early as 2  days postoperatively. Mittal et  al.
sue (conjunctival, keratolimbal, or mucosal), showed that epithelialization stems from the lim-
the use of autologous or allogeneic tissue bal transplants over the amniotic membrane
(cadaveric or living-related) source, as well as transplant and progresses over 2 weeks. The lim-
cell culture techniques, such as cultivated lim- bal transplants disappear within 1–2 months [8],
bal epithelial transplantation (CLET) [2, 3]. in some cases it can take upto 4–6 months.
14  Simple Limbal Epithelial Transplantation: An Update 215

14.4 Advantages of Simple 14.5 Indications for Simple


Limbal Epithelial Limbal Epithelial
Transplantation Transplantation

SLET offers several advantages over other LSCD The indication for a limbal transplantation proce-
surgical approaches [7]. Conjunctival limbal dure largely relies on the status of the limbal stem
autograft (CLAU) requires the removal of three cells and the conjunctiva. Staging paradigms
to six clock hours of limbus and adjacent con- have classified ocular surface disease based upon
junctiva from a donor eye that is then transplanted the percentage of involvement of the limbus,
to the recipient eye [9]. While CLAU is a one-­ involvement of the conjunctiva, and level of
time surgery similar to SLET and has a reported active inflammation of the conjunctiva [14]. Stem
success rate of 77–100%, it requires harvesting a cell transplantation is typically difficult in
much larger area of donor limbus and conjunctiva patients with a dry, inflamed ocular surface with
as compared with SLET to achieve success [10, conjunctival scarring [13].
11]. The procedure itself is also technically SLET, although a relatively new surgical pro-
demanding and poses the risk of destabilizing the cedure, has shown increasing promise in success-
ocular surface of the donor eye [12]. fully treating LSCD.  The procedure was
Alternative limbal transplantation procedures originally described for unilateral 360° LSCD,
can be classified according to the source and but it has also shown benefit in patients with par-
anatomic location of the donor tissue. These can tial LCSD [7]. It has been most successful in
include cadaveric conjunctival limbal autografts patients with unilateral LSCD secondary to
(c-CLAL), conjunctival limbal autografts from chemical burns or previous surgery (i.e., ocular
living-related donors (lr-CLAL), and finally surface tumor removal) with a well-lubricated
cadaveric keratolimbal allografts (KLAL) [13]. ocular surface (Figs.  14.2, 14.3, and 14.4). A
These procedures are typically reserved for cases large retrospective, multicenter, interventional
of bilateral LSCD where autologous transplanta- series by Vazirani et al. found SLET was success-
tion is unable to be performed. These procedures ful in 83.8% of cases of unilateral LSCD with a
in particular require patients to use long-term wet ocular surface, with a survival probability of
immunosuppression, typically three agents greater than 80% at a median of 1 year of follow-
including a corticosteroid, an antimetabolite, ­up [18]. Outcomes are limited in patients with a
and a T-cell inhibitor, in order to decrease the poorly optimized ocular surface in the setting of
chances of immunological rejection. The use of active chronic inflammatory conditions such as
immunosupression comes with increased sys- Stevens-Johnson syndrome or ocular cicatricial
temic side effects, and these patients need to be pemphigoid [18].
monitored by an expert in the field of immuno- Sangwan first implemented SLET surgery in a
suppression [14]. cohort of six patients with unilateral or total
In comparison to the above techniques, culti- LSCD secondary to ocular surface burns. The
vated limbal epithelial transplantation (CLET) unaffected fellow eye served as the donor eye in
allows for the ex vivo expansion of less than one all cases. Postoperatively, all six eyes had stable,
clock hour of harvested donor limbus cells [15]. avascular, epithelialized corneal surfaces by 6
Success rates range from 50 to 100%; however, weeks, which were maintained at a mean of 9
the feasibility of this technique is limited by the months. Visual acuity improved to 20/60 from
time and cost needed to cultivate cells as well as worse than 20/200 on presentation in 66.6% of
the need for facilities to allow for cell harvesting cases [7]. In Amescua et al.’s series of four patients
and expansion. In the United States, there is no with unilateral LCSD treated with SLET using a
Food and Drug Administration-approved facility double-layer cryopreserved amniotic membrane,
that will allow for the ex vivo cultivation of lim- 100% of cases had graft success. In addition to
bal stem cells at this time [16, 17]. achieving a stable epithelium, 100% of patients
216 N. Venkateswaran and G. Amescua

a b

Fig. 14.2 (a) Clinical photograph of a patient with lim- epithelium with corneal clearing and no evidence of
bal stem cell deficiency after a chemical burn. (b) Two recurrence of limbal stem cell deficiency
years postoperatively, the patient has achieved a stable

a b c

Fig. 14.3 (a) Clinical photograph of a patient with 360° operatively, the patient has achieved corneal clarity with
limbal stem cell deficiency after a chemical burn. (b) mild residual haze and no further corneal
Clinical photograph of a patient postoperatively after sim- neovascularization
ple limbal epithelial transplantation. (c) Two years post-

a b c

Fig. 14.4 (a) Clinical photograph of a young patient who plant on the corneal surface. (c) Five months postopera-
developed limbal stem cell deficiency following at alkali tively, the patient has noted marked improvement in visual
burn. (b) Postoperatively, pieces of the donor limbal graft acuity with a stable ocular surface
are evenly distributed on an amniotic membrane trans-
14  Simple Limbal Epithelial Transplantation: An Update 217

had improvement in visual acuity to 20/50 or bet- facilitate anatomical and visual success of subse-
ter from initial acuities of 20/200 or worse and quent deep anterior lamellar keratoplasties in
had complete resolution of ocular symptoms. pediatric patients with severe chemical injuries
Donor eyes notably had no damage [6]. [25]. Arya et al. have also reported the success of
Basu et al., in a prospective case series of 125 SLET in restoring a stable, avascular epithelium
patients including adults and children with in patients with acid injuries as well as severe
LSCD, found that 76% of cases maintained epi- dry eye disease [26].
thelialized, avascular corneas at a median follow-
­up of 1.5 years after undergoing SLET surgery,
and 67% of cases achieved a final vision of 14.6 Complications and Failures
greater than 20/60, which was statistically sig- of Simple Limbal Epithelial
nificant [12]. SLET can also be used as an effec- Transplantation
tive treatment modality for unilateral ocular
surface burns, with a success rate of 70% at a SLET is an overall well-tolerated procedure.
median follow-up of 1.1 years as noted by Gupta Complications to the donor eye are infrequent
et al. [19]. In children particularly, SLET has also and can include subconjunctival hemorrhages,
shown promise in treating severe, unilateral pyogenic granuloma formation, or LCSD. In the
chemical burns [20]. recipient eyes, SLET surgery can cause compli-
Several case reports have also described the cations including corneal perforation at the time
success of SLET in various other clinical set- of fibrovascular pannus debridement, postoper-
tings. Cases of concomitant SLET at the time of ative keratitis, detached amniotic membrane,
extensive ocular surface squamous neoplasia loss of limbal transplants, or symblepharon for-
excision have shown complete restoration of mation [12].
limbal stem cells and maintenance of a stable SLET, although largely successful, can also be
ocular surface without evidence of LSCD [21, associated with failures. Basu et  al. found pro-
22]. The use of SLET to treat LCSD refractory to gressive conjunctivalization in 18.4% of cases,
CLET in pediatric patients suggests the benefits and factors associated with failure included his-
of using SLET in this population as a first- or tory of acid injury, thermal burns, severe sym-
second-line treatment approach, as compared blepharon, SLET combined with keratoplasty,
with CLET which has shown a suboptimal suc- and postoperative loss of transplants (Fig. 14.5)
cess rate in children as compared with adults [12]. Gupta et  al. as well as Vazirani et  al. also
[23, 24]. Studies have also demonstrated that sta- found that the presence of symblepharon and
bilization of the ocular surface with SLET can combined SLET with penetrating keratoplasty

a b c

Fig. 14.5 (a) Clinical photograph of a patient who sus- toplasty and simple limbal epithelial transplantation with
tained limbal stem cell deficiency after Pseudomonas subsequent corneal clearing and restoration of visual
scleritis. (b, c) The patient underwent a penetrating kera- function
218 N. Venkateswaran and G. Amescua

a b

Fig. 14.6  Clinical photographs of a patient who devel- brane graft, there was only marginal improvement in the
oped limbal stem cell deficiency and conjunctival scarring ocular surface. (a) Limbal stem cell deficiency and cor-
after a corneal burn from a fireworks injury. Despite lysis neal scarring after chemical/thermal burn. (b) Recurrance
of symblepharon, simple limbal epithelial transplantation, of Limbal stem cell deficiency after ocular surface recon-
and ocular surface reconstruction with a mucous mem- struction with mucous membrane graft and SLET

was associated with SLET failure [18, 19]. These low-up can help clinicians better understand the
risk factors for failure should be taken into con- utility of this procedure in the treatment of LCSD.
sideration when selecting the patient candidates
for SLET procedure surgery (Fig. 14.6). Acknowledgments We would like to thank Dr. Anat
Galor MD MSPH for assistance in obtaining clinical pho-
tographs used in this publication.

14.7 Conclusions
Compliance with Ethical Requirements
Conflict of Interest Nandini Venkateswaran and
SLET surgery has proven to be a safe, reproduc- Guillermo Amescua declare that they have no conflict of
ible, and effective surgical intervention that allows interest.
for the long-term regeneration of the ocular sur- Informed Consent  No human studies were carried out
by the authors for this article.
face in cases of unilateral LCSD in patients with a No animal studies were carried out by the authors for
wet ocular surface and adequate control of the this article.
ocular surface inflammation. The outcomes of
SLET have also shown that by just restoring the
corneal epithelial phenotype, patients can achieve References
significant improvements in visual acuity [6]. The
procedure is easy to replicate, does not require 1. Haagdorens M, Van Acker SI, Van Gerwen V, Ni
laboratory support for cell cultivation, poses mini- Dhubhghaill S, Koppen C, Tassignon MJ, et al. Limbal
stem cell deficiency: current treatment options and
mal risk to the donor eye, and has proven to be emerging therapies. Stem Cells Int. 2016;2016:9798374.
successful in both pediatric and adult cases of 2. Atallah MR, Palioura S, Perez VL, Amescua
LSCD. SLET offers several advantages to CLAU G.  Limbal stem cell transplantation: current per-
and CLET, especially in the pediatric population; spectives. Clinical Ophthalmol (Auckland, NZ).
2016;10:593–602.
however, there have been no randomized control 3. Yin J, Jurkunas U.  Limbal stem cell transplan-
trials comparing these different treatment modali- tation and complications. Semin Ophthalmol.
ties. Ultimately, assessment of larger cohort of 2018;33(1):134–141. https://doi.org/10.1080/088205
patients treated with SLET with long-term fol- 38.2017.1353834. Epub 2017 Nov 27.
14  Simple Limbal Epithelial Transplantation: An Update 219

4. Nubile M, Lanzini M, Miri A, Pocobelli A, Calienno 16. Shortt AJ, Tuft SJ, Daniels JT. Ex vivo cultured limbal
R, Curcio C, et  al. In vivo confocal microscopy epithelial transplantation. A clinical perspective. Ocul
in diagnosis of limbal stem cell deficiency. Am J Surf. 2010;8(2):80–90.
Ophthalmol. 2013;155(2):220–32. 17.
Ramachandran C, Basu S, Sangwan VS,
5. Espana EM, Di Pascuale M, Grueterich M, Balasubramanian D.  Concise review: the coming of
Solomon A, Tseng SC.  Keratolimbal allograft in age of stem cell treatment for corneal surface damage.
corneal reconstruction. Eye (London, England). Stem Cells Transl Med. 2014;3(10):1160–8.
2004;18(4):406–17. 18. Vazirani J, Ali MH, Sharma N, Gupta N, Mittal V,
6. Amescua G, Atallah M, Nikpoor N, Galor A, Perez Atallah M, et  al. Autologous simple limbal epithe-
VL. Modified simple limbal epithelial transplantation lial transplantation for unilateral limbal stem cell
using cryopreserved amniotic membrane for unilat- deficiency: multicentre results. Br J Ophthalmol.
eral limbal stem cell deficiency. Am J Ophthalmol. 2016;100(10):1416–20.
2014;158(3):469–75.e2. 19. Gupta N, Joshi J, Farooqui JH, Mathur U.  Results
7. Sangwan VS, Basu S, MacNeil S, Balasubramanian of simple limbal epithelial transplantation in uni-
D.  Simple limbal epithelial transplantation (SLET): lateral ocular surface burn. Indian J Ophthalmol.
a novel surgical technique for the treatment of uni- 2018;66(1):45–52.
lateral limbal stem cell deficiency. Br J Ophthalmol. 20. Mittal V, Jain R, Mittal R, Vashist U, Narang

2012;96(7):931–4. P.  Successful management of severe unilateral
8. Mittal V, Jain R, Mittal R.  Ocular surface epitheli- chemical burns in children using simple limbal epi-
alization pattern after simple limbal epithelial trans- thelial transplantation (SLET). Br J Ophthalmol.
plantation: an in  vivo observational study. Cornea. 2016;100(8):1102–8.
2015;34(10):1227–32. 21. Mittal V, Narang P, Menon V, Mittal R, Honavar

9. Kenyon KR, Tseng SC.  Limbal autograft transplan- S.  Primary simple limbal epithelial transplantation
tation for ocular surface disorders. Ophthalmology. along with excisional biopsy in the management of
1989;96(5):709–22; discussion 22–3. extensive ocular surface squamous neoplasia. Cornea.
10. Tan DT, Ficker LA, Buckley RJ. Limbal transplanta- 2016;35(12):1650–2.
tion. Ophthalmology. 1996;103(1):29–36. 22.
Guarnieri A, Moreno-Montanes J.  Concomitant
11.
Arora R, Dokania P, Manudhane A, Goyal simple limbal epithelial transplantation after surgical
JL.  Preliminary results from the comparison of excision of ocular surface squamous neoplasia. Am J
simple limbal epithelial transplantation with con- Ophthalmol. 2017;179:205.
junctival limbal autologous transplantation in severe 23. Bhalekar S, Basu S, Lal I, Sangwan VS.  Successful
unilateral chronic ocular burns. Indian J Ophthalmol. autologous simple limbal epithelial transplanta-
2017;65(1):35–40. tion (SLET) in previously failed paediatric limbal
12. Basu S, Sureka SP, Shanbhag SS, Kethiri AR, Singh transplantation for ocular surface burns. BMJ Case
V, Sangwan VS. Simple limbal epithelial transplanta- Reports. 2013;2013:bcr2013009888. https://doi.
tion: long-term clinical outcomes in 125 cases of uni- org/10.1136/bcr-2013-009888.
lateral chronic ocular surface burns. Ophthalmology. 24. Nair D, Mohamed A, Sangwan VS. Outcome of cata-
2016;123(5):1000–10. ract surgery following simple limbal epithelial trans-
13. Holland EJ, Schwartz GS.  The evolution of epithe- plantation for lime injury-induced limbal stem cell
lial transplantation for severe ocular surface dis- deficiency. BMJ Case Rep. 2015;2015.
ease and a proposed classification system. Cornea. 25. Singh D, Vanathi M, Gupta C, Gupta N, Tandon

1996;15(6):549–56. R.  Outcomes of deep anterior lamellar keratoplasty
14. Biber JM, Skeens HM, Neff KD, Holland EJ. The cin- following autologous simple limbal epithelial trans-
cinnati procedure: technique and outcomes of com- plant in pediatric unilateral severe chemical injury.
bined living-related conjunctival limbal allografts and Indian J Ophthalmol. 2017;65(3):217–22.
keratolimbal allografts in severe ocular surface fail- 26. Arya SK, Bhatti A, Raj A, Bamotra RK.  Simple
ure. Cornea. 2011;30(7):765–71. limbal epithelial transplantation in acid injury
15. Tsai RJ, Li LM, Chen JK. Reconstruction of damaged and severe dry eye. J Clin Diagn Res: JCDR.
corneas by transplantation of autologous limbal epi- 2016;10(6):Nd06–7.
thelial cells. N Engl J Med. 2000;343(2):86–93.
Cell Therapy Using Ex Vivo
Cultured Limbal Cells: CLET
15
and Equivalent

Paolo Rama and Giulio Ferrari

15.1 Introduction: Epidemiology visual impairment, because the conjunctival epi-


of LSCD thelium is opaque and vascularized.
The estimated prevalence of LSCD is 1–9 sub-
Limbal stem cell deficiency (LSCD) is defined as jects in 100,000. As such, it is considered a rare
a loss or deficiency of the corneal epithelial cells, disease, and it is listed on the Orphanet database
which are located at the limbus [1–4]. Many ocular (ORPHA: 171673) in Europe. There is evidence,
diseases can induce LSCD. These include congen- however, for a higher prevalence when considering
ital anomalies, such as aniridia; traumatic diseases, different geographical areas. For instance, in India
such as chemical/thermal injuries; inflammatory it has been estimated that prevalence of severe ocu-
disease, such as Stevens-­Johnson or Lyell syn- lar surface disorders associated with LSCD is
drome; toxic injuries, such as factitious keratitis or 1.25% [7], perhaps as a consequence of a higher
contact lens overuse; and infections [5, 6]. It incidence of industry/domestic ocular injury.
should be noted that, although these disorders
eventually induce stem cell damage, the entire
ocular surface is also affected (typically the eye- 15.1.1 Treatments for LSCD
lids, conjunctiva, corneal nerves, stroma, and lac-
rimal system). Such a complex pathophysiology 15.1.1.1 Non-surgical Treatments
of ocular surface disease in LSCD should be kept Medical therapy cannot heal LSCD, but it is an
in mind when treating LSCD. adjunctive treatment of the ocular surface that might
Because the corneal epithelium is constantly improve clinical findings in the early stage. Therapy
renewed by shedding older cells, LSCD causes includes frequent topical application of lubricants,
corneal epithelial turnover breakdown and bandage contact lenses, and topical steroids in case
delayed wound healing at first. Then, secondary of chronic or persistent inflammation.
ingrowth of the conjunctiva onto the cornea
occurs. Conjunctival migration, or “conjunctival- 15.1.1.2 Surgical Treatments
ization”, protects the cornea from infections and Amniotic membrane grafting can work only in
stromal melting, although at the cost of severe partial LSCD by promoting recovery from the
undamaged area of the limbus.
P. Rama (*) · G. Ferrari Corneal transplantation has also been
San Raffaele Hospital, Department of attempted, although LSCD is now unanimously
Ophthalmology—Cornea and Ocular Surface, considered a contraindication to corneal trans-
Milan, Italy plantation, either penetrating or lamellar, because
e-mail: rama.paolo@hsr.it

© Springer Nature Switzerland AG 2019 221


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_15
222 P. Rama and G. Ferrari

donor graft re-epithelialization will not take place, ment of corneal burns (Holoclar; Holostem
which results in defective epithelization, graft Terapie Avanzate, Modena, Italy).
conjunctivalization, and eventually graft failure. The protocol for CLET includes biopsy on
Unilateral limbal stem cell deficiency has the fellow eye and grafting of the ex  vivo
been successfully treated for years by directly expanded limbal stem cells on a fibrin layer. The
grafting a portion of the healthy limbal tissue biopsy is usually taken from the superior quad-
taken from the contralateral eye, although poten- rant of the limbus, but it can be done in any
tial risks to the unaffected eye are associated quadrant when we deal with partial LSCDs. The
with this procedure, especially if the donor eye is procedure is done under topical or local anaes-
also affected by partial LSCD, as it is frequently thesia, and no bandaging is required. The biopsy
the case. In addition, no quality control can be is then shipped to the laboratory, where it is pro-
provided on these implants other than clinical cessed and stem cells are extracted and culti-
inspection. In summary, LSCD used to be an vated. The stem cell product undergoes a number
area of significant and unmet medical need. of quality checks (for sterility, expression of
Simple limbal epithelial transplantation stemness markers, etc.). Expression of p63 is
(SLET) is a recently described surgical technique used to distinguish LSCs from transiently ampli-
[8] where autologous limbal stem cells are col- fying cells. Deductively, potency is addressed by
lected from the unaffected eye and implanted on quantification of p63-bright cells. Release speci-
the LSCD eye in one surgical session. Specifically, fications were set at 2.5–10.0% p63-bright cells.
the donor tissue is cut into small pieces and trans- In fact, a previous study showed that cultures in
planted using amniotic membrane scaffold and which p63 cells accounted for more than 3%
fibrin glue. SLET has shown good results at the resulted in successful clinical outcome in nearly
1.5-year follow-up; it is a low-cost procedure and 80% of cases [12]. The stem cells are then placed
remains an option where tissue shipping and lab- on a fibrin sheet and shipped back to the hospi-
oratory processing may not be feasible. tal. The surgical procedure is done under locore-
gional or general anaesthesia. First, the pannus is
15.1.1.3 CLET removed with careful cauterization to prevent
To overcome risks for the donor eye, and ensure bleeding under the fibrin sheet. Then, stem cells
that the correct amount of stem cells is adminis- are transferred on the cornea on their fibrin glue
tered, autologous cultivated limbal epithelial layer and the lids temporarily closed with a
transplantation (CLET) has been developed. Steri-Strip. Postsurgical therapy includes corti-
CLET was originated from the work of costeroid treatment, progressively tapered, to
Rheinwald and Green, who managed to obtain a keep the inflammation under control; persistent
layer of stratified epithelium by cultivating a sin- or recurrent inflammation has been the first
gle keratinocyte on a layer of murine fibroblasts cause of failure in previous studies [12, 13].
(3T3 cells) [9]. Later, Pellegrini et al. showed that CLET has several advantages compared with
autologous grafts of cultivated cells obtained in vivo grafting of limbal tissue.
from a 1 mm2 limbal biopsy were able to restore First, less invasive surgery is required on the
the corneal surface in two patients with complete contralateral eye, which is frequently partially
loss of the corneal limbus epithelium [10]. The involved by the same disease-inducing
culture procedure was then standardized [11], LSCD. This greatly reduces potential risks aris-
and to date more than 270 grafts have been trans- ing from the surgical procedure, including infec-
planted in various centres throughout Italy, with tions and induction of LSCD in the fellow eye.
long-term stability reported in more than 150 Second, it allows regrafting if failure of the
patients and with a success rate in 70–80% of primary implant should occur.
cases [12, 13]. The procedure was then further Third, stringent quality controls on CLET
refined to fulfil regulatory standards and was warrant that the minimal amount of stem cells
finally approved by the European Medicines required to achieve clinical success is trans-
Agency (EMA) in February 2015 for the treat- planted each time.
15  Cell Therapy Using Ex Vivo Cultured Limbal Cells: CLET and Equivalent 223

Finally, banking of autologous stem cells 15.2 C


 onclusion: CLET and Future
could pave the way to genetic manipulation of Perspectives
stem cells so to correct genetic anomalies (i.e.
aniridia, Meesmann dystrophy) or reduce chances The recent approval of CLET from the European
of graft rejection (major histocompatibility com- Medicines Agency has paved the way to stem cell
plex modulation). therapy for ocular surface disorders. Now that a
In summary, monolateral or partial bilateral regulatory pathway has been established, further
LSCD can be successfully treated with CLET. developments can be foreseen. First, transplanta-
Notably, CLET presents a number of advan- tion of induced pluripotent stem cells (iPSCs)
tages over allogenic stem cell transplantation (i.e. could allow treatment of bilateral stem cell defi-
when cell recipient and donor are not the same ciency patients, where no stem cells are available
individual). Specifically, CLET does not require to biopsy. Second, gene editing of stem cells
systemic, lifelong immunosuppression. (autologous or allogenic) can be envisaged. This
Moreover, published data suggest long-term sta- approach would be ideal to definitively heal cer-
bility of transplanted epithelium in CLET [12], tain corneal epithelial dystrophies (e.g.
while questions remain over long-term efficacy Meesmann epithelial dystrophy). In addition,
of allogenic stem cell transplantation, since a modulation of the expression of major histocom-
number of studies [14–16] have failed to detect patibility complex (MHC) molecules could allow
persistent expression of donor DNA in corneal to control/modulate graft rejection.
implants already 3 months after allogenic trans- In summary, CLET discloses a number of
plantation. It should be noted that CLET is not an opportunities in areas where there is a current
option in case of bilateral LSCD, which can significant and unmet medical need.
instead be addressed with allogenic Technology advancements are making cell
transplantation. reprogramming and gene editing easier, faster,
and less toxic. Nevertheless, long-term risks for
15.1.1.4 Restoration of the Ocular toxicity or malignant transformation and the high
Surface Ecosystem costs of GMP production deserve close attention
We would like to highlight the importance of and will eventually determine future develop-
restoring the appropriate ocular surface milieu ments of CLET.
in LSCD patients, as this is instrumental to the Figure 15.1 shows severe limbal stem cell
success of any cell therapy. Indeed, stem cell deficiency secondary to chemical burn, and
deficiency is concomitant with extensive sub- Fig. 15.2 shows stable epithelium with clear cor-
version of the entire ocular surface, including nea 1 year after CLET.
the conjunctiva, lids, and lachrymal systems.
Hence, conjunctival scarring causing ankylo-
blepharon/symblepharon should be surgically
corrected, as well as lid position anomalies. It
is also fundamental to maintain a good lubrica-
tion of the ocular surface, as severe dry eye is
the norm in these patients and can severely
impact any cellular implant. Finally, ocular sur-
face inflammation should be controlled at all
times. Recalcitrant inflammation is a frequent
occurrence in LSCD patients, typically in those
affected with chemical burns, Stevens-Johnson,
and ocular cicatricial pemphigoid. Bursts of
inflammation should be promptly controlled
with appropriate topical immunomodulating Fig. 15.1  Severe limbal stem cell deficiency secondary
agents. to chemical burn
224 P. Rama and G. Ferrari

The limbus is the transition area placed between the


conjunctiva and the cornea.
5. Dua HS, Saini JS, Azuara-Blanco A. Limbal stem cell
deficiency: concept, aetiology, clinical presentation,
diagnosis and management. Indian J Ophthalmol.
2000;48(2):83–92. Review
6. Daniels JT, Notara M, Shortt A, et  al. Limbal epi-
thelial stem cell therapy. Expert Opin Biol Ther.
2007;7(1):1–3. Review.
7. Vemuganti GK, Sangwan VS. Interview: affordability
at the cutting edge: stem cell therapy for ocular sur-
face reconstruction. Regen Med. 2010;5:337–4.
8. Basu S, Sureka SP, Shanbhag SS, et al. Simple lim-
bal epithelial transplantation: long-term clinical
Fig. 15.2  Stable epithelium with clear cornea 1  year outcomes in 125 cases of unilateral chronic ocular
after CLET surface burns. Ophthalmology. 2016;123(5):1000–10.
9. Rheinwald JG, Green H. Serial cultivation of strains
of human epidermal keratinocytes: the formation
of keratinizing colonies from single cells. Cell.
15.3 Compliance with Ethical 1975;6(3):331–43.
Requirements 10. Pellegrini G, Traverso CE, Franzi AT, et  al. Long-­
term restoration of damaged corneal surfaces with
autologous cultivated corneal epithelium. Lancet.
Paolo Rama and Giulio Ferrari declare that they 1997;349(9057):990–3.
have no conflict of interest. No human or animal 11. Pellegrini G, Golisano O, Paterna P, et  al. Location
studies were carried out by the authors for this and clonal analysis of stem cells and their differenti-
ated progeny in the human ocular surface. J Cell Biol.
chapter. 1999;145(4):769–82.
12. Rama P, Matuska S, Paganoni G, et al. Limbal stem-­
cell therapy and long-term corneal regeneration. N
References Engl J Med. 2010;363(2):147–55.
13. Pellegrini G, Rama P, Matuska S, et  al. Biological
parameters determining the clinical outcome of
1. Notara M, Alatza A, Gilfillan J, et al. In sickness and
autologous cultures of limbal stem cells. Regen Med.
in health: corneal epithelial stem cell biology, pathol-
2013;8(5):553–67.
ogy and therapy. Exp Eye Res. 2010;90(2):188–95.
14. Chen P, Zhou Q, Wang J, et  al. Characterization of
2. Ordonez P, Chow S, Wakefield D, et  al. Human
the corneal surface in limbal stem cell deficiency and
limbal epithelial progenitor cells express
after transplantation of cultured allogeneic limbal
alphavbeta5-­ integrin and the interferon-induc-
epithelial cells. Graefes Arch Clin Exp Ophthalmol.
ible chemokine CXCL10/IP-10. Stem Cell Res.
2016;254:1765–77.
2013;11(2):888–901.
15. Henderson TR, Coster DJ, Williams KA, et  al. The
3. Cotsarelis G, Cheng SZ, Dong G, et al. Existence of
long term outcome of limbal allografts: the search for
slow-cycling limbal epithelial basal cells that can be
surviving cells. Br J Ophthalmol. 2001;85:604–9.
preferentially stimulated to proliferate: implications
16. Daya SM, Watson A, Sharpe JR, et al. Outcomes and
on epithelial stem cells. Cell. 1989;57(2):201–9.
DNA analysis of ex vivo expanded stem cell allograft
4. Thoft RA, Wiley LA, Sundarraj N. The multipotential
for ocular surface reconstruction. Ophthalmology.
cells of the limbus. Eye (Lond). 1989;3(Pt 2):109–13.
2005;112:470–7.
Cell Therapy Using Cultivated Oral
Mucosal Epithelial Transplant
16
(COMET)

Roberto Fernández Buenaga and Sajjad Ahmad

16.1 Main Principles 3. Managing dry eye disease: LSCF is not



in the Management of LSCF uncommonly associated with tear film abnor-
malities. This can be as a result of obstruction
There are five main principles in the management of lacrimal duct outflow such as in mucous
of LSCF: membrane pemphigoid, loss of goblet cells
such as in Stevens-Johnson syndrome, or mei-
1. Understanding and controlling inflammation: bomian gland dysfunction resulting from any
It is important that ocular surface inflamma- cicatrizing conjunctivitis. These abnormalities
tion is controlled prior to any surgical man- need to be improved as much as possible prior
agement, using either topical or systemic to consideration of any ocular surface recon-
immune suppression. After an acute chemical structive procedure. Surgical measures such
burn, inflammation can take as long as as punctal occlusion and permanent partial
12–18  months to control. Patients with tarsorrhaphies may also be needed to achieve
mucous membrane pemphigoid will require this.
systemic immune suppression. 4. Partial versus total LSCF: One of the main
2. Correcting lid malposition and fornix adhe- goals of ocular surface reconstruction is to
sions: These are important factors that con- improve vision. In both partial and total
tribute to ocular surface trauma and involvement of the limbus and therefore cor-
inflammation that exacerbate the disease pro- neal surface, central corneal involvement is an
cess and prevent successful ocular surface important reason for surgical reconstruction.
reconstruction. These therefore need to be There are many patients with partial LSCF
corrected in the first instance, prior to any sur- who do not have central corneal involvement
gical procedure for the LSCF. and some patients with total LSCF with cen-
tral corneal sparing who do not need ocular
surface reconstruction procedures. It must
also be noted that in partial LSCF, healthy
portions of the limbus can be utilised for sur-
R. F. Buenaga (*) gical reconstructive methods (either auto-­
Cornea, Cataract and Refractive Surgery Department, CLET or auto-SLET).
Vissum Madrid, Madrid, Spain 5. Unilateral versus bilateral LSCF: Knowledge
S. Ahmad of aetiology in LSCF is important in deter-
Cornea and External Diseases Department, mining the surgical options, mainly whether
Moorfields Eye Hospital, London, UK

© Springer Nature Switzerland AG 2019 225


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_16
226 R. F. Buenaga and S. Ahmad

autograft or allograft procedures should be a


employed. In unilateral disease, most com-
monly in injurious or iatrogenic causes of
LSCF, an autograft procedure would be b
employed. In bilateral disease, such as genetic
and inflammatory diseases, COMET or
allograft procedures would be utilised.
Fig. 16.1  Schematic illustration of culture methods for
Following the principles above in managing oral mucosal epithelial cells. Figure (a) shows an explant
each patient individually enables successful culture on amniotic membrane. The arrows indicate out-
growth of expanding epithelial cells around the explant.
determination of which surgical ocular surface Figure (b) shows a cell suspension culture with single epi-
reconstructive procedure to employ. thelial cells on the amniotic membrane proliferating to
give rise to multiple colonies or clumps of expanding cells

16.2 Early Oral Mucosal Studies epithelialised 10  days after surgery. The oral
mucosal epithelial cells showed 5–6 layers of
The use of oral mucosa for ocular surface recon- stratification and appeared very similar to in vivo
struction was first attempted in 1963 by Ballen normal corneal epithelium. The authors detected
and co-workers [1]. These oral mucosal grafts the presence of nonkeratinised, mucosal-specific
included both epithelium and subepithelial tis- keratin 4, keratin 13, and importantly corneal
sues and developed early fibrosis and intense cor- epithelial specific keratin 3, whereas keratinisa-
neal vascularisation. Later on, in 1986, Gipson tion-related keratin 1 or keratin 10 was not
and co-workers [2] used oral epithelium (freed of detected. Epithelial cells were very similar in
connective tissue to avoid vascularisation), but appearance to those of normal corneal epithe-
they found that these epithelial cells could not lium, had numerous desmosomes, and were
survive in central avascular areas of the cornea. attached to a basement membrane with hemides-
mosomes. These phenotypic characteristics were
also confirmed in other studies later on [4, 5].
16.3 The Development of COMET Other investigators have also reported a
method of oral mucosa culture without the use of
Taking into consideration the issues of these pre- animal feeder cells (like fibroblasts) and/or
vious studies, Nakamura and co-workers [3] animal-­ derived products in the culture media,
developed a method to culture rabbit oral muco- thus avoiding the risk of animal pathogen trans-
sal epithelial cells on amniotic membrane as a mission and reducing the risk of rejection [6].
carrier (Fig. 16.1). Small oral biopsies (approxi-
mately 2–3  mm) were obtained from the oral
cavity. The biopsy specimens were then incu- 16.4 COMET: Surgical Procedure
bated with enzymatic reagents to separate epi-
thelial cells from the underlying connective A 360° conjunctival peritomy 3  mm from the
tissue. The resultant single-cell suspensions of limbus is performed. All perilimbal scarred or
oral mucosal epithelial cells were cocultured for inflamed subconjunctival tissue down to bare
2–3  weeks on a denuded amniotic membrane sclera is removed. The corneal pannus can be
carrier, with inactivated 3T3 fibroblasts. Toward completely removed by blunt dissection or
the end of the culture period, an air-lifting tech- superficial keratectomy using surgical scissors
nique was used to facilitate epithelial differentia- or a blade. The subconjunctival spaces can be
tion and stratification. The resulting cultured treated with MMC 0.04% for 5 min followed by
grafts were transplanted in rabbit studies. The vigorous wash with saline solution. The culti-
eight rabbit corneas studied were all clear and vated oral mucosal epithelial sheet is placed over
16  Cell Therapy Using Cultivated Oral Mucosal Epithelial Transplant (COMET) 227

the corneal surface and secured in place with medical or surgical therapy (including repeated
10-0 nylon sutures at the limbus. The integrity of COMET or tectonic grafts) requiring eviscera-
the cultivated epithelium is confirmed by fluores- tion finally.
cein staining at the end of the surgery, and the Nakamura and co-workers [9] reported better
ocular surface is protected with a silicone hydro- results in 19 eyes after a longer mean follow-up
gel bandage contact lens. of 55 months. All of the cases demonstrated total
re-epithelialisation of the corneal surface within
2–7  days after surgery. BCVA was improved
16.5 Clinical Results of COMET more than two lines in 15 eyes (79%), and VA at
the postoperative 36th month was improved in
Re-establishment of a stable and transparent cor- eight eyes (42%). All eyes manifested various
neal epithelium, regression of corneal conjuncti- degrees of postoperative corneal neovascularisa-
valisation/vascularisation, and resolution of tion, but it gradually abated, and its activity was
persistent epithelial defects (PEDs) have been stable at 6 months after surgery. Regarding com-
considered as criteria for clinical success in plications, seven eyes (37%) manifested PEDs
LSCF.  In a short-term follow-up study, ten eyes once during the follow-up periods. Ocular hyper-
with PED due to LSCF from several ocular sur- tension was detected in three eyes (16%). Finally,
face chronic conditions were transplanted with one eye developed a postoperative corneal infec-
COMET. At the fourth postoperative week, seven tion caused by a methicillin-resistant S. aureus.
eyes (70%) had achieved complete epithelialisa- Other authors [10] reported similar outcomes in
tion and three eyes showed a small epithelial 20 eyes with a mean follow-up of 31.9 months.
defect. At the 24th postoperative week, PED had Clinical success (defined as stable ocular surface)
improved in all patients except one patient who was 70.5% at 4  years and BCVA improved in
did not undergo the 24th week visit [7]. In another 70% of the eyes. Corneal perforation occurred in
study with longer mean follow-up of 25.5 months, one patient with SJS complicated by severe lid
40 eyes underwent COMET surgery [8]. Corneal abnormality that induced a persistent epithelial
surface stability revealed an early decline in trans- defect. This patient did not comply with the fol-
planted COMET over the first 6 months, remain- low-­up protocol and the cornea perforated requir-
ing stable thereafter (1 year, 64.8%; 2 years, 59%; ing a tectonic graft which was successful and the
and 3 years, 53.1%). Persistent epithelial failure situation settled.
despite COMET showed a very strong correlation
with preoperative PED (most of them eyes with
SJS, chemical and thermal injuries). Median sur- 16.6 Sequential Procedures:
vival in these eyes was only 27 days. Fibrovascular COMET + Keratoplasty
invasion of the corneal surface extending over the
pupillary was observed in eight eyes after surgery. Severe stem cell deficiency is sometimes accom-
This was especially remarkable in eyes with panied by significant corneal opacity and/or
OCP.  Reconstruction of a functional fornix was endothelial dysfunction. When the COMET pro-
combined with the corneal reconstruction at the cedure has been successful to restore the ocular
time of surgery in 15 eyes. This was achieved in surface, a corneal graft, usually penetrating kera-
11 out of the 15 cases at the final examination toplasty (PKP), may be considered as a two-step
visit. In terms of vision, visual acuity was approach in order to achieve a better corneal clar-
improved in many patients; however, the survival ity. In these patients, a keratoplasty without a pre-
curve showed a gradual decline afterwards. vious epithelial transplantation (COMET in this
Complications like melting, perforation, or case) would result in persistent epithelial defect
infectious keratitis occurred in ten eyes, and all of leading to many possible complications such as
these eyes had PED despite the COMET treat- graft melting, perforation, and/or conjunctival
ment. Finally, two eyes showed no response to invasion compromising the visual axis.
228 R. F. Buenaga and S. Ahmad

There are a few publications reporting neal epithelialisation occurred 5–7  days after
COMET+ keratoplasty outcomes. Inatomi et al. PKP in all eyes. Thirteen eyes had stable ocular
[11] reported two cases treated with COMET fol- surface without epithelial defect at final examina-
lowed approximately 6  months later by a PKP tion. One eye (case 6) developed persistent cor-
triple procedure. The mean follow-up was neal epithelial defect followed by bacterial
22.5 months. No complications such as persistent keratitis 3 months after surgery. It was success-
epithelial defects, rejections, or recurrence of fully treated by topical antibiotics and finally
cicatrisation were observed. In the first case healed with graft vascularisation. Overall graft
(chemical burn), the BCVA improved from hand survival rate was 92.6%, and rejection-free graft
motion preoperatively to 20/100 after both proce- survival rate at 12, 24, and 40 months was 76.9,
dures and remained stable during the follow-up. 69.2, and 69.2%, respectively. Mild stable periph-
In the second case (SJS), the preoperative visual eral superficial vascularisation (grade 1) began
acuity was hand motion and improved to 20/125 1–2 months after surgery in most of the cases and
after both procedures and remained stable after- peaked at 3–6 months. During follow-up period,
wards. Satake et al. [8], in their series of 40 eyes four eyes had episodes of endothelial graft rejec-
undergoing COMET surgery, performed a sec- tion, which was controlled with treatment.
ondary keratoplasty in 7 eyes (4 eyes had PKP, 2
eyes underwent anterior lamellar keratoplasty,
and 1 eye underwent deep anterior lamellar kera- 16.7 Conclusions
toplasty). The mean period between COMET and
keratoplasty and the mean follow-up after kerato- COMET is a valid alternative in the management
plasty were 12.6 months (range, 6–19.6 months) of bilateral LSCF with visual axis involvement.
and 22.2  months (range, 9.4–42.5  months), For a sub-group of patients with LSCF, such as in
respectively. Corneal epithelium was maintained graft versus host disease, where immune suppres-
in six eyes; however, two of these eyes showed sion is not possible, autologous treatment modali-
gradual conjunctival invasion and neovasculari- ties such as COMET are necessary. It is therefore
sation of the corneal surface at 18 months. Clarity an important management option for LSCF. The
of the cornea with epithelial stability was major benefits of COMET are its autologous
observed in 4 eyes (57.1%) after keratoplasty. nature and its mucosal origin. In bilateral eye dis-
Visual acuity in four eyes with no recurrence of eases, such as mucous membrane pemphigoid,
conjunctival invasion was still better than 20/125 Stevens-Johnson syndrome, and graft-versus-­host
at the end of the follow-up. Complications disease are most likely indications for COMET. In
included recurrent herpes keratitis in one eye and these diseases, where fluctuating inflammation is
an endothelial rejection in another one. a key component of the disease process, alloge-
Baradaran-Rafii et  al. [12] recently reported neic limbal transplantation is a less favourable
their results in 14 patients with previous success- treatment than COMET which is autologous. The
ful COMET surgery who underwent a PKP main indications for COMET are therefore bilat-
7.6  ±  1.3  months later. Mean follow-up period eral LSCF with visual axis involvement where
after PKP was 28.2 ± 8.0 months. Complete cor- immune suppression is not possible (Fig. 16.2).
16  Cell Therapy Using Cultivated Oral Mucosal Epithelial Transplant (COMET) 229

Corrected lid malposition and significant


fornix adhesions

Optimise the tear film

Visual axis NO Reconstruction not


involved? necessary

Yes

Total or NO Consider partial


more than epitheliectomy with AM
1/3 LSCF? overlay

Yes

Bilateral NO
Perform CLAU or auto-SLET
disease?

Failure

Yes
Use commercial available
auto-CLET

Can systemic
immune- NO
COMET
suppression
be used?

Yes

Kerato-limbal allograft or allo- Failure


Allo-CLET
SLET from cadaveric tissue

Fig. 16.2  Proposed decision-making flow chart


230 R. F. Buenaga and S. Ahmad

Compliance with Ethical Requirements  Sajjad Ahmad oral mucosal epithelium for the treatment of total
and Roberto Fernandez Buenaga declare that they have no bilateral limbal stem cell deficiency. Stem Cells.
conflict of interest. No human or animal studies were car- 2014;32(8):2135–46.
ried out by the authors for this chapter. 7. Sotozono C, Inatomi T, Nakamura T, et al. Cultivated
oral mucosal epithelial transplantation for persistent
epithelial defect in severe ocular surface diseases
with acute inflammatory activity. Acta Ophthalmol.
References 2014;92(6):e447–53.
8. Satake Y, Higa K, Tsubota K, et  al. Long-term out-
1. Ballen PH.  Mucous membrane grafts in chemical come of cultivated oral mucosal epithelial sheet
(eye) burns. Am J Ophthalmol. 1963;55:302–12. transplantation in treatment of total limbal stem cell
2. Gipson IK, Geggel HS, Spurr-Michaud SJ. Transplant deficiency. Ophthalmology. 2011;118:1524–30.
of oral mucosal epithelium to rabbit ocular surface 9. Nakamura T, Takeda K, Inatomi T, et  al. Long-term
wounds in vivo. Arch Ophthalmol. 1986;104:1529–33. results of autologous cultivated oral mucosal epithelial
3. Nakamura T, Endo K, Cooper L, et al. The success- transplantation in the scar phase of severe ocular sur-
ful culture and autologous transplantation of rabbit face disorders. Br J Ophthalmol. 2011;95(7):942–6.
oral mucosal epithelial cells on amniotic membrane. 10. Prabhasawat P, Ekpo P, Uiprasertkul M, et al. Long-­
Invest Ophthalmol Vis Sci. 2003;44:106–16. term result of autologous cultivated oral mucosal
4. Nakamura T, Inatomi T, Cooper LJ, et al. Phenotypic epithelial transplantation for severe ocular surface
investigation of human eyes with transplanted disease. Cell Tissue Bank. 2016;17(3):491–503.
autologous cultivated oral mucosal epithelial sheets 11. Inatomi T, Nakamura T, Kojyo M, et al. Ocular sur-
for severe ocular surface diseases. Ophthalmology. face reconstruction with combination of cultivated
2007;114(6):1080–8. autologous oral mucosal epithelial transplantation
5. Satake Y, Dogru M, Yamane GY, et al. Barrier function and penetrating keratoplasty. Am J Ophthalmol.
and cytologic features of the ocular surface epithelium 2006;142(5):757–64.
after autologous cultivated oral mucosal epithelial 12. Baradaran-Rafii A, Delfazayebaher S, Aghdami N,
transplantation. Arch Ophthalmol. 2008;126(1):23–8. et  al. Midterm outcomes of penetrating keratoplasty
6. Kolli S, Ahmad S, Mudhar HS, et  al. Successful after cultivated oral mucosal epithelial transplantation
application of ex  vivo expanded human autologous in chemical burn. Ocul Surf. 2017;15(4):789–94.
Cell Therapy Using Extraocular
Mesenchymal Stem Cells
17
Teresa Nieto-Miguel, Sara Galindo,
Marina López-­Paniagua, Inmaculada Pérez,
José M. Herreras, and Margarita Calonge

Abbreviations CCL2/MCP-1 Chemokine (C-C motif) ligand


2/monocyte chemoattractant
ABCG2 ATP-binding cassette subfam- protein-1
ily G member 2 CD Cluster of differentiation
ALDH3A1 Aldehyde dehydrogenase 3 CINC-1/CXCL1 Cytokine-induced neutrophil
family member A1 chemoattractant 1
AM Amniotic membrane CK Cytokeratin
APCs Antigen-presenting cells CLET  Cultivated limbal epithelial
AT-MSCs Adipose tissue-derived mesen- transplantation
chymal stem cells Cox-2 Cyclooxygenase-2
BM-MSCs Bone marrow-derived mesen- Cx43 Connexin 43
chymal stem cells CXCR4 C-X-C chemokine receptor type 4
C/EBPδ Cytosine-cytosine-adenosine-­ DP-MSCs Dental pulp-derived MSCs
adenosine-thymidine/ EGF Epidermal growth factor
enhancer-binding protein-δ GM-CSF  G r a n u l o cy t e - m a c r o p h a g e
CAT Catalase colony-­stimulating factor
CCL Chemokine (C-C motif) ligand GMP Good manufacturing practices

I. Pérez
These authors contributed equally: Teresa Nieto-Miguel, IOBA (Institute of Applied Ophthalmobiology),
Sara Galindo and Marina López-Paniagua University of Valladolid, Valladolid, Spain
T. Nieto-Miguel · S. Galindo · M. López-Paniagua · J. M. Herreras
M. Calonge (*) CIBER-BBN (Biomedical Research Networking
CIBER-BBN (Biomedical Research Networking Centre in Bioengineering, Biomaterials and
Centre in Bioengineering, Biomaterials and Nanomedicine), Carlos III National Institute of
Nanomedicine), Carlos III National Institute of Health, Madrid, Spain
Health, Valladolid, Spain
IOBA (Institute of Applied Ophthalmobiology),
IOBA (Institute of Applied Ophthalmobiology), University of Valladolid, Valladolid, Spain
University of Valladolid, Valladolid, Spain
Department of Ophthalmology, Clinic University
e-mail: calonge@ioba.med.uva.es
Hospital, Valladolid, Spain

© Springer Nature Switzerland AG 2019 231


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_17
232 T. Nieto-Miguel et al.

GPX Glutathione peroxidase SDF-1α/CXCL12 Stromal cell-derived factorα1/


GvHD Graft versus host disease C-X-C motif chemokine 12
HLA-DR Human leukocyte antigen-DR SGPT Serum glutamic-pyruvic
ICAM-1  Intercellular adhesion mole- trans­aminase
cule 1 SOD Superoxide dismutase
IDO Indoleamine-2,3-dioxygenase SSEA4 Stage-specific embryonic
IFN-γ Interferon gamma antigen-4
Ig Immunoglobulin TER Transepithelial electrical
IGF-I Insulin-like growth factor-I resistance
IL Interleukin TGF-β Transforming growth factor
iNOS Inducible nitric oxide synthase beta
iPSC Induced pluripotent stem cells TLR Toll-like receptors
iPSC-MSCs Induced pluripotent stem cell-­ TNF-α Tumor necrosis factor alpha
derived mesenchymal stem Treg Regulatory T cells
cells TSG-6 Tumor necrosis factor-a-stim-
IVCM In vivo confocal microscopy ulated gene/protein-6
KGF-2 Keratinocyte growth factor-2 TSP-1 Thrombospondin-1
LESCs Limbal epithelial stem cells UC-MSCs Umbilical cord-derived mes-
LSCD Limbal stem cell deficiency enchymal stem cells
M1 Macrophages type 1 VCAM-1 Vascular cell adhesion protein 1
M2 Macrophages type 2 VEGF Vascular endothelial growth
MCP-1 Monocyte chemotactic protein 1 factor
MDA Malondialdehyde WJ-MSCs Mesenchymal stem cells
MHC Major histocompatibility derived from the Wharton’s
complex jelly of the umbilical cord
MIP-1α Macrophage inflammatory XOX Xanthine oxidase
protein-1 alpha ZO-1 Zonula occludens-1
MMP Matrix metallopeptidase
MPO Myeloperoxidase
MSCs Mesenchymal stem cells
MSCT Mesenchymal stem cell 17.1 Introduction
transplantation
NaOH Sodium hydroxide Maintenance of corneal epithelium integrity is
NF-kB Nuclear factor-kappa beta essential for preserving corneal transparency and
NK Natural killer cells visual function. For that reason, a population of
NO Nitric oxide stem cells residing in the basal epithelial layer of
NT Nitrotyrosine the corneoscleral limbus, the so-called limbal
PanCK Pan-cytokeratin epithelial stem cells (LESCs), is continuously
Pax6 Paired box 6 renewing the corneal epithelium layers [1–4].
PCNA Proliferating cell nuclear Destruction or dysfunction of these limbal epi-
antigen thelial stem cells or their niche induces a syn-
PD-1 Programmed death-1 drome known as limbal stem cell deficiency
PDGF Platelet-derived growth factor (LSCD), which can be caused by a wide variety
PD-L1 Programmed death ligand-1 of ocular surface injuries and disorders such as
PEDF Pigment epithelium-derived chemical or thermal burns, multiple previous eye
factor surgeries, cicatrizing-autoimmune pathologies,
PGE2 Prostaglandin E2 severe dry eye syndrome, infections, congenital
RT-PCR Reverse transcription-poly- disorders, etc. LSCD is characterized by a defi-
merase chain reaction cient regeneration of the corneal epithelium that
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 233

eventually leads to persistent epithelial defects, tissue but also on their immunomodulatory and
ingrowth of conjunctival tissue onto the cornea anti-inflammatory properties, their capability to
surface, neovascularization, and persistent ocular migrate to injured and inflamed tissues, and their
surface inflammation. All these events usually capacity to modulate local environment, activate
result in loss of vision and a chronic pain syn- endogenous progenitor cells, and secrete trophic
drome [5, 6]. factors. In addition, compared with other types of
Transplantation of cultivated limbal epithelial stem cells, MSCs have several advantages, such
cells (CLET) harvested from the contralateral as availability and easy harvesting from a variety
healthy eye is the current treatment of choice for of tissue types, few ethical concerns, elevated
unilateral cases of LSCD, in which it is possible proliferative capacity in vitro, and low immuno-
to harvest autologous limbal tissue from the genicity [12, 14, 15].
healthy fellow eye and cultured it to expand the For all these reasons, MSCs have emerged as
stem cell population before transplantation [7]. In very attractive candidates for cell-based therapies
bilateral cases, which in fact are more frequent, it in numerous clinical applications including auto-
is necessary to resort to allogeneic limbal tissue immune and inflammatory diseases (e.g., graft
from deceased donors due to the scarcity of versus host disease (GvHD), Crohn’s disease, and
healthy tissue in patient’s eyes, transforming an rheumatoid arthritis), skeletal disorders, heart dis-
autologous transplant into an allogeneic one. eases, diabetes, neurological disorders, etc. [12,
Allogeneic transplantations require 1 year of sys- 15]. In ophthalmology, the potential therapeutic
temic immunosuppression in order to avoid applications are also many, and, according to the
immune rejection, which consequently increases ClinicalTrial.gov research database, there are cur-
the risk of the patient’s morbidity and the medical rently several ongoing clinical trials to test the
expenditures, although it avoids the costs related efficacy of MSCs in the treatment of a variety of
with the extraction of a biopsy [7, 8]. The use of ocular disorders such as retinitis pigmentosa, age-
alternative sources of extraocular cells could help related macular degeneration, glaucoma,
to overcome the dependence on and the limita- non-arteritic ischemic optic neuropathy, and dif-
tions of limbal epithelial cells, which are difficult ferent ocular surface disorders [16]. This book
to obtain and culture and that, in case of being chapter is specifically focused on the preclinical
allogeneic, can induce immune rejection [9–11]. and clinical advancements on the use of extraocu-
Mesenchymal stem cells (MSCs) constitute a lar MSCs for corneal epithelium regeneration.
subset of stromal cells that can be isolated from a
variety of tissues including bone marrow, adipose
tissue, dental pulp, umbilical cord, and limbal 17.2 E
 xtraocular MSCs for Corneal
stroma of the human eye, among others [9, 12]. Epithelium Regeneration
They are partially defined by their adherence to
plastic supports when cultured in standard condi- 17.2.1 Therapeutic Potential
tions; their multipotent differentiation potential of Extraocular MSCs
to bone, cartilage, and adipose tissue in vitro; and in Corneal Epithelium
the expression of a specific profile of surface Regeneration
antigens, including positivity for cluster of dif-
ferentiation (CD) 73, CD90, and CD105 and There are abundant in vitro and in vivo studies on
negativity for CD34, CD45, CD11b or CD14, the use of extraocular MSCs for corneal epithe-
CD19 or CD79α, and HLA-DR markers [13]. lium regeneration. All published studies have
The use of human MSCs in cell-based thera- revealed promising results in animal models and
pies has tremendously increased over the last have shown significant corneal regeneration,
decade in the field of regenerative medicine. improved corneal transparency, and a rapid heal-
There is a growing body of literature supporting ing process associated with the restoration of
that the therapeutic effects of MSCs not only rely vision [10, 17–19]. However, the studies per-
on their differentiation ability to repair damaged formed to unravel the mechanisms underlying the
234 T. Nieto-Miguel et al.

Fig. 17.1  Properties of Differentiation toward


mesenchymal stem cells other cell types Support of tissue-specific
(MSCs). MSCs have the progenitor cells
capacity to differentiate
into other cell types; to Selective migration
support tissue-specific to the site of injury
progenitor cells; to
provide a regenerative
microenvironment; to be
non-immunogenic; to
have
immunomodulatory,
Regenerative
anti-fibrotic, and
microenvironment
anti-apoptotic
properties; and to
migrate to the site of
MSCs
injury

Anti-fibrotic and
anti-apoptotic effect lmmunomodulation

Non-immunogenic cells

beneficial effects of MSCs on the damaged ocu- Transdifferentiation Capacity of Bone


lar surface have shown that multiple mechanisms Marrow-Derived MSCs (BM-MSCs) into
might contribute simultaneously to their thera- Corneal Epithelial Cells
peutic action. Although it remains uncertain if There are several studies published on the trans-
MSCs can transdifferentiate into corneal epithe- differentiation capacity of human BM-MSCs into
lial cells [20], these cells have shown a capacity corneal epithelial cells. It has been shown that
of secreting trophic and growth factors capable of human BM-MSCs cultured on amniotic mem-
stimulating resident stem cells, reducing tissue brane (AM) and transplanted onto the surface of
injury, an ability to exert anti-inflammatory and an alkali-burned rabbit cornea migrated to the
immunomodulatory effects, and a capability to cornea and expressed the corneal epithelial cell
migrate into injured tissues [11, 17, 19–21] marker cytokeratin (CK) 12 when located in the
(Fig. 17.1). corneal epithelium [26]. Subsequently, the poten-
tial of human BM-MSCs to differentiate into cor-
17.2.1.1 Transdifferentiation neal epithelial cells was confirmed when they
Capacity into Corneal were cultured first on pig Bowman’s membrane
Epithelial Cells and later under air-lifting conditions for creating
Corneal epithelium derives from the surface ecto- a multilayer of cells. After 4 weeks of differentia-
derm during embryonic development [22]. tion induction, Hou et  al. saw that part of the
Although it seems pretty clear that MSCs can dif- BM-MSCs expressed the corneal and limbal epi-
ferentiate not only into mesodermal but also into thelial markers CK12 and CK19, respectively,
non-mesodermal cell lineages including neuroec- but not the limbal stem cell marker ATP-binding
todermal and epithelial cells [23–25], the poten- cassette subfamily G member 2 (ABCG2) [27].
tial transdifferentiation capacity of MSCs into Afterward, Rohaina et al. also showed that while
corneal epithelial cells is unclear and remains CK3 and p63 were not detected in human
under investigation [20]. BM-MSCs, upon being cultured in limbal media
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 235

for 10  days, they significantly increased their highly specific for differentiated corneal epithe-
expression for the corneal and limbal epithelial lium [32].
markers CK3 and p63, respectively [28]. They Concerning BM-MSCs isolated from rats,
also demonstrated that transplantation of induced Jiang et al. demonstrated that co-cultivation of rat
BM-MSCs stratified on AM onto a nude LSCD BM-MSCs with corneal stromal cells induced the
rat model remarkably improved corneal regener- expression of the corneal epithelial-specific
ation in terms of corneal transparency and vascu- marker CK12 and the acquisition of morphology
larization. Although they did not track and cell structures (tight junctions) typical of epi-
transplanted cells, they observed CK3 and p63 thelial cells. Transplantation of the induced rat
protein expression in the newly regenerated cor- BM-MSCs cultured on AM for 7 days ­demonstrated
nea [28]. A recent study has also revealed that that those cells significantly improved the recon-
human BM-MSCs positive for the early embry- struction of the cornea surface in an alkali-­burned
onic stem cell marker stage-specific embryonic LSCD rat model [33]. Zhang et al. also supported
antigen-4 (SSEA4) had higher potential to differ- that rat BM-MSCs had the ability to transdifferen-
entiate into corneal epithelial cells. Even though tiate into corneal epithelial cells when they cul-
differentiated cells did not reach the typical tran- tured them on a xenogeneic acellular corneal
sepithelial electrical resistance (TER) of corneal matrix in  vitro for 7  days and showed that cells
epithelium, they presented an epithelial-like mor- acquired an epithelium-like shape and expressed
phology and were positive for the corneal and the corneal epithelial marker CK3 [34].
limbal epithelial markers CK3, CK12, CK8, In regard to mouse BM-MSCs, a recent
CK14, CK15, β1-integrin, and E-cadherin [29]. in vitro study demonstrated that when they were
In addition, another in  vivo study has demon- cultured with corneal extract in the presence of
strated too that human BM-MSCs administered insulin-like growth factor-I (IGF-I), mouse
to mice by subconjunctival injection migrated to BM-MSCs differentiated to cells with features of
the corneal tissues and displayed morphological corneal epithelial cells and keratocytes and main-
characteristics of epithelial, stromal, and endo- tained MSC properties. They observed that the
thelial cells [30]. expression of the corneal-specific markers CK12,
Regarding rabbit BM-MSCs, Gu et al. showed keratocan, and lumican was upregulated after
that when they were co-cultured with limbal epi- culture with corneal extract and that adding IGF-I
thelial stem cells, or their conditioned medium, to the culture medium significantly increased the
they changed their fibroblastic morphology to a expression of those genes [35].
more flattened epithelial-like shape and tran- In contrast to the evidence supporting the dif-
siently increased their expression of the corneal ferentiation potential of BM-MSCs into corneal
epithelial-specific marker CK3 [31]. They also epithelial cells, a study conducted by Ma et  al.
saw that when they were suspended in fibrin gels revealed that human BM-MSCs transplanted
and further transplanted onto alkali-burned rabbit over AM onto a chemically burned rat cornea sur-
corneas, rabbit BM-MSCs induced a successful vived and successfully reconstructed the dam-
ocular surface reconstruction, and some of those aged rat corneal surface but that the therapeutic
that located in the corneal epithelium expressed effect did not come from epithelial differentiation
the corneal epithelial-specific marker CK3, of BM-MSCs but as result of inhibition of neo-
implying BM-MSC differentiation into corneal vascularization and inflammation [36]. Likewise,
epithelial cells [31]. Interestingly, another study another study showed that rabbit BM-MSCs sys-
evidenced that, under standard culture condi- temically administered in a rabbit corneal alkali-­
tions, rabbit BM-MSCs were negative for the burned model, engrafted to injured cornea,
limbal and corneal epithelial markers ABCG2, promoted wound healing by stimulating and syn-
p63, and connexin 43 (Cx43), but positive for ergizing with native cells and by differentiating
CK3/12 and α-enolase, suggesting, for the first not into corneal or limbal epithelia cells but into
time, that CK3 and CK12 markers were not myofibroblast [37].
236 T. Nieto-Miguel et al.

Although the outcome of many in vitro experi- induction of the expression of the corneal
ments supports the idea that BM-MSCs are able epithelium-­specific marker CK12 [41].
to transdifferentiate into corneal epithelial cells
under certain conditions, in  vivo data published Transdifferentiation Capacity of MSCs
up to date has shown contradictory and inconclu- Derived from the Wharton’s Jelly
sive results pointing out that further investiga- of the Umbilical Cord (WJ-MSCs) into
tions are required to confirm whether BM-MSCs Corneal Epithelial Cells
can or cannot acquire a corneal epithelial cell Base on the accessibility, abundance, lack of ethi-
phenotype. cal issues, and capability of transdifferentiating
into epithelial cells [42], Garzón et  al. chose
Transdifferentiation Capacity human Wharton’s jelly MSCs to generate a three-­
of Subcutaneous Adipose Tissue-Derived dimensional heterotypic human cornea with
MSCs (AT-MSCs) into Corneal Epithelial fibrin-agarose scaffolds. They showed that
Cells WJ-MSCs were able to differentiate into corneal
Up to now, results from the few studies specifi- epithelial-like cells with results similar to the
cally performed on AT-MSCs differentiation into native corneas in the expression of corneal epi-
corneal epithelial cells do not bring enough evi- thelial key markers such as CK3/12, plakoglobin,
dence to support the hypothesis of their real zonula occludens 1 (ZO-1) and Cx43 [43]. On
transdifferentiation into functional corneal epi- the other hand, Sidney et al. studied the expres-
thelial cells since all of them are only based on sion of specific epithelial cell markers at different
in vitro experiments. While a first study showed passages and activity stages of corneal stromal
that orbital fat-derived stem cells had more poten- cell and compared it to the one in WJ-MSCs.
tial to differentiate into corneal epithelial cells Results demonstrated that in vitro culture of stro-
than MSCs derived from subcutaneous AT due to mal cells not only caused phenotypical changes
their same developmental origin during embry- but also induced expression of specific cytokera-
onic development [38], a subsequent study found tins. Despite not being from epithelial origin,
that human AT-MSCs expressed the progenitor WJ-MSCs were positively stained for CK3/12,
markers p63 and ABCG2 and the cytokeratins for CK19, panCK, and E-caherin, while corneal stro-
differentiated epithelial cells CK12 and CK76. mal cells only were stained for CK3/12 and
Furthermore, the study showed that CK12 expres- CK19. Reverse transcription-polymerase chain
sion spontaneously and progressively increased reaction (RT-PCR) analyses revealed signifi-
by cell adhesion over time in culture, suggesting cantly decreased gene expression of CK3, CK12,
that AT-MSCs had the potential to acquire cor- and CK19  in WJ-MSCs compared to corneal
neal epithelial-like characteristics under the stromal cells and showed detectable levels of
appropriate conditions [39]. Those results were both CK5 and E-cadherin genes, suggesting that
further confirmed by Nieto-Miguel et  al. when there might be some specificity to the cytokera-
they demonstrated that human AT-MSC grown tins expressed in stromal cells isolated from dif-
under basal culture conditions expressed the cor- ferent tissues [44].
neal epithelial markers CK3 and CK12 and that
AT-MSCs acquired a more epithelial-like mor- Transdifferentiation Capacity of Dental
phology and an upregulated expression of CK12 Pulp-Derived MSCs (DP-MSCs) into
when they were cultured in corneal epithelial Corneal Epithelial Cells
cell-conditioned medium for 15 days [40]. Later, There are only few studies in which it has been
the transdifferentiation capacity of AT-MSCs was explored the potential capacity of DP-MSCs to
also advocated when another in  vitro study differentiate into corneal epithelial cells [45–48].
showed that paired box 6 (Pax6) stable transfec- In the first published study, Monteiro et al. dem-
tion stimulated differentiation of murine onstrated that human immature dental pulp stem
AT-MSCs into corneal epithelial-like cells by cells expressed in vitro limbal epithelial stem cell
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 237

markers such as ABGC2, Cx43, integrin-β1, p63, even before an official definition about MSCs
and vimentin, while they did not express or was established, it was known that mesenchymal
weakly expressed the corneal epithelial-specific progenitor cells obtained from bone marrow
markers CK3 and CK12, respectively [45]. The could secrete different cytokines which had spe-
subsequent transplantation of DP-MSC sheets cific functions inside tissues under physiological
onto the ocular surface of a mild and severe conditions and whose expression was modified
LSCD rabbit model also confirmed that DP-MSCs under inflammation situations [49]. This fact,
successfully induced corneal epithelium recon- together with results obtained from different
struction and had the capacity to transdifferenti- studies in which BM-MSCs enhanced corneal
ate into corneal epithelial cells in vivo [45, 46]. wound healing in spite of the absence of MSC
Later, it was demonstrated that stem cells from migration to damaged tissue [50, 51], contributed
human exfoliated deciduous teeth also had the to establish the concept currently known as “tro-
potential to be induced to differentiate into cor- phic effect of MSCs.”
neal epithelial-like cells in  vitro. Tsai et  al. At the moment, it is known that MSCs secrete
showed that both direct and indirect co-culturing a large panel of soluble bioactive molecules
with corneal epithelial cells upregulated the which can cause direct intracellular signaling
expression of the corneal epithelial marker CK3 (autocrine effect), direct extracellular effects, or
and the corneal epithelial progenitor marker indirect effects by causing another cell in the
CK19 [48]. Furthermore, when Kushnerev et al. vicinity to secrete the functionally active agent
used soft contact lenses to deliver MSCs derived [52, 53]. The main functions of these bioactive
from pulp tissue explants to an ex vivo model of molecules are (1) to inhibit apoptosis and limit
human epithelial-debrided corneas in organ cul- the field of damage or injury (anti-apoptosis
ture, they saw that, once the cells had transferred, mechanism), (2) anti-inflammatory action, (3) to
they began to express the corneal epithelial-­ regulate angiogenesis (pro- and anti-angiogenic),
specific markers CK3 and CK12, supporting that (4) to stimulate and support the growth and dif-
DP-MSCs differentiated into corneal epithelial-­ ferentiation of local stem cell, and (5) oxidative
like cells once they attached to Bowman’s mem- stress protection [52, 54, 55]. The trophic effects
brane [47]. of MSCs have been also detected in corneal epi-
Although the results derived from several thelium regeneration, showing that different fac-
in vitro experiments support the idea that MSCs tors secreted by MSCs can be implicated in a set
are able to differentiate into a corneal epithelial-­ of processes, such as corneal re-epithelialization,
like cell phenotype under certain conditions, it and anti-inflammatory and anti-angiogenic
remains uncertain whether the expression of cer- actions that finally end in total or partial corneal
tain epithelial markers like cytokeratins repre- tissue regeneration.
sents a real functional differentiation or just the
result of MSC culturing [20, 44]. Moreover, the Corneal Wound Healing,
fact that results from several in  vivo studies do Re-epithelialization, and Anti-apoptosis
not support the data shown by in  vitro experi- Mechanism
ments suggests the need of performing additional Corneal wound healing is a complex multistage
in vivo studies to address whether MSCs can or process. For corneal wounds to heal optimally, a
cannot transdifferentiate into functional corneal cascade of multicellular interactions and tissue
epithelial cells. remodeling must proceed in an orderly fashion.
This process can be enhanced by several tissue
17.2.1.2 Trophic Activity repair growth factors secreted by cells and those
The Mesenchymal and Tissue Stem Cell recruited to the wound immediately after tissue
Committee of the International Society for injury [34, 56]. Currently, it is known that
Cellular Therapy proposed the minimal criteria BM-MSCs secrete factors as vascular endothelial
to define human MSCs in 2006 [13]. Nevertheless, growth factor (VEGF), epidermal growth factor
238 T. Nieto-Miguel et al.

(EGF), pigment epithelium-derived factor of the wounded cornea [64], this fact supports the
(PEDF), and transforming growth factor beta hypothesis that MSCs transplanted after corneal
(TGF-β) 1 [34], with the EGF and TGF-β1 play- and limbal epithelium debridement could pro-
ing an important role in the proliferation and mote the proliferation of LESCs and decrease the
migration of corneal epithelial cells [57–59] and, induced apoptosis in corneal epithelial cells.
therefore, in the stimulation of corneal epithe- Recently, it has been also shown that tumor
lium regeneration. When the effect of rat MSCs necrosis factor-a-stimulated gene/protein-6
versus rat LESCs was compared for in vitro heal- (TSG-6), secreted by murine BM-MSCs, pro-
ing corneal epithelium wounds, it was observed moted corneal epithelial wound healing in dam-
that the secretion of EGF and TGF-β1 was higher aged corneal epithelium of diabetic mice through
by MSCs than by LESCs. This difference would activation of LESCs [65].
indicate that MSCs could be better than LESCs
for corneal wound healing [34]. However, differ- Anti-inflammatory Action
ent studies showed that rabbit BM-MSCs, In 2006, it was reported the first evidence that
AT-MSCs, and LESCs secreted, spontaneously human BM-MSCs could be used for the treat-
or after stimulation with lipopolysaccharides, ment of damaged corneal epithelium through an
levels of TGF-β without significant differences anti-inflammatory mechanism, since pro-­
among them [60, 61]. In addition, when these inflammatory molecules which were detected in
three types of cells were cultured on nanofiber injured corneas treated with AM alone, such as
scaffold and transplanted onto a de-epithelialized CD45, interleukin (IL) 2, and MMP-2, were not
rabbit corneal surface, the transfer of stem cell-­ detected in the damaged corneal surfaces treated
seeded nanofiber scaffolds onto the damaged with AM plus MSCs [36]. Interestingly, the ther-
ocular surface significantly improved corneal apeutic effect of the transplantation could be
healing, in part because the number of apoptotic associated with the anti-inflammatory trophic
caspase 3+ cells decreased in the groups treated effect of MSCs, although this hypothesis must
with stem cell-seeded nanofibers with respect to be contrasted. Some years after, it was observed
untreated injured corneas or with respect to cor- that topical application of MSC-derived condi-
neas treated with cell-free nanofiber scaffolds tioned media contributed to restoration anti-­
[60]. This fact support the important implication inflammatory property of the rat corneal
of numerous trophic and growth factors secreted epithelium in damaged animals [66]. These
by stem cells, such as MSCs or LESCs, on wound authors suggested that the exogenous high con-
healing. In addition, the best re-epithelialization centration of IL-6 secreted by MSCs to extracel-
was observed in corneas treated with nanofiber lular medium, and subsequently administrated to
scaffolds cultured with BM-MSCs or LESCs, damaged corneas, could suppress dendritic cell
reporting that MSCs from different sources can activation in injured ocular surface [66–68]. In
have different trophic effects on wounded corneal accordance with this, several research groups
epithelium [60]. On the other hand, it is known have supported that soluble factors secreted by
that MSCs promote the survival and proliferation the MSCs played an important role as anti-­
of LESCs in the tissue, and that these effects may inflammatory therapy in wounded corneas, since
be mediated in a paracrine manner, due to the fact decreased corneal inflammation can be observed
that cell proliferation rate and expression of in spite of the absence of MSC migration to
EGFs were significantly higher in LESCs co-­ damaged tissue [51, 69], establishing that the
cultured with MSCs than in LESCs cultured anti-­inflammatory mechanism of MSCs could be
alone [62]. In addition, the expression of matrix mediated by paracrine pathways. Several works
metallopeptidase (MMP)-9 by damaged human have supported that trophic factors secreted by
corneal epithelial cells is strongly inhibited by MSCs could suppress the infiltration of inflam-
human MSCs [63]. Taking into account that matory cells and CD68+ macrophages by sup-
MMP-9 interferes with normal re-­epithelialization pressing the expression of the macrophage
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 239

inflammatory protein-1 alpha (MIP-1α) [51, 70, as pro-angiogenic molecules in the cornea, while
71] and that factors secreted by MSCs are capa- TSP-1 is a powerful anti-angiogenic factor that
ble of inhibiting production of pro-inflammatory acts inhibiting VEGF-induced angiogenesis [76–
tumor necrosis factor alpha (TNF-α) by macro- 78], maintaining the characteristic avascular cor-
phages in locally burned corneas [51, 72, 73]. In neal situation. Another molecule secreted by
addition, it is known that MSCs, activated by MSCs that has been proposed as anti-angiogenic
inflammatory signals, secreted the anti-­factor after corneal damaged is the TSG-6, since
inflammatory protein TSG-6 [69, 74]. it can decrease the neovascularization in corneas
Specifically, it is known that chemical injury to with their epithelium debrided [75]. On the other
corneal epithelial cells activated human MSCs to hand, it is known that MSCs and LESCs secrete,
secrete anti-inflammatory protein TNF-α that spontaneously or after stimulation with lipo-
stimulated TSG-6 expression, molecule which polysaccharides, levels of VEGF without signifi-
interacts through the CD44 receptor on resident cant differences among them, suggesting that
macrophages, attenuating the inflammatory cas- there is not additional risk of inducing corneal
cade that is initiated by them [74] and reducing neovascularization by MSCs [34, 60]. Therefore,
inflammatory damage to the cornea [69]. A when cultured rabbit BM-MSCs, AT-MSCs, or
meticulous study showed that intraocular injec- LESCs on nanofiber scaffold were transplanted
tion of TSG-6 into the anterior chamber reduced on a de-­epithelized rabbit corneal surface, the
markedly the expression of different pro-­ neovascularization decreased with all three types
inflammatory cytokines and chemokines (IL-6, of stem cells. Surprisingly, however, the greatest
IL-1β, cytokine-induced neutrophil chemoat- decrease was found in injured corneas treated
tractant 1 (CINC-1/CXCL1), and monocyte che- with nanofiber scaffolds seeded with BM-MSCs
moattractant protein-­ 1 (CCL2/MCP-1)), and or LESCs [60].
levels of MMP-9, as well as the infiltration of
neutrophils and proteases into the cornea after Stimulation and Support of the Growth
corneal and limbal epithelial debridement, being and Differentiation of Local Stem Cells
the effects of TSG-6 dose-dependent [75]. These Several studies have reported that MSCs can
results strongly support the hypothesis that induce re-epithelialization on wounded corneas.
TSG-6 produced by human BM-MSCs in This fact can be explained by the transdifferentia-
response to injury signals can protect the corneal tion of MSCs transplanted or by other mecha-
surface from the excessive inflammatory nism. A group of authors observed that when
response [75]. de-epithelialized rat corneas were treated with
rabbit BM-MSCs, AT-MSCs or LESCs cultured
Anti-angiogenic Function on nanofiber scaffold, the wound healing and the
In 2008, it was proposed that MSCs could have epithelial expression of CK3 and CK12 signifi-
anti-angiogenic actions that could be mediated cantly enhanced on the cornea [60]. They thought
by paracrine pathways, in accordance with a that MSC transdifferentiation is not the main
work in which it was shown that thrombospon- mechanism of the re-epithelialization effect, sug-
din 1 (TSP-1), a powerful anti-angiogenic factor, gesting that important effects represented by the
was upregulated at high levels in response to cor- production of numerous trophic and growth fac-
neal epithelium injury treated with topical MSC-­ tors that can support the growth of residual cor-
derived conditioned media [66]. Currently, it is neal epithelial cells and LESCs [60].
known that human BM-MSCs secreted different
molecules in basal conditions, such as VEGF, Oxidative Stress Protection
MMP-2, and TSP-1 [75]. In addition, MSCs Currently, it is known that reactive oxygen species
secrete high amounts of VEGF when these cells are increased in corneal epithelium after alkali cor-
are stimulated under inflammatory environment neal injury [79, 80]. However, rabbit BM-MSCs
[66, 68, 72, 75]. VEGF and MMP-2 play a role cultured on nanofiber scaffold can reduce alkali-
240 T. Nieto-Miguel et al.

induced oxidative and nitrosative stress in de-epi- Table 17.1  Trophic activity of mesenchymal stem cells
(MSCs)
thelized corneas. It is believed that MSCs cultured
on nanofiber scaffolds protect the formation of Trophic
factors
toxic peroxynitrite, lower apoptotic cell death, and secreted by
decrease matrix metalloproteinase and pro-inflam- MSCs Effect on corneal regeneration
matory cytokine production, resulting in reduced EGF Induction of proliferation and
corneal inflammation as well as neovascularization migration of corneal epithelial cells
and significantly accelerated corneal healing [80]. IL-6 Anti-inflammatory effect, suppressing
dendritic cells
In addition, it is known that the expression of anti- SOD3 Antioxidant effect
oxidant enzymes superoxide dismutase (SOD) and TGF-β Induction of proliferation and
xanthine oxidase (XOX) is restored in the regener- migration of corneal epithelial cells
ated corneal e­ pithelium similar to healthy corneas TNF-α Stimulation of TSG-6 expression
after BM-MSCs or LESCs nanofiber treatment TSG-6 Activation of LESCs. Inhibition of
local activated macrophages, decreased
[61]. In accordance with these results, some authors
expression of pro-inflammatory
postulated that MSCs can exert direct antioxidant cytokines, chemokines, and MMP-9.
activities through the secretion of antioxidant mol- Decreased neutrophil infiltration on the
ecules, due MSCs secrete the extracellular antioxi- cornea and decreased corneal
neovascularization
dant molecule SOD3 [81].
TSP-1 Anti-angiogenic effect
In summary, paracrine mechanisms of MSCs Trophic Effect of inhibition action on corneal
may exert a significant impact in promoting cor- factors regeneration
neal wound repair, which involves the joint par- inhibited by
ticipation of different soluble factors that MSCs
IL-2 Anti-inflammatory effect
modulate inflammation and angiogenesis as well
MMP-2 Anti-inflammatory effect
as improve tissue regeneration (Table  17.1).
MMP-9 Increase of proliferation of LESCs and
However, the involved biophysiological factors decreased apoptosis of corneal
and the underlying mechanism in corneal wound epithelial cells
healing remain unclear [82]. VEGF Anti-angiogenic effect
Recently, important reviews reporting the tro- Abbreviations: EGF epidermal growth factor, IL interleu-
phic effects of MSCs, obtained from different kin, LESCs limbal epithelial stem cells, MMP matrix
metallopeptidase, MSCs mesenchymal stem cells, SOD
sources, had been published [53, 83]. Concretely, superoxide dismutase, TGF-β transforming growth factor
in corneal regeneration, the potential less thera- beta, TNF-α tumor necrosis factor alpha, TSG-6 tumor
peutic effects of AT-MSCs in comparison with necrosis factor-a-stimulated gene/protein-6, TSP-1 throm-
BM-MSCs or LSCs could be due to the lower dif- bospondin 1, VEGF vascular endothelial growth factor
ferentiation potential of AT-MSCs and to the dif-
ferent spectrum of growth and immunoregulatory somes are secreted by a variety of cell types;
factors produced by these cells [53, 83], although between them we can find the MSCs, transport-
this fact is currently in controversy. On the other ing genetic material and proteins [87, 88]. They
hand, several authors have reviewed that com- are thought to mimic the roles played by MSCs
monly referred to as the MSC secretome, describ- from which they originate [88, 89]. Currently, the
ing exhaustibly soluble factors, and reporting effect of MSC exosomes on tissue wound healing
new mechanisms related with factors released in had been reported [90]. Although currently there
extracellular vesicles, for example, exosomes and are no studies in which this action was checked in
microvesicles [83–85]. Exosomes are one of sev- corneal epithelium, it could be other possible via
eral groups of secreted vesicles, which have a for trophic action of MSCs [83].
size of 40–100  nm, and can be distinguished
from microvesicles and apoptotic bodies by size 17.2.1.3 Immunomodulatory Ability
and morphology [86]. They can fuse with cellular Over time, several studies have confirmed that
membranes and it has been suggested that exo- MSCs have important immunomodulatory prop-
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 241

erties [82]. This question was proposed some Moreover, MSCs promote expansion and differ-
years ago thought studies in which MSCs directly entiation of regulatory T cells (Treg), which
inhibited proliferation of T cells [91, 92]. The inhibit T cell response [82]. On the other hand,
immunomodulatory action of MSCs is often MSCs have pattern-recognition receptors which
attributed at the action of different components can regulate specific immunostimulatory proper-
that this type of cells can secrete. To effectively ties of these cells, such as toll-like receptors
influence immunoregulation, MSCs require to be (TLR). For example, in response to TLR3 signal-
activated by an inflammatory microenvironment ing, MSCs maintain an anti-inflammatory pheno-
and stimulation by pro-inflammatory cytokines, type, while in response to signaling through
such as interferon gamma (IFN-γ) and TNF-α TLR4, MSCs adopt a pro-inflammatory pheno-
[82]. This immunomodulatory action can be type, upregulating production of IL-8, IL-1, and
characterized by a final immunosuppression CCL-5 [93, 108–110].
action or by a final immunostimulatory effect Finally, MSCs show reduced expression of
[82, 93]. In addition, MSCs can suppress mecha- major histocompatibility complex (MHC) class I
nisms regulated by innate and adaptive immune and do not express MHC II neither the classic co-­
cells [82, 93]. Regarding innate immune reaction stimulatory molecules [72]. In addition, MSCs
regulation, MSCs secrete molecules such as reduce the expression of MHC class II and co-­
granulocyte-macrophage colony-stimulating fac- stimulatory molecules such as CD80, CD86, and
tor (GM-CSF), IL-6, and prostaglandin E2 CD40 on the surface of other cells [82], facts that
(PGE2). GM-CSF elicits the anti-inflammatory wake it possible for MSCs to likely have a non-­
state of macrophages type 2 (M2) which have immunogenic phenotype, showing immunoprivi-
decreased the capacity of antigen presentation leged features. Therefore, MSCs can potentially
function and do not secrete proinflamatory mol- be used allogenically or xenogenically in a vari-
ecules [73, 94]. IL-6 secreted by MSCs inhibit ety of tissue disease states [54].
neutrophil action, preventing respiratory bursts
[95, 96], while PGE2 secreted by these cells Immunosuppression
inhibit mast cell degranulation [97]. In addition, and Immunostimulation Induced by MSCs
PGE2 together with TGF-β or indoleamine-­2,3-­ on Corneal Epithelium
dioxygenase (IDO) can decrease natural killer All these immune modulatory characteristics of
cell proliferation and the pro-inflammatory cyto- MSCs play a determinant role in corneal epithe-
kine production carried out by these cells, while lium regeneration. It is known that when human
PGE2 with IL-6 inhibit dendritic cell maturation, corneal epithelial cells are cultured under a pro-­
cytokine expression, and migration to lymph inflammatory environment (IFN-γ and TNF-α
nodes [98, 99]. In adaptive immune response, the stimulation), they increased their expression of
programmed death-1 (PD-1) and programmed intercellular adhesion molecule 1(ICAM-1),
death ligand-1 (PD-L1) factors, both secreted by MHC-I, and HLA-DR (MHC-II) [111]. However,
MSCs, can inhibit B cell proliferation and reduce MSCs effectively decreased IFN-γ/TNF-α-­
production of immunoglobulin (Ig) G, IgM, and induced ICAM-1, MHC-I, and MHC-II expres-
IgA [100–103]. In addition, MSCs inhibit T cell sion on this type of cells [111]. Since these
proliferation and regulate cytokine expression, adhesion molecules play an important role in the
being both mechanisms regulated through mole- infiltration of activated leucocytes in the injured
cules secreted by MSCs, such as nitric oxide corneal epithelium [112, 113], this fact could be
(NO), IL-10, PD-1, PD-L1, TGF-β, etc. [72, 91, one of the mechanisms through MSCs decrease
104, 105]. For example, MSCs can decrease the the infiltration of leukocytes in damaged corneas.
expression level of IFN-γ (proinflamatory mole- Currently, it is suggested that MSCs reduce the
cule) from Th1 cells and increase the expression expression of adhesion molecules in corneal epi-
levels of IL-4 and IL-10 from Th2 cells, both thelial cells through nuclear factor-kappa beta
with anti-inflammatory effects [72, 106, 107]. (NF-kB) signaling pathways [111], due to the
242 T. Nieto-Miguel et al.

specific inhibition of NF-kB nuclear transloca- mune ocular inflammation (autoimmune uveitis)
tion which decreases the expression of molecules has been tested, observing an increase of Treg
such as ICAM-1 and vascular cell adhesion pro- cells and M2  in peripheral blood and lung,
tein 1 (VCAM-1) on epithelial cells [66, 114– respectively [119]. However, since high levels of
116]. On the other hand, contrary to the M2 lasted longer than Treg, it was reported that
expectations, MSCs significantly attenuated M2 were effectively protected against experi-
cytokine-induced IDO expression on human cor- mental autoimmune uveitis independently of
neal epithelial cells [111], molecule which is Treg, indicating a direct role of M2 in the MSC-­
strongly linked to prolong corneal allograft sur- induced tolerance [119].
vival [117]. Since some years, the This fact was confirmed by other authors, who
­immunomodulatory effect of MSCs on corneal co-cultured MSCs with macrophage isolates
epithelium also was studied by in  vivo experi- from mouse peritoneum. When macrophages
ments. Currently, it is known mechanical injury where cultured alone, they differentiated into
on mice corneal-­limbus epithelium induced an classical macrophages type 1 (M1) (pro-­
inflammatory reaction characterized by produc- inflammatory); however, when these polarized
tion of IFN-γ and inducible nitric oxide synthase M1 were exposed to human umbilical cord-­
(iNOS) in the ocular surface [118]. In addition, derived MSCs (UC-MSCs), these macrophages
IL-6 is also expressed when damaged ocular sur- began to express markers typical of M2 (anti-­
faced is treated with allogeneic limbal transplan- inflammatory) [120], although the effect of
tation. However, both inflammatory situations UC-MSCs on M1 was inhibited when UC-MSCs
are decreased when nanofiber scaffolds with were modified, removing from their surface
BM-MSCs are transplanted in a mice model of extracellular matrix, dermatan sulfate, and hyal-
damaged corneal/limbal epithelium, suggesting uronan [120]. In addition, currently, it is known
that inflammation inhibition was partially regu- that the presence of extracellular matrix and hyal-
lated by MSC immunomodulatory effects, since uronan in the surface of UC-MSCs is vital for
MSCs inhibited IFN-γ secretion and T cell prolif- these MSCs to suppress inflammation and escape
eration in  vitro [118]. On the other hand, it has host rejection when they were transplanted in a
been reported the immunoregulatory effect of model of epithelial corneal alkali burn [120].
rabbit BM-MSCs and AT-MSCs cultured on Another action that the hyaluronan of MSCs can
nanofiber scaffold in a rabbit model of corneal regulate is the induction of regulatory T cells
de-epithelialization, which showed a strong tis- [120, 121]. In conclusion, currently it is sug-
sue infiltration with executor cells of adaptive (T gested that UC-MSC glycocalyx play an impor-
lymphocytes) and innate (iNOS-expressing cells) tant role in host immune modulation.
immunity [60]. The MSC transplantation signifi-
cantly decreased all of the harmful manifesta- Effect of Immunomodulatory Properties
tions of the injury; although less pronounced of MSCs in Corneal Transplantation
therapeutic effects of AT-MSCs were observed in Currently, MSCs have been widely used as
comparison with BM-MSCs. The effects of MSC immunotherapy in solid organ transplantation.
transplantation were related with the expression Over time, several authors have reported that
of genes for the immunoregulatory molecules MSCs show potential therapeutic effect for cor-
IDO-2, cyclooxygenase-2 (Cox-2), and iNOS by neal allograft transplantation, decreasing the
these cells, since these molecules secreted by this immune rejection and promoting corneal
type of cells like to inhibit proliferation of acti- allograft survival [82]. The first study developed
vated T cells, inhibit natural killer (NK) cells, and in this field was carried out in 2009, supporting
suppress mast cells in the presence of low con- that the survival of corneal xenografts was not
centration of IFNα/IL-1, respectively [93, 97]. significantly prolonged by rat commercial MSC
On the other hand, the effect of BM-MSC intra- application [122]. However, some years later, a
venous injection in a mouse model of autoim- research group found that intravenous injection
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 243

of rat BM-MSCs prolonged corneal allograft cells showed in animals treated with allo- or
survival when cells were administered immedi- third strain-MSCs compared to animal corneas
ately after transplantation [123], while preopera- treated with syn-MSCs. It was also shown a
tive infusion was ineffective, in accordance with decrease of CD4 T and B cell activated in cor-
the results previously reported [122, 123]. In this neas from animals receiving either of the MSC
work, when mononuclear cells were isolated types compared with untreated animals [125].
from the spleen and lymph nodes of both MSC-­ All these results induce to conclude that the
treated and vehicle-treated rats, previously trans- effect of MSCs on inhibiting the immune rejec-
planted with allogenic cornea, the proliferation tion response can depend of MSC source.
of T cells from MSC-treated rats was signifi- However, in spite of the data previously
cantly reduced compared with T cells from reported, the in vivo effect of MSCs on inhibiting
vehicle-­ treated postoperative rats, suggesting the immune rejection response is controversial
that MSCs may prevent rejection at least in part [82], and it is known that other parameters, as the
by reducing effector T cell proliferation in vivo. moment of MSC application with respect to tis-
It was also observed that media obtained from T sue transplant or the total cell doses applied, can
lymphocytes isolated from rats treated postop- play an important role about the effect of MSCs
eratively with MSCs showed a decrease quantity on corneal transplantation [82, 93]. When the
of Th1 pro-inflammatory cytokines and elevated effect of local and systemic injection of human
IL-4 secretion. In addition, Treg were upregu- AT-MSCs into two rabbit models of corneal
lated in mononuclear cultures isolated from rats allograft rejection (normal- and high-risk model)
treated with MSCs. Consequently, it was sug- was studied, in contrast to the expectation, it was
gested that MSCs can decrease corneal allograft observed that local injection of AT-MSCs, at the
rejection, postulating that this effect could be time of surgery, induced an increase of inflamma-
mediated by immunomodulatory reactions, as tion, corneal edema, and higher levels of infiltra-
inhibition of pathogenic T cell responses, an tion of lymphocytes in both models, inducing a
anti-­inflammatory shift in the Th1/Th2 balance, decrease in graft survival [126]. This results
and activation of Treg cells [123]. In addition, could not be explained by to the use of human
currently, it is known that intravenous-injected MSCs could create a xenogeneic environment,
mice BM-MSCs to allografted mice recipients, because the results were similar when trans-
3 h after surgery, increase allograft corneal sur- planted MSCs were isolated from human or rab-
vival by inhibiting antigen-presenting cell (APC) bit. It has been speculated that MSCs could be
maturation and induction of alloreactive T cells recognized by the adaptive immune system [127].
[119, 124]. On the other hand, the ability of rat In the current scenario in which an inflammatory
BM-MSCs from three distinct sources to pro- environment is already established, the secretion
long rat corneal allograft survival has been of IL-6 and IL-8 by AT-MSCs can exacerbate the
tested. MSCs obtained from same strain of recip- chemotaxis of leukocytes to the injected zone,
ient (syngeneic transplant, syn-MSCs), from thus increasing inflammation. In this case, the
same strain of donor (Allo-MSCs) or from a action of pro-inflammatory cytokines IL-6 and
third-type strain (third-­ MSCs) were injected IL-8 seems to be higher than the immunosup-
intravenously in rat recipient. A high percentage pressive effects of IDO and NO also secreted by
of untreated and syn-MSC-treated allografts AT-MSCs [126].
were rejected (80% and 100%, respectively). In conclusion, although the immunomodula-
Conversely, corneal allograft survival was sig- tory effect of MSCs has been shown in repeated
nificantly prolonged in allo-MSC (90%)- and occasions and some action mechanisms are
third-MSC (80%)-treated allograft recipients known at the moment (Table 17.2), more research
[125]. This fact could be explained by an inferior is necessary in order to clarify all immunomodu-
number of corneal natural killer T cells coupled latory actions that MSCs can perform in dam-
with a higher proportion of splenic regulatory T aged corneal epithelia.
244 T. Nieto-Miguel et al.

Table 17.2  Immunomodulatory ability of mesenchymal inflammation induce stem cell mobilization,
stem cells (MSCs)
migration, and colonization [82, 128]. Although
Immunomodulatory the tissue homing capacity of MSCs has been
factors secreted by General effects on corneal
MSCs regeneration
robustly demonstrated by several authors along
Cox-2 Inhibition of natural killer the time [55, 82, 83, 129–133], at the moment,
GM-CSF Induction of anti-­inflammatory the mechanisms involved in the regulation of
state of macrophages MSC homing capacity still remain unclear, while
IL-6 Inhibition of neutrophil action or it is thought that this property is mainly regulated
exacerbation of the chemotaxis
of leucocytes to damaged cornea
by growth factors and/or chemokines expressed
IL-10 Inhibition of T cell proliferation by damaged tissues and by extracellular matrix
iNOS Inhibition of mast cells receptors expressed on the surface of MSCs [55,
NO Inhibition of T cell proliferation 82, 134]. For example, the interactions of stromal
PD-1 Decrease of production of IgG, cell-derived factor 1/C-X-C motif chemokine 12
IgM, and IgA (SDF-1α/CXCL12) and C-X-C chemokine
Inhibition of B and T cell
proliferation receptor type 4 (CXCR4) directly mediate the
PD-L1 Decrease of production of IgG, trafficking of transplanted MSCs, while the pres-
IgM, and IgA ence of inflammatory cytokines as platelet-­
Inhibition of B and T cell derived growth factor (PDGF), TGF-β1, TNF-α,
proliferation
and SDF-1α/CXCL12 increases MSC motility
PGE2 Inhibition of mast cells
degranulation. PGE2 + TGFβ or [83, 135–137]. One of the several tissues where
IDO: decrease of natural killer MSCs can migrate is to the inflamed corneal epi-
cell proliferation. PGE2 + IL-6: thelium; due to the fact that damaged corneal epi-
inhibition of dendritic cells
thelium expresses specific chemoattractants that
TGF-β Inhibition of T cell proliferation
can mobilize MSCs to this region.
Immunomodulatory Effect of inhibition action on
factors inhibited or general immune response BM-MSC migration into peripheral blood,
downregulated by detected on corneal regeneration bone marrow, and cornea was monitored after
MSCs intravenous administration in a rabbit model of
IDO Increase of T cell proliferation alkali burn, observing that successful engraft-
IFN-γ Decrease of ICAM1 and MHCI
and II expression on corneal
ment of MSCs occurred in all treated animals,
epithelial cells. Consequently, although different cell migration patterns were
decrease of leucocyte infiltration observed. When MSCs were administrated to no
on corneal inflamed epithelium immunosuppressed animals, MSCs rapidly
TNF-α Decrease of ICAM1 and MHCI
migrated to bone marrow and corneo-limbal
and II expression on corneal
epithelial cells. Consequently, wound (epithelium, stroma, and corneal endothe-
decrease of leucocyte infiltration lium), although, subsequently, the presence of
on corneal inflamed epithelium MSCs decreased over time. However, when
Abbreviations: Cox-2 cyclooxygenase-2, GM-CSF MSCs were transplanted in immunosuppressed
granulocyte-­macrophage colony-stimulating factor, IDO animals, MSCs slowly migrated to bone marrow.
indoleamine-2,3-dioxygenase, IFN-γ interferon gamma,
Ig immunoglobulin, IL interleukin, iNOS inducible nitric In this case, MSCs remained long time in periph-
oxide synthase, MSCs mesenchymal stem cells, NO nitric eral blood, and presence of MSCs in cornea,
oxide, PGE2 prostaglandin E2, PD-1 programmed death- mainly in stomal area, slowly increased over
­1, PD-L1 programmed death ligand-1, TGF-β transform- time, observing high infiltration of MSCs after
ing growth factor beta, TNF-α tumor necrosis factor alpha
1 month. Consequently, currently it is known that
bone marrow state plays an important role in
17.2.1.4 Tissue Homing Capacity BM-MSC migration capacity when these cells are
Currently, it is known that MSCs have capacity to systematically injected. Other experiments showed
migrate specifically to inflamed tissues, in accor- that intravenously injected mice BM-MSCs, in a
dance with general the premise that injury and mouse model of corneal epithelial thermal
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 245

c­auterization, could migrate specifically to the injured corneal epithelium was also shown by
injured corneas, persisting MSC infiltration at another research group, who administrated
least for 50  days, while MSCs did not migrate human AT-MSCs in suspension (topical drops) in
into healthy contralateral corneas [138]. In addi- a rat model of corneal burn [143]. On the other
tion, it was shown that when exogenous MSCs hand, when human MSCs obtained from amni-
were not applied in damaged animals, increased otic membrane were administrated subconjuncti-
levels of endogenous MSCs were found in the vally on damaged corneal epithelia of rabbits, it
peripheral blood 48  h after the stimulus of cor- was observed that MSCs migrated to limbal-­
neal injury, a fact that correlated with the pres- corneal tissues at day 28 after administration
ence in blood circulation and corneal damaged [144]. In addition, human BM-MSCs in damaged
tissues of high levels of SDF-1/CXCL12 factor corneal tissue of mice after its subconjunctival
and substance P molecules which reported to administration could migrate to corneal epithe-
possess stem cell chemoattractant properties lium, in which MSCs appeared at different epi-
[139–141]. It was speculated that the injured cor- thelial layers. In addition, these cells showed
neal tissue was the source of these MSC che- different morphologies from cubic to flattened,
moattractants [138]. Recently, a new study has depending on their position within the tissue
supported the important role of SDF-1/ [30]. Also, when MSCs were cultured on amni-
CXCL12  in corneal epithelium healing when otic membrane, with or without keratinocyte
MSCs are implicated in this repair function. In an growth factor-2 (KGF-2) and/or with human
in  vitro model of wound healing, in which a serum, and this cell-substratum complex was
scratch was made through rat limbal MSC mono- transplanted on rat with damaged corneal epithe-
layer, the presence of SDF-1/CXCL12 stimulated lium, MSCs generally could migrate to injured
MSC migration from wound edge to damaged cornea [145]. In a recently developed study, in
culture area in order to repair cell monolayer which human AT-MSCs cultured on amniotic
[89]. In another study in which systemically membrane were transplanted on rabbits in which
MSC administration was used as therapy for the a partial or a total limbal stem cell deficiency had
treatment of corneal damage, specifically in cor- been previously induced, AT-MSCs could migrate
neal transplantation, it was shown that BM-MSCs, from amniotic membrane to ocular surface, since
intravenously administrated, preferentially home MSCs were found in the inflamed areas of the
to the inflamed ocular surface and draining lymph superior and inferior limbal stroma and corneal
nodes, while these cells did not migrate to the 8  weeks after transplantation [18]. This cell
healthy corneas [124]. migration could be regulated by the signaling
Some authors have suggested that after sys- pathway mediated by SDF-1/CXCL12 and its
temic MSC administration, most of the trans- receptor CXCR4 expressed by MSCs [146].
planted cells migrate to the main organs, such as However, MSC migration to inflamed tissues has
the lung, liver, kidney, and spleen, and, therefore, not been supported by all studies performed in
MSCs have lower capacity to migrate to areas this area [50, 51].
with tissue damage [82, 142], as the damaged At the moment, cell mechanisms involved in
corneal epithelium. Consecutively, some MSC migration to damaged tissues remain
researchers began to perform topical MSC trans- unclear, as well as the different physiological and
plantations. Some years ago, rat BM-MSC migra- pathological situations that regulate MSC migra-
tion to damaged corneal epithelium after topical tion [82]; although in general, results reported by
application was evaluated, finding MSCs in cor- several authors suggested that MSCs have the
neal epithelium 3 weeks after the injury. This fact capacity to migrate to injured corneal epithelium,
support that MSCs were successfully engrafted enhancing epithelial wound healing and reducing
underneath the epithelial cell layers up to 3 weeks corneal opacification and neovascularization,
after topical application [66]. The migration independently of the administration route. In
capacity of topically administrated MSCs to conclusion, scientific community accepts that
246 T. Nieto-Miguel et al.

homing capacity of MSCs, together with their 17.2.2.2 Mobilization of Endogenous


tropic and immunomodulatory actions, plays an MSCs
important role in corneal epithelium regeneration. Different researchers have demonstrated that
endogenous MSCs can carry out a therapeutic
effect in the cornea when a corneal epithelial
17.2.2 Extraocular MSCs damage has been induced. In a partial LSCD
in Experimental Models model, endogenous MSCs could reduce the epi-
of Corneal Epithelial Damage thelial corneal defects and opacity in rabbits with
normal bone marrow function [147], while this
17.2.2.1 Experimental Models did not happen in bone marrow-suppressed rab-
of Corneal Epithelial bits. Furthermore, MSCs and hematopoietic stem
Damage cells were able to migrate from bone marrow to
In order to study different potential therapies to the injured cornea in rabbits with competent bone
treat diseases, it is necessary to deepen in the marrow, but it did not occur in the bone marrow-­
knowledge and the understanding of diseases. suppressed animals [147]. Additionally, a signifi-
Experimental models represent one of the best cant increment in the number of circulating
tools that help to accomplish this goal. endogenous MSCs was observed in animals with
There are a lot of different kinds of experi- a partial LSCD [138]. This increment in circulat-
mental models of corneal epithelial damage with ing MSCs was accompanied by an increase of
diverse characteristics, such as the size of the SDF-1/CXCL12 and substance P levels both in
affected area in the cornea and/or the limbus, the peripheral blood and in corneal tissue. These fac-
severity of the model, and the methodology used tors then contribute to the mobilization of the
to develop it, among others. Experimental mod- endogenous cells [138]. Therefore, these data
els with central corneal damage but without lim- demonstrate that endogenous MSCs can migrate
bal injury are considered corneal burn or corneal from the bone marrow to the cornea in order to
injury models. Corneal burn models are induced trigger their therapeutic effect in corneal epithe-
by a chemical agent, whereas corneal injury is lial damage.
induced by mechanical damage. Experimental
models that include limbal damage are desig- 17.2.2.3 Intravenous Administration
nated as LSCD models. The most used agent dur- of MSCs
ing the induction of both corneal burn and LSCD There is evidence that MSCs administered by
models is sodium hydroxide (NaOH). However, intravenous injection can help during the wound
ethanol, n-heptanol, and mechanical scraping healing process and reduce the clinical signs of
have been also used to induce corneal epithelial the corneal epithelial damage in different experi-
damage. mental models. BM-MSCs intravenously admin-
There are several studies that describe the istered reduced significantly the epithelial defects
use of MSCs to treat corneal epithelial injuries in corneal injury and partial LSCD models [19,
using different experimental models and meth- 138]. Furthermore, the neovascularization devel-
odology. One of the principal factors that could oped in a partial LSCD model was also reduced
condition the results of these works is the by the intravenous administration of BM-MSCs
administration route: systemically (intravenous [37]. In addition, corneal opacity has been widely
or intraperitoneal), using different carriers studied in experimental models of corneal epithe-
(amniotic membrane, fibrin, biopolymers, etc.), lial damage, showing that intravenous MSCs
topically, or even by subconjunctival injection. were able to diminish it in both corneal injury
In addition, some researchers have studied the and LSCD models [19, 37, 69, 148]. Some of
therapeutic effect of endogenous MSCs in the these studies showed that MSCs effectively
ocular surface after induced corneal epithelial reduce corneal opacity in corneal epithelial dam-
damage. age, but their effect was variable depending on
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 247

the cell donor; MSCs from some donors were inflammatory cells in the corneal stroma and less
more efficient at reducing corneal opacity than disorganized cornea were observed when
cells from other donors [148]. Additionally, intra- BM-MSCs were intravenously administered in
venous induced pluripotent stem cells (iPSC)- LSCD models [150]. Furthermore, the level of
derived MSCs (iPSC-MSCs) could also reduce myeloperoxidase (MPO), a molecule released by
corneal opacity in a LSCD model [149]. activated neutrophils, was lower in the injured
Therefore, all these data together indicate that the corneas treated with intravenous BM-MSCs than
intravenous administration of MSCs in experi- in the untreated corneas [69, 148], indicating less
mental models of corneal epithelial damage can infiltration of neutrophils in the cornea of the
contribute to the reduction of clinical signs. treated animals [149]. Nevertheless, a variable
However, MSCs administration in those studies effect on the BM-MSC capacity to reduce the
was performed in the acute phase of the corneal MPO levels has been observed depending on the
epithelial damage. This fact could contribute to donor of the cells [148]. Moreover, a direct rela-
restrain the progression of the damage in an early tion between the level of TSG-6 expression in the
phase. In contrast, there are a lot of patients suf- BM-MSCs and the capacity of these cells to
fering LSCD or corneal epithelial damage in a reduce MPO has been described [148]. In addi-
chronic phase; hence it would be necessary to tion, the expression of the pro-inflammatory
study the effect of the MSCs intravenously cytokines TNF-α, IL-1β, and IL-6, CCL2, CCL3,
administered in corneal epithelial damage or and CCL4 was reduced in the corneas with LSCD
LSCD models well established in a chronic and treated with MSCs [69, 149]. Also the expres-
phase. sion of tenascin C, a proangiogenic and profi-
To trigger the therapeutic effect of MSCs in brotic molecule, was reduced in the corneas
the corneal epithelium when they are intrave- administered with MSCs [69]. Additionally, the
nously administered, it would be necessary that anti-inflammatory factors TGF-β and IL-1Ra
the cells migrate from the peripheral blood to the grew up in the corneal tissues after the intrave-
injured cornea or that they carry out their actions nous administration of BM-MSCs in LSCD mod-
from the distance. Some authors observed the els [138]. Thus, MSCs seem to have the capacity
presence of intravenously administered MSCs in to release these anti-inflammatory factors in the
the injured cornea [37, 138, 150]. Furthermore, injury site. Furthermore, in order to demonstrate
they also described that the systemic adminis- if the therapeutic effect of MSCs is mediated by
tered cells were found in the peripheral blood TSG-6, some authors knocked down the expres-
and migrated to the bone marrow [37]. Although sion of TSG-6  in MSCs, and these cells were
some authors advocate the migration of the intravenously administered without showing
MSCs intravenously administered to the injured therapeutic effect in the cornea [69, 149]. These
cornea, others have located fewer than 10 iPSC- data could indicate that the TSG-6 is responsible
MSCs in the cornea [149] or have not found data for the beneficial effect of the MSCs in the cor-
that supports the migration of MSCs [69]. In this nea. With all these data, we can conclude that the
case, the authors postulated that the therapeutic administration of intravenous MSCs promotes an
effect of the MSCs was produced from the dis- anti-inflammatory effect in the corneal epithe-
tance [69]. lium when it has been damaged. Furthermore,
The anti-inflammatory properties of MSCs this effect seems to be induced to a large degree
have been widely described in many tissues, such by the factor TSG-6, secreted by MSCs.
as the cornea. This anti-inflammatory effect Regarding corneal epithelium phenotype, there
could be beneficial for the regeneration of the are data to support that MSCs can help to the pro-
corneal epithelium after damage. In this regard, genitor residence cells to recover their phenotype
some authors have reported the anti-­inflammatory and function through the release of trophic and
role of the MSCs after their intravenous adminis- paracrine factors. The expression of the stem cell
tration in LSCD models. Less infiltration of markers ABCG2, p63, Hes1, and C/EBPδ, as well
248 T. Nieto-Miguel et al.

as the epithelial cell markers CK12 and CK19, studies applying the cells in a chronic phase of
increased in the corneas of partial LSCD models the corneal epithelial damage or in chronic LSCD
treated with intravenous BM-MSCs in compari- models are necessary to deepen on the knowl-
son with the non-­administered animals [37, 138]. edge of the beneficial effects of the MSCs sys-
In addition, the cell proliferative activity marker temically administered.
proliferating cell nuclear antigen (PCNA) showed
its expression not only in the basal limbus but 17.2.2.5 Ocular Topical
also in the peripheral corneal epithelium in the Administration of MSCs
MSC-treated group [37]. Therefore, MSCs pro- Topical administration is the easiest way to apply
mote a positive effect in the recovery of the phe- a drug on the ocular surface; however, it has some
notype of the corneal and limbal cells after a drawbacks, such as low retention time, high
corneal epithelial damage. washing rate, and low permeability of the corneal
epithelium, among others. The administration of
17.2.2.4 Intraperitoneal topical AT-MSCs in corneal burn models reduced
Administration of MSCs the epithelial defects similarly to the effect of the
The knowledge about the effect of MSCs intra- treatment with topical serum [143]. The lack of
peritoneally administered in experimental mod- differences in the therapeutic effect between the
els of corneal epithelial damage is limited. As in serum group and the AT-MSC-treated group
the case of the intravenous administration, a could represent a drawback. However, the mild-
decrease in corneal opacity and MPO levels was ness of the corneal burn model, along with the
observed after intraperitoneal administration of small number of animals and the short follow-up
MSCs in a rat LSCD model [69]. The expression time of the study, could have contributed to the
of pro-inflammatory cytokines and chemokines lack of differences between the treatment groups.
IL-1β, CXCL1, CCL2, CCL3, and CCL4 as well In addition, the corneas of a LSCD model topi-
as tenascin C was also decreased in the group cally treated with BM-MSCs did not show posi-
with intraperitoneal administration of MSCs. As tive healing results [80]. Thus, the rapid washing
in the intravenous administration, the MSCs away of the topically administered BM-MSCs to
administered intraperitoneally were not detected the ocular surface could have avoided their thera-
in the cornea [69]. Therefore, the therapeutic peutic action. In contrast, neovascularization and
effect of the MSCs was produced from the dis- corneal opacity decreased in the corneas treated
tance. However, IL-17 levels in corneal tissues with BM-MSCs or with BM-MSCs conditioned
were lower in a mice model of LSCD treated with medium in a rat LSCD model, mainly in the
intraperitoneal BM-MSCs in comparison with BM-MSC-treated group [66]. Additionally,
untreated animals [151]. Thus, the authors con- whereas the anti-angiogenic factor TSP-1
cluded that BM-MSCs have an anti-inflammatory increased, the pro-angiogenic factor MMP-2
effect on the IL-17 secreting cells by blocking decreased only in the BM-MSC-treated corneas
both non-Th17 cells and Th17 cells [151]. These [66]. All these data indicate a poor beneficial
data demonstrate that the MSCs intraperitoneally effect of the MSCs topically administered on the
administered have an anti-inflammatory effect in clinical signs of a corneal epithelial damage.
corneal epithelial damage. However, there is no AT-MSCs topically administered were found
data to support the migration from the peritoneal in the corneal epithelium and stroma of a corneal
area to the damage cornea, so their effect could burn model developed in rat [143]. Therefore,
be produced from the distance. It is important to cells were able to migrate from the corneal sur-
note that, as in the case of the intravenous admin- face to deeper layers of the corneal epithelium
istration, all the studies with MSCs intraperitone- and stroma. However, this migration experiment
ally administered were performed in an acute was only performed in both eyes of one rat; more
phase of the epithelial damage, and this fact data should be obtained from more animals to
could condition the results. Therefore, further draw any valid conclusions.
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 249

The inflammation of the corneal tissue was plantation of MSCs seeded on AM to the ocular
lower in corneal burn and LSCD models treated surface [18, 28, 33, 36, 145]. Additionally, the
with AT-MSCs or BM-MSCs than in other exper- expression of the pro-angiogenic factor MMP-2
imental groups [66, 143]. In addition, the expres- was reduced in the corneas treated with
sion of the pro-inflammatory factors IL-2 and BM-MSCs [36]. Furthermore, in order to check
IFN-γ decreased in the corneas treated with visual function of the rats, a head tracking method
BM-MSCs. However, the expression of the anti-­ showed that the animals transplanted with
inflammatory factors IL-10 and TGF-β increased BM-MSCs had better visual capacity than the
in the BM-MSC-treated corneas. Although, the untreated groups [36]. Nevertheless, transplanta-
level of IL-6 was mild, it was higher in the tion of BM-MSCs on AM did not show signifi-
BM-MSC-treated group than in the vehicle-­ cant improvement of the clinical signs in a
treated group [66]. This data point out that IL-6 corneal burn model, whereas BM-MSCs subcon-
could mediate the immunomodulatory action of junctivally injected showed therapeutic action
MSCs. In general, BM-MSCs applied topically [50]. However, it is important to note that the cell
to the damaged corneas for 2  h showed better dose administered by subconjunctival injection
results than conditioned medium administration. was higher than the cell dose seeded on the AM,
It could indicate that cell-to-cell contact has addi-thus it could have influenced the results. In gen-
tive effects to the action of soluble factors or thateral, the transplantation of MSCs seeded on AM
the continuous secretion of factors induces better to the ocular surface of experimental models with
outcomes than the administration of these factors corneal epithelial damage not only improves the
by the conditioned medium. These data along wound healing process but also reduces the neo-
with the poor therapeutic effect of the topical vascularization and corneal opacity.
MSCs observed in the clinical signs could indi- Some authors have demonstrated that MSCs
cate that the low retention time, the high washing migrate from AM to corneal tissues in experi-
rate, and the low permeability of the corneal epi- mental models of corneal epithelial damage [18,
thelium are responsible for these results. 36, 145]. Therefore, MSCs transplanted to the
Furthermore, there are not enough data about the ocular surface using AM as carrier are able to
migration capacity of the MSCs topically admin- migrate to the corneal tissues and exert its benefi-
istered to the ocular surface; thus, further investi-cial effect there.
gations in this field are necessary. The anti-inflammatory properties of the MSCs
have been also confirmed when they are trans-
17.2.2.6 MSCs Seeded on Amniotic planted on AM to the ocular surface of experi-
Membrane mental models with corneal epithelial damage.
The AM is the innermost layer of the placenta, Less inflammatory cells and less disorganization
and it is increasingly used for ocular surface of the cornea were observed in LSCD models
reconstruction in a variety of ocular pathologies, after the MSC transplantation using AM as car-
even as carrier for the ex vivo expansion of stem rier [18, 28, 33]. In addition, the expression of the
cells in order to treat corneal epithelial damage or leukocyte marker CD45 and the pro-­inflammatory
LSCD.  AM have several beneficial properties, cytokine IL-2 was significantly lower in eyes
such as low immunogenicity, stimulation of the transplanted with BM-MSCs [36]. All these
epithelialization, and anti-fibrotic, anti-­results indicate that the MSCs seeded on AM and
inflammatory, anti-angiogenic, and antimicrobial transplanted to the ocular surface can reduce the
effects [152]. In this field, the AM has widely infiltration of inflammatory cells in the experi-
used as carrier to transplant MSCs to the ocular mental models of corneal epithelial damage.
surface of experimental models with corneal epi- However, the mechanism that triggers the anti-­
thelial damage. The improvement of epithelial inflammatory action of the MSCs transplanted on
defects, neovascularization, and corneal opacity AM to the ocular surface has not been deeply
was observed in LSCD models after the trans- studied.
250 T. Nieto-Miguel et al.

Regarding corneal markers, the recovery of the untreated group. Moreover, the expression level
corneal epithelial markers CK3 and E-cadherin in of the epithelial marker Cx43 and the limbal stem
damaged corneas has been demonstrated after the cell markers ABCG2 and β1-integrin was higher
transplantation of AT-MSCs or BM-MSCs seeded in the cornea of the BM-MSC-treated group than
on AM [18, 28]. Nevertheless, other authors did in the untreated group [32].
not find the expression of the corneal epithelial
markers CK3, CK12, and panCK in the epithe- 17.2.2.8 M  SCs Seeded on Fibrin
lium of the BM-MSCs treated eyes [33, 36]. Matrix
Therefore, they postulated that the therapeutic Fibrin gel is prepared from fibrinogen and throm-
effect of BM-MSCs may not come from differen- bin, obtaining a matrix able to house cells and
tiation of these cells into corneal epithelial cells. suitable for transplantation. The transplantation of
Moreover, the expression of the limbal stem cell fibrin matrix could reconstruct the corneal epithe-
markers CK15 and p63 was partially restored in lium in a LSCD model not only when it contained
the limbal epithelium of the AT-MSC-treated eyes BM-MSCs but also without cells [31].
[18, 28]. Additionally, the rats transplanted with Furthermore, BM-MSCs were located in the cor-
BM-MSCs showed positive expression of neal epithelium 4 weeks after transplantation. In
CK19 in the limbal epithelium, while the untreated addition, the expression of the corneal epithelial
rats did not express this marker [145]. However, marker CK3 was observed in both experimental
the location of the CK19 marker is controversial groups; however the expression was more regular
because its expression has been described not and continuous in the cornea treated with
only in the limbal epithelium but also in the con- BM-MSCs. In addition, some of the BM-MSCs
junctival epithelium [153]. These works alto- transplanted to the ocular surface expressed CK3
gether show poor evidence of MSCs having effect [31]. However, other authors have described that
in the recovery of the corneal and limbal pheno- MSCs cultured in normal conditions express CK3
types. Therefore, further works are necessary to [32, 40]; thus the expression of CK3 in the trans-
elucidate this issue. planted BM-MSCs might not be indicating a real
In the case of MSC transplantation using AM differentiation of MSCs into corneal epithelium.
as carrier, the cell dose is lower than in the admin-
istration by other routes. Furthermore, the dose of 17.2.2.9 Contact Lens Carriers
the transplanted MSCs is not totally controlled in Other carriers used to transplant MSCs to the ocu-
several of the studies performed using AM as car- lar surface are contact lenses. The effect of human
rier, because some authors seed a number of cells AT-MSCs seeded on scleral contact lenses applied
on the AM, but then they keep them in culture for over the cornea of a LSCD model has been stud-
some days. Additionally, MSCs seeded on AM are ied [154]. Epithelial defects, neovascularization,
administered in a less acute phase of the disease and corneal opacity were reduced in the animals
than with other administration ways; this fact treated with AT-MSCs in contrast to the group
could help to know the effect of the MSCs in a with contact lenses without cells. The authors also
chronic phase of the corneal epithelial damage. observed less disorganization and fewer inflam-
matory cells in the stroma of the AT-MSC-treated
17.2.2.7 Amniotic Membrane Pocket corneas [154]. Thus, contact lenses could be an
AM has also been used in order to create a pocket adequate substrate to seed stem cells and to trans-
in which BM-MSCs were administered in a plant them to the ocular surface in order to treat
LSCD model developed in rabbits [32]. However, corneal epithelial damage.
there were no differences in clinical signs among
the different experimental groups [32]. The cor- 17.2.2.10 Poly-L-Lactic Acid Carriers
neal epithelial markers CK3/12 were more Biopolymers are one of the alternative carriers
strongly expressed in the corneas of the that have been investigated in order to find repro-
BM-MSC-treated group than in those of the ducible substrates that allow cell adhesion and
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 251

proliferation, as well as the re-epithelialization of [60, 61, 80, 155]. Furthermore, the expression of
the ocular surface when they are transplanted to XOX and the levels of the antioxidants SOD,
the ocular surface of experimental models with glutathione peroxidase (GPX), and catalase
corneal epithelial damage. The most studied bio- (CAT) were similar between the BM-MSC-
polymer in this kind of experimental model is treated corneas and the healthy control corneas,
based on poly-L-lactic acid. Epithelial defects, while SOD, CAT, and GPX showed lower levels
neovascularization, corneal opacity, and corneal in AT-MSC-treated corneas than healthy corneas
thickness showed a reduction in corneal burn and [61]. Moreover, in the injured corneas treated
LSCD models transplanted with poly-L-lactic with BM-MSCs the expression of MMP-9 was
acid carriers containing MSCs, both BM-MSCs reduced or suppressed [61, 80, 155]. Therefore,
and AT-MSCs [60, 61, 80, 155]. However, all these results demonstrate the anti-inflamma-
AT-MSCs showed less pronounced therapeutic tory properties of MSCs, as well as their benefi-
effect in LSCD models than BM-MSCs [60, 61]. cial role against the oxidative stress produced in
In addition, the expression of the pro-angiogenic a corneal epithelial damage. Nevertheless, the
factor VEGF was also reduced in the BM-MSC-­ treatment was applied always in an acute phase
treated corneas [60, 61, 80, 155]. All these data of the damage. In this regard, it has been
demonstrate that MSCs seeded on poly-L-lactic described that the sooner the treatment is applied
acid nanofibers and transplanted to the ocular the better the healing is, mainly in an inflamma-
surface can mitigate the clinical signs in experi- tory and oxidative environment. However, in the
mental models of corneal epithelial damage. clinical setting, there are a lot of patients in a
Furthermore, BM-MSCs show higher therapeutic chronic phase of the disease, and then it is neces-
effect than AT-MSCs when they are transplanted sary to know the behavior of these cells in those
using poly-L-lactic acid carriers. patients. Therefore, further experiments in a
There are no data about the migration capacity chronic phase of corneal damage are necessary
of the MSCs from the poly-L-lactic acid carriers to overcome this issue. In addition, as in the AM
to the ocular surface. It would be very interesting experiments, the number of stem cells that can
to perform further experiments to elucidate the be transplanted using this kind of substrates is
migration of the cells transplanted in this kind of very low in contrast to the cell dose that can be
substrates. administered by injection. It is important to note
Regarding corneal inflammation, MSCs that the cell dose transplanted in these carriers to
transplanted to the ocular surface on poly-L-lac- the ocular surface was not totally controlled
tic acid carriers were able to decrease the pres- because the cells were kept in culture for some
ence of CD3+ cells and macrophages in the days before transplantation.
corneal tissues of LSCD models [60, 80]. Regarding corneal phenotype, the expression
Furthermore, the expression of the pro-inflam- of the corneal epithelial markers CK3 and CK12
matory cytokines TGF-β, IL-6, IL-8, IL-1β, increased in the MSC-treated groups, mainly in
IL-2, and IFN-γ also decreased in the corneas the BM-MSC-treated group [60].
with epithelial damage and treated with
BM-MSCs seeded on poly-L-­lactic acid carriers 17.2.2.11 Polyamide 6/12 Carriers
[61, 80, 155]. On the other hand, the increment Polyamide nanofiber scaffolds containing LESCs
of the oxidative stress in corneal epithelial dam- + BM-MSCs have been transplanted to the ocular
age has been demonstrated [80]. In this regard, surface of a LSCD model [118]. A decrease of
the BM-MSCs seeded on nanofibers reduced the the iNOS, IL-2, and IFN-γ levels was observed in
expression of the oxidative stress markers iNOS, the treated corneas in comparison with the
caspase-3, nitrotyrosine (NT), and malondialde- untreated eyes [118]. This anti-inflammatory
hyde (MDA) and on the other hand promoted the effect could be originated by both cell types, or
increment of the protective enzyme aldehyde just by one of them, but the design of the study
dehydrogenase 3 family member A1 (ALDH3A1) does not allow to know.
252 T. Nieto-Miguel et al.

17.2.2.12 Subconjunctival Injection tein-1 (MCP-1), the pro-inflammatory cytokines


of MSCs TNF-α and IL-1β, and the protein MMP-2 was
Subconjunctival injection is a technique widely lower in the corneas of the BM-MSC-treated
used in ophthalmology. Although this way of groups in contrast to the untreated groups [51, 65,
administration allows to apply a local treatment 156]. Additionally, the anti-inflammatory factor
with a high cell dose, the volume administered TGF-β showed an increment in the BM-MSC-­
have to be low. A reduction in epithelial defects, treated corneas [156]. Moreover, the expression
neovascularization and corneal opacity has been of TSG-6 was elevated in the corneas treated with
observed after subconjunctival administration of BM-MSCs, whereas no changes were observed in
MSCs in corneal burn experimental models [50, other factors that could be also related with the
51, 156, 157]. Furthermore, BM-MSCs adminis- immunomodulatory effect of MSCs [65]. In addi-
tered by subconjunctival injection immediately tion, BM-MSCs knocked down for TSG-6 did not
after injury produced a decrease of the epithelial show therapeutic effect in the diabetic corneas
defects in the diabetic mice, showing similar reep- with epithelial damage [65]. Therefore, TSG-6
ithelialization rate of the nondiabetic mice [65]. It seems to be one of the mediators of the therapeu-
has also been demonstrated that the expression of tic effect of the MSCs when they are subconjunc-
the pro-angiogenic factors VEGF and MMP-9 is tivally injected into a damaged cornea.
lower in the BM-MSC-treated corneas, whereas An increase in the expression of the epithelial
the anti-angiogenic TSP-1 factor increase its markers Cx43 and β-catenin in the AT-MSC-­
expression in the corneas of the BM-MSC-treated treated corneas was described, whereas differ-
groups [50, 51, 156]. Therefore, MSCs, both ences were not seen in the epithelial marker
BM-MSCs and AT-MSCs, subconjunctivally E-cadherin and in the limbal stem cell marker p63
administered have a positive effect in the reduc- [157]. Therefore, the role of the MSCs subcon-
tion of the clinical signs in experimental models junctivally administered in the recovery of the
with corneal epithelial damage. corneal and limbal phenotypes is not yet clear.
In regard to the migration of the MSCs sub-
conjunctivally administered, although some 17.2.2.13 Combined Administration
authors have located the BM-MSCs in the wound of MSCs
edge of the cornea [65], others located the AT-MSCs have also been administered to the cor-
BM-MSCs in the injection site but not in the cor- nea of a partial LSCD model using combined
nea [50, 51]. These data seem to indicate that the routes: topically, subconjunctivally, and in a stro-
MSCs subconjunctivally administered tent to mal pocket [158]. The AT-MSC-treated group
keep in the injection site and do not migrate to the showed attenuated epithelial defects, neovascu-
central cornea. Nonetheless, the therapeutic larization, and cornea opacity in contrast with the
effect of the cells has been demonstrated; thus the untreated corneas. The expression of the prolif-
MSCs trigger their beneficial role from the injec- eration marker Ki67 increased in the AT-MSC-­
tion site without migrating to the damaged area. treated corneas, whereas a reduction of the
The MSCs subconjunctivally administered secreted VEGF factor was observed in these cor-
also showed an anti-inflammatory role in the ocu- neas. Twenty-eight days after damage and treat-
lar surface. Less infiltration of inflammatory cells, ment, the AT-MSCs administered were found in
macrophages (CD68+ cells), and leukocytes the cornea. In regard to inflammation, the authors
(CD45+ cells) was observed in the MSC-­treated noted inflammatory cells in the untreated cor-
eyes of experimental models with corneal epithe- neas, while the animals treated with AT-MSCs
lial damage [50, 51, 65]. In addition, the expres- presented a normal corneal stroma without
sion of MPO was lower in the BM-MSC-­treated inflammation. Furthermore, the indicator of dam-
corneas than in the PBS-treated corneas [65]. age serum glutamic-pyruvic transaminase
Furthermore, the expression of the chemotactic (SGPT) was reduced in the corneas treated with
factor MIP-1α, the monocyte chemotactic pro- AT-MSCs. Therefore, the combined administra-
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 253

tion of MSCs, using three local applications, is of the corneal epithelium by secreting factors
effective in the treatment of corneal epithelial that act at a paracrine level promoting prolif-
damage in experimental models. eration and differentiation of the resident stem
In conclusion, both BM-MSCs and cells in the tissues.
AT-MSCs have shown therapeutic effects at
reducing clinical signs in corneal epithelial
damage or LSCD experimental models using 17.2.3 Extraocular MSCs for Corneal
different administration routes. Nevertheless, Epithelial Damage in Humans
all of these ways of administration have limi-
tations; the intravenous or intraperitoneal MSCs have been used successfully to repair and
administrations inject the MSCs very far from regenerate tissues [159]. In the 1980s, it was con-
the cornea, the transplantation of MSCs on cluded that MSCs could differentiate into osteo-
carriers do not allow high cell dose, in the top- blasts, chondrocytes, adipocytes, and myoblasts,
ical administration the cells are washed away and, consequently, they have been used to repair
very quickly, and the volume administered by infracted myocardium [160], in hepatic cirrhosis
subconjunctival injection has to be low. [161], in gastrointestinal cancer [162], in peri-
Despite these drawbacks, all these administra- odontal therapy [163], or for bone deformities
tion routes of MSCs are effective in the treat- and fractures [164]. It was also demonstrated that
ment of experimental corneal epithelial these cells could contribute to the regeneration of
damage. Additionally, migration of MSCs ectodermal tissues, such as bronchial and alveo-
from the application site to the cornea has lar epithelial cells [165], glomerular and tubular
been confirmed using intravenous and intra- cells, or gastrointestinal cells [166]. This process
peritoneal injection or transplantation of the of differentiation will depend on induced and
cells on a carrier. However, subconjunctival control factors that cohabit in the environment
injection does not allow the migration of the where they are located [167].
cells to the central cornea, and further experi- In the field of ophthalmology and vision-­
ments are needed to know the migration of related therapies, some approaches have been
topically administered MSCs. Furthermore, made using MSCs, mainly at the ocular surface.
the anti-inflammatory properties of these But even though there are publications using
cells, as well as their beneficial role against these cells in animal models [10, 17–19, 60, 82],
the oxidative stress triggered by the epithelial there is practically no studies in the clinical area.
corneal damage, have been demonstrated. In MSCs have been used to treat the deleterious
addition, one of the principal factors impli- consequences of chronic GvHD, one of the most
cated in the anti-inflammatory effect of MSCs dreaded complications of otherwise successful
is TSG-­6. Although some authors advocate the allogeneic hematopoietic stem cell transplanta-
transdifferentiation of extraocular MSCs to tion to treat hematological malignancies [168,
corneal epithelial cells, only one of them have 169]. Some years ago, a study published the
demonstrated that MSCs expressed CK3 in the beneficial effect of intravenous injection of
corneal epithelium when they were trans- MSCs to treat the severe ocular surface involve-
planted to the ocular surface of an experimen- ment that often occurs in this disease. This study
tal corneal epithelial damage [31]. But since described improvement in clinical dry eye
some authors have shown that MSCs express scores in 54.55% of the 22 patients enrolled,
CK3  in basal conditions [18, 32], the CK3 also accompanied by in increased levels of regu-
expression might not indicate a real transdif- latory CD8+CD28− lymphocytes but not
ferentiation. Therefore, transdifferentiation of CD4+CD25+, which led authors to conclude that
MSCs into corneal epithelial cells has not be the immunomodulatory effect by which MSC
confirmed in  vivo, but all the results seem to transfusion improved the refractory dry eye dis-
indicate that MSCs contribute to the recovery ease could be due to regulating the Th1/Th2 bal-
254 T. Nieto-Miguel et al.

ance by triggering the generation of CD8+CD28−T 17.3 Future Perspectives


lymphocytes [169]. and Conclusions
There is also a single case report in which a
patient suffering from keratoconus and a persis- Extraocular MSCs represent an available, non-­
tent corneal epithelial defect unresponsive to immunogenic source of stem cells that has proved
conventional therapy was treated by topical to possess a potential therapeutic value in corneal
administration of adipose tissue-derived MSCs, epithelium regeneration based on results obtained
achieving corneal epithelial healing after in both preclinical and clinical studies, although
1 month [170]. only a few clinical studies have been performed
MSCs have been experimentally used to treat yet, as this approach is quite novel. However,
not only the acute phase of ocular chemical MSCs isolated from different tissues by different
burns [171] but also the chronic devastating con- techniques may possess different properties and
sequences of severe ocular chemical injuries. To levels of activity, which may directly influence
this later aim, there exists now enough experi- their efficacy. In addition, most of the MSC isola-
mental evidence showing that MSCs are benefi- tion, propagation, and characterization protocols
cial in dealing with the chronic sequelae of vary among laboratories, and clinical translation
chemical injuries, helping to provide corneal of MSC-based therapies implies that their manu-
epithelium with the lost regenerative capacity facturing process should comply with good man-
due to the destruction of the limbal stem cells ufacturing practices (GMP) in order to preserve
and usually their niche entirely [10, 17–19]. But, the quality and safety standards of the final cell
although MSCs are shown useful in animal mod- product. Therefore, the development of standard-
els of LSCD, they have not been yet fully ized protocols for the preparation, characteriza-
approved to clinical use, as there is only one tion, and evaluation of the biological potential
clinical trial on the use of MSCs for LSCD in activity (potency assays) of MSCs is essential to
humans. This exploratory proof-of-concept clin- better understand MSCs function and clinical
ical trial, performed in our institution [172], utility and to improve comparison of results
explored the safety and efficacy of bone marrow- between research and clinical centers. To achieve
derived MSCs for bilateral LSCD of any etiol- this goal, cooperation between basic researchers,
ogy in which the more widely accepted physicians, industry, and regulatory authorities is
autologous cultivated limbal epithelial trans- fundamental.
plantation (CLET) was not feasible due to the Although several scientific groups, including
complete absence of healthy limbus where to ours, have reported promising results on the use
remove a biopsy from [7–9]. In this randomized of MSCs for corneal epithelial regeneration, the
controlled clinical trial, we compared patients precise mechanism of how these cells exert their
subjected to CLET, as we had enough previous therapeutic effects on the ocular surface remains
experience in this procedure [8, 173], with unclear and deserves further investigation.
patients subjected to MSC transplantation Additional research in LSCD animal models is
(MSCT), enrolling though a small but significant needed to address whether MSCs can or cannot
number of patients due to its exploratory nature. transdifferentiate into corneal epithelial cells, to
Using an extensive evaluation endpoint analysis, determine how long MSCs can survive in the cor-
there was enough evidence showing that MSCT nea secreting factors that reduce tissue injury,
was as safe and as effective as CLET, with a and to answer questions such as which is the
73–86% success rate after 12  months [172] most appropriate cell dose and best route of
(Fig. 17.2). These results, although must be rep- administration for applying MSCs onto the ocu-
licated in a larger clinical trial, provide hope to lar surface.
those patients that are not eligible to have autolo- Our proof-of concept clinical trial have shown
gous limbal transplantation procedures due to that MSCs transplantation onto the ocular sur-
extensive bilateral disease. face of patients suffering from LSCD can safely
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 255

Fig. 17.2  This 48-year-old man had a bilateral chemical 4  weeks before cell transplantation. The middle row
injury 4 years before, for which multiple amniotic mem- shows transplant in place with the cells facing down and
brane transplantations were performed, as per standard of the amniotic membrane (arrows) facing the scleral contact
care. Both eyes show the typical end-stage picture of a lens 2 days after surgery. Lower row displays successful
severe chemical burn: residual corneal vascularized opaci- transplant after 12 months. In vivo confocal microscopy
fication that is due to a full-thickness stromal scarring. showed a conjunctival-like phenotype in the central cor-
Not seen in these pictures are bilateral cataracts, many nea before transplantation (upper row). Twelve months
areas of synechiae between the iris and posterior cornea after surgery, a corneal-like phenotype was present in his
and glaucoma. Thus, three surgical procedures, i.e., stem left central cornea (lower row). The continued improve-
cell transplantation, further penetrating keratoplasty, and ment of the central corneal epithelial phenotype 12 month
another non-corneal surgery (cataract removal), would be after transplantation indicates that this patient would have
required to recover vision. He underwent MSC transplan- a better chance for a successful penetrating corneal trans-
tation in his left eye. The upper row displays left eye plant to improve his vision
256 T. Nieto-Miguel et al.

and effectively restore the corneal epithelium. 4. Schermer A, Galvin S, Sun TT.  Differentiation-­
related expression of a major 64K corneal kera-
Nevertheless, to keep moving forward it is tin in  vivo and in culture suggests limbal location
needed further research and to decipher the of corneal epithelial stem cells. J Cell Biol.
mechanism by which transplanted MSCs 1986;103(1):49–62.
improve corneal epithelial cell phenotype. To 5. Dua HS, Saini JS, Azuara-Blanco A, Gupta
P.  Limbal stem cell deficiency: concept, aetiology,
accomplish this, it is first essential to agree on clinical presentation, diagnosis and management.
LSCD diagnostic criteria and clinical evaluation Indian J Ophthalmol. 2000;48(2):83–92.
endpoints so that the data collected from differ- 6. Nakamura T, Inatomi T, Sotozono C, Koizumi N,
ent clinical centers applying cell-based therapies Kinoshita S.  Ocular surface reconstruction using
stem cell and tissue engineering. Prog Retin Eye
for LSCD treatment can be effectively Res. 2016;51:187–207.
compared. 7. Rama P, Matuska S, Paganoni G, Spinelli A, De
Luca M, Pellegrini G.  Limbal stem-cell therapy
Acknowledgment  Financial Support: Carlos III National and long-term corneal regeneration. N Engl J Med.
Institute of Health, Spain (CIBER-BBN, CB06/01/003 2010;363(2):147–55.
MINECO/FEDER; Spanish Network on Cell Therapy, 8. Ramírez BE, Sánchez A, Herreras JM, Fernández
TerCel RD12/0019/0036); Ministry of Economy and I, García-Sancho J, Nieto-Miguel T, et al. Stem cell
Competitiveness and European Regional Development therapy for corneal epithelium regeneration follow-
Fund, Spain (SAF2015-63594-R MINECO/FEDER, EU); ing good manufacturing and clinical procedures.
Regional Center for Regenerative Medicine and Cell Biomed Res Int. 2015;2015:408495.
Therapy, Castilla y León, Spain. 9. Singh V, Shukla S, Ramachandran C, Mishra DK,
Katikireddy KR, Lal I, et al. Science and art of cell-­
based ocular surface regeneration. Int Rev Cell Mol
Compliance with Ethical Requirements  T Nieto- Biol. 2015;319:45–106.
Miguel, S Galindo, M López-Paniagua, I Pérez, JM 10. Zhang L, Coulson-Thomas VJ, Ferreira TG, Kao
Herreras and M Calonge declare that they have no WWY.  Mesenchymal stem cells for treating ocular
surface diseases. BMC Ophthalmol. 2015;15(Suppl
conflict of interest. 1(155)):55–65.
All procedures followed were in accordance 11. Yao L, Bai H. Review: mesenchymal stem cells and
with the ethical standards of the responsible com- corneal reconstruction. Mol Vis. 2013;19:2237–43.
mittee on human experimentation (institutional 12. Rohban R, Pieber TR.  Mesenchymal stem and
progenitor cells in regeneration: tissue specific-
and national) and with the Helsinki Declaration ity and regenerative potential. Stem Cells Int.
of 1975, as revised in 2000. Informed consent 2017;2017:1–16.
was obtained from all patients for being included 13. Dominici M, Le Blanc K, Mueller I, Slaper-­
in the study. Cortenbach I, Marini FC, Krause DS, et  al.
Minimal criteria for defining multipotent mesen-
All institutional and national guidelines for chymal stromal cells. The International Society for
the care and use of laboratory animals were Cellular Therapy position statement. Cytotherapy.
followed. 2006;8(4):315–7.
14. Fu Y, Karbaat L, Wu L, Leijten J, Both SK,
Karperien M. Trophic effects of mesenchymal stem
cells in tissue regeneration. Tissue Eng Part B Rev.
2017;23(6):515–28.
15. Si YL, Zhao YL, Hao HJ, Fu XB, Han WD. MSCs:
References biological characteristics, clinical applications
and their outstanding concerns. Ageing Res Rev.
1. Cotsarelis G, Cheng SZ, Dong G, Sun TT, Lavker 2011;10:93–103.
RM.  Existence of slow-cycling limbal epithelial 16. Higuchi A, Suresh Kumar S, Benelli G, Alarfaj AA,
basal cells that can be preferentially stimulated to Munusamy MA, Umezawa A, et al. Stem cell thera-
proliferate: implications on epithelial stem cells. pies for reversing vision loss. Trends Biotechnol.
Cell. 1989;57(2):201–9. 2017;35(11):1102–17.
2. Li W, Hayashida Y, Chen YT, Tseng SCG.  Niche 17. Holan V, Javorkova E.  Mesenchymal stem cells,
regulation of corneal epithelial stem cells at the lim- nanofiber scaffolds and ocular surface reconstruc-
bus. Cell Res. 2007;17:26–36. tion. Stem Cell Rev Rep. 2013;9(5):609–19.
3. Schlötzer-Schrehardt U, Kruse FE.  Identification 18. Galindo S, Herreras JM, López-Paniagua M, Rey E,
and characterization of limbal stem cells. Exp Eye de la Mata A, Plata-Cordero M, et  al. Therapeutic
Res. 2005;81:247–64. effect of human adipose tissue-derived mesen-
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 257

chymal stem cells in experimental corneal failure mesenchymal stem cells in experimental limbal
due to limbal stem cell niche damage. Stem Cells. stem cell deficiency in rabbits. Acta Ophthalmol.
2017;35(10):2160–74. 2011;89(8):741–8.
19. Mittal SK, Omoto M, Amouzegar A, Sahu A, 33. Jiang T-S, Cai L, Ji W-Y, Hui Y-N, Wang Y-S, Hu D,
Rezazadeh A, Katikireddy KR, et al. Restoration of et al. Reconstruction of the corneal epithelium with
corneal transparency by mesenchymal stem cells. induced marrow mesenchymal stem cells in rats.
Stem Cell Rep. 2016;7(4):583–90. Mol Vis. 2010;16:1304–16.
20. Harkin DG, Foyn L, Bray LJ, Sutherland AJ, Li 34. Zhang J, Huang C, Feng Y, Li Y, Wang
FJ, Cronin BG.  Concise reviews: can mesenchy- W.  Comparison of beneficial factors for corneal
mal stromal cells differentiate into corneal cells? wound-healing of rat mesenchymal stem cells and
a systematic review of published data. Stem Cells. corneal limbal stem cells on the xenogeneic acellular
2015;33:785–91. corneal matrix in vitro. Mol Vis. 2012;18:161–73.
21. Li F, Zhao S. Control of cross talk between angio- 35. Trosan P, Javorkova E, Zajicova A, Hajkova M,
genesis and inflammation by mesenchymal stem Hermankova B, Kossl J, et  al. The supportive role
cells for the treatment of ocular surface diseases. of insulin-like growth factor-I in the differentiation
Stem Cells Int. 2016;2016:1–8. of murine mesenchymal stem cells into corneal-like
22. Graw J.  Eye development. Curr Top Dev Biol. cells. Stem Cells Dev. 2016;25(11):774–81.
2010;90(C):343–86. 36. Ma Y, Xu Y, Xiao Z, Yang W, Zhang C, Song E, et al.
23. Phinney DG, Prockop DJ.  Concise review: mes- Reconstruction of chemically burned rat corneal sur-
enchymal stem/multipotent stromal cells: the state face by bone marrow-derived human mesenchymal
of transdifferentiation and modes of tissue repair-­ stem cells. Stem Cells. 2006;24(2):315–21.
current views. Stem Cells. 2007;25(11):2896–902. 37. Ye J, Yao K, Kim JC. Mesenchymal stem cell trans-
24. Baer PC.  Adipose-derived stem cells and their plantation in a rabbit corneal alkali burn model:
potential to differentiate into the epithelial lineage. engraftment and involvement in wound healing. Eye.
Stem Cells Dev. 2011;20(10):1805–16. 2006;20(4):482–90.
25. Barui A, Chowdhury F, Pandit A, Datta P. Rerouting 38. Ho JH-C, Ma W-H, Tseng T-C, Chen Y-F, Chen
mesenchymal stem cell trajectory towards epithelial M-H, Lee OK-S.  Isolation and characterization of
lineage by engineering cellular niche. Biomaterials. multi-potent stem cells from human orbital fat tis-
2018;156:28–44. sues. Tissue Eng Part A. 2011;17(1–2):255–66.
26. Guo T, Wang W, Zhang J, Chen X, Li B, Li 39. Martínez-Conesa EM, Espel E, Reina M, Casaroli-­
L. Experimental study on repairing damage of cor- Marano RP. Characterization of ocular surface epi-
neal surface by mesenchymal stem cells transplanta- thelial and progenitor cell markers in human adipose
tion. Zhonghua Yan Ke Za Zhi. 2006;42(3):246–50. stromal cells derived from lipoaspirates. Invest
27. Hou GH, Ye N, Wu J, Xu JT, Shi WJ, Chen Y, Ophthalmol Vis Sci. 2012;53(1):513–20.
et  al. Preliminary study on human bone mar- 40. Nieto-Miguel T, Galindo S, Reinoso R, Corell A,
row mesenchymal stem cells differentiation into Martino M, Pérez-Simón JA, et  al. In vitro simu-
epithelial-like cells. Zhonghua Yan Ke Za Zhi. lation of corneal epithelium microenvironment
2010;46(8):719–24. induces a corneal epithelial-like cell phenotype from
28. Rohaina CM, Then KY, Ng AMH, Wan Abdul Halim human adipose tissue mesenchymal stem cells. Curr
WH, Zahidin AZM, Saim A, et  al. Reconstruction Eye Res. 2013;38(9):933–44.
of limbal stem cell deficient corneal surface 41. Liu W, Liu Y, Liu H, Luo Y, Xu J. Differentiation of
with induced human bone marrow mesenchymal adipose-derived mesenchymal stem cells after trans-
stem cells on amniotic membrane. Transl Res. fection with Pax6 gene. Zhongguo Xiu Fu Chong
2014;163(3):200–10. Jian Wai Ke Za Zhi. 2014;28(8):1004–8.
29. Katikireddy KR, Dana R, Jurkunas 42. Garzon I, Miyake J, Gonzalez-Andrades M,
UV.  Differentiation potential of limbal fibroblasts Carmona R, Carda C, Sanchez-Quevedo M, Del C,
and bone marrow mesenchymal stem cells to corneal et al. Wharton’s jelly stem cells: a novel cell source
epithelial cells. Stem Cells. 2014;32(3):717–29. for oral mucosa and skin epithelia regeneration.
30. Sanchez-Abarca LI, Hernandez-Galilea E, Lorenzo Stem Cells Transl Med. 2013;2(8):625–32.
R, Herrero C, Velasco A, Carrancio S, et al. Human 43. Garzón I, Martín-Piedra MA, Alfonso-Rodríguez
bone marrow stromal cells differentiate into corneal C, Gonźalez-Andrades M, Carriel V, Martínez-­
tissue and prevent ocular graft-versus-host disease in Gómez C, et al. Generation of a biomimetic human
mice. Cell Transplant. 2015;24(12):2423–33. artificial cornea model using wharton’s jelly mes-
31. Gu S, Xing C, Han J, Tso MOM, Hong enchymal stem cells. Invest Ophthalmol Vis Sci.
J. Differentiation of rabbit bone marrow mesenchy- 2014;55(7):4073–83.
mal stem cells into corneal epithelial cells in  vivo 44. Sidney LE, McIntosh OD, Hopkinson
and ex vivo. Mol Vis. 2009;15:99–107. A.  Phenotypic change and induction of cyto-
32. Reinshagen H, Auw-Haedrich C, Sorg RV, keratin expression d­ uring in vitro culture of cor-
Boehringer D, Eberwein P, Schwartzkopff J, neal stromal cells. Invest Ophthalmol Vis Sci.
et  al. Corneal surface reconstruction using adult 2015;56(12):7225–35.
258 T. Nieto-Miguel et al.

45. Monteiro BG, Serafim RC, Melo GB, Silva MCP, cross-induction by other EGF family members. Mol
Lizier NF, Maranduba CMC, et  al. Human imma- Vis. 2007;13:2119–28.
ture dental pulp stem cells share key character- 60. Holan V, Trosan P, Cejka C, Javorkova E, Zajicova
istic features with limbal stem cells. Cell Prolif. A, Hermankova B, et  al. A comparative study of
2009;42(5):587–94. the therapeutic potential of mesenchymal stem
46. Gomes JÁP, Monteiro BG, Melo GB, Smith RL, cells and limbal epithelial stem cells for ocular
da Silva MCP, Lizier NF, et al. Corneal reconstruc- surface reconstruction. Stem Cells Transl Med.
tion with tissue-engineered cell sheets composed 2015;4(9):1052–63.
of human immature dental pulp stem cells. Invest 61. Cejka C, Holan V, Trosan P, Zajicova A, Javorkova
Ophthalmol Vis Sci. 2010;51(3):1408–14. E, Cejkova J. The favorable effect of mesenchymal
47. Kushnerev E, Shawcross SG, Sothirachagan S, stem cell treatment on the antioxidant protective
Carley F, Brahma A, Yates JM, et  al. Regeneration mechanism in the corneal epithelium and renewal of
of corneal epithelium with dental pulp stem cells corneal optical properties changed after alkali burns.
using a contact lens delivery system. Investig Oxidative Med Cell Longev. 2016;2016:5843809.
Opthalmology Vis Sci. 2016;57(13):5192–9. 62. Hu N, Zhang Y-Y, Gu H-W, Guan H-J.  Effects
48. Tsai C-L, Chuang P-C, Kuo H-K, Chen Y-H, Su W-H, of bone marrow mesenchymal stem cells on cell
et al. Differentiation of stem cells from human exfo- proliferation and growth factor expression of
liated deciduous teeth toward a phenotype of corneal limbal epithelial cells in  vitro. Ophthalmic Res.
epithelium in vitro. Cornea. 2015;34:1471–7. 2012;48(2):82–8.
49. Haynesworth SE, Baber MA, Caplan AI. Cytokine 63. Oh JY, Kim MK, Shin MS, Wee WR, Lee
expression by human marrow-derived mes- JH.  Cytokine secretion by human mesenchymal
enchymal progenitor cells in  vitro: effects of stem cells cocultured with damaged corneal epithe-
dexamethasone and IL-1 alpha. J Cell Physiol. lial cells. Cytokine. 2009;46(1):100–3.
1996;166(3):585–92. 64. Taheri F, Bazan HEP.  Platelet-activating factor
50. Ghazaryan E, Zhang Y, He Y, Liu X, Li Y, Xie J, et al. overturns the transcriptional repressor disposi-
Mesenchymal stem cells in corneal neovasculariza- tion of Sp1  in the expression of MMP-9  in human
tion: comparison of different application routes. Mol corneal epithelial cells. Invest Ophthalmol Vis Sci.
Med Rep. 2016;14(4):3104–12. 2007;48(5):1931–41.
51. Yao L, Li ZR, Su WR, Li YP, Lin ML, Zhang WX, 65. Di G, Du X, Qi X, Zhao X, Duan H, Li S, et  al.
et  al. Role of mesenchymal stem cells on cornea Mesenchymal stem cells promote diabetic corneal
wound healing induced by acute alkali burn. PLoS epithelial wound healing through TSG-6-dependent
One. 2012;7(2):e30842. stem cell activation and macrophage switch. Investig
52. Caplan AI, Dennis JE.  Mesenchymal stem cells as Opthalmol Vis Sci. 2017;58(10):4344–54.
­trophic mediators. J Cell Biochem. 2006;98:1076–84. 66. Oh JY, Kim MK, Shin MS, Lee HJ, Ko JH, Wee WR,
53. Kocan B, Maziarz A, Tabarkiewicz J, Ochiya T, et  al. The anti-inflammatory and anti-angiogenic
Banaś-Ząbczyk A.  Trophic activity and phenotype role of mesenchymal stem cells in corneal wound
of adipose tissue-derived mesenchymal stem cells as healing following chemical injury. Stem Cells.
a background of their regenerative potential. Stem 2008;26(4):1047–55.
Cells Int. 2017;2017:1653254. 67. Sotozono C, He J, Matsumoto Y, Kita M, Imanishi
54. Caplan AI.  Why are MSCs therapeutic? New data: J, Kinoshita S.  Cytokine expression in the alkali-­
new insight. J Pathol. 2009;217:318–24. burned cornea. Curr Eye Res. 1997;16(7):670–6.
55. da Silva ML, Fontes AM, Covas DT, Caplan 68. Djouad F, Charbonnier L-M, Bouffi C, Louis-Plence
AI. Mechanisms involved in the therapeutic proper- P, Bony C, Apparailly F, et  al. Mesenchymal stem
ties of mesenchymal stem cells. Cytokine Growth cells inhibit the differentiation of dendritic cells
Factor Rev. 2009;20:419–27. through an interleukin-6-dependent mechanism.
56. Singer AJ, Clark RAF.  Cutaneous wound Stem Cells. 2007;25(8):2025–32.
healing. Epstein FH, editor. N Engl J Med. 69. Roddy GW, Oh JY, Lee RH, Bartosh TJ, Ylostalo
1999;341(10):738–46. J, Coble K, et  al. Action at a distance: systemi-
57. Mohan RR, Wilson SE. Ex vivo human corneal epi- cally administered adult stem/progenitor cells
thelial cells express membrane-bound precursor and (MSCs) reduce inflammatory damage to the cor-
mature soluble epidermal growth factor (EGF) and nea without engraftment and primarily by secretion
transforming growth factor (TGF) alpha proteins. of TNF-α stimulated gene/protein 6. Stem Cells.
Exp Eye Res. 1999;68(1):129–31. 2011;29(10):1572–9.
58. Yang H, Sun X, Wang Z, Ning G, Zhang F, Kong J, 70. DiPietro LA, Burdick M, Low QE, Kunkel SL,
et al. EGF stimulates growth by enhancing capaci- Strieter RM. Mip-1α as a critical macrophage che-
tative calcium entry in corneal epithelial cells. J moattractant in murine wound repair. J Clin Invest.
Membr Biol. 2003;194(1):47–58. 1998;101(8):1693–8.
59. Morita S, Shirakata Y, Shiraishi A, Kadota Y, 71. Choi JA, Choi J-S, Joo C-K.  Effects of amniotic
Hashimoto K, Higashiyama S, et  al. Human cor- membrane suspension in the rat alkali burn model.
neal epithelial cell proliferation by epiregulin and its Mol Vis. 2011;17(February):404–12.
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 259

72. Aggarwal S, Pittenger MF.  Human mesenchy- 86. Raposo G, Stoorvogel W.  Extracellular vesicles:
mal stem cells modulate allogeneic immune cell exosomes, microvesicles, and friends. J Cell Biol.
responses. Blood. 2005;105(4):1815–22. 2013;200:373–83.
73. Maggini J, Mirkin G, Bognanni I, Holmberg J, 87. Baglio SR, Rooijers K, Koppers-Lalic D, Verweij FJ,
Piazzón IM, Nepomnaschy I, et  al. Mouse bone Lanzón MP, Zini N, et al. Human bone marrow- and
marrow-derived mesenchymal stromal cells turn adipose-mesenchymal stem cells secrete exosomes
activated macrophages into a regulatory-like profile. enriched in distinctive miRNA and tRNA species.
PLoS One. 2010;5(2):e9252. Stem Cell Res Ther. 2015;6(1):127.
74. Choi H, Lee RH, Bazhanov N, Oh JY, Prockop 88. Katsuda T, Ochiya T.  Molecular signatures of
DJ.  Anti-inflammatory protein TSG-6 secreted by mesenchymal stem cell-derived extracellular
activated MSCs attenuates zymosan-induced mouse vesicle-­mediated tissue repair. Stem Cell Res Ther.
peritonitis by decreasing TLR2/NF-κB signaling in 2015;6:212.
resident macrophages. Blood. 2011;118(2):330–8. 89. Tang Q, Luo C, Lu B, Fu Q, Yin H, Qin Z, et  al.
75. Oh JY, Roddy GW, Choi H, Lee RH, Ylöstalo JH, Thermosensitive chitosan-based hydrogels releas-
Rosa RH, et  al. Anti-inflammatory protein TSG-6 ing stromal cell derived factor-1 alpha recruit MSC
reduces inflammatory damage to the cornea follow- for corneal epithelium regeneration. Acta Biomater.
ing chemical and mechanical injury. Proc Natl Acad 2017;61:101–13.
Sci U S A. 2010;107(39):16875–80. 90. Hu L, Wang J, Zhou X, Xiong Z, Zhao J, Yu R, et al.
76. Cursiefen C, Masli S, Ng TF, Dana MR, Bornstein Exosomes derived from human adipose mensenchy-
P, Lawler J, et  al. Roles of thrombospondin-1 and mal stem cells accelerates cutaneous wound healing
-2 in regulating corneal and iris angiogenesis. Invest via optimizing the characteristics of fibroblasts. Sci
Ophthalmol Vis Sci. 2004;45(4):1117–24. Rep. 2016;6:32993.
77. Cursiefen C.  Immune privilege and angiogenic 91. Di Nicola M, Carlo-Stella C, Magni M, Milanesi M,
privilege of the cornea. Chem Immunol Allergy. Longoni PD, Matteucci P, et al. Human bone marrow
2007;92:50–7. stromal cells suppress T-lymphocyte proliferation
78. Zak S, Treven J, Nash N, Gutierrez LS.  Lack of induced by cellular or nonspecific mitogenic stimuli.
thrombospondin-1 increases angiogenesis in a Blood. 2002;99(10):3838–43.
model of chronic inflammatory bowel disease. Int J 92. Krampera M, Glennie S, Dyson J, Scott D, Laylor
Color Dis. 2008;23(3):297–304. R, Simpson E, et  al. Bone marrow mesenchymal
79. Downes JE, Swann PG, Holmes RS. Ultraviolet light-­ stem cells inhibit the response of naive and memory
induced pathology in the eye: associated changes in antigen-­specific T cells to their cognate peptide.
ocular aldehyde dehydrogenase and alcohol dehy- Blood. 2003;101(9):3722–9.
drogenase activities. Cornea. 1993;12(3):241–8. 93. Glenn JD, Whartenby KA. Mesenchymal stem cells:
80. Cejkova J, Trosan P, Cejka C, Lencova A, Zajicova emerging mechanisms of immunomodulation and
A, Javorkova E, et al. Suppression of alkali-induced therapy. World J Stem Cells. 2014;6(5):526–39.
oxidative injury in the cornea by mesenchymal stem 94. Zhang QZ, Su WR, Shi SH, Wilder-Smith P, Xiang
cells growing on nanofiber scaffolds and transferred AP, Wong A, et  al. Human gingiva-derived mesen-
onto the damaged corneal surface. Exp Eye Res. chymal stem cells elicit polarization of M2 macro-
2013;116:312–23. phages and enhance cutaneous wound healing. Stem
81. Kemp K, Gray E, Mallam E, Scolding N, Wilkins Cells. 2010;28(10):1856–68.
A.  Inflammatory cytokine induced regulation of 95. Raffaghello L, Bianchi G, Bertolotto M, Montecucco
superoxide dismutase 3 expression by human mesen- F, Busca A, Dallegri F, et al. Human mesenchymal
chymal stem cells. Stem Cell Rev. 2010;6(4):548–59. stem cells inhibit neutrophil apoptosis: a model for
82. Li F, Zhao S-Z. Mesenchymal stem cells: potential neutrophil preservation in the bone marrow niche.
role in corneal wound repair and transplantation. Stem Cells. 2008;26(1):151–62.
World J Stem Cells. 2014;6(3):296–304. 96. Maqbool M, Vidyadaran S, George E, Ramasamy
83. Konala VBR, Mamidi MK, Bhonde R, Das AK, R.  Human mesenchymal stem cells protect neutro-
Pochampally R, Pal R.  The current landscape of phils from serum-deprived cell death. Cell Biol Int.
the mesenchymal stromal cell secretome: a new 2011;35(12):1247–51.
paradigm for cell-free regeneration. Cytotherapy. 97. Brown JM, Nemeth K, Kushnir-Sukhov NM,
2016;18:13–24. Metcalfe DD, Mezey E. Bone marrow stromal cells
84. Bermudez MA, Sendon-Lago J, Eiro N, Trevino M, inhibit mast cell function via a COX2-dependent
Gonzalez F, Yebra-Pimentel E, et al. Corneal epithe- mechanism. Clin Exp Allergy. 2011;41(4):526–34.
lial wound healing and bactericidal effect of condi- 98. Sotiropoulou PA, Perez SA, Gritzapis AD, Baxevanis
tioned medium from human uterine cervical stem CN, Papamichail M.  Interactions between human
cells. Invest Ophthalmol Vis Sci. 2015;56(2):983–92. mesenchymal stem cells and natural killer cells.
85. Bermudez MA, Sendon-Lago J, Seoane S, Eiro N, Stem Cells. 2006;24(1):74–85.
Gonzalez F, Saa J, et al. Anti-inflammatory effect of 99. Spaggiari GM, Capobianco A, Becchetti S, Mingari
conditioned medium from human uterine cervical MC, Moretta L.  Mesenchymal stem cell-natural
stem cells in uveitis. Exp Eye Res. 2016;149:84–92. killer cell interactions: evidence that activated NK
260 T. Nieto-Miguel et al.

cells are capable of killing MSCs, whereas MSCs 112. Kao WWY, Zhu G, Benza R, Kao CWC, Ishizaki
can inhibit IL-2-induced NK-cell proliferation. M, Wander AH.  Appearance of immune cells and
Blood. 2006;107(4):1484–90. expression of MHC II DQ molecule by fibroblasts in
100. Augello A, Tasso R, Negrini SM, Amateis A, Indiveri alkali-burned corneas. Cornea. 1996;15(4):397–408.
F, Cancedda R, et  al. Bone marrow mesenchymal 113. Li Z, Burns AR, Smith CW. Lymphocyte function-­
progenitor cells inhibit lymphocyte proliferation by associated antigen-1-dependent inhibition of corneal
activation of the programmed death 1 pathway. Eur J wound healing. Am J Pathol. 2006;169(5):1590–600.
Immunol. 2005;35(5):1482–90. 114. Yomogida S, Hua J, Sakamoto K, Nagaoka
101. Corcione A, Benvenuto F, Ferretti E, Giunti D, I. Glucosamine suppresses interleukin-8 production
Cappiello V, Cazzanti F, et  al. Human mesenchy- and ICAM-1 expression by TNF-alpha-stimulated
mal stem cells modulate B-cell functions. Blood. human colonic epithelial HT-29 cells. Int J Mol
2006;107(1):367–72. Med. 2008;22(2):205–11.
102. Tabera S, Pérez-Simón JA, Díez-Campelo M, 115. Kim JY, Kang JS, Kim HM, Ryu HS, Kim
Sánchez-Abarca LI, Blanco B, López A, et al. The HS, Lee HK, et  al. Inhibition of bone marrow-­
effect of mesenchymal stem cells on the viability, derived dendritic cell maturation by glabridin. Int
proliferation and differentiation of B-lymphocytes. Immunopharmacol. 2010;10(10):1185–93.
Haematologica. 2008;93(9):1301–9. 116. Lee YJ, Moon MK, Hwang SM, Yoon JJ, Lee SM,
103. Franquesa M, Hoogduijn MJ, Bestard O, Grinyó Seo KS, et al. Anti-Inflammatory effect of Buddleja
JM.  Immunomodulatory effect of mesenchymal officinalis on vascular inflammation in human
stem cells on B cells. Front Imunol. 2012;3:212. umbilical vein endothelial cells. Am J Chin Med.
104. Ren G, Zhang L, Zhao X, Xu G, Zhang Y, Roberts AI, 2010;38(3):585–98.
et al. Mesenchymal stem cell-mediated immunosup- 117. Zaher SS, Germain C, Fu H, Larkin DFP, George
pression occurs via concerted action of chemokines AJT. 3-hydroxykynurenine suppresses CD4+ T-cell
and nitric oxide. Cell Stem Cell. 2008;2(2):141–50. proliferation, induces T-regulatory-cell develop-
105. Qu X, Liu X, Cheng K, Yang R, Zhao ment, and prolongs corneal allograft survival. Invest
RCH.  Mesenchymal stem cells inhibit Th17 cell Ophthalmol Vis Sci. 2011;52(5):2640–8.
differentiation by IL-10 secretion. Exp Hematol. 118. Zajicova A, Pokorna K, Lencova A, Krulova M,
2012;40(9):761–70. Svobodova E, Kubinova S, et al. Treatment of ocu-
106. English K, Ryan JM, Tobin L, Murphy MJ, Barry FP, lar surface injuries by limbal and mesenchymal
Mahon BP. Cell contact, prostaglandin E2 and trans- stem cells growing on nanofiber scaffolds. Cell
forming growth factor beta 1 play non-redundant Transplant. 2010;19(10):1281–90.
roles in human mesenchymal stem cell induction 119. Ko JH, Lee HJ, Jeong HJ, Kim MK, Wee WR, Yoon
of CD4+CD25Highforkhead box P3+ regulatory T S, et al. Mesenchymal stem/stromal cells precondi-
cells. Clin Exp Immunol. 2009;156(1):149–60. tion lung monocytes/macrophages to produce toler-
107. Fiorina P, Jurewicz M, Augello A, Vergani A, Dada ance against allo- and autoimmunity in the eye. Proc
S, La Rosa S, et  al. Immunomodulatory function Natl Acad Sci. 2016;113(1):158–63.
of bone marrow-derived mesenchymal stem cells 120. Coulson-Thomas VJ, Gesteira TF, Hascall V, Kao
in experimental autoimmune type 1 diabetes. J W. Umbilical cord mesenchymal stem cells suppress
Immunol. 2009;183(2):993–1004. host rejection: the role of the glycocalyx. J Biol
108. Traggiai E, Volpi S, Schena F, Gattorno M, Ferlito Chem. 2014;289(34):23465–81.
F, Moretta L, et  al. Bone marrow-derived mesen- 121. Bollyky PL, Evanko SP, Wu RP, Potter-Perigo
chymal stem cells induce both polyclonal expansion S, Long SA, Kinsella B, et  al. Th1 cytokines pro-
and differentiation of B cells isolated from healthy mote T-cell binding to antigen-presenting cells via
donors and systemic lupus erythematosus patients. enhanced hyaluronan production and accumula-
Stem Cells. 2008;26(2):562–9. tion at the immune synapse. Cell Mol Immunol.
109. Romieu-Mourez R, François M, Boivin M-N, 2010;7(3):211–20.
Bouchentouf M, Spaner DE, Galipeau J.  Cytokine 122. Oh JY, Kim MK, Ko JH, Lee HJ, Lee JH, Wee
modulation of TLR expression and activation in WR.  Rat allogeneic mesenchymal stem cells did
mesenchymal stromal cells leads to a proinflamma- not prolong the survival of corneal xenograft in a
tory phenotype. J Immunol. 2009;182(12):7963–73. pig-to-rat model. Vet Ophthalmol. 2009;12(SUPPL.
110. Waterman RS, Tomchuck SL, Henkle SL, Betancourt 1):35–40.
AM.  A new mesenchymal stem cell (MSC) para- 123. Jia Z, Jiao C, Zhao S, Li X, Ren X, Zhang L, et al.
digm: polarization into a pro-inflammatory MSC1 Immunomodulatory effects of mesenchymal stem
or an immunosuppressive MSC2 phenotype. PLoS cells in a rat corneal allograft rejection model. Exp
One. 2010;5(4):e10088. Eye Res. 2012;102:44–9.
111. Wen L, Zhu M, Madigan MC, You J, King NJC, 124. Omoto M, Katikireddy KR, Rezazadeh A, Dohlman
Billson FA, et  al. Immunomodulatory effects of TH, Chauhan SK. Mesenchymal stem cells home to
bone marrow-derived mesenchymal stem cells on inflamed ocular surface and suppress allosensitiza-
pro-inflammatory cytokine-stimulated human cor- tion in corneal transplantation. Invest Ophthalmol
neal epithelial cells. PLoS One. 2014;9(7):e101841. Vis Sci. 2014;55(10):6631–8.
17  Cell Therapy Using Extraocular Mesenchymal Stem Cells 261

125. Treacy O, O’Flynn L, Ryan AE, Morcos M, Lohan cells in corneal injury. Invest Ophthalmol Vis Sci.
P, Schu S, et  al. Mesenchymal stem cell therapy 2012;53(7):3638–44.
promotes corneal allograft survival in rats by local 139. Ceradini DJ, Kulkarni AR, Callaghan MJ, Tepper
and systemic immunomodulation. Am J Transplant. OM, Bastidas N, Kleinman ME, et  al. Progenitor
2014;14(9):2023–36. cell trafficking is regulated by hypoxic gradi-
126. Fuentes-Julián S, Arnalich-Montiel F, Jaumandreu ents through HIF-1 induction of SDF-1. Nat Med.
L, Leal M, Casado A, García-Tuñon I, et al. Adipose-­ 2004;10(8):858–64.
derived mesenchymal stem cell administration does 140. Hong HS, Lee J, Lee E, Kwon YS, Lee E, Ahn W,
not improve corneal graft survival outcome. PLoS et al. A new role of substance P as an injury-­inducible
One. 2015;10(3):e0117945. messenger for mobilization of CD29 + stromal-like
127. Crop MJ, Korevaar SS, de Kuiper R, Ijzermans cells. Nat Med. 2009;15(4):425–35.
JNM, van Besouw NM, Baan CC, et  al. Human 141. Hannoush EJ, Sifri ZC, Elhassan IO, Mohr AM,
mesenchymal stem cells are susceptible to lysis Alzate WD, Offin M, et  al. Impact of enhanced
by CD8+ T cells and NK cells. Cell Transplant. mobilization of bone marrow derived cells to site of
2011;20(10):1547–59. injury. J Trauma. 2011;71(2):283–91.
128. Kang SK, Shin IS, Ko MS, Jo JY, Ra JC.  Journey 142. Gao J, Dennis JE, Muzic RF, Lundberg M, Caplan
of mesenchymal stem cells for homing: strategies AI.  The dynamic in  vivo distribution of bone
to enhance efficacy and safety of stem cell therapy. marrow-­derived mesenchymal stem cells after infu-
Stem Cells Int. 2012;2012:342968. sion. Cells Tissues Organs. 2001;169(1):12–20.
129. Devine SM, Cobbs C, Jennings M, Bartholomew 143. Zeppieri M, Salvetat ML, Beltrami AP, Cesselli
A, Hoffman R.  Mesenchymal stem cells dis- D, Bergamin N, Russo R, et  al. Human adipose-­
tribute to a wide range of tissues following sys- derived stem cells for the treatment of chemically
temic infusion into nonhuman primates. Blood. burned rat cornea: preliminary results. Curr Eye Res.
2003;101(8):2999–3001. 2013;38(4):451–63.
130. Rojas M, Xu J, Woods CR, Mora AL, Spears W, 144. Zeng W, Li Y, Zeng G, Yang B, Zhu
Roman J, et al. Bone marrow-derived mesenchymal Y. Transplantation with cultured stem cells derived
stem cells in repair of the injured lung. Am J Respir from the human amniotic membrane for corneal
Cell Mol Biol. 2005;33(2):145–52. alkali burns: an experimental study. Ann Clin Lab
131. Rüster B, Göttig S, Ludwig RJ, Bistrian R, Müller Sci. 2014;44(1):73–81.
S, Seifried E, et al. Mesenchymal stem cells display 145. Pınarlı FA, Okten G, Beden U, Fışgın T, Kefeli M,
coordinated rolling and adhesion behavior on endo- Kara N, et  al. Keratinocyte growth factor-2 and
thelial cells. Blood. 2006;108(12):3938–44. autologous serum potentiate the regenerative effect
132. Segers VFM, Van Riet I, Andries LJ, Lemmens K, of mesenchymal stem cells in cornea damage in rats.
Demolder MJ, De Becker AJML, et al. Mesenchymal Int J Ophthalmol. 2014;7(2):211–9.
stem cell adhesion to cardiac microvascular endo- 146. Sohni A, Verfaillie CM.  Mesenchymal stem cells
thelium: activators and mechanisms. Am J Physiol migration homing and tracking. Stem Cells Int.
Heart Circ Physiol. 2006;290:H1370–7. 2013;2013:130763.
133. da Silva ML, Caplan AI, Nardi NB. In search of the 147. Ye J, Lee SY, KooK KH, Yao K.  Bone marrow-­
in  vivo identity of mesenchymal stem cells. Stem derived progenitor cells promote corneal wound
Cells. 2008;26(9):2287–99. healing following alkali injury. Graefes Arch Clin
134. Ponte AL, Marais E, Gallay N, Langonné A, Delorme Exp Ophthalmol. 2008;246(2):217–22.
B, Hérault O, et al. The in vitro migration capacity of 148. Lee RH, Yu JM, Foskett AM, Peltier G, Reneau JC,
human bone marrow mesenchymal stem cells: com- Bazhanov N, et al. TSG-6 as a biomarker to predict
parison of chemokine and growth factor chemotactic efficacy of human mesenchymal stem/progenitor
activities. Stem Cells. 2007;25(7):1737–45. cells (hMSCs) in modulating sterile inflammation
135. Ries C, Egea V, Karow M, Kolb H, Jochum M, Neth in vivo. Proc Natl Acad Sci. 2014;111(47):16766–71.
P. MMP-2, MT1-MMP, and TIMP-2 are essential for 149. Yun YI, Park SY, Lee HJ, Ko JH, Kim MK, Wee
the invasive capacity of human mesenchymal stem WR, et  al. Comparison of the anti-inflammatory
cells: differential regulation by inflammatory cyto- effects of induced pluripotent stem cell–derived and
kines. Blood. 2007;109(9):4055–63. bone marrow–derived mesenchymal stromal cells
136. Karp JM, Leng Teo GS.  Mesenchymal stem cell in a murine model of corneal injury. Cytotherapy.
homing: the devil is in the details. Cell Stem Cell. 2017;19(1):28–35.
2009;4:206–16. 150. Ahmed SK, Soliman AA, Omar SMM, Mohammed
137. Baek SJ, Kang SK, Ra JC. In vitro migration capac- WR.  Bone marrow mesenchymal stem cell trans-
ity of human adipose tissue-derived mesenchymal plantation in a rabbit corneal alkali burn model (a
stem cells reflects their expression of receptors for histological and immune histo-chemical study). Int J
chemokines and growth factors. Exp Mol Med. Stem Cells. 2015;8(1):69–78.
2011;43(10):596–603. 151. Lee JY, Jeong HJ, Kim MK, Wee WR. Bone marrow-­
138. Lan Y, Kodati S, Lee HS, Omoto M, Jin Y, Chauhan derived mesenchymal stem cells affect immuno-
SK.  Kinetics and function of mesenchymal stem logic profiling of interleukin-17-secreting cells in a
262 T. Nieto-Miguel et al.

chemical burn mouse model. Korean J Ophthalmol. I, first in human, first in class trial. Oncotarget.
2014;28(1011–8942 (Print)):246–56. 2017;8(46):80156–66.
152. Jirsova K, Jones GLA. Amniotic membrane in oph- 164. Yin X, Li P, Li Y, Cai Y, Wen J, Luan Q.  Growth/
thalmology: properties, preparation, storage and differentiation factor-5 promotes in  vitro/vivo peri-
indications for grafting—a review. Cell Tissue Bank. odontal specific differentiation of induced pluripo-
2017;18:193–204. tent stem cell-derived mesenchymal stem cells. Exp
153. Krenzer KL, Freddo TF.  Cytokeratin expression Ther Med. 2017;14(5):4111–7.
in normal human bulbar conjunctiva obtained by 165. Prasadam I, Akuien A, Friis TE, Fang W, Mao
impression cytology. Invest Ophthalmol Vis Sci. X, Crawford RW, et  al. Mixed cell therapy of
1997;38(1):142–52. bone marrow-derived mesenchymal stem cells
154. Espandar L, Caldwell D, Watson R, Blanco-­ and articular cartilage chondrocytes amelio-
Mezquita T, Zhang S, Bunnell B.  Application of rates osteoarthritis development. Lab Investig.
adipose-derived stem cells on scleral contact lens 2018;98(1):106–16.
carrier in an animal model of severe acute alkaline 166. Kotton DN, Ma BY, Cardoso WV, Sanderson EA,
burn. Eye Contact Lens. 2014;40(4):243–7. Summer RS, Williams MC, et  al. Bone marrow-­
155. Cejka C, Cejkova J, Trosan P, Zajicova A, Sykova derived cells as progenitors of lung alveolar epithe-
E, Holan V. Transfer of mesenchymal stem cells and lium. Development. 2001;128:5181–8.
cyclosporine A on alkali-injured rabbit cornea using 167. Okamoto R, Yajima T, Yamazaki M, Kanai T, Mukai
nanofiber scaffolds strongly reduces corneal neovas- M, Ikeda Y, et al. Damaged epithelia regenerated by
cularization and scar formation. Histol Histopathol. bone marrow-derived cells in the human gastrointes-
2016;31(9):969–80. tinal tract. Nat Med. 2002;8(9):1011–7.
156. Ke Y, Wu Y, Cui X, Liu X, Yu M, Yang C, et  al. 168. Liu Z-J, Zhuge Y, Velazquez OC.  Trafficking and
Polysaccharide hydrogel combined with mesenchy- differentiation of mesenchymal stem cells. J Cell
mal stem cells promotes the healing of corneal alkali Biochem. 2009;106(6):984–91.
burn in rats. PLoS One. 2015;10(3):e0119725. 169. Cocho L, Fernández I, Calonge M, de la Maza
157. Lin HF, Lai YC, Tai CF, Tsai JL, Hsu HC, Hsu RF, MS, Rovira M, Stern ME, et  al. Prehematopoietic
et  al. Effects of cultured human adipose-derived stem cell transplantation tear cytokines as potential
stem cells transplantation on rabbit cornea regenera- susceptibility biomarkers for ocular chronic graft-­
tion after alkaline chemical burn. Kaohsiung J Med versus-­ host disease. Invest Ophthalmol Vis Sci.
Sci. 2013;29(1):14–8. 2017;58(11):4836–46.
158. Almaliotis D, Koliakos G, Papakonstantinou 170. Weng J, He C, Lai P, Luo C, Guo R, Wu S, et  al.
E, Komnenou A, Thomas A, Petrakis S, et  al. Mesenchymal stromal cells treatment attenuates dry
Mesenchymal stem cells improve healing of the eye in patients with chronic graft-versus-host dis-
cornea after alkali injury. Graefes Arch Clin Exp ease. Mol Ther. 2012;20(12):2347–54.
Ophthalmol. 2015;253(7):1121–35. 171. Agorogiannis GI, Alexaki VI, Castana O, Kymionis
159. Trounson A, McDonald C.  Stem cell therapies in GD.  Topical application of autologous adipose-­
clinical trials: progress and challenges. Cell Stem derived mesenchymal stem cells (MSCs) for persis-
Cell. 2015;17:11–22. tent sterile corneal epithelial defect. Graefes Arch
160. Jackson L, Jones DR, Scotting P, Sottile V.  Adult Clin Exp Ophthalmol. 2012;250(3):455–7.
mesenchymal stem cells: differentiation poten- 172. Calonge M, Herreras JM, Pérez I, Galindo S, Nieto-
tial and therapeutic applications. J Postgrad Med. Miguel T, López-Paniagua M, Alberca M, García-
2007;53(2):121–7. Sancho J, Sánchez A. A randomized controlled trial
161. Miao C, Lei M, Hu W, Han S, Wang Q.  A brief of cultivated limbal epithelial cells compared to
review: the therapeutic potential of bone marrow mesenchymal stem cells for the treatment of cor-
mesenchymal stem cells in myocardial infarction. neal failure due to limbal stem cell deficiency. Invest
Stem Cell Res Ther. 2017;8(1):242. Ophthalmol Vis Sci. 2017;58(8):3372. ARVO 2017
162. Dai L-J, Li HY, Guan L-X, Ritchie G, Zhou JX. The Annual Meeting.
therapeutic potential of bone marrow-derived mes- 173. Ramírez BE, Victoria DA, Murillo GM, Herreras
enchymal stem cells on hepatic cirrhosis. Stem Cell JM, Calonge M. In vivo confocal microscopy assess-
Res. 2009;2(1):16–25. ment of the corneoscleral limbal stem cell niche
163. von Einem JC, Peter S, Günther C, Volk H-D, Grütz before and after biopsy for cultivated limbal epi-
G, Salat C, et  al. Treatment of advanced gastroin- thelial transplantation to restore corneal epithelium.
testinal cancer with genetically modified autologous Histol Histopathol. 2015;30(2):183–92.
mesenchymal stem cells  - TREAT-ME-1  - a phase
Cell-based Therapy Using Induced
Plutipotent Stem Cell
18
Ricardo Pedro Casaroli-Marano

18.1 Introduction maintaining ocular surface integrity and renewal


of the corneal epithelium. In addition, the sclero-
The structural integrity and the maintenance of corneal limbus not only serves as a reservoir for
the physiology of the cornea are key elements for LSCs but also functions as a structural barrier
its transparency. Corneal epithelium has an against uncontrolled and abnormal growth of
important role in corneal transparency, contribut- conjunctival epithelial cells replacing the corneal
ing not only to the absence of blood vessels but epithelium and inducing neovascularization
also to a process of permanent repair, which is [1–3].
essential to preserve its normal status [1–3]. The microenvironment of the sclerocorneal
Thus, homeostasis of the corneal epithelium limbus provides a specialized niche for ocular
maintains a state of structural integrity of the surface homeostasis, making the limbic ocular
ocular surface that is essential for vision. region of special interest in the cell renewal pro-
Continuous renewal of the corneal epithelium cess. Thus, aggressions that affect the integrity of
has been shown to be possible thanks to a cellular the sclerocorneal limbic region, leading to the
subpopulation located in the basal layers of the disappearance, reduction, or even functional
sclerocorneal limbic region. These cells – known deterioration of LSCs, may produce a pathologi-
as limbal stem cells (LSCs)  – have progenitor cally maintained condition of limbal stem cell
characteristics and can maintain a constant epi- deficiency (LSCD) with significant changes to
thelial renewal dynamic. Their key properties are the ocular surface. These changes include the
that they are oligopotent cells that present a high appearance of persistent corneal defects, epithe-
proliferative potential with a slow cell cycle, lial keratinization, and conjunctivalization phe-
while showing great capacity for self-renewal by nomena, with corneal neovascularization under a
asymmetric division [2, 3]. Due to these particu- persistent inflammatory process that reduces cor-
larities, they are attributed a crucial function in neal transparency and promotes vision loss [1–4].
In these situations, replacing the diseased cornea
by keratoplasty would result in graft failure
R. P. Casaroli-Marano (*)
Department of Surgery, School of Medicine and because the physiological mechanism of corneal
Hospital Clínic de Barcelona, University of regeneration by epithelial renewal is damaged by
Barcelona, Barcelona, Spain the functional absence of the LSCs [4].
Institute of Biomedical Research (IIB-Sant Pau) and Currently, cell-based therapy is the most
Barcelona Tissue Bank, Banc de Sang i Teixits, appropriate approach for the replacement and
Barcelona, Spain repair of the ocular surface in cases of LSCD
e-mail: rcasaroli@ub.edu

© Springer Nature Switzerland AG 2019 263


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_18
264 R. P. Casaroli-Marano

[5, 6]. The cultured limbal epithelial transplanta- stroma; its lowermost aspect, which is farthest
tion (CLET) or the cultured oral mucosal epithe- from the surface, is probably where the largest
lial transplantation (COMET) approaches population of LSCs is located [9, 11].
involve obtaining healthy autologous tissue by The niche provides a variety of specific factors
minimally invasive biopsy after ex  vivo cell that contribute to the survival and dynamics of
expansion. Both approaches present very encour- LSCs. On the other hand, they provide certain
aging clinical results for the recovery of the ocu- components of the extracellular matrix and con-
lar surface and may contribute to the restoration tribute to the growth factors and cytokines that
of vision in appropriately selected patients. In determine a special proliferative state and a pro-
unilateral disease, autologous treatment with cess of organized cell growth. The niche is also
CLET is preferred, although this source of pro- responsible for maintaining a state of stemness in
genitor cells is not always available. If both eyes LSCs, with controlled inhibition of cell differen-
present with severe disease, allogeneic LSCs or tiation observed together with asymmetric divi-
the COMET technique can be used [5–7]. sion [1, 12]. This asymmetric division allows one
The development of cell therapies with adult of the daughter cells to retain its stemness and
tissue progenitors, such as adult mesenchymal replenish the stem cell pool, while the other
stem cells (MSCs) or induced pluripotent stem daughter cell enters differentiation.
cells (IPSCs), represents a significant advance in It has been demonstrated that LSCs can dif-
the treatment of certain eye diseases [2, 8]. ferentiate into postmitotic cell populations
Whether associated with the emerging approaches through an oriented process of cell differentia-
of tissue bioengineering or not, these offer a tion. This obeys a continuous centripetal move-
physiological treatment option that is more ratio- ment from the deepest peripheral epithelial layers
nal and less invasive for the regeneration of the (basal layers of the sclerocorneal limbus) to the
ocular surface. In this chapter, future perspectives superficial and more central layers (superficial
regarding the use of IPSCs as a cell source in corneal epithelium). This ordered mechanism of
these new approaches in regenerative medicine cellular proliferation and differentiation ensures
for the ocular surface are outlined and discussed. constant renewal of the corneal epithelium and
maintenance of its integrity [10, 13–15].
However, a cellular subpopulation may be located
18.2 L
 imbal Stem Cells (LSCs): in the central cornea that has the capacity to gen-
Epithelial Renewal erate undifferentiated progenitor cells with
and Molecular Mechanisms regenerative epithelial capacity [10, 15, 16]. This,
in turn, indicates the possibility that accessory
One of the essential elements for maintaining an mechanisms exist for the renewal and mainte-
adequate renewal and regeneration process in the nance of the corneal epithelium.
corneal epithelium, apart from maintaining the Research has also demonstrated that the clo-
structural integrity of the sclerocorneal limbus, is nogenic capacity of epithelial cells in the
the balance of the microenvironment, or niche, sclerocorneal limbic region is organized under
where the LSCs are located. The palisades of a system of progenitor cells stratified in com-
Vogt – a series of radially oriented fibrovascular partments [10, 12, 15, 17]. In the initial
ridges in the region of the human sclerocorneal ­compartment, the undifferentiated cell subpop-
limbus – and, more recently, the limbal epithelial ulation presents characteristics of progenitor
crypts have been described as the natural niche cells with high self-renewing capacity; how-
for LSCs [9, 10]. Histologically, palisades of ever, this property decreases as cells migrate
Vogt correspond to projections of the corneal through the following compartments. In the
stroma into the basal epithelial layer. Interspersed final compartment, a cellular subpopulation
between them, limbal epithelial crypts are regions exists with terminal differentiation characteris-
where the epithelium projects deep into the tics associated with low or no self-­ renewal
18  Cell-based Therapy Using Induced Pluripotent Stem Cell 265

capacity. These cells are incorporated as differ- Taking the above into consideration, it is
entiated corneal epithelial cells on the s­ urface accepted that LSCs give rise to a subpopulation
of the central cornea. In this sense, and in con- of transient amplifying cells (TACs) and that
trast to epithelial cells of the central cornea, it these cells migrate and differentiate under an
is considered that epithelial cells of the sclero- ordered centripetal movement from the limbic
corneal limbic region can form holoclones with region toward the central corneal surface, where
greater clonogenic potential [12]. Likewise, they generate differentiated corneal epithelial
epithelial cells isolated from the basal layers in cells that eventually disappear. This coordinated
the limbal region exhibit a high in vitro prolif- mechanism is responsible for the renewal of the
erative potential during expansion or in corneal epithelium and its continued regenera-
response to corneal injury and show an undif- tion (Fig. 18.1).
ferentiated phenotype lacking expression of At present, the identification of specific mark-
corneal epithelial differentiation markers [17]. ers and the detailed study of cellular and molecu-

Conjunctiva (A) Limbus (B) Peripheral cornea (C) Central cornea (D)

LSC Corneal epithelial cells

LSCs TACs PMCs TDCs


Conjuntival
epithelial
and Proliferation zone Differentiation zone
goblet cells

C1 C2 C3

CK5, CK14 ABCG2, ABCG5 ABCG2, p63 Nestin, Involucrin CK3


Muc5AC p63, p75, Bmi-1 CK19, a9 Cx43, a2, a6 CK12
C/EPBδ, Notch-1 Vimentin E-cadherin
CK15, CK19, a9

Fig. 18.1  Differentiation of limbal stem cells (LSCs) C2). The proliferation zone (compartments C1 and C2) is
into corneal epithelial cells during the repair of corneal also characterized by a specialized environment (niche)
epithelium, according to Thoft hypothesis (XYZ hypoth- responsible for maintaining LSC oligopotency. Postmitotic
esis) [13]. LSCs differentiate in an ordered manner that is cells (PMCs) and terminally differentiated cells (TDCs)
carried out progressively in compartments and regions of are in the differentiation zone (compartment C3) and are
cellular “maturation.” The sclerocorneal limbic region (B) characterized by their terminal differentiation state that
corresponds to the LSC “reservoir” and works as a spe- lacks self-renewal capability. Repair of the corneal epithe-
cialized microenvironment (niche), providing a barrier lium is established by events in which LSCs proliferate,
against the growth of the conjunctival epithelial cells (A) migrate, and differentiate from the suprabasal layers of
and their blood vessels in peripheral (C) and central (D) the sclerocorneal limbus (epithelial crypts), resulting in
corneal tissue. LSCs are characterized as a cell subpopula- TACs (XY movement, B to C), and with centripetal
tion with stemness, high proliferative capacity, and self-­ migration forming TDCs as cell differentiation proceeds
renewal proprieties (compartment C1). Transient toward to the superficial layers of the central cornea
amplifying cells (TACs) appear to be another cell subpop- (movement YZ; to D). The main putative cell markers are
ulation presenting some degree of differentiation and lim- shown in the right part of the diagram
ited capacity for self-renewal (intermediate compartment
266 R. P. Casaroli-Marano

lar relationships in the niche are of great interest stroma mesenchymal cells, melanocytes, vascu-
to improving our understanding of the therapeutic lar cells, and nerve endings, as well as extracel-
potential of LSCs. The transcription factor p63 lular elements, such as extracellular matrix
(ΔNp63α), specifically its α-isoform, has been (ECM) proteins and intracellular signaling mol-
described as one of the most characteristic mark- ecules [10, 14, 15]. The precise and coordinated
ers for the identification of LSCs in humans [4, function of all cellular and extracellular compo-
18]. Based on the presence of p63 cells, it has also nents in the niche are essential for its correct
been possible to establish clinically relevant qual- physiology when regulating and modulating the
ity criteria for the success of CLET techniques [4, proliferation, migration, and differentiation of
18, 19]. It has been concluded that the clinical the LSCs. Thus, homeostasis of the sclerocorneal
success of transplantation is dependent on the limbic region should involve a complex interrela-
existing p63+ cell population in the cultures tionship between niche elements (i.e., the ECM
obtained by ex vivo expansion. Successful clini- and cellular components) and the LSCs.
cal results have been observed more frequently in Corneal stroma mesenchymal cells appear to
cases in which transplanted cell cultures con- exert a regulatory role on LSCs, contributing to
tained between 5% and 10% p63+ cells. Cases in the maintenance of their stemness [10, 15]. The
which the percentage of p63+ cells was slower niche is also rich in vascular and nerve endings
than 3% were associated with clinical failure [19]. that are essential for maintaining the survival and
Recently, our group has been able to verify that, trophic function of LSCs [15, 23]. Also, certain
in limbic biopsies obtained from donors older than major ECM proteins, such as laminin, fibronectin,
65 years, cell cultures can be generated with a high and collagens [24], are not only essential for the
percentage of p63+ cells, making them viable for ultrastructural conformation of the niche but also
allogeneic transplant [20]. However, because the for participating in the events related to the regu-
LSC population in the cornea is scarce, character- lation and control of LSC migration and differen-
ization and cellular behavior studies imply an tiation. Thus, the structural and molecular
added difficulty, both for in vitro methodologies as characteristics of the limbal niche make it a
well as for experimental models. Thus, there is still unique environment for LSC behavior and stem-
no consensus on the definition of specific markers ness. Likewise, the intracellular signaling mediat-
for this cell population. Despite this, different puta- ing molecules and the molecular pathways
tive markers have been described based on pro- involved in cell behavior are crucial for maintain-
teomic and immunolocalization studies. These ing proliferation and differentiation of LSCs [24].
have used a reduced panel of possible markers, The signals that trigger the activation, prolif-
with ΔNp63α, ATP-binding cassette subfamily G eration, and migration of LSCs into TAC com-
member 2 (ABCG2), member 5 (ABCG5), Ccaat- partments are not fully understood. It has been
(cytosine-­c ytosine-­a denosine-adenosine- suggested that certain growth factors and cyto-
thymidine)-enhancer-binding protein delta-isoform kines, as well as chemotactic molecules that
(C/EBPδ), polycomb complex protein B lymphoma include ECM degradation products, all contribute
Mo-MLV (Bmi-1), cytokeratin 15 (CK15), and to the mechanisms of LSC corneal epithelial
Notch homolog 1, translocation-associated renewal and homeostasis.
(Drosophila) (Notch-1) being the main candidates.
The exclusion of cellular expression of cytokeratin
13 (CK3) and cytokeratin 12 (CK12), considered 18.3 Induced Pluripotent Stem
markers of corneal epithelial differentiation, is also Cells (IPSCs): Mechanisms
associated [21, 22] (Fig. 18.1). for Reprogramming
The LSC niche (i.e., the palisades of Vogt and
the limbal epithelial crypts) consists of various Together with clustered regularly interspaced
elements. These include cellular elements from short palindromic repeats (CRISPR)-associated
different characteristics, such as antigen-­ systems (i.e., Cas) for genome editing  – also
presenting and immune system cells, corneal called “genome surgery” [25, 26] – the technol-
18  Cell-based Therapy Using Induced Pluripotent Stem Cell 267

ogy for inducing pluripotent stem cells has been (PBMCs), together with fibroblasts and dermal
one of the most significant advances in the con- keratinocytes, as previously mentioned, represent
text of advanced therapies in regenerative medi- the preferred cellular candidates for inducing
cine over recent decades. Through progressive pluripotency [29, 33–35].
development, it has been shown that transfection The methodology for generating IPSC lines is
and overexpression of a specific set of transcrip- based on the approaches for cellular reprogram-
tion factors  – Oct4/Klf4/Sox2/cMyc or Oct4/ ming (Fig.  18.2). Since the early work of
Sox2/Lin28/Nanog  – in differentiated somatic Yamanaka [27], which used a highly efficient
cells from adult tissue (e.g., skin keratinocytes or viral reprogramming method (the Moloney
fibroblasts) can reprogram cell fate from an murine leukemia retrovirus) that incorporated a
immature cell state to recover its capacity for plu- retrovirus into the cell genome, variations of this
ripotency: IPSCs, with high clonal and prolifera- method have been developed. However, this tech-
tive potential [27–29]. These cells can be nology presents restrictions for clinical applica-
differentiated into several other cell types of dif- tion in humans, partly due to the unknown
ferent embryonic origin, such as endo-, meso-, or problems associated with using IPSCs, such as
ectodermal lineages. This property has opened the long-term repercussion of incorporating viral
possibilities to research different cellular events, elements in host DNA.  Cellular transfection
including the mechanisms of differentiation, pro- methods with high efficiency (>0.1%) are note-
liferation, senescence, and toxicity, as well as worthy, such as integrating retrovirus and lentivi-
other processes related to cell behavior. Unlike rus in the cellular genome (integrative
embryonic stem cells, for which there are estab- reprogramming) or using the adenovirus and
lished ethical problems related to their collec- Sendai virus (non-integrative reprogramming)
tion and use, IPSCs can readily be collected [29, 36, 37]. Other approaches for non-­integrative
from differentiated cells in adult tissue to create cellular reprogramming have been developed that
an autologous source of progenitor cell popula- do not use viruses as cellular vectors, but these
tions. In turn, these can be used in both regen- have variable transfection efficiency. Methods
erative therapy and research into the mechanisms based on DNA plasmids (non-episomal) and epi-
of rare pathologies. Thus, cell modeling  – somes are non-integrative approaches that show
through patient-specific IPSCs – can be used for medium transfection efficiency (0.01–0.1%) [29,
certain rare disorders to help better understand 36]. In recent years, synthetic RNA or micro-­
their pathophysiology and develop treatments RNA (miRNA) has been developed with supe-
[30–32]. rior efficiency to the average non-integrative
IPSCs represent a promising resource for methodologies [29, 37]. Moreover, non-integra-
future clinical application in advanced cell-based tive reprogramming methods appear to be pre-
therapy. Currently, the methodologies and ferred for clinical application [38] because they
approaches related to IPSCs are being expanded limit the possibilities of internal mutagenesis,
to increase the efficiency of their production and cell cycle dysregulation, and tumor formation
to reduce their immunogenic and tumorigenic [29, 38].
potential. Cell therapy with patient-specific Direct cellular reprogramming in different
autologous IPSCs has been shown not only to states, whether pluripotent or somatic, has recently
mitigate potential immune reactions but also to been developed. These are among the most promis-
provide a large and continuous supply of cells for ing approaches in regenerative medicine, with
transplantation [32, 33]. For clinical application great potential for secure and therapeutic clinical
of IPSCs in humans, choosing the correct donor application. Direct reprogramming is characterized
cell type and the most appropriate methodology by a process in which mature and completely dif-
for reprogramming is critical. The donor somatic ferentiated somatic cells or immature cells that
cell must be easily obtained and reprogrammed retain some potential for differentiation are induced
and ideally be from the patient requiring treat- to other cell types without needing to pass through
ment. Thus, peripheral blood mononuclear cells a state of pluripotency [36, 39]. For this purpose,
268 R. P. Casaroli-Marano

OKSM
Delivery
OSLN
methods

Virus

Retrovirus

Lentivirus

Adenovirus

Sendai virus
IPSC
DNA Differentiated cell
Adult somatic cell
Plasmids

Episomes

RNA

Synthetic RNA
Direct reprogramming
miRNA
Protein

Peptides STF

Fig. 18.2  Strategies for cell reprogramming. The intro- ming approaches (adenovirus, Sendai virus, DNA, epi-
duction (transfection) and overexpression of certain tran- somes, RNA, and peptides). Recently, direct
scription factors in partial or terminally differentiated reprogramming strategies have been described that permit
adult (somatic) cells can induce pluripotency (i.e., IPSCs). the transfection of selective and specific transcription fac-
The first transcription factors shown to be capable of tors (STF) that allow somatic cells to acquire differentia-
inducing pluripotency were Oct4/Klf4/Sox2/cMyc tion capacity (partial induced pluripotency) to other cell
(OKSM) and Oct4/Sox2/Lin28/Nanog (OSLN). The types, without necessarily going through a specific state
methods used for cellular transfection are based on the use of pluripotency, thereby avoiding the risk of mutagenesis
of viral and non-viral vectors, giving rise to integrative and predisposition to teratogenesis
(retrovirus and lentivirus) and non-integrative reprogram-

cells can be reprogrammed by transient overex- be limiting factors. However, there is still only
pression of transcriptional factors during a rela- limited experience with generating LSCs by
tively short interval. Cells in this state are therefore transdifferentiation of readily accessible adult
called partial IPSCs and can be modulated to somatic cells. The first evidence of the potential
respond to different signaling environments (e.g., for transdifferentiation was achieved by convert-
growth factors, cytokines, or inducing agents). In ing progenitor cells from the bulb of the hair fol-
appropriate environments, these cells have the licle into corneal epithelial cells. These cells
capacity to direct cell fate in reprogramming [39] acquired several markers of LSCs, including
(Fig.  18.2). For corneal repair, direct reprogram- CK15, p63, and the low-affinity nerve growth
ming would be of great advantage, not only elimi- factor receptor (p75; neurotrophin receptor),
nating the need for the potentially risky pluripotency when the niche of the sclerocorneal limbic region
stages but also reducing time and resource require- was mimicked; for this, the presence of corneal
ments when producing a specific IPSC line. stromal fibroblasts and laminin-5 were crucial
Another advantage of direct reprogramming is [40]. The limbal stem-like cells that were
that very few cells would be needed for cell ther- obtained using this method were successfully
apy of the ocular surface, meaning that difficulty used to repair the damaged cornea of an experi-
obtaining and producing an IPSC line would not mental model in which autologous stem cells
18  Cell-based Therapy Using Induced Pluripotent Stem Cell 269

were used [41]. In other research murine adult tive capacity and their specificity (autologous
stem cells from a hair follicle bulb were transdif- cells) can provide sufficient quantities of cells,
ferentiated into corneal epithelial-like pheno- while theoretically minimizing the risks of
types by forcing the overexpression of Pax6 and immune reactions and graft versus host disease
were maintained in supplemented culture associated with allogeneic transplant. This could
medium with soluble factors obtained from the also mitigate or obviate the undesirable adverse
sclerocorneal limbic region [42]. Also, corneal effects resulting from immunosuppression of the
epithelial cells could be generated from progeni- receptor. For example, in neurodegenerative and
tor skin cells and cultured in the presence of heart diseases, which have limited treatment
growth factors [43]. Corneal epithelial-like phe- options because neuronal and cardiac tissues are
notypes can also be obtained using MSCs from nonrenewable, IPSC-based strategies could be
adipose tissue when cultivated with an LSC cul- encouraging [29, 33].
ture medium supplemented with inducing agents Clinically, the first trials of cell-based therapy
for cellular differentiation [44]. These are encour- with autologous and allogenic IPSCs in humans
aging results that may open the door to novel have recently been carried out for the treatment
sources of autologous LSC supply. of age-related macular degeneration [46, 47]. In
Interestingly, the results obtained from a these clinical trials, retinal pigment epithelium
recently developed direct transdifferentiation (RPE) cells derived from autologous IPSCs gen-
approach have been described for LSCs and cor- erated from dermal fibroblasts were first gener-
neal epithelium and was compared with a deriva- ated from the patient to avoid a potential
tion method from IPSCs [45]. Human fibroblasts immunological rejection. Once implanted there
were directly transdifferentiated into LSCs and were no side effects, but the clinical trial was sus-
corneal epithelial cells by overexpressing three pended before treating the second patient,
limbal-specific transcription factors in lentivirus because mutations were detected in the IPSCs.
transfection and maintaining them in a corneal-­ This finding, coupled with the fact that the gen-
specific culture medium. They used ΔNp63α, C/ eration of patient-specific IPSCs is long and
EBPδ, and transcriptional factor 4 (TCF4 or expensive, has led to the use of allogeneic IPSCs;
immunoglobulin transcription factor 2; ITF-2) as these have not only been shown to be safe but
limbal-specific transcription factors. The resulting also to be HLA compatible with the haplotype of
differentiated cells exhibited morphological and the recipient. For HLA-A, HLA-B, and HLA-­
marker changes consistent with corneal epithelium, DRB1 antigens, it has recently been demon-
demonstrating the potential for direct reprogram- strated that human lymphocytes recognize RPE
ming from human dermal fibroblasts, bypassing cells derived from IPSCs that directly express
the intermediate state of pluripotency. Although HLA class I/II antigens, though this has not been
cells derived from the IPSCs also showed specific observed with homozygous RPE donor cells with
characteristics and markers, the expression of plu- identical haplotypes [48, 49]. This approach is
ripotency markers (e.g., Oct4 and Klf4) could not highly innovative, translational, and cost-­
be eliminated. This remains an advantage of direct effective because it can generate a bank of IPSCs
programming over using undifferentiated states of from cells with homozygous haplotypes, thereby
pluripotency (IPSC state). covering a greater number of candidate recipients
for cell-based therapies.
As discussed, LSCs make it possible to repair
18.4 Translational Application and constantly regenerate the corneal epithelium.
on the Ocular Surface These cells have the capacity for continuous self-­
renewal and to produce TACs that differentiate
In the context of clinical application for the ocu- terminally after a brief period of proliferation [1,
lar surface, therapeutic approaches based on 2, 10, 14, 15, 50]. However, there are limitations
IPSCs are promising. Their unlimited prolifera- to its therapeutic use in LSCD.  When a patient
270 R. P. Casaroli-Marano

has unilateral LSCD, even obtaining a biopsy tion of miRNA profiles as molecular switches for
through a minimally invasive procedure can ocular epithelial differentiation. Interestingly, an
cause damage to healthy tissue in either the con- increase in miR-450b-5p was shown to inhibit
tralateral or affected eye. In cases with bilateral PAX6 expression, and consequently the in  vitro
LSCD, however, obtaining allogeneic cell popu- development of corneal epithelium, while a
lations can increase the risk of immune rejection detectable decrease in miR-­184 during early ocu-
in the recipient. Thus, IPSCs represent a real lar development was shown to inhibit PAX6 and
alternative, providing a permanent cellular CK3 expressions [53]. These findings allow us to
reserve that could be obtained directly from the conclude that miRNA – a group of small noncod-
patient, thereby avoiding problems of immune ing RNA molecules  – also participate in deter-
rejection and cell availability. mining the fate of IPSCs. Overexpression or
Initial studies have shown that it is possible to depletion of individual miRNAs was associated
generate corneal epithelial cell lineages from with human disease, suggesting that dysregula-
IPSCs by blocking the TGF-β/Nodal and Wnt/β-­- tion of the miRNA system and alterations in ele-
catenin signaling pathways to induce the early ments of their biogenesis could also be involved
developmental events of the ocular ectoderm [51]. in LSC dysfunction [54, 55].
To this end, Mikhailova and colleagues [52] stud- Recently, it has become possible to produce a
ied a differentiation protocol without either a pure and maintained population of LSCs derived
feeder-layer system for cell growth or serum sup- from IPSCs, with phenotypes and molecular
plements in cell culture medium. The two relevant characteristics that are comparable to those of
cellular signaling pathways were blocked with somatic LSCs. Moreover, these preserve terminal
small molecule inhibitors (SB431542 and GSK3β, differentiation capacity under cell culture conflu-
respectively), and intracellular signaling related to ence conditions [56]. Four lines of human IPSCs
the fibroblast growth factor (FGF) pathways was were used – AnaW04, DFC, B5CRE, and 29.3 –
stimulated. After a long period of induction to dif- the first obtained from hair follicle keratinocytes
ferentiation, they generated relatively pure popu- and the last three derived from human dermal
lations of corneal epithelial progenitor cells fibroblasts. To initiate limbal compromise, the
capable of expressing the specific markers IPSC lines were cultured in an environment simi-
ABCG2, CK15, p63, and PAX6. The terminal dif- lar to the human sclerocorneal limbal niche by
ferentiation toward mature corneal epithelial cells means of a feeder-layer system with irradiated
(CK3 and CK12) was also achieved. fibroblasts isolated from the peripheral cornea in
It has been shown that a reproducible protocol culture plates coated with collagen IV. The cul-
can restate the canonic pathway of corneal epithe- tures were maintained in modified SHEM culture
lial embryogenesis and reproduce it by inducing medium that is usually used for the ex  vivo
pluripotency in the keratinocytes of the hair folli- expansion of human LSCs [57]. The study dem-
cles and dermal fibroblasts. This can be achieved onstrated that PAX6, CK14, and E-cadherin were
by infection with a lentiviral vector containing a expressed after differentiation [56]. Although the
polycistronic cassette of Oct4/Sox2/c-­Myc/Klf4, research aim had been to develop differentiated
with a conditioned culture medium derived from cell lines for in vitro cytotoxicity tests for use in
corneal fibroblasts to induce differentiation, and preclinical validations, they clearly showed the
supplemented with bone morphogenic protein 4 potential for IPSCs to differentiate into pheno-
(BMP4) in the presence of collagen IV substrate types similar to LSCs, allowing their clinical use
[53]. The organotypic induction assays, with rela- to be considered.
tively pure cell populations that exhibit corneal Recent efforts have made significant advances
epithelial phenotypes, suggest an ability to strat- in the development of self-organized three-­
ify into epithelium that is positive for the specific dimensional organoid approaches from IPSCs,
corneal markers p63, CK14, and CK12. and these will surely have translational implica-
Additionally, this model allowed the identifica- tions in the future [58]. These studies may affect
18  Cell-based Therapy Using Induced Pluripotent Stem Cell 271

toxicology and drug testing, human ocular dis- Finally, the possibility of differentiating IPSCs
ease modeling, and replacement treatments based in other corneal cell lines, such as stromal kerato-
on cell therapies [59–61]. cytes or corneal endothelium, has also been veri-
Hayashi and colleagues [62] described an fied. Due to its mesenchymal origin, the
in  vitro model of ocular morphogenesis using differentiation protocols must pass through an
human IPSCs maintained in differentiation media, intermediate cellular subpopulation, neural crest
which spontaneously and progressively formed cells (NCCs). These NCCs are embryonic migra-
an ocular primordium comprising four morpho- tory cells with the ability to differentiate into a
logically identifiable concentric zones  – self- wide variety of cell types in different tissues,
formed ectodermal autonomous multi-zone including the cornea, the peripheral nervous sys-
(SEAM). In this interesting research, the concen- tem, and the cutaneous melanocytes [65]. Using
tric SEAMs mimicked the development of a com- cell pellets compacted by centrifugation, NCCs
plete primitive eye because the cell locations in derived from IPSCs and cultured in a 3D matrix
different zones corresponded to lineages of the were shown to form satellite cells with pheno-
ocular surface ectoderm: the lens, the neurosen- types similar to keratocytes of the corneal stroma,
sory retina, and the RPE. Cellular characterization which expresses keratocan. However, a more
studies showed that the innermost SEAMs were detailed analysis by real-time qPCR suggested
positive for class III β-tubulin (TUBB3), a marker that they were different to in vivo stromal kerato-
for neuronal lineage, and for PAX6. Surprisingly, cytes [66]. Interestingly, when NCCs were intro-
the outermost SEAMs were positive for the epi- duced in the appropriate niche in  vitro
thelial/superficial ectodermal markers p63 and environment (human corneal tissue), they actively
E-cadherin. Cells isolated from the ectodermal migrated into the stromal collagen fibrils,
zone corresponding to the ocular surface could be acquired a fibroblastic-like phenotype, and pre-
expanded ex vivo to create corneal epithelium that sented an identical transcriptional behavior to
specifically expressed CK12 [63]. These epithe- in vivo corneal keratocytes [66]. Fukuta and col-
lial sheets, once implanted in an experimental leagues [67] also described a simple and robust
LSCD model, were able to recover the function of derivation method based on using defined cell
a damaged ocular surface. culture medium supplemented with insulin. They
An efficient approach for generating complex induced small molecules – blocking the Wnt/β-­-
three-dimensional corneal organoids from human catenin (GSK3β) and TGF-β/Avidin/Nodal
IPSCs has been described recently [64]. In this (SB431542) signaling pathways  – allowing
research, the F2-3F1 human IPSC line formed NCCs to be generated as an intermediate of
typical embryoid body colonies both on mouse IPSCs, retaining the ability to produce differenti-
embryonic fibroblast feeder layers and on ated cells in terminal stages. Induced NCCs were
matrigel-­ coated substrates. As the three-­successfully differentiated into corneal endothe-
dimensional structures matured progressively lial cells, among other cell types, using a condi-
in vitro, oval-shaped eye field primordial clusters tioned cell culture medium from human corneal
developed in self-organized miniature structures endothelial cell cultures and a selective inhibitor
that consisted of retinal cells, a corneal primor- of Rho-kinase (ROCK, Y-27632).
dium, and an outer covering resembling a primi-
tive eyelid. The mature organoids were developed
in complex 3D tissues, like mini-corneas, and 18.5 New Perspectives
simulated the events of early development, and Challenges for Clinical
in  vitro. These 3D structures, in more terminal Application
stages, expressed specific markers, such as PAX6,
p63, and CK12 [64]. These approaches could In recent years, the great potential of advanced
represent an alternative source of tissue for sev- cell-based therapies for the repair and regenera-
eral ocular cell types. tion of different ocular structures in a variety of
272 R. P. Casaroli-Marano

ophthalmic diseases has started to be realized, from allogeneic tissues have the potential to
mainly for corneal and degenerative retinal cause immune reactions in recipient tissue, recent
pathology [68, 69]. The advances focus on new advances in adult tissue autologous stem cells
technologies for embryonic, adult, and induced (IPSCs and induced MSCs) herald the possibility
pluripotent cells that are differentiated to various of compatible therapy based on deriving different
cell phenotypes. Improved understanding of the types of ocular tissues [68, 71–73]. The required
molecular mechanisms of pluripotency and its progenitor cells can be obtained easily from the
clinical implications are producing ever richer skin, adipose tissue, or peripheral blood by mini-
data on the way in which the signals of the limbal mally invasive methods. Although IPSCs have
niche promote differentiation of stem cells. not yet been used clinically to reconstruct the
Particularly in the cornea, stem cell-based ther- ocular surface, several recent studies have dem-
apy offers the most physiological, most rational, onstrated the successful generation of LSC and
and least invasive approach for regeneration. For corneal epithelial cell populations from IPSCs.
LSCD, this is already considered an advanced However, further refinement of the corneal cell
therapeutic option. derivation from IPSCs will be required before
There remain some important problems with this potentially unlimited source of corneal epi-
IPSCs that need to be resolved before they can be thelial stem cells meets the requirements for clin-
used freely in cell therapy. Notably, strict guide- ical application.
lines on their safety and tolerability are required to Generally, the methods used to induce cellular
minimize the potential for harmful effects in recip- pluripotency and the ex vivo expansion of differ-
ients. The main concerns for the clinical use of entiated cell lines requires the use of xenobiotics
IPSCs surround their nature and the methodology or allogeneic products. These may include cells
used to obtain pluripotent cell lines. These must be of nonhuman origin for the co-culture approaches
thoroughly evaluated before it can be judged in feeder-layer techniques, the supplementation
whether their clinical use will be safe. Assessment of cell culture media with serum of animal origin,
should ensure the following: (a) they should not or the use of other molecules for the induction or
produce tumors or cancers, (b) they should not inhibition of in  vitro molecular phenomena.
proliferate uncontrollably, (c) they should not gen- Unfortunately, these may not meet the minimum
erate immunogenic responses in recipient tissue quality specifications required for clinical appli-
when using vectors (viral or not), (d) they should cation. These products potentially risk disease
not produce an aggressive response in the recipient transmission, tumorigenesis, or precipitating an
tissue, and (e) they must be located exclusively at immune response in the recipient [4, 18, 33, 74,
the specific application site of the target regener- 75]. They also show an idiosyncratic biological
ated tissue [70]. Consistent with these, the most variability that can negatively affect the quality
recent methodologies for inducting pluripotency of cell cultures and the final results of transplan-
by non-­integrative reprogramming – transient and tation. Therefore, there is currently a need to
controlled transfection vectors, using DNA or investigate the options for using chemically
interference RNA, episomes – probably represent defined xenofree culture media and recombinant
the best options for creating a safe final differenti- molecules of human origin that meet clinical
ated product. The methods of direct reprogram- grade specifications and protocols [57]. Thus, the
ming, which avoid pure pluripotent states, may production of clinical grade cell lines, in the
also offer significant advantages for the control of absence of xenobiotics, and guided by Good
cellular behavior in host tissue. Manufacturing Procedures (GMP) regulations
An interesting advantage that favors wide- and standard operating procedures, is important
spread clinical application of IPSCs is that autol- and mandatory before we can produce IPSCs and
ogous ocular tissue may not always be available differentiated cells for clinical purposes. Another
where there is severe bilateral ocular disease. important concept is the balance between cost
Whereas cell populations obtained and applied and safety [32]. To this end, a significant step for-
18  Cell-based Therapy Using Induced Pluripotent Stem Cell 273

ward in establishing the definitive clinical appli- arisen in progenitor cells from adult tissue.
cability of this cell type may be to develop IPSC These provide high proliferative potential and
biobanks in which cells originate from homozy- lower immunogenicity, although with certain
gous haplotypes that show high frequency in the limitations for complete terminal differentiation.
population [76]. Ex vivo cell expansion techniques based on the
As confirmed, the in vivo or in vitro microen- patient’s own tissue represent a rational, physio-
vironment (niche) is a crucial element for main- logical, and minimally invasive approach for the
taining the homeostasis and stemness of treatment of some prevalent ocular surface dis-
progenitor cells. Improving our knowledge of the eases, including LSCD.  Indeed, autologous tis-
properties of this niche for LSCs is a key issue as sue provides key advantages over allogenic
we move forward, since this could contribute to approaches, most notably in relation to immune
understand the mechanisms related to corneal tolerance in the recipient tissue. These tissue
repair and the events associated with cell differ- engineering techniques continue to evolve, with
entiation. Indeed, the balance between cells and methodological improvements continuing to
their environment is due, in large part, to the increase their efficacy and clinical applicability.
presence of sustained production and secretion of There have also been significant increases in our
paracrine factors, which are specific to each understanding of the mechanisms of regenera-
microenvironment. Intercellular communication tion and repair of the ocular surface. In turn, bet-
is partially mediated by signals that come from ter understanding of LSC behavior and the
the ECM and secreted microvesicles (SMV) – the limbal niche has shed light on improvements in
exosomes are SMVs of endosomal origin that are translational observations for clinical results. A
released as multivesicular bodies [77]  – which central concept is that the paracrine factors pro-
are enriched by cytoskeletal molecules, signal duced and secreted in the sclerocorneal limbic
transduction proteins, and cytoplasmic enzymes region constitute specialized elements that are
[77, 78]. Surprisingly, they can also transfer required for cell survival and LSC self-renewal,
functional RNA and miRNA to recipient cells making them essential to the regenerative pro-
[78]. Exosomes and miRNA also have pleiotro- cess of corneal epithelial cells.
pic biological functions that are still poorly The advent of IPSCs holds great promise for
understood, but that may include roles in immune future clinical application. The eye, and the ocu-
activation, angiogenesis modulation, and cell lar surface (cornea) in particular, appears to have
protection and repair [55, 79]. In addition, the the optimal characteristics for the application of
regenerative and cell-inducing plasticity propri- advanced therapy based on IPSC techniques.
eties have been proven based on exosomes Being a well-compartmentalized structure of
obtained from MSCs [80] and ESCs [72]. Thus, a small dimensions, the eye requires only a small
key future area of study in the field of corneal number of cells for treatment and enjoys an
regeneration will surround the corneal microen- “immune privilege” state. Another advantage is
vironment, and particularly the role of exosomes that processes are easy to monitor by image
secreted by corneal cells to maintain the homeo- ­capture and recording after implantation on ocu-
stasis and regeneration of the ocular surface. lar surfaces. Likewise, the ocular surface of the
contralateral eye can serve as a control for the
evolution of outcomes in cases of bilateral dis-
18.6 Conclusions ease. These are incontestable facts because, at
the time of writing, the only ongoing human
Advanced cell-based therapy, whether associ- clinical trials using IPSCs were for the treatment
ated with tissue bioengineering approaches or of age-related macular degeneration. Specifically,
not, has experienced some surprising advances this was through the implantation of RPE cells
over the last decade. An alternative emerging derived from IPSCs that originated from the
source of cells for ocular surface treatment has patient’s own dermal fibroblasts [46].
274 R. P. Casaroli-Marano

Despite the promise of these new approaches, example of limbal stem cell-based regenerative medi-
cine. Stem Cells. 2014;32:26–34.
there remain some unresolved points regarding 5. Rama P, Ferrari G, Pellegrini G.  Cultivated limbal
the safety and tolerability of differentiated cell epithelial transplantation. Curr Opin Ophthalmol.
lines from IPSCs. Among these, concerns of the 2017;28:387–9.
cell cycle and the potential for tumorigenesis are 6. Fuest M, Yam GH-F, Peh GS-L, Mehta
JS. Advances in corneal cell therapy. Regen Med.
the most worrisome. Approaches for inducing 2016;11:601–15.
pluripotency by non-integrative or even direct 7. Kim K, Mian S. Diagnosis of corneal limbal stem cell
reprogramming have been integrated into the deficiency. Curr Opin Ophthalmol. 2017;24:355–62.
methodological arsenal to allow for a more con- 8. Casaroli-Marano R, Nieto-Nicolau N, Martínez-­
Conesa E, Edel M, Álvarez-Palomo AB.  Potential
trolled clinical application. There is also a need role of induced pluripotent stem cells (IPSCs) for
for cell culture techniques that produce cell cell-based therapy of the ocular surface. J Clin Med.
expansion and maintenance with defined culture 2015;4:318–42.
media and supplements that are of clinical grade 9. Dua HS, Shanmuganathan VA, Powell-Richards AO,
Tighe PJ, Joseph A. Limbal epithelial crypts: a novel
requirements and free of xenoproducts. This anatomical structure and a putative limbal stem cell
would not only improve the quality of the final niche. Br J Ophthalmol. 2005;89:529–32.
therapeutic product but would also increase its 10. Gonzalez G, Sasamoto Y, Ksander BR, Frank MH,
safety in the recipient. These problems are cur- Frank NY.  Limbal stem cells: identity, developmen-
tal origin, and therapeutic potential. Wiley Interdiscip
rently the focus of ongoing research. Once Rev Dev Biol. 2017. https://doi.org/10.1002/
resolved, we could see a new era based on the wdev.303.
controlled clinical use of IPSCs for certain ocu- 11. Yeung AMH, Schlötzer-Schrehardt U, Kulkarni B,
lar diseases, particularly those of the ocular Tint NL, Hopkinson A, Dua HS.  Limbal epithelial
crypt: a model for corneal epithelial maintenance and
surface. novel limbal regional variations. Arch Ophthalmol.
2008;126:665–9.
Acknowledgments The author acknowledges support 12. Pellegrini G, Golisano O, Paterna P, Lambiase
from Fundació Marató TV3 (120630-31) and Fondo de A, Bonini S, Rama P, et  al. Location and clonal
Investigaciones Sanitarias del Instituto de Salud Carlos analysis of stem cells and their differentiated
III (FIS14-PI00196 and FIS18-PI00355), from the progeny in the human ocular surface. J Cell Biol.
European Regional Development Fund (FEDER) of 1999;145:769–82.
European Union. 13. Thoft RA, Friend J. The X, Y, Z hypothesis of corneal
Compliance with ethical requirements: Ricardo Pedro epithelial maintenance. Invest Ophthalmol Vis Sci.
Casaroli-Marano declares that he has no conflict of inter- 1983;24:1442–3.
est. All procedures followed were in accordance with the 14. Dua HS, Miri A, Alomar T, Yeung AM, Said DG. The
ethical standards of the responsible committee on human role of limbal stem cells in corneal epithelial main-
experimentation (institutional and national) and with the tenance. Testing the dogma. Ophthalmology.
Helsinki Declaration of 1975, as revised in 2000. Informed 2009;116:856–63.
consent was obtained from all patients for being included 15. Yazdanpanah G, Jabbehdari S, Djalilian AR. Limbal
in the study. No animal studies were carried out by the and corneal epithelial homeostasis. Curr Opin
authors for this article. Ophthalmol. 2017;28:348–54.
16. Majo F, Rochat A, Nicolas M, Jaoudé GA, Barrandon
Y. Oligopotent stem cells are distributed throughout the
mammalian ocular surface. Nature. 2008;456:250–4.
References 17. Zhang Y, Sun H, Liu Y, Chen S, Cai S, Zhu Y, et al.
The limbal epithelial progenitors in the limbal niche
1. Osei-Bempong C, Figueiredo FC, Lako M. The lim- environment. Int J Med Sci. 2016;13:835–40.
bal epithelium of the eye–a review of limbal stem 18. Pellegrini G, Rama P, Matuska S, Lambiase A, Bonini
cell biology, disease and treatment. BioEssays. S, Pocobelli A, et al. Biological parameters determin-
2013;35:211–9. ing the clinical outcome of autologous cultures of lim-
2. Casaroli-Marano RP, Nieto-Nicolau N, Martinez-­ bal stem cells. Regen Med. 2013;8:553–67.
Conesa EM. Progenitor cells for ocular surface regen- 19. Rama P, Matuska S, Paganoni G, Spinelli A, De
erative therapy. Ophthalmic Res. 2013;49:115–21. Luca M, Pellegrini G.  Limbal stem-cell therapy
3. Utheim TP. Limbal epithelial cell therapy: past, pres- and long-term corneal regeneration. N Engl J Med.
ent, and future. Methods Mol Biol. 2013;1014:3–43. 2010;363:147–55.
4. Pellegrini G, Rama P, Di Rocco A, Panaras A, De 20. Nieto-Nicolau N, Martínez-Conesa E, Casaroli-­
Luca M.  Concise review: hurdles in a successful Marano R. Limbal stem cells from aged donors are a
18  Cell-based Therapy Using Induced Pluripotent Stem Cell 275

suitable source for clinical application. Stem Cells Int. 37. Schlaeger T. Nonintegrating human somatic cell repro-
2016;2016:3032128. gramming methods. Adv Biochem Eng Biotechnol.
21. Joe AW, Yeung SN. Concise review: identifying lim- 2018;163:1–21.
bal stem cells: classical concepts and new challenges. 38. Parameswaran S, Balasubramanian S, Rao MS,

Stem Cells Transl Med. 2014;3:318–22. Ahmad I.  Concise review: non-cell autonomous
22. Ljubimov AV, Saghizadeh M.  Progress in corneal
reprogramming: a nucleic acid-free approach to
wound healing. Prog Retin Eye Res. 2015;49:17–45. induction of pluripotency. Stem Cells. 2011;29:
23. Huang M, Wang B, Wan P, Liang X, Wang X, Liu Y, 1013–20.
et  al. Roles of limbal microvascular net and limbal 39. Kelaini S, Cochrane A, Margariti A.  Direct repro-
stroma in regulating maintenance of limbal epithelial gramming of adult cells: avoiding the pluripotent
stem cells. Cell Tissue Res. 2015;359:547–63. state. Stem Cells Cloning. 2014;7:19–29.
24. Polisetti N, Zenkel M, Menzel-Severing J, Kruse FE, 40. Blazejewska EA, Schlötzer-Schrehardt U, Zenkel M,
Schlötzer-Schrehardt U. Cell adhesion molecules and Bachmann B, Chankiewitz E, Jacobi C, et al. Corneal
stem cell-niche-interactions in the limbal stem cell limbal microenvironment can induce transdifferentia-
niche. Stem Cells. 2016;34:203–19. tion of hair follicle stem cells into corneal epithelial-­
25. Mali P, Yang L, Esvelt KM, Aach J, Guell M, DiCarlo like cells. Stem Cells. 2009;27:642–52.
JE, et al. RNA-guided human genome engineering via 41. Meyer-Blazejewska EA, Call MK, Yamanaka O, Liu
Cas9. Science. 2013;339:823–6. H, Schlötzer-Schrehardt U, Kruse FE, et al. From hair
26. Cabral T, DiCarlo JE, Justus S, Sengillo JD, Xu
to cornea: toward the therapeutic use of hair follicle-­
Y, Tsang SH.  CRISPR applications in ophthal- derived stem cells in the treatment of limbal stem cell
mologic genome surgery. Curr Opin Ophthalmol. deficiency. Stem Cells. 2011;29:57–66.
2017;28:252–9. 42. Yang K, Jiang Z, Wang D, Lian X, Yang T. Corneal
27. Takahashi K, Yamanaka S.  Induction of pluripotent epithelial-like transdifferentiation of hair follicle stem
stem cells from mouse embryonic and adult fibro- cells is mediated by pax6 and β-catenin/Lef-1. Cell
blast cultures by defined factors. Cell. 2006;126: Biol Int. 2009;33:861–6.
663–76. 43. Saichanma S, Bunyaratvej A, Sila-Asna M.  In vitro
28. Yu J, Vodyanik MA, Smuga-Otto K, Antosiewicz-­ transdifferentiation of corneal epithelial-like cells
Bourget J, Frane JL, Tian S, et  al. Induced pluripo- from human skin-derived precursor cells. Int J
tent stem cell lines derived from human somatic cells. Ophthalmol. 2012;5:158–63.
Science. 2007;318:1917–20. 44. Martínez-Conesa EM, Espel E, Reina M, Casaroli-­
29. Yamanaka S.  Induced pluripotent stem cells: past, Marano RP.  Characterization of ocular surface
present, and future. Cell Stem Cell. 2012;10:678–84. epithelial and progenitor cell markers in human adi-
30. Compagnucci C, Bertini E.  The potential of iPSCs pose stromal cells derived from lipoaspirates. Invest
for the treatment of premature aging disorders. Ophthalmol Vis Sci. 2012;53:513–20.
Int J Mol Sci. 2017;18. https://doi.org/10.3390/ 45. Cieślar-Pobuda A, Rafat M, Knoflach V, Skonieczna
ijms18112350. M, Hudecki A, Małecki A, et al. Human induced plu-
31. Li L, Chao J, Shi Y.  Modeling neurological dis-
ripotent stem cell differentiation and direct transdiffer-
eases using iPSC-derived neural cells: iPSC mod- entiation into corneal epithelial-like cells. Oncotarget.
eling of neurological diseases. Cell Tissue Res. 2016;7:42314–29.
2018;371:143–51. 46. Mandai M, Watanabe A, Kurimoto Y, Hirami Y,

32. Grskovic M, Javaherian A, Strulovici B, Daley
Morinaga C, Daimon T, et  al. Autologous induced
GQ. Induced pluripotent stem cells--opportunities for stem-cell–derived retinal cells for macular degenera-
disease modelling and drug discovery. Nat Rev Drug tion. N Engl J Med. 2017;376:1038–46.
Discov. 2011;10:915–29. 47. Takashima K, Inoue Y, Tashiro S, Muto K.  Lessons
33. Diecke S, Diecke S, Diecke S, Jung SM, Lee J, Lee J, for reviewing clinical trials using induced pluripotent
et al. Recent technological updates and clinical appli- stem cells: examining the case of a first-in-human trial
cations of induced pluripotent stem cells. Korean J for age-related macular degeneration. Regen Med.
Intern Med. 2014;29:547–57. 2018;13:123–8.
34. Chou BK, Mali P, Huang X, Ye Z, Dowey SN, Resar 48. Sugita S, Iwasaki Y, Makabe K, Kimura T, Futagami
LMS, et al. Efficient human iPS cell derivation by a T, Suegami S, et  al. Lack of T cell response to
non-integrating plasmid from blood cells with unique iPSC-­ derived retinal pigment epithelial cells
epigenetic and gene expression signatures. Cell Res. from HLA homozygous donors. Stem Cell Rep.
2011;21:518–29. 2016;7:619–34.
35. Takahashi K, Tanabe K, Ohnuki M, Narita M, Ichisaka 49. Sugita S, Iwasaki Y, Makabe K, Kamao H, Mandai
T, Tomoda K, et al. Induction of pluripotent stem cells M, Shiina T, et al. Successful transplantation of reti-
from adult human fibroblasts by defined factors. Cell. nal pigment epithelial cells from MHC homozygote
2007;107:861–72. iPSCs in MHC-matched models. Stem Cell Rep.
36. Pietronave S, Prat M.  Advances and applications
2016;7:635–48.
of induced pluripotent stem cells. Can J Physiol 50. Morikawa S, Mabuchi Y, Niibe K, Suzuki S, Nagoshi
Pharmacol. 2012;90:317–25. N, Sunabori T, et  al. Development of mesenchymal
276 R. P. Casaroli-Marano

stem cells partially originate from the neural crest. 65. Zhu Q, Lu Q, Gao R, Cao T. Prospect of human plurip-
Biochem Biophys Res Commun. 2009;379:1114–9. otent stem cell-derived neural crest stem cells in clini-
51. Fuhrmann S.  Wnt signaling in eye organogenesis.
cal application. Stem Cells Int. 2016;2016:7695836.
Organogenesis. 2008;4:60–7. 66. Naylor RW, McGhee CNJ, Cowan CA, Davidson

52. Mikhailova A, Ilmarinen T, Uusitalo H, Skottman
AJ, Holm TM, Sherwin T.  Derivation of corneal
H.  Small-molecule induction promotes corneal keratocyte-­like cells from human induced pluripotent
epithelial cell differentiation from human induced stem cells. PLoS One. 2016;11:e0165464.
pluripotent stem cells. Stem Cell Rep. 2014;2: 67. Fukuta M, Nakai Y, Kirino K, Nakagawa M, Sekiguchi
219–31. K, Nagata S, et al. Derivation of mesenchymal stro-
53. Shalom-Feuerstein R, Serror L, De La Forest Divonne mal cells from pluripotent stem cells through a neural
S, Petit I, Aberdam E, Camargo L, et al. Pluripotent crest lineage using small molecule compounds with
stem cell model reveals essential roles for miR-­ defined media. PLoS One. 2014;9:e112291.
450b-­5p and miR-184  in embryonic corneal lineage 68. Aharony I, Michowiz S, Goldenberg-Cohen N.  The
specification. Stem Cells. 2012;30:898–909. promise of stem cell-based therapeutics in ophthal-
54. Xu S. microRNA expression in the eyes and their sig- mology. Neural Regen Res. 2017;12:173–80.
nificance in relation to functions. Prog Retin Eye Res. 69. Zarbin M. Cell-based therapy for degenerative retinal
2009;28:87–116. disease. Trends Mol Med. 2016;22:115–34.
55. Rassi D, De Paiva C, Dias L, Modulo C, Adriano 70. Palomo ABA, Lucas M, Dilley RJ, McLenachan S,
L, Fantucci M, et  al. Review: microRNAs in ocular Chen FK, Requena J, et al. The power and the prom-
surface and dry eye diseases. Ocul Surf. 2017;14: ise of cell reprogramming: personalized autologous
660–9. body organ and cell transplantation. J Clin Med.
56. Aberdam E, Petit I, Sangari L, Aberdam D. Induced 2014;3:373–87.
pluripotent stem cell-derived limbal epithelial cells 71. Kao WWY, Coulson-Thomas VJ. Cell therapy of cor-
(LiPSC) as a cellular alternative for in  vitro ocular neal diseases. Cornea. 2016;35:S9–19.
toxicity testing. PLoS One. 2017;12:e0179913. 72.
Erbani J, Aberdam D, Larghero J, Vanneaux
57. Gonzalez S, Chen L, Deng S. Comparative study of V. Pluripotent stem cells and other innovative strate-
xenobiotic-free media for the cultivation of human gies for the treatment of ocular surface diseases. Stem
limbal epithelial stem/progenitor cells. Tissue Eng Cell Rev. 2016;12:171–8.
Part C Methods. 2017;23:219–27. 73. Angunawela RI, Mehta JS, Daniels JT. Ex-vivo ocular
58. Lancaster MA, Knoblich JA. Organogenesis in a dish: surface stem cell therapies: current techniques, appli-
modeling development and disease using organoid cations, hurdles and future directions. Expert Rev Mol
technologies. Science. 2014;345:1247125. Med. 2013;15:e4.
59. Kuwahara A, Ozone C, Nakano T, Saito K, Eiraku M, 74. Menzel-Severing J, Kruse FE, Schlötzer-Schrehardt
Sasai Y. Generation of a ciliary margin-like stem cell U.  Stem cell–based therapy for corneal epithelial
niche from self-organizing human retinal tissue. Nat reconstruction: present and future. Can J Ophthalmol.
Commun. 2015;6:6286. 2013;48:13–21.
60.
Völkner M, Zschätzsch M, Rostovskaya M, 75. Di Iorio E, Ferrari S, Fasolo A, Böhm E, Ponzin D,
Overall RW, Busskamp V, Anastassiadis K, et  al. Barbaro V. Techniques for culture and assessment of
Retinal organoids from pluripotent stem cells effi- limbal stem cell grafts. Ocul Surf. 2010;8:146–53.
ciently recapitulate retinogenesis. Stem Cell Rep. 76. Morrison M, Bell J, George C, Harmon S, Munsie
2016;6:525–38. M, Kaye J.  The European General Data Protection
61. Foster JW, Wahlin K, Adams SM, Birk DE, Zack DJ, Regulation: challenges and considerations for
Chakravarti S. Cornea organoids from human induced iPSC researchers and biobanks. Regen Med.
pluripotent stem cells. Sci Rep. 2017;7:41286. 2017;12:693–703.
62. Hayashi R, Ishikawa Y, Sasamoto Y, Katori R, Nomura 77. Colombo M, Raposo G, Théry C. Biogenesis, secre-
N, Ichikawa T, et al. Co-ordinated ocular development tion, and intercellular interactions of exosomes and
from human iPS cells and recovery of corneal func- other extracellular vesicles. Annu Rev Cell Dev Biol.
tion. Nature. 2016;531:376–80. 2014;30:255–89.
63. Hayashi R, Ishikawa Y, Katori R, Sasamoto Y,
78. Zhang J, Li S, Li L, Li M, Guo C, Yao J, et al. Exosome
Taniwaki Y, Takayanagi H, et al. Coordinated genera- and exosomal microRNA: trafficking, sorting, and
tion of multiple ocular-like cell lineages and fabrica- function. Genomics Proteomics Bioinformatics.
tion of functional corneal epithelial cell sheets from 2015;13:17–24.
human iPS cells. Nat Protoc. 2017;12:683–96. 79. Klingeborn M, Dismuke WM, Bowes Rickman C,
64. Susaimanickam PJ, Maddileti S, Pulimamidi VK,
Stamer WD.  Roles of exosomes in the normal and
Boyinpally SR, Naik RR, Naik MN, et al. Generating diseased eye. Prog Retin Eye Res. 2017;59:158–77.
minicorneal organoids from human induced pluripo- 80. Yu B, Zhang X, Li X. Exosomes derived from mesen-
tent stem cells. Development. 2017;144:2338–51. chymal stem cells. Int J Mol Sci. 2014;15:4142–57.
Cultivated Limbal Stem Cell
Transplantation: Indications
19
and Technique

Joséphine Behaegel, Sorcha Ní Dhubhghaill,
and Marie-José Tassignon

Table 19.1  Aetiologies of limbal stem cell deficiency


19.1 Introduction Genetic Acquired
Aniridia [3] Chemical and thermal burns [4]
Limbal stem cell deficiency (LSCD) is a condi- Turner syndrome [5] Stevens-Johnson syndrome [6]
tion that results from a loss of limbal stem cells Dyskeratosis Ocular cicatricial pemphigoid
or a dysfunctional limbal microenvironment. congenita [7] [8]
The term refers to a highly heterogeneous group Xeroderma Graft versus host disease [10]
pigmentosum [9]
of pathologies ranging from acquired diseases
LADD syndrome [11] Infectious keratitis [12]
such as chemical burns to severe autoimmune KID syndrome [13] Iatrogenic causes:
conditions, iatrogenic trauma and genetic abnor- chemotherapy, radiation,
malities such as aniridia [1, 2] (Table 19.1). In surgery, cryotherapy [14–16]
all of these conditions, the reserve of corneal Dominantly inherited Ocular surface tumours [18]
keratitis [17]
limbal stem cells buried in the limbal niches is
Congenital epidermal Chronic contact lens wear [20]
dysplasia [19]
Bullous keratopathy [21]

J. Behaegel
Faculty of Medicine and Health Sciences,
Department of Ophthalmology, Visual Optics and
Visual Rehabilitation, University of Antwerp,
Campus Drie Eiken, Antwerp, Belgium
Department of Ophthalmology, Antwerp University
Department of Ophthalmology, Antwerp University Hospital, Edegem, Belgium
Hospital, Edegem, Belgium
Center for Cell Therapy and Regenerative Medicine,
Center for Cell Therapy and Regenerative Medicine, Antwerp University Hospital, CCRG-Oogheelkunde,
Antwerp University Hospital, CCRG-Oogheelkunde, Edegem, Belgium
Edegem, Belgium e-mail: Sorcha.NiDhubhghaill@uza.be
Department of Ophthalmology, Brussels University M.-J. Tassignon (*)
Hospital, Jette, Belgium Faculty of Medicine and Health Sciences,
e-mail: josephine.behaegel@uzbrussel.be Department of Ophthalmology, Visual Optics and
Visual Rehabilitation, University of Antwerp,
S. N. Dhubhghaill
Campus Drie Eiken, Antwerp, Belgium
Faculty of Medicine and Health Sciences,
Department of Ophthalmology, Visual Optics and Department of Ophthalmology, Antwerp University
Visual Rehabilitation, University of Antwerp, Hospital, Edegem, Belgium
Campus Drie Eiken, Antwerp, Belgium e-mail: Marie-jose.tassignon@uza.be
© Springer Nature Switzerland AG 2019 277
J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_19
278 J. Behaegel et al.

depleted. In the absence of these cells, lost epi- Cultivated in  vivo expansion has several
thelial cells cannot be replenished resulting in advantages when compared with large-tissue
corneal epithelial defects. This breakdown of transplantations of donor tissue:
the epithelial barrier results in a status of chronic
inflammation with eventual migration of con- 1. There is a reduction of the risks to donor eye
junctival cells to cover the defects and subse- since only a small amount of donor tissue is
quent superficial vascularization of the cornea needed.
[4, 22]. 2. If necessary, a regraft can be performed [28].
As a result, patients suffer from decreased 3. There is less risk of donor rejection in case of
visual acuity, pain and photophobia. Traditional allogeneic transplantation since the antigen-­
corneal grafting techniques tend to fail in these presenting macrophages do not survive the
patients for two reasons: cultivation process [29].
4. The dose and the duration of systemic immu-
1. The epithelial growth over a new transplant nosuppression are significantly reduced in
is typically slow, and the new graft stroma case of allogeneic transplantation since the
may be exposed by a persistent epithelial donor cells show limited long-term survival
defect. Even when the epithelium does close, [30–32].
this is frequently due to conjunctival epithe-
lial cell overgrowth with compromised As a result, many ocular centres have pivoted
transparency. to a cultivated approach.
2. The increased corneal vascularization pres-
ent in these eyes will significantly increase
the risk of graft rejection over the long term 19.2 Preoperative Assessment
[23].
19.2.1 Indications for Surgery
For both of these reasons, standard corneal and Preoperative Evaluation
transplantation is not advised in these high-risk
patients. Cultivated limbal stem cell transplantation is
The primary goal, therefore, in LSCD treat- indicated for patients suffering from either total
ment is to restore the limbal microenvironment or subtotal LSCD.  Prior to transplantation, the
and to regain a corneal epithelial phenotype that eye should be carefully evaluated and prepared to
can self-sustain. This can be achieved by the optimize the environment and to ensure the best
transplantation of healthy limbal epithelial cells. chance of a successful grafting as not all LSCD
The main difference between surgical approaches patients are likely to benefit.
is the quantity, source and postharvest manipula-
tion of the cells. Direct transplantation of “large-­ 19.2.1.1 Patient
tissue” grafts includes keratolimbal allograft, Management of LSCD begins with a careful eval-
conjunctival-limbal autografts and living-related uation of the extent of deficiency and associated
conjunctival-limbal allografts, all of which abnormalities [33]. Because LSCD is a term that
require large sections of limbal donor tissue [24– applies to the end result of a variety of patholo-
26]. While autologous grafts perform better than gies, it rarely presents as a single, isolated epithe-
allogenic, there is a safety limit to how much tis- lial problem. Patients typically have a complex
sue the healthy fellow eye can donate. Large-­ history of disease progression, treatments and
tissue autologous grafts therefore can typically surgeries. Depending on aetiology and severity,
only be performed once. Cultivated graft tech- there can be involvement of the surrounding ocu-
niques were developed to reduce this risk to lar tissues including the corneal nerves, corneal
donor eyes by reducing the amount of donor tis- stroma, endothelium, conjunctiva, eyelids and
sue required [27]. lacrimal system.
19  Cultivated Limbal Stem Cell Transplantation: Indications and Technique 279

Tear Film
Similarly, a healthy tear film is essential for sur-
Tears vival of the stem cells. A tear film supports the
ocular surface by providing oxygen and nutrients
and protects against desiccating environments
[36]. Tear film can be optimized by the use of
nonpreservative teardrops or punctal occlusion
and by treating eyelid inflammation such as
Epithelium
blepharitis. While a mild dry eye is not an exclu-
sion criteria, a complete absence of any tear pro-
duction and/or the presence of corneal epithelial
keratinization identifies an extremely hostile
Stem cells environment, and as a result these eyes perform
very poorly and are typically excluded from lim-
bal stem cell surgery.
Ocular surface Additionally, any inflammation of the ocular
surface should be evaluated and treated if neces-
sary. Inflammation disturbs the normal milieu of
Fig. 19.1 Successful limbal stem cell transplantation is the remnants of the limbal niche and may lead to
similar to growing a plant from seed (Analogy adapted
from Tseng and Tsubota [34])
dysfunction or aberrant differentiation of the
stem cells [37].

Limbal stem cells may be considered as seeds, Corneal Epithelium


while the ocular surface is the soil in which they At this point we examine the cornea itself. While
are embedded [34] (Fig.  19.1). The seeds will classification schemes for LSCD clinical trials
only grow in an optimal environment with the are currently debated, it is useful to note the
right nutrients. Therefore, prior to stem cell trans- degree of clock hours that have been conjuncti-
plantation, it is important to prepare the best valized or vascularized. The presence and mea-
recipient bed for the cultivated cells in order to surements of any epithelial defects should also be
restore the ocular surface defences. noted. This can easily be documented by means
The preoperative examination therefore begins of ocular surface photography. It is important to
with the eyelids. determine whether the visual axis is involved or
threatened, as no surgical treatment may be
Eyelids required in case the central axis is free from
The eyelids are crucial for the preservation of abnormal tissue. In case of partial LSCD, a
ocular surface integrity, as well as for maintain- repeated debridement of the conjunctival epithe-
ing a healthy preocular tear film [35]. Many of lium can be performed. This simple surgery peels
the pathologies that result in LSCD such as back the migrating conjunctival cells to buy time
chemical injury or mucous membrane pemphi- for the healthy corneal epithelial cells to cover
goid will also cause severe conjunctival and eye- the defect in a technique known as sequential
lid disease. Even the highest-quality cultivated sector conjunctival epitheliectomy (SSCE) [38,
cells will not be able to resist repeated mechani- 39]. This has the advantage of not requiring
cal abrasions experienced with trichiasis, entro- donor tissues but is only an option for those cases
pion or exposure and will not engraft. Contact of milder disease where there are some remain-
lenses used to protect the new grafts can also be ing stem cells to sustain an epithelial layer. In
difficult to fit around symblepharon, so prior to total LSCD with central corneal involvement, a
CLET all lid abnormalities should be corrected to limbal stem cell graft remains the treatment of
the fullest extent possible. choice.
280 J. Behaegel et al.

Corneal Stroma ened corneal stroma on preoperative imaging. If


In the simplest cases of LSCD where the pathol- this is the case, it may also be an argument for
ogy is limited to the epithelium, the thick fibro- performing a PK at the time of CLET.
vascular pannus can be peeled away to reveal a
clear stromal bed. These are the ideal patients, Corneal Nerves
where successful CLET surgery can result not Corneal sensitivity should be tested and docu-
only in an improvement in the epithelial layer but mented (in clinical studies with a Cochet-Bonnet
also an increase in visual acuity. In reality, how- esthesiometer). In health, corneal epithelial cells
ever, this represents the minority of LSCD are highly innervated by the terminal nerve fibres
patients. Injuries that cause LSCD are often of the sub-basal plexus [41]. This innervation is
accompanied by deeper damage in the stromal not only essential for the provision of nerve
layer resulting in scarring. It is important for growth factor (NGF), which plays a crucial role
these patients to understand that if there is signifi- in corneal healing processes, but is also a key
cant corneal stromal scarring, CLET surgery will input to the lacrimal functional unit (LFU) gov-
not be a sight-restoring procedure. The aim of erning tear film production [42, 43]. Loss of cor-
CLET in these cases is to improve the epithelium neal sensitivity can therefore cause a “double-hit”
as a first step and to facilitate a sight-restoring to the newly seeded stem cells by depriving them
corneal transplant as a second step. of NGF and damaging the tear film.
Persistent epithelial defects and stromal lysis
can also result in regions of corneal thinning Glaucoma
which can be of particular concern during CLET As mentioned previously, LSCD typically occurs
surgery. If the donor stromal bed is too thin, in the larger context of severe anterior segment
removing the fibrovascular pannus and placing a disease. Raised intraocular pressure (IOP) and
CLET can result in a cornea with severely dam- glaucoma as a result are frequent comorbidities,
aged biomechanics leading to further thinning particularly in cases of aniridia. The postopera-
and perforation. A residual stromal bed of at least tive management of the transplant requires a pro-
350 μm is advised to avoid this serious adverse longed period of topical corticosteroid use, and
outcome. When significant stromal thinning is even in normotensive patients, the possibility of a
noted preoperatively, the surgical plan may have cortisone IOP response should be considered.
to be adapted to include stromal replacement by Any increase in IOP should be treated prior to
PK or lamellar surgery (deep anterior lamellar considering CLET.  Topical glaucoma medica-
keratoplasty, DALK) combined with the CLET tions can cause superficial punctate keratopathy
surgery, though this is not without its own risks. in normal corneas [44] and are even more toxic to
cultured epithelial transplants. We recommend
Corneal Endothelium preservative-free drops, laser or surgery for glau-
A functional corneal endothelium is also known coma control prior to CLET.
to be essential for epithelial cell adhesion. Indications and contraindications may vary
Without endothelial active pump function, even among various centres. Bilateral LSCD is often
healthy corneal epithelium can form bullae and described as a contraindication since this requires
slough off [40]. Transplanted epithelium does not allogeneic stem cells. In our centre we include
fare any better. Clinical evaluation of the endo- both uni- and bilateral diseases and include all
thelium can be difficult in these patients, as the aetiologies of LSCD.  Our contraindications are
corneal pannus and stromal scarring can often listed in Table 19.2.
prevent visualization of the anterior segment and
a covering of conjunctival epithelium will not 19.2.1.2 Donor
form the same characteristic bullae as corneal Both autologous and allogeneic donor eyes may be
epithelium to aid diagnosis. Severe endothelial included for limbal tissue donation. Autologous tis-
dysfunction can be suspected by a highly thick- sue is always preferred as it confers no
19  Cultivated Limbal Stem Cell Transplantation: Indications and Technique 281

Table 19.2  Contraindications for CLET for traceability purposes. Unlike autologous tis-
Contraindications: sue, allogenic cells do confer additional infec-
relative Contraindications: absolute tious risks to the recipient patient. As a result all
Clear visual axis Corneal keratinization allogenic tissue donors require additional screen-
Inflammation of ocular Phthisis bulbi
surface (ocular surface/
ing for human T-cell leukaemia-lymphoma virus
eyelids) 1 and testing with nucleic acid tests (NAT) for
Conjunctival pathology Absence of tear film HBV, HCV and HIV. Allogeneic donors undergo
(symblepharon, screening questionnaires for Creutzfeldt-Jakob
scarring)
disease (CJD) and rabies among others.
Malposition of the Inability to follow post-op
eyelids treatment schedule/any A maximum of two biopsies are permitted per
conditions hampering donor eye. This is due to the surgical damage to
treatment compliance the harvest site. Removing more than two biop-
Glaucoma Uncontrollable raised IOP sies may compromise the donor limbus and con-
Hypersensitivity to post-op
fer a risk of developing localized iatrogenic
treatment
LSCD. Therefore, a previous history of two biop-
sies is an exclusion criterion for donation in our
i­mmunoreactivity and does not require systemic protocol.
immunosuppression postoperatively. In case of One final but important consideration in the
bilateral disease or other conditions prohibiting case of living-related allogeneic tissue donation
autologous tissue donation, living-related or cadav- is the ethics of donation. This is often considered
eric donors can be considered. While the benefit of in organ transplantation but frequently forgotten
human leucocyte antigen (HLA) testing for CLET in stem cell surgery. It can be difficult to verify
and limbal tissue donation has not been validated, whether the donation is truly a voluntary act.
we nevertheless perform it in our centre as we Family pressures, emotions and external influ-
believe that it could confer an advantage. We ences can play a role in manipulating the donor’s
require a match of at least 50% on loci HLA-A, right to choose. They may be afraid to say that
HLA-B and HLA-DR for tissue donation in order they actually do not want to donate. We advise
to reduce graft rejection. Only a small piece of discussing the donation with the possible donor
healthy limbal tissue is required for donation; in a quiet space, without the recipient or other
1 × 2 mm is sufficient for most culture protocols. family members. Should the possible donor not
We exclude donation from eyes suffering from par- wish to donate, we respect their wishes and com-
tial limbal stem cell deficiency, to avoid additional municate to the patient that the possible donor
risks to an already compromised eye. Opinions was “not appropriate” for donation. Informing
vary, as other groups do harvest stem cells from the family that the possible donor simply did not
compromised eyes with the goal of providing want to donate could have serious ramifications
autologous grafts even in bilateral disease [45]. in the family. The confidentiality of the possible
When considering a limbal region for biopsy, donor must be respected, and it is therefore not
we estimate the overall health of the region by ethical to let the details of their decision be
visualizing the limbal palisades of Vogt (POV) known.
using slit lamp biomicroscopy and non-invasive Our indications and contraindications for tis-
imaging modalities. In addition, all tissue donors, sue donation are listed in Table 19.3.
whether autologous or allogeneic, are screened
for human immunodeficiency virus 1 and 2
(HIV-1 and HIV-2), hepatitis B (HBV) and C 19.2.2 Supplementary Examinations
(HCV) and syphilis prior to surgery. While autol-
ogous tissue will not transmit disease to the In addition to slit lamp biomicroscopic evalua-
recipient, this is performed to determine status of tion, we perform a number of additional preop-
the tissue that is to be handled and must be known erative tests during patient and donor screening.
282 J. Behaegel et al.

Table 19.3 (Contra-)indications for limbal tissue 19.2.2.3 Optical Coherence


donation
Tomography (OCT)
Indications Contraindications OCT is a non-invasive, non-contact imaging
Healthy cornea and Signs of limbal stem cell technique that we use for cross-sectional images
limbus deficiency
Age ≥ 18y Age < 18y
of the anterior segment including the cornea and
≥50% HLA matching <50% HLA matching on limbus. This imaging modality is an added value
on locus HLA-A, locus HLA-A, HLA-B and for preoperative assessment of the patient and the
HLA-B and HLA-DRB1a donor eye, intraoperatively during pannus dissec-
HLA-DRB1a
tion and postoperatively during the follow-up
Voluntary donationb History of two limbal biopsy
donations/large limbal graft period. OCT has been shown to be a more reli-
Positive serology testing a able tool for visualization of the palisades of
Questionnaire/history Vogt, when compared with slit lamp biomicros-
suggestive for potential copy [48], but when compared with in vivo con-
transmittable diseasesb focal microscopy (IVCM), the resolution of OCT
External pressure for tissue
donation b images is significantly lower. OCT examination
is, however, less time-consuming, non-contact
a
In case of an allogeneic (living-related and cadaveric)
donor and less operator dependent and has a wider field
b
In case of an allogeneic living-related donor of view making it easier to implement in practice.
Using Fourier domain-OCT (FD-OCT), we have
developed a technique for quantification of lim-
19.2.2.1 Schirmer Test bal POV, which can aid in targeting the biopsy
Schirmer filter paper testing is used to estimate site for tissue donation [48] (Fig. 19.2).
the quantity of the tear film [46]. This quantita- During preoperative evaluation, OCT images
tive method provides little information about the are also used to examine the stromal involvement
quality of the tear film, which may be more and thickness and to estimate the residual corneal
important than quantity, but it is a useful preop- stroma after pannus removal. If the thinnest
erative measure. If the wetting strips measure less region measures less than 350  μm, CLET may
than <10  mm/5  min, punctal plugs should be need to be combined with either a PK or DALK
inserted prior to stem cell surgery to enhance the in order to prevent postoperative corneal thinning
tear film stability. If the wetting strips show no and perforation. Knowing this preoperatively
moisture whatsoever, this may be a contraindica- also allows the opportunity to discuss the adapted
tion to surgery. surgery with the patient.
Intraoperatively, OCT can be used to assist the
19.2.2.2 Impression Cytology surgeon in the removal of the fibrovascular pan-
Immunohistochemistry carried out on corneal nus. This aids the surgeon in a more precise
impression cytology allows to distinguish removal of the diseased tissue, which augments
between corneal and conjunctival epithelial cells chances of sparing any viable corneal stroma and
and therefore aids in diagnosing LSCD.  While may reduce risks of extreme thinning and perfo-
the procedure is relative simple, the test is ration [49]. During the postoperative follow-up
regarded as painful by the patients and risks dis- period, OCT is useful for evaluation of the
rupting the epithelium. Moreover, the false-­ ­corneal epithelium, graft tracking and assessment
negative rate of impression cytology is high of the corneal thickness.
partly because of coexisting goblet cell defi-
ciency in many other surface disorders [47]. We 19.2.2.4 In Vivo Confocal Microscopy
no longer routinely perform impression cytology (IVCM)
during patient screening and postoperative evalu- IVCM is a contact imaging technique that pro-
ation, preferring to use in vitro confocal micros- vides high-resolution images of the ocular sur-
copy, as described below. face at the cellular level. This technique has
19  Cultivated Limbal Stem Cell Transplantation: Indications and Technique 283

a b

Fig. 19.2  Evaluation of a donor limbus: (a) Slit lamp through the inferior limbal rim, identifying the limbal
image of an inferior limbal rim. The black arrow indicates niches (grey arrows). (c) OCT selection of the region of
the palisades of Vogt. (b) OCT image of the same subject interest

ICVM signs of LSCD include the presence


of goblet cells on the cornea, changes in the epi-
thelial phenotype, loss or abnormal palisades of
Vogt, neovascularization of the cornea and loss
of sub-basal nerves (Fig.  19.4). These signs
assist ophthalmologists in making the clinical
diagnosis of LSCD, and as such, ICVM can pro-
vide an alternative diagnostic technique to
impression cytology [54–56]. Postoperatively,
IVCM assists in evaluation and can show graft
integration and possible disease progression
Fig. 19.3  IVCM findings of a stem cell-rich region of the prior to it being clinically visible at the slit
inferior limbus showing limbal palisade ridges (arrows) lamp.
and focal stromal projections (arrowheads)
All supplementary tests during patient and
donor assessment are listed in Table 19.4.
gained wide popularity among corneal specialists
to evaluate various structures of the cornea, con-
junctiva and limbus [50–53]. It can be used in 19.3 Procedure
preoperative patient and donor screening and in
the assessment of limbal stem cell deficiency. Different protocols for cultivation and transplan-
Because of its high resolution, IVCM enables tation of limbal stem cells have been described
differentiation between the stem cell niches. worldwide. Variability is seen in patient selection,
Imaging of these niches helps to ensure the health culture techniques, donor materials, surgical
of the donor limbus and indicates the regions approaches and the postoperative treatment and
containing highest density of stem cells. In prac- follow-up. Good clinical results have been
tice, the donor is imaged at inferior and superior reported using different approaches, and to date
limbus, as stem cells are predominantly located there is no consensus on the optimal protocol. The
in these regions. Documentation of the most stem complexity and expense of cell-based approaches
cell-rich regions enables a targeted biopsy har- as well as the orphan nature of the disease make
vesting, which could lead to improved clinical large multicentre comparative trials effectively
outcomes (Fig. 19.3). impossible. In this chapter we will describe our
284 J. Behaegel et al.

a b

c d

Fig. 19.4  IVCM findings of LSCD: goblet cells on the cornea (a), corneal neovascularization (b), abnormal palisades
of Vogt (c) and metaplastic epithelial phenotype (d)

standardized approach, but the reader should con- ocular surface reconstruction. The membranes
sider the alternative approaches [29, 57]. are obtained from the Amniotic Membrane
Bank University Hospital Antwerp and are pre-
pared in a standardized way [58]. Briefly, 24 h
19.3.1 Culture Method before use, the HAM is thawed and washed
three times with 0.9% sterile NaCl for 30 min.
19.3.1.1 Scaffold They are then treated with thermolysin enzyme
We use human amniotic membrane (HAM) as a (Sigma-Aldrich) in order to remove the devital-
scaffold for cellular expansion of our limbal ized amniotic epithelium [58, 59].
stem cells. HAM is known to have anti-inflam- Before placing the cells in culture, the HAM is
matory and antimicrobial properties, and its low stretched into two interlockable amnion rings
immunogenicity makes it an ideal carrier for (Lumox cultureware, Greiner Bio-One, BeLux).
19  Cultivated Limbal Stem Cell Transplantation: Indications and Technique 285

Table 19.4  Supplementary tests tres of medium is placed on the fixed HAM and
Patient Donor 2,5  ml in the culture dish. These conditions are
Biomicroscopic Biomicroscopic evaluation maintained for the duration of the entire culture
evaluation period of 14  days, with the medium changed
Tear film (Schirmer Medical questionnaire tissue
test) donation
every 2–3 days. The medium is routinely checked
IOP IOP for aerobic bacteria, anaerobic bacteria, fungi,
OCT OCT (limbus) yeast, mycoplasma and endotoxin [58].
IVCM IVCM (limbus)
Flash visual evoked
potentiala
19.3.2 Surgical Procedure
Serology Serology
Nephrology screeningb
19.3.2.1 Biopsy
a
In case visual acuity is less than hand movements or sus-
pected optic neuropathy The biopsy harvesting is performed under topical
b
In case immunosuppression is required anaesthesia. Superficial limbal tissue of 1 mm by
2 mm is taken from either the superior or the infe-
rior limbus in a stem cell-rich region, as indicated
These rings are made of plastic and provide a by preoperative imaging. The edges of the biopsy
taut, wrinkle-free construct. are marked off with a 45° diamond blade, and a
The spongy layer of the HAM is removed with crescent knife is used to tunnel a depth of approx-
the help of two 25  mm cell scrapers (Sarsted, imately 100 microns. We do not use any forceps
Essen, Belgium). This provides a thinner, more during this procedure in order to avoid crush
transparent scaffold, with a more standardized injury to the limbal stem cells or niches within
proteomic profile [59, 60]. After removal of the the biopsy (Fig. 19.5). Immediately after dissec-
spongy layer, the inner ring (diameter, 32 mm) is tion, the harvested tissue is placed in culture
placed on top of the HAM, and the edges of the medium and transported to the laboratory where
membrane are pulled over the outer surface of the it is rinsed with culture medium supplemented
ring. The HAM attached to the inner ring is with antibiotics [58].
flipped over so that this is correctly oriented.
Then, the outer ring (diameter, 35 mm) is locked 19.3.2.2 Graft Transplantation
over it, sandwiching the membrane in place. The After the 14-day culture period, the limbal stem
final position of the HAM is with its basement cell graft is transplanted under general anaes-
membrane facing upwards [58]. thesia onto the stem cell-deficient eye. Surgery
begins with a 360 ̊ conjunctival periotomy at
19.3.1.2 Technique the limbus using conjunctival scissors
The 1  ×  2 mm biopsies are cultured using a (Fig.  19.6a). The fibrovascular pannus is then
xenogeneic-­ free culture protocol. The fixed
HAM are placed in a 55 mm culture dish, con-
taining 3.5 mL CnT-20 medium (progenitor cell-­
targeted, corneal epithelial medium; CELLnTEC,
Bern, Switzerland), supplemented with 1%
human AB serum (Sigma-Aldrich) and incubated
at 37 °C with 5% CO2 for 24 h. Before the limbal
biopsy is placed in culture, it is rinsed six times in
CnT-Prime medium containing antibiotics and
placed at the centre of the prepared HAM.
The cells are then cultured in the CnT-Prime
medium supplemented with 1% human AB serum
at the air liquid interface. One hundred microli- Fig. 19.5  Biopsy harvest of a donor limbus
286 J. Behaegel et al.

dissected (Fig. 19.6b), under guidance of ante- 19.4 Postoperative Management


rior OCT imaging. After pannus dissection, the
composite graft is attached to the cornea by 19.4.1 Early Postoperative Treatment
applying fibrin glue (Tissucol®) to the surface
of the cornea (Fig. 19.6c). The graft is cut out Preservative-free drops are the gold standard in
of the rings just beyond the limbs using a tre- the management of limbal stem cell patients. The
phine and m ­ icrosurgical scissors (Fig.  19.6d). cellular toxicity associated with preservatives
A secondary HAM is then placed over the graft includes a risk for both the stem cells and the sur-
with its epithelial side facing downwards. This face microenvironment [61]. Therefore eye drops
membrane serves as a biological patch offering containing preservatives should be avoided wher-
additional mechanical protection to the graft ever possible. We prescribe all patients with 0.3%
preventing primary dehiscence. The patch is ofloxacin drops 4×/day, 0.1% dexamethasone
secured with four 10–0 nylon sutures to the drops 8×/day and 20% autologous serum drops
conjunctiva and is removed 5–7  days after 16×/day for 1  month postoperatively and then
transplantation. To end, a bandage contact lens gradually taper the drops down. In cases the
is applied to the eye [58]. patient has any transmittable disease prohibiting

a b

c d

Fig. 19.6  Different0 steps during CLET: conjunctival using a trephine (not shown) and scissors (d), releasing
periotomy (a), dissection of the pannus tissue (b), applica- the graft using a 13 mm trephine in diameter and end the
tion of the graft using fibrin glue (c), releasing the graft trephination with Vanas scissors
19  Cultivated Limbal Stem Cell Transplantation: Indications and Technique 287

serum drops (e.g. HIV), these are replaced by eral it is accepted that a two-stage approach is pre-
preservative-free sodium hyaluronate eye drops. ferred above a combined procedure since the latter
Allogeneic limbal transplant patients need is associated with poorer clinical outcomes [62].
additional immunosuppressive treatment in order In our opinion certain ocular conditions necessi-
to prevent graft rejection. We use cyclosporine A tate a combined approach. As mentioned previ-
3 mg/kg/day which is commenced 1 week prior ously, in patients with a residual stroma of less
to transplantation and continued for 2  months than 350 μm as assessed by OCT imaging, a CLET
thereafter. Blood trough levels should be main- will be combined with either a full-thickness cor-
tained at approximately 100–150  ng/ml (unless neal transplant or a DALK in order to reduce the
renal or liver function impairment occurs) and risk on postoperative complications. In addition,
then tapered to 25–50 mg/kg/day and maintained evidence of a non-functional endothelium may be
at this dosage for a period of at least 1  year. If an indication to perform PK during stem cell sur-
required, perioperative IV methylprednisolone gery. It is important to note that every patient is
can be administered (1–2 mg/kg/day, depending different, and treatment should be adapted based
on the degree of vascularization of the cornea and on their individual needs and expectations.
limbus) and continued daily for 3 days. Depending In cases of graft failure, a second stem cell
on host inflammatory response, the corticosteroid transplantation can be offered to the patient. We
treatment can be prolonged orally if required, but limit treatment to a maximum of two transplanta-
it is generally recommended that the preoperative tions per patient. For safety reasons, we do not
inflammation is well controlled prior to surgery. harvest more than two biopsies from the same
donor eye. In the case of allotransplants, there is
also the possibility of host sensitization to the
19.4.2 Additional (Surgical) alloantigens, and as a result we also limit these
Management transplants to a maximum of two per patient.

The primary goal in limbal epithelial stem cell


transplantation is to restore the ocular surface. 19.5 Conclusion
While visual acuity may already improve follow-
ing CLET if the underlying stroma is clear, further LSCD is a complex disorder requiring an individu-
surgery is often required to restore corneal clarity alized approach and a step-by-step r­econstruction.
and improve visual outcomes (Fig. 19.7). In gen- Transplantation of ex  vivo cultivated limbal stem

a b

Fig. 19.7  Total limbal stem cell-deficient eye after an alkali burn before (a) and after a two-stage approach: 25 months
post-CLET; 9 months post-PK (b)
288 J. Behaegel et al.

cells can offer corneal rehabilitation in both unilat- 4. Ahmad S.  Concise review: limbal stem cell defi-
ciency, dysfunction, and distress. Stem Cells Transl
eral and bilateral (sub)total stem cell-deficient eyes. Med. 2012;1(2):110–5.
A comprehensive preoperative evaluation and care- 5. Strungaru MH, Mah D, Chan CC.  Focal limbal
ful preparation of the ocular surface are crucial to stem cell deficiency in turner syndrome. Cornea.
provide an optimal environment for the transplanted 2014;33(2):207–9.
6. Catt CJ, Hamilton GM, Fish J, Mireskandari K,
cells. A well-­prepared ocular surface together with Ali A.  Ocular manifestations of Stevens-Johnson
a rigorous postoperative management and follow- Syndrome and toxic epidermal necrolysis in children.
up will ensure the best chance on successful graft- Am J Ophthalmol. 2016;166:68–75.
ing. Despite the safety of the CLET technique, the 7. Aslan D, Akata R, Holm H, Vulliamy T, Dokal
I. Limbal stem cell deficiency in patients with inher-
treatment of LSCD remains challenging because of ited stem cell disorder of dyskeratosis congenita. Int
the heterogeneity of the population and associated Ophthalmol. 2012;32(6):615–22.
comorbidities. In case of graft failure, there is still 8. Eschle-Meniconi ME, Ahmad SR, Foster CS. Mucous
the possibility to perform a second transplantation. membrane pemphigoid: an update. Curr Opin
Ophthalmol. 2005;16(5):303–7.
9. Fernandes M, Sangwan VS, Vemuganti GK.  Limbal
Acknowledgement  We would like to acknowl- stem cell deficiency and xeroderma pigmentosum: a
edge the important input of Nadia Zakaria who case report. Eye. 2004;18(7):741–3.
initiated the culturing method as used in the clini- 10. Sivaraman KR, Jivrajka RV, Soin K, Bouchard CS,
Movahedan A, Shorter E, et  al. Superior limbic
cal trial and realised the GMP accreditation of the keratoconjunctivitis-like inflammation in patients
ophthalmology branch of the CCRG unit. She is with chronic graft-versus-host disease. Ocul Surf.
employee at Novartis Institute of Biomedical 2016;14(3):393–400.
Research since 2017. 11. Cortes M, Lambiase A, Sacchetti M, Aronni S, Bonini
S. Limbal stem cell deficiency associated with LADD
syndrome. Arch Ophthalmol. 2005;123(5):691.
Compliance with Ethical Requirements  12. Bobba S, Di Girolamo N, Mills R, Daniell M, Chan
Joséphine Behaegel, Sorcha Ní Dhubhghaill and E, Harkin DG, et al. Nature and incidence of severe
Marie-José Tassignon declare that they have no limbal stem cell deficiency in Australia and New
Zealand. Clin Exp Ophthalmol. 2017;45(2):174–81.
conflict of interest. 13. Messmer EM, Kenyon KR, Rittinger O, Janecke

AR, Kampik A.  Ocular manifestations of keratitis-­
Informed Consent  All procedures followed ichthyosis-­deafness (KID) syndrome. Ophthalmology.
were in accordance with the ethical standards of 2005;112(2):3–8.
14. Lichtinger A, Pe’er J, Frucht-Pery J, Solomon

the responsible committee on human experimen- A. Limbal stem cell deficiency after topical mitomy-
tation (University Hospital of Antwerp) and with cin C therapy for primary acquired melanosis with
the Helsinki Declaration of 1975, as revised in atypia. Ophthalmology. 2010;117(3):431–7.
2000. Informed consent was obtained from all 15. Nghiem-Buffet M-H, Gatinel D, Jacquot F, Chaine
G, Hoang-Xuan T.  Limbal stem cell deficiency fol-
patients for being included in the study. lowing phototherapeutic keratectomy. Cornea.
2003;22(5):482–4.
No animal studies were carried out by the 16. Schwartz GS, Holland EJ. Iatrogenic limbal stem cell
authors for this article. deficiency. Cornea. 1998;17(1):31–7. A.
17. Lim P, Fuchsluger TA, Jurkunas UV.  Limbal stem
cell deficiency and corneal neovascularization. Semin
Ophthalmol. 2009;24(3):139–48.
18. Gupta N, Sachdev R, Tandon R. Ocular surface squa-
References mous neoplasia in xeroderma pigmentosum: clini-
cal spectrum and outcome. Graefes Arch Clin Exp
1. Kruse FE. Stem cells and corneal epithelial regenera- Ophthalmol. 2011;249(8):1217–21. A.
tion. Eye. 1994;8:170–83. 19. Merchant A, Zhao T-Z, Foster CS.  Chronic kerato-
2. Puangsricharern V, Tseng SC.  Cytologic evidence conjunctivitis associated with congenital dyskeratosis
of corneal diseases with limbal stem cell deficiency. and erythrokeratodermia variablis. Ophthalmology.
Ophthalmology. 1995;102(10):1476–85. 1998;105(7):1286–91.
3. Le Q, Deng SX, Xu J.  In vivo confocal microscopy 20. Rossen J, Amram A, Milani B, Park D, Harthan J,
of congenital aniridia-associated keratopathy. Eye. Joslin C, et al. Contact Lens-induced limbal stem cell
2013;27(6):763–6. deficiency. Ocul Surf. 2016;14(4):419–34.
19  Cultivated Limbal Stem Cell Transplantation: Indications and Technique 289

21. Fdos P, Goncalves ED, Barros JN, Campos MS, Sato cell disease clinical features and management strate-
EH, Gomes JAP. Impression cytology findings in bul- gies. Ophthalmology. 2014;121(10):2053–8.
lous keratopathy. Br J Ophthalmol. 2010;94(6):773–6. 38. Dua HS. The conjunctiva in corneal epithelial wound
22. Dua HS, Saini JS, Azuara-Blanco A, Gupta P. Limbal healing. Br J Ophthalmol. 1998;82(12):1407–11.
stem cell deficiency: concept, aetiology, clinical 39. Dua HS.  Sequential sector conjunctival epitheliec-
presentation, diagnosis and management. Indian J tomy. In: Holland EJ, Mannis M, editors. Ocular
Ophthalmol. 2000;48(2):83–92. surface disease, medical and surgical management.
23. Djalilian AR, Holland EJ, Schwartz GS.  Limbal
New York: Springer; 2002. p. 168–74.
stem cell deficiency. Ophthalmology. 2003;110(10): 40. Waring GO, Bourne WM, Edelhauser HF, Kenyon
2071–2. KR. The corneal endothelium: normal and pathologic
24. Holland EJ, Djalilian AR, Schwartz GS. Management structure and function. Ophthalmology. 1982;89(6):
of aniridic keratopathy with keratolimbal allograft: 531–90.
a limbal stem cell transplantation technique. 41. Oliveira-Soto L, Efron N.  Morphology of cor-

Ophthalmology. 2003;110(1):125–30. neal nerves using confocal microscopy. Cornea.
25.
Javadi M-A, Baradaran-Rafii A.  Living-related 2001;20(4):374–84.
conjunctival-­limbal allograft for chronic or delayed- 42. Lambiase A, Manni L, Bonini S, Rama P, Micera A,
onset mustard gas keratopathy. Cornea. 2009;28(1): Aloe L. Nerve growth factor promotes corneal heal-
51–7. ing: structural, biochemical, and molecular analyses
26. Meallet MA, Espana EM, Grueterich M, Ti S-E, Goto of rat and human corneas. Invest Ophthalmol Vis Sci.
E, Tseng SC. Amniotic membrane transplantation with 2000;41(5):1063–9.
conjunctival limbal autograft for total limbal stem cell 43. Stern ME, Gao J, Siemasko KF, Beuerman RW,

deficiency. Ophthalmology. 2003;110(8):1585–92. Pflugfelder SC.  The role of the lacrimal functional
27. Pellegrini G, Traverso CE, Franzi AT, Zingirian M, unit in the pathophysiology of dry eye. Exp Eye Res.
Cancedda R, De Luca M.  Long-term restoration of 2004;78(3):409–16.
damaged corneal surfaces with autologous cultivated 44. Aguayo Bonniard A, Yeung JY, Chan CC, Birt

corneal epithelium. Lancet. 1997;349:990–3. CM.  Ocular surface toxicity from glaucoma topical
28. Behaegel J, Dhubhghaill SN, Koppen C, Zakaria
medications and associated preservatives such as ben-
N.  Safety of cultivated limbal epithelial stem cell zalkonium chloride (BAK). Expert Opin Drug Metab
transplantation for human corneal regeneration. Stem Toxicol. 2016;12(11):1279–89.
Cells Int. 2017;2017:1. 45. Rama P, Matuska S, Paganoni G, Spinelli A, De

29. Shortt AJ, Secker GA, Notara MD, Limb GA, Khaw Luca M, Pellegrini G.  Limbal stem-cell therapy
PT, Tuft SJ, et al. Transplantation of ex vivo cultured and long-term corneal regeneration. N Engl J Med.
limbal epithelial stem cells: a review of techniques and 2010;363(2):147–55.
clinical results. Surv Ophthalmol. 2007;52(5):483–502. 46. Cho P, Yap M. Schirmer test. I. A review. Optom Vis
30. Henderson TR, Coster DJ, Williams KA.  The long Sci. 1993;70(2):152–6.
term outcome of limbal allografts: the search for sur- 47. Sejpal K, Bakhtiari P, Deng SX.  Presentation, diag-
viving cells. Br J Ophthalmol. 2001;85(5):604–9. nosis and management of limbal stem cell deficiency.
31. Daya SM, Watson A, Sharpe JR, Giledi O, Rowe Middle East Afr J Ophthalmol. 2013;20(1):5–10.
A, Martin R, et  al. Outcomes and DNA analysis of 48. Haagdorens M, Behaegel J, Rozema J, Van Gerwen
ex vivo expanded stem cell allograft for ocular surface V, Michiels S, Ní Dhubhghaill S, et al. A method for
reconstruction. Ophthalmology. 2005;112(3):470–7. quantifying limbal stem cell niches using OCT imag-
32. Sharpe JR, Daya SM, Dimitriadi M, Martin R,
ing. Br J Ophthalmol. 2017;101(9):1250.
James SE.  Survival of cultured allogeneic limbal 49. Zakaria N, Ni Dhubhghaill S, Taal M, Berneman Z,
epithelial cells following corneal repair. Tissue Eng. Koppen C, Tassignon M-J. Optical coherence tomog-
2007;13(1):123–32. raphy in cultivated limbal epithelial stem cell trans-
33. Dua HS, Miri A, Said DG. Contemporary limbal stem plantation surgery. J Ophthalmol. 2015;4(6):339–45.
cell transplantation – a review. Clin Exp Ophthalmol. 50. Villani E, Baudouin C, Efron N, Hamrah P, Kojima
2010;38(2):104–17. T, Patel SV, et al. In vivo confocal microscopy of the
34. Tseng SCG, Tsubota K.  Important concepts for
ocular surface: from bench to bedside. Curr Eye Res.
treating ocular surface and tear disorders. Am J 2014;39(3):213–31.
Ophthalmol. 1997;124(6):825–35. 51. Miri A, Al-Aqaba M, Otri AM, Fares U, Said DG,
35. Knop E, Korb D, Blackie C, Knop N. The lid margin Faraj LA, et al. In vivo confocal microscopic features
is an underestimated structure for preservation of ocu- of normal limbus. Br J Ophthalmol. 2012;96(4):
lar surface health and development of dry eye disease. 530–6.
Dev Ophthalmol. 2010;45:108–22. 52. Patel DV, Sherwin T, McGhee CNJ.  Laser scanning
36. Willcox MDP, Argüeso P, Georgiev GA, Holopainen in  vivo confocal microscopy of the normal human
JM, Laurie GW, Millar TJ, et al. TFOS DEWS II tear corneoscleral limbus. Investig Ophthalmol Vis Sci.
film report. Ocul Surf. 2017;15(3):366–403. 2006;47(7):2823–7.
37. Kim BY, Riaz KM, Bakhtiari P, Chan CC, Welder JD, 53. Messmer EM, Mackert MJ, Zapp DM, Kampik A. In
Holland EJ, et  al. Medically reversible limbal stem vivo confocal microscopy of pigmented ­conjunctival
290 J. Behaegel et al.

tumors. Graefes Arch Clin Exp Ophthalmol. epithelial stem cell graft generation and transplanta-
2006;244(11):1437–45. tion. Tissue Eng Part C Methods. 2010;16(5):921–7.
54. Nubile M, Lanzini M, Miri A, Pocobelli A, Calienno 59.
Hopkinson A, Shanmuganathan VA, Gray T,
R, Curcio C, et  al. In vivo confocal microscopy Yeung AM, Lowe J, James DK, et al. Optimization
in diagnosis of limbal stem cell deficiency. Am J of amniotic membrane (AM) denuding for tis-
Ophthalmol. 2013;155(2):220–32. sue engineering. Tissue Eng Part C Methods.
55. Miri A, Alomar T, Nubile M, Al-aqaba M, Lanzini 2008;14(4):371–81.
M, Fares U, et al. In vivo confocal microscopic find- 60. Hopkinson A, McIntosh RS, Tighe PJ, James DK,
ings in patients with limbal stem cell deficiency. Br J Dua HS.  Amniotic membrane for ocular surface
Ophthalmol. 2012;96(4):523–9. reconstruction: donor variations and the effect of han-
56. Deng SX, Sejpal K, Tang Q, Aldave AJ, Lee OL, dling on TGF-beta content. Invest Ophthalmol Vis
Yu F.  Characterization of limbal stem cell defi- Sci. 2006;47(10):4316–22.
ciency by in  vivo laser scanning confocal micros- 61. Baudouin C, Labbé A, Liang H, Pauly A. Preservatives
copy: a microstructural approach. Arch Ophthalmol. in eyedrops: The good, the bad and the ugly. Prog
2012;130(4):440–5. Retin Eye Res. 2010;29(4):312–34.
57. Zhao Y, Ma L. Systematic review and meta-analysis 6 2. Basu S, Mohamed A, Chaurasia S, Sejpal K,

on transplantation of ex vivo cultivated limbal epithe- Vemuganti GK, Sangwan VS.  Clinical out-
lial stem cell on amniotic membrane in limbal stem comes of penetrating keratoplasty after autolo-
cell deficiency. Cornea. 2015;34(5):592–600. gous cultivated limbal epithelial transplantation
58. Zakaria N, Koppen C, Van Tendeloo V, Berneman Z, for ocular surface burns. Am J Ophthalmol.
Hopkinson A, Tassignon M-J.  Standardized limbal 2011;152(6):917–924.e1.
Optimizing the Ocular Surface
for Regenerative Surgery:
20
What Is Important and What Is
Essential for the Outcome

Kai B. Kang and Ali R. D’jalilian

Summary • Eyelid abnormalities should also be


• Approaches to the optimization of ocu- repaired to fix exposure and misdirected
lar surface health for regenerative sur- lashes prior to regenerative surgery.
gery include improving the quality of
the tear film, controlling of local and
systemic inflammation, and repairing
eyelid abnormalities. 20.1 Introduction
• Various tear supplements, hemoderiva-
tives, treatment for Meibomian gland The preoperative planning for regenerative sur-
dysfunction, usage of contact lenses, gery requires the optimization of patients’ ocular
amniotic membrane, and punctal occlu- surface health. A dry and hostile surface environ-
sion as well as reduction in the toxicity ment will lead to failure of regenerative surgery.
of glaucoma drops are essential in In this chapter, we outline several important
improving tear film health. approaches the surgeon should take to maximize
• It is crucial to control local and system the ocular surface health of patients. These
inflammation prior to regenerative sur- approaches include optimization of the tear film,
gery. Approaches to the control of local control of local and systemic inflammation, and
inflammation may include the use of topi- repair of eyelid abnormalities.
cal steroids, cyclosporine, tacrolimus, and Figure 20.1 shows key components in the
lifitegrast. Approaches to the control of optimization of ocular surface health.
systemic inflammation include various
steroids and steroid-sparing agents.
20.2 Improving Tear Film Health

20.2.1 Tear Supplementation


Kai B. Kang and Ali R. D’jalilian declare that they have
no conflict of interest. No human or animal studies were To improve the success of regenerative surgery, it
carried out by the authors for this article. is important to return the tear film and the ocular
K. B. Kang · A. R. D’jalilian (*) surface to the normal homeostatic state by opti-
University of Illinois at Chicago, Department of mizing the quality of the tear film. Several steps
Ophthalmology and Visual Sciences, may be taken to improve tear quality depending
Chicago, IL, USA on the root cause. One important step is to
e-mail: adjalili@uic.edu

© Springer Nature Switzerland AG 2019 291


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_20
292 K. B. Kang and A. R. D’jalilian

Optimization of ocular surface

Improving tear film health Controlling inflammation Repairing eyelid abnormalities

1. Local inflammation
1. Tear supplements 1. Exposure
a. Steroids
2. MGD treatment 2. Misdirected lashes
b. Cyclosporine
3. Autologous serum
c. Lifitegrast
4.Soft and scleral lenses
2. Systemic inflammation
5. Punctal occlusion
a. Steroids
6. Minimizing drop toxicity
b. Steroid-sparing agents

Amniotic membrane

Fig. 20.1  Key components in the optimization of ocular surface health

s­upplement tear with lubricants, which include Several antibiotics with anti-inflammatory
artificial tears, ointments, and gels. A dry ocular property are also available in the management of
surface may cause further osmotic stress and con- Meibomian gland disease [9]. These antibiotics
sequent elevation in electrolyte concentration include tetracyclines, e.g., doxycycline and
leading to activation of cellular stress signaling minocycline, and macrolides, e.g., azithromycin.
pathways and cellular damage [1]. Ocular lubri- Several studies comparing the effectiveness of
cants may provide both electrolytes and hypoto- topical or oral azithromycin to oral doxycycline
nicity to the tear film in order to overcome these in the management of Meibomian gland dysfunc-
problems. For regenerative surgery, preservative-­ tion found that despite different mechanisms of
free lubricants are the best as preservatives action, either antibiotics relieved the signs and
(detergent or oxidative) may further irritate and symptoms and restored normal lipid properties of
damage epithelial cells and decrease cellular Meibomian gland secretions [10, 11]. However,
survival. azithromycin may stimulate Meibomian gland
epithelial cell differentiation and may be prefer-
able in certain instances as a short 5-day course
20.2.2 MGD and Blepharitis showed overall better clinical response compared
Treatment to a 1-month treatment with doxycycline [12].

In addition, many therapeutic options are avail-


able and essential in improving Meibomian gland 20.2.3 Hemoderivatives
health. For example, omega-3 fatty acid supple-
mentation (e.g., from fish oil and flaxseed oil) has Various hemoderivatives produced from the patient’s
been shown to alter the composition of Meibomian own blood are valuable options to help in optimizing
gland lipids and decrease pro-­inflammatory medi- the ocular surface health for patient undergoing
ators, e.g., IL-1 and TNF-­alpha [2, 3]. Clinically, regenerative surgery. These include autologous
omega-3 fatty acid supplementation showed serum, platelet-rich plasma, and plasma rich in
improved tear production, TBUT, Schirmer score, growth factors (PRGF). Hemoderivatives are made
and ocular surface disease index [2, 4–8]. preservative-free and provide natural ­substitutes for
20  Optimizing the Ocular Surface for Regenerative Surgery: What Is Important and What Is Essential… 293

many proteins, vitamins, and lipids active in the example, the Boston scleral prosthetic device
human tear. Autologous serum drops have been (Prosthetic Replacement of the Ocular Surface
shown to suppress apoptosis in the conjunctival and Ecosystem, PROSE, Boston, MA, USA) allows
corneal epithelium [13]. Concentrations between for a highly customizable device that precisely
20% and 100% of serum tears have been used. fits the curvature of the patient’s eye and may
Recently, a study of long-term outcomes of 50% be beneficial especially in the regenerative sur-
serum tears demonstrated improvement in fluores- gery patient population with highly irregular
cein staining, Schirmer scores, and ocular surface ocular surface [20].
index scores [14]. Autologous serum tears are espe-
cially important in patients with severe aqueous tear
deficiency and deficiency in reflex tearing. This 20.2.5 Punctal Occlusion
includes patients with scarring of the lacrimal glands
as seen in Stevens-Johnson syndrome, mucous Another approach to the management of a dry
membrane pemphigoid, or following alkali or ther- ocular surface due to tear insufficiency is punctal
mal injuries. Recent studies have also shown that occlusion. Either intracanalicular placement of
platelet-rich plasma and plasma rich in growth fac- punctal plugs or placement at the punctal level is
tors can be superior than autologous serum espe- available. We recommend a trial with temporary
cially when used in patients with persistent epithelial punctal plugs (e.g., collagen plugs) first, and if
defects [15]. proven effective for the specific patient, the clini-
cian may consider the placement of permanent
plugs (e.g., silicone or acrylic plugs).
20.2.4 Soft and Scleral Contact
Lenses
20.2.6 Minimizing Drop Toxicity
Therapeutic soft contact lenses are often used
to protect and hydrate a dry corneal surface. It is also important to minimize the toxic effects
Soft contact lenses often provide immediate of topical agents in preparation for patients
relief of symptoms in patients with severe dry undergoing regenerative surgery. Many patients
eyes and help to protect the ocular surface using glaucoma medications experience toxic
against mechanical trauma. For example, in conjunctivitis after chronic use. For example,
patients suffering from lid margin keratiniza- topical beta-adrenergic blockers such as timolol
tion, friction from blinking may cause damage have been associated with punctate epithelial
to the ocular surface. One way to prevent cor- keratopathy and epithelial erosions [21]; dorzol-
neal damage is to use a bandage contact lens amide has been associated with mucopurulent
chronically. To achieve a better contact lens fit, sterile conjunctivitis [22]; and brimonidine can
the clinician may consider a larger diameter commonly cause allergic blepharoconjunctivitis
Kontur lens. Additionally, scleral lenses are and follicular conjunctivitis [23]. For glaucoma
typically fluid-filled and vault over the cornea. patients who are on multiple drops, the clinician
They rest over the limbus and provide the ocu- may consider surgery or oral agents to minimize
lar surface with constant lubrication. These drop toxicity.
lenses may protect the ocular surface by mini-
mizing mechanical and frictional forces on the
epithelium and maintaining a constant tear film 20.2.7 Amniotic Membrane
over the epithelium. Studies have shown that
scleral lenses improved patient comfort and Amniotic membranes were first used in skin
ocular surface health in those with severe ocu- transplants to treat ulcers and burns [24]. The
lar surface diseases [16–20]. Different options basement membrane of amniotic membrane has
of scleral lenses are currently available. For collagen and laminin structures that are similar to
294 K. B. Kang and A. R. D’jalilian

the conjunctiva [24]. It has a basement membrane mia, vascular congestion, edema, and pain
that facilitates epithelial cell migration, adhesion, associated with acute inflammation. They also
proliferation, and differentiation [25–27]. The suppress fibroblast proliferation, collagen depo-
basement membrane of amniotic membrane also sition, and scarring associated with late inflam-
has anti-inflammatory, antiangiogenic, and anti-­ matory response. For short-term or pulse therapy,
fibrotic properties. These properties make amni- one may use commercial available steroids
otic membrane a great substrate to promote (dexamethasone, prednisolone acetate, lotepred-
ocular surface healing. nol, difluprednate, fluorometholone, etc.).
There are two major types of amniotic mem- Preservative-free forms, e.g., methylpredniso-
branes: cryopreserved and dehydrated. lone 1%, also act as good alternatives to commer-
PROKERA biological bandage (Bio-Tissue, cially available steroid preparations and are less
Doral, Florida) is held in place with a PMMA toxic to the epithelium [29, 30].
ring. It has anti-inflammatory effect to quiet
acute phase ocular surface inflammation. In Cyclosporine/Tacrolimus
patients who cannot tolerate PROKERA’s Calcineurin inhibitors act to reduce pro-­
PMMA ring, or in those with glaucoma filtering inflammatory cytokines and activated lympho-
blebs or tubes that make fitting difficult, dehy- cytes [31]. Prepared in a glycerin, castor oil, and
drated amniotic grafts, e.g., AmbioDisk (IOP polysorbate 80 emulsion, the vehicle for the
Ophthalmics, Costa Mesa, California), may topical preparation cyclosporine 0.05% itself
work [28]. This dehydrated form is placed on the provides good lubrication to the ocular surface.
ocular surface and then covered with a bandage Twice-daily dosing with this agent has negligi-
contact lens. Its use is indicated in patient with ble systemic absorption and side effects. A recent
severely dry ocular surface. In addition, sutured systemic review on efficacy of cyclosporine
surgical amniotic membrane grafts are often 0.05% showed lower ocular surface disease
used in conjunction with regenerative surgery, index scores, longer tear film breakup time,
e.g., KLAL in inflamed eyes, as it has been improved Schirmer I scores, reduced corneal
shown to suppress inflammation and facilitate fluorescein staining, and higher goblet cell den-
epithelization. sities [31]. Although it is known that this agent
may only benefit a subgroup of patient, it contin-
ues to act as an important nonsteroidal agent in
20.3 Controlling Inflammation the management of ocular surface disease.
Tacrolimus is another calcineurin inhibitor, and
Another major step in optimizing patients’ ocular the topical preparation of tacrolimus 0.03% and
surface is to control local and systemic inflamma- other concentrations has shown reduced side
tion. Regenerative surgery will not be successful effect profile and good efficacy in control of
in an overall inflamed eye. Many agents are avail- inflammation, especially in patients with atopic
able to help control inflammation. diseases [32].

Lifitegrast
20.3.1 Local Inflammation Topical lifitegrast is one of newest addition to the
current management regimen to reduce ocular
Steroids surface inflammation. It works to block lympho-
Topical glucocorticoids are still the most effec- cyte function-associated antigen-intracellular
tive and fast-acting therapy to suppress inflam- adhesion molecule-1 interaction, thus decreasing
mation on the ocular surface. Glucocorticoids T-cell recruitment [33]. We hope that in the
work to suppress arachidonic acid release from future, we will gain more experience and conduct
phospholipids. At the tissue level, glucocorti- studies that look at the efficacy of this agent for
coids may prevent or suppress local hyperther- various disease processes.
20  Optimizing the Ocular Surface for Regenerative Surgery: What Is Important and What Is Essential… 295

20.3.2 Systemic Inflammation class of immunosuppressive agent is sirolimus


(rapamycin), a macrolide immunosuppressant
Steroids that prolongs the cell cycle by inhibiting mTOR,
Uncontrolled ocular surface inflammation is a which acts on phosphorylation of many cell
threat to any regenerative surgery. Systemic cycle-dependent kinases. It is important to moni-
immunosuppressive therapy is often necessary to tor serum lipids and blood counts periodically in
further suppress inflammation prior to regenera- patients who use sirolimus, as hyperlipidemia,
tive surgery. Oral steroid is still the most rapid leukopenia, and thrombocytopenia may develop
and effective systemic immunosuppressant avail- [38]. In addition, various biological drugs that
able. Due to its adverse effects when used in the target specific components of the immune sys-
long term, including osteoporosis, hypertension, tem may help patients who fail traditional immu-
hyperglycemia, weight changes, and mood nosuppressive agents. For example, TNF
changes, only short courses of steroids are rec- inhibitors such as infliximab (Remicade) or
ommended prior to surgery to further decrease monoclonal CD-20 antibody rituximab (Rituxan)
ocular surface inflammation. have been shown to be effective in controlling
ocular inflammation.
Steroid-Sparing Immunosuppression
In the preparation for regenerative surgery, ade-
quate control of systemic inflammation in 20.4 Repair Lid Abnormalities
patients with autoimmune diseases such as
mucous membrane pemphigoid and Stevens-­ Abnormalities in the eyelids can lead to disrup-
Johnson syndrome is essential. Long-term tion of ocular surface health as normal eyelid
immunosuppression can be achieved in these blinking, apposition, and closure are essential for
patients with steroid-sparing agents. It is recom- ocular surface protection. The ophthalmologist
mended that these patients be managed in col- often encounters abnormalities in eyelid struc-
laboration with a local organ transplant team tures and functions especially in patients under-
who has a broad experience with available going regenerative surgeries as these patients
­immunosuppressive agents. Examples of avail- often suffer from cicatricial conditions or have a
able immunosuppressive agents include calci- history of chemical or thermal injuries. Prior to
neurin inhibitors which mainly act on T cells by any regenerative surgery, it is essential to repair
inhibiting transcription of lymphokines, includ- eyelid abnormalities in order to provide a pro-
ing IL-2 [34]. Cyclosporine A (CsA) and tacroli- tected and wet ocular surface.
mus (Prograf) are the most commonly used
agents. Adverse effects of calcineurin inhibitors
include hypertension, nephrotoxicity, neurotox- 20.4.1 Exposure
icity, hyperglycemia, and hepatotoxicity [34,
35]. One advantage of tacrolimus over CsA is Several eyelid abnormalities may cause chronic
that hirsutism and gingival hyperplasia are not exposure of the ocular surface. For example,
seen [34]. Another class of immunosuppressive patients may suffer from lagophthalmos due to
agents is antimetabolite that blocks the prolifera- paralysis from facial nerve palsy, ectropion, and
tion of dividing cells, including azathioprine malpositioned and keratinized lid margins.
(Imuran) and mycophenolate mofetil (CellCept). Lagophthalmos can cause chronic desiccation of
Complete blood counts and liver function tests the cornea; thus it is essential to repair lagoph-
should be monitored on a regular basis for thalmos. The management of lagophthalmos
patient on these medications as myelosuppres- depends on the underlying cause and requires the
sive effects can be observed in these patients [36, participation of the oculoplastic surgeon. For
37]. Mycophenolate mofetil is often used as part example, there are various causes for lagophthal-
of a triple therapy with CsA and steroids. A third mos due to ectropion. Involutional ectropion,
296 K. B. Kang and A. R. D’jalilian

caused by horizontal lid laxity, requires a lateral 3. Lemp MA. Advances in understanding and managing
dry eye disease. Am J Ophthalmol. 2008;146(3):350–
tarsal strip procedure [39]. Paralytic ectropion 6. https://doi.org/10.1016/j.ajo.2008.05.016.
secondary to facial nerve palsy may require a lat- 4. Brignole-Baudouin F, Baudouin C, Aragona P,
eral tarsorrhaphy. Lagophthalmos from facial et  al. A multicentre, double-masked, randomized,
nerve palsy may require the placement of gold or controlled trial assessing the effect of oral supple-
mentation of omega-3 and omega-6 fatty acids on a
platinum weight [40]. Cicatricial ectropion may conjunctival inflammatory marker in dry eye patients.
require scar revision and anterior lamella length- Acta Ophthalmol. 2011;89(7):e591–7. https://doi.
ening with skin grafts [41]. org/10.1111/j.1755-3768.2011.02196.x.
5. Deinema LA, Vingrys AJ, Wong CY, Jackson DC,
Chinnery HR, Downie LE.  A randomized, double-­
masked, placebo-controlled clinical trial of two forms
20.4.2 Misdirected Lashes of omega-3 supplements for treating dry eye disease.
Ophthalmology. 2017;124(1):43–52. https://doi.
Misdirected lashes due to entropion, trichiasis, org/10.1016/j.ophtha.2016.09.023.
6. Epitropoulos AT, Donnenfeld ED, Shah ZA,
and distichiasis may cause mechanical injury to et  al. Effect of oral re-esterified omega-3 nutri-
the conjunctival and corneal epithelium, lead- tional supplementation on dry eyes. Cornea.
ing to a chronic inflammatory state. It is essen- 2016;35(9):1185–91. https://doi.org/10.1097/
tial to reduce misdirected lashes prior to ICO.0000000000000940.
7. Georgakopoulos CD, Makri OE, Pagoulatos D, et al.
regenerative surgery. Long-term management Effect of omega-3 fatty acids dietary supplementation
plan including cryotherapy, argon laser abla- on ocular surface and tear film in diabetic patients
tion, full-thickness wedge resections, and repair with dry eye. J Am Coll Nutr. 2017;36(1):38–43.
of eyelid entropion would be beneficial prior to https://doi.org/10.1080/07315724.2016.1170643.
8. Sheppard JD, Singh R, McClellan AJ, et al. Long-term
regenerative surgery. supplementation with n-6 and n-3 PUFAs improves
moderate-to-severe keratoconjunctivitis sicca: a
randomized double-blind clinical trial. Cornea.
20.5 Conclusion 2013;32(10):1297–304. https://doi.org/10.1097/
ICO.0b013e318299549c.
9. Thulasi P, Djalilian AR. Update in current diagnostics
One of the first steps in regenerative surgeries is and therapeutics of dry eye disease. Ophthalmology.
to ensure the presence of a healthy ocular surface 2017;124(11S):S27–33. https://doi.org/10.1016/j.
to allow for the long-term success of surgeries. It ophtha.2017.07.022.
10. Foulks GN, Borchman D, Yappert M, Kakar S. Topical
is essential to avoid a dry and hostile surface azithromycin and oral doxycycline therapy of meibo-
environment. In this chapter, we outline various mian gland dysfunction: a comparative clinical and
approaches the surgeon should take to optimize spectroscopic pilot study. Cornea. 2013;32(1):44–53.
the ocular surface health. Many medications are https://doi.org/10.1097/ICO.0b013e318254205f.
11. Kashkouli MB, Fazel AJ, Kiavash V, Nojomi M,

available in our management armamentarium; Ghiasian L. Oral azithromycin versus doxycycline in
however, the surgeons should individualize the meibomian gland dysfunction: a randomised double-­
treatment course based on each patient’s disease masked open-label clinical trial. Br J Ophthalmol.
mechanism. 2015;99(2):199–204. https://doi.org/10.1136/
bjophthalmol-2014-305410.
12. Liu Y, Kam WR, Ding J, Sullivan DA.  Can tetracy-
cline antibiotics duplicate the ability of azithromycin
Bibliography to stimulate human meibomian gland epithelial cell
differentiation? Cornea. 2015;34(3):342–6. https://
1. Dry Eye WorkShop. Management and therapy of doi.org/10.1097/ICO.0000000000000351.
dry eye disease: report of the management and 13. Kojima T, Higuchi A, Goto E, Matsumoto Y,

therapy subcommittee of the International Dry Eye Dogru M, Tsubota K.  Autologous serum eye drops
WorkShop. Ocul Surf. 2007;5(2):163–78. for the treatment of dry eye diseases. Cornea.
2. Pinna A, Piccinini P, Carta F.  Effect of oral linoleic 2008;27(Suppl 1):S25–30. https://doi.org/10.1097/
and gamma-linolenic acid on meibomian gland dys- ICO.0b013e31817f3a0e.
function. Cornea. 2007;26(3):260–4. https://doi. 14. Hussain M, Shtein RM, Sugar A, et al. Long-term use
org/10.1097/ICO.0b013e318033d79b. of autologous serum 50% eye drops for the treatment
20  Optimizing the Ocular Surface for Regenerative Surgery: What Is Important and What Is Essential… 297

of dry eye disease. Cornea. 2014;33(12):1245–51. 29. Marsh P, Pflugfelder SC.  Topical nonpreserved

https://doi.org/10.1097/ICO.0000000000000271. methylprednisolone therapy for keratoconjuncti-
15. Soni NG, Jeng BH.  Blood-derived topical ther-
vitis sicca in Sjögren syndrome. Ophthalmology.
apy for ocular surface diseases. Br J Ophthalmol. 1999;106(4):811–6. https://doi.org/10.1016/
2016;100(1):22–7. https://doi.org/10.1136/ S0161-6420(99)90171-9.
bjophthalmol-2015-306842. 30. Hong S, Kim T, Chung S-H, Kim EK, Seo

16. La Porta Weber S, Becco de Souza R, Gomes JÁP, KY.  Recurrence after topical nonpreserved meth-
Hofling-Lima AL. The use of the esclera scleral con- ylprednisolone therapy for keratoconjunctivitis
tact lens in the treatment of moderate to severe dry sicca in Sjögren’s syndrome. J Ocul Pharmacol
eye disease. Am J Ophthalmol. 2016;163:167–173.e1. Ther. 2007;23(1):78–82. https://doi.org/10.1089/
https://doi.org/10.1016/j.ajo.2015.11.034. jop.2006.0091.
17. Alipour F, Kheirkhah A, Jabarvand BM. Use of mini 31. Wan KH, Chen LJ, Young AL. Efficacy and safety of
scleral contact lenses in moderate to severe dry eye. topical 0.05% cyclosporine eye drops in the treatment
Cont Lens Anterior Eye. 2012;35(6):272–6. https:// of dry eye syndrome: a systematic review and meta-­
doi.org/10.1016/j.clae.2012.07.006. analysis. Ocul Surf. 2015;13(3):213–25. https://doi.
18. Carracedo G, Blanco MS, Martin-Gil A, Zicheng W, org/10.1016/j.jtos.2014.12.006.
Alvarez JC, Pintor J. Short-term effect of scleral lens 32.
Shoughy SS.  Topical tacrolimus in ante-
on the dry eye biomarkers in keratoconus. Optom rior segment inflammatory disorders. Eye Vis
Vis Sci. 2016;93(2):150–7. https://doi.org/10.1097/ (Lond). 2017;4:7. https://doi.org/10.1186/
OPX.0000000000000788. s40662-017-0072-z.
19. Harthan J, Nau CB, Barr J, et al. Scleral lens prescrip- 33. Holland EJ, Luchs J, Karpecki PM, et al. Lifitegrast
tion and management practices: the SCOPE study. for the treatment of dry eye disease: results of a
Eye Contact Lens. 2017. https://doi.org/10.1097/ phase III, randomized, double-masked, placebo-
ICL.0000000000000387. controlled trial (OPUS-3). Ophthalmology.
20.
Heur M, Bach D, Theophanous C, Chiu 2017;124(1):53–60. https://doi.org/10.1016/j.
GB.  Prosthetic replacement of the ocular surface ophtha.2016.09.025.
ecosystem scleral lens therapy for patients with ocu- 34.
Vanrenterghem YF.  Which calcineurin inhibi-
lar symptoms of chronic Stevens-Johnson syndrome. tor is preferred in renal transplantation: tacrolimus
Am J Ophthalmol. 2014;158(1):49–54. https://doi. or cyclosporine? Curr Opin Nephrol Hypertens.
org/10.1016/j.ajo.2014.03.012. 1999;8(6):669–74.
21. Van Buskirk EM.  Adverse reactions from timolol
35. Jain A, Nalesnik M, Reyes J, et  al. Posttransplant
administration. Ophthalmology. 1980;87(5):447–50. lymphoproliferative disorders in liver transplantation:
22. Schnyder CC, Tran VT, Mermoud A, Herbort
a 20-year experience. Ann Surg. 2002;236(4):429–
CP.  Sterile mucopurulent conjunctivitis associ- 36; discussion 436–437. https://doi.org/10.1097/01.
ated with the use of dorzolamide eyedrops. Arch SLA.0000033429.89424.F8.
Ophthalmol. 1999;117(10):1429–31. 36. Liang L, Sheha H, Tseng SCG. Long-term outcomes
23. Katz LJ.  Twelve-month evaluation of brimonidine-­ of keratolimbal allograft for total limbal stem cell
purite versus brimonidine in patients with glaucoma or deficiency using combined immunosuppressive
ocular hypertension. J Glaucoma. 2002;11(2):119–26. agents and correction of ocular surface deficits. Arch
24. John T.  Human amniotic membrane transplantation: Ophthalmol. 2009;127(11):1428–34. https://doi.
past, present, and future. Ophthalmol Clin N Am. org/10.1001/archophthalmol.2009.263.
2003;16(1):43–65, vi. 37. European Mycophenolate Mofetil Cooperative Study
25. Rahman I, Said DG, Maharajan VS, Dua HS. Amniotic Group. Mycophenolate mofetil in renal transplanta-
membrane in ophthalmology: indications and limita- tion: 3-year results from the placebo-controlled trial.
tions. Eye (Lond). 2009;23(10):1954–61. https://doi. Transplantation. 1999;68(3):391–6.
org/10.1038/eye.2008.410. 38. Kruh J, Foster CS.  Corticosteroid-sparing agents:

26. Optimizing the ocular surface with amniotic mem- conventional systemic immunosuppressants.
brane therapy|Ophthalmology Magazine. https:// Dev Ophthalmol. 2012;51:29–46. https://doi.
www.eyeworld.org/optimizing-ocular-surface-amni- org/10.1159/000336185.
otic-membrane-therapy. Accessed 1 Jan 2018. 39. Jordan DR, Anderson RL.  The lateral tarsal strip

27. Paolin A, Cogliati E, Trojan D, et al. Amniotic mem- revisited. The enhanced tarsal strip. Arch Ophthalmol.
branes in ophthalmology: long term data on transplan- 1989;107(4):604–6.
tation outcomes. Cell Tissue Bank. 2016;17(1):51–8. 40. Reanimation of the paretic eyelid using gold weight
https://doi.org/10.1007/s10561-015-9520-y. implantation. A new approach and prospective evalua-
28. Vlasov A, Sia RK, Ryan DS, et al. Sutureless cryopre- tion. PubMed – NCBI. https://www.ncbi.nlm.nih.gov/
served amniotic membrane graft and wound healing pubmed/?term=gilbard+paretic+eyelid+god+weight.
after photorefractive keratectomy. J Cataract Refract Accessed 31 Dec 2017.
Surg. 2016;42(3):435–43. https://doi.org/10.1016/j. 41. Brown BZ, Beard C. Split-level full-thickness eyelid
jcrs.2015.11.045. graft. Am J Ophthalmol. 1979;87(3):388–92.
Stem Cell Spheres for Corneal
Regeneration
21
Salim Ismail, Jennifer J. McGhee, Ye Li,
Jeremy John Mathan, Jinny Jung Yoon,
Himanshu Wadhwa, Stephanie U-Shane Huang,
and Trevor Sherwin

21.1 S
 tem Cell Spheres acterization of the behaviour and abilities of cul-
for Corneal Repair tured stem cell-enriched spheres and their true
potential as a stem cell therapy.
Limbal cells isolated from human corneoscleral
tissue and cultured in vitro as stem cell-enriched
spheres have incredible potential as transplant- 21.2 L
 imbal Stem Cells/
able elements for corneal repair and regeneration. Keratocyte Progenitor Cells/
Our laboratory has championed the use of mixed Stromal Stem Cells
populations of cells isolated from the limbus as
transplantable elements capable of reforming and The human eye is one of a number of organs
restoring the limbal niche. However, as these within the body that harbour a repository of adult
spheres inherently contain mixed cell types of stem cells. Although there is increasing evidence
different lineages in addition to the stem cells, that these adult stem cells may possess the ability
their characterization and regenerative abilities to differentiate into other cell types, their main
need to be carefully assessed. We have extensive purpose is considered to be the long-term mainte-
experience of both in vitro and in situ evaluation nance of the tissues that make up the eye.
of these spheres in our laboratory. In this chapter, Within the cornea, the limbal stem cell (LSC)
we discuss the differing cell types that make up has long been accepted as the driver of corneal sur-
our spheres and the roles they each play in the face repair and maintenance. Their identification
onset and treatment of corneal dystrophies within however has proved difficult due to the lack of a
human tissue. We also outline our extensive char- single definitive limbal stem cell marker or even a
definitive panel of markers [1]. The well-­
characterized attributes of limbal stem cells are that
S. Ismail · J. J. McGhee Y. Li · J. J. Mathan they first and foremost reside at the limbus [2], the
J. J. Yoon · H. Wadhwa S. U.-S. Huang
T. Sherwin (*) region at the periphery of the transparent cornea
Department of Ophthalmology, New Zealand and the opaque sclera. They are quiescent and slow
National Eye Centre, Faculty of Medical and Health cycling but, when required, have the ability to
Sciences, The University of Auckland, respond quickly and positively to ensure mainte-
Auckland, New Zealand
e-mail: s.ismail@auckland.ac.nz; nance of vision. Within the human limbal region,
c.mcghee@auckland.ac.nz; j.mcghee@auckland.ac.nz; LSCs have been reported to exist in the palisades of
yli912@aucklanduni.ac.nz; Vogt [3, 4] and limbal epithelial crypts [5]. During
stephanie.huang@hudson.org.au;
normal homoeostatic progression, LSCs give rise
t.sherwin@auckland.ac.nz

© Springer Nature Switzerland AG 2019 299


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_21
300 S. Ismail et al.

to transient amplifying cells which migrate centrip- the cornea, the limbus contains a niche [15]
etally along the basal epithelium and differentiate which perpetuates the regeneration and repair
to replace lost epithelial cells [6]. Therefore the pri- ability of these stem and progenitor cells [16, 17]
mary purpose of LSCs is considered to be long- and houses a host of other cell types including
term maintenance of the corneal surface. melanocytes, immune cells (Langerhans cells,
Corneal stromal stem cells have proven even macrophages), vascular endothelial cells and lim-
more difficult to identify and isolate than corneal bal stromal fibroblasts [18], all or some of which
epithelial stem cells. Despite the stroma making may play a role in preserving the niche through-
up 90% of the thickness of the cornea, the major- out adulthood.
ity of its structure is composed of collagen
fibrils, glycoproteins and proteoglycans with a
consistent distribution of differentiated but qui- 21.3 Limbal Stem Cell Deficiency
escent keratocytes [7]. In 2005, a report demon-
strated the presence of a side population (SP) of Limbal stem cell deficiency (LSCD) arises when
cells from within the corneal peripheral stroma the normal population of limbal stem cells are
that appeared to possess stem cell properties either partially or completely destroyed and/or
similar to that of limbal epithelial stem cells and become dysfunctional; the phenotypic outcome
could be induced to express differentiation mark- of which is a loss of corneal transparency and
ers consistent with a keratocyte-like phenotype ultimately blindness.
[8] indicating the presence of a specific popula- This destruction of LSCs has been reported to
tion of adult stem cells for stromal repair and/or occur in a number of ways including mechanical
maintenance. Corneal stromal stem cells are injury, chemical or thermal burns as well as infec-
considered to be mostly located in the anterior tions and autoimmune diseases that cause scar-
stroma subadjacent to the basal side of the pali- ring [19]. In addition, inflammatory diseases
sades of Vogt [9, 10]. Current evidence appears such as Stevens-Johnson syndrome, contact lens
to suggest that this population of stromal stem wear, UV irradiation and chemotherapy have also
cells concentrated at the peripheral cornea play a been implicated in some cases of LSCD [20, 21].
role in maintaining the keratocyte population The resultant characteristic absence or dys-
within the stroma. In addition, there is limited function of LSCs severely impacts the regenera-
evidence that they may also contribute to epithe- tive ability of the corneal surface leading to a
lial regeneration by mesenchymal-­ epithelial range of persistent epithelial defects [22].
transdifferentiation [11]. Commonly, conjunctival invasion is observed
Keratocyte progenitor cells (KPCs) have a which results in a loss of transparency coupled
limited ability to self-renew and differentiate with vascularization and inflammation [19, 22–
[12]. This potency differentiates them from true 24]. Patients presenting with this condition suffer
stem cells although they do possess the ability to from chronic pain, redness, irritation, scarring
grow clonally in vitro and phenotypically become and decreased vision due to opacity and irregular
keratocytes; however they differ from keratocytes astigmatism [25]. In severe cases, LSCD can
in that they do not rapidly lose their phenotype progress to stromal scarring including the loss of
with in vitro expansion. KPCs are found through- resident keratocytes, ulceration and, on rare occa-
out the stroma [8, 13, 14] but appear to be par- sions, perforation [25].
ticularly concentrated at the peripheral cornea Given the vast spectrum of progression and
[13] and therefore are often not easily distin- severity of LSCD symptoms that patients present
guishable from true stromal stem cells. The pri- with, diagnosis is not straightforward [26]. The
mary role of KPCs is considered to be keratocyte most obvious and reliable clinical feature used
replenishment in corneal homoeostasis. as a diagnostic indicator is conjunctivalization
In addition to stem and progenitor cells that identified by slit-lamp examination [27, 28].
play a direct role in the repair and maintenance of Impression cytology is commonly used to col-
21  Stem Cell Spheres for Corneal Regeneration 301

lect cells from the corneal surface which are then to corneal scraping to restrict further conjuncti-
examined for the presence of goblet cells to pro- valization and assist epithelial recovery is amni-
vide evidence of conjunctivalization [29]. This otic membrane transplantation. Immediately post
technique, however, is of limited sensitivity and scraping, biologically sourced donor amniotic
can result in false negatives [25]. Imaging tech- membrane is attached to the affected epithelial
niques such as in vivo laser scanning confocal region [42]. The nutrient-rich amniotic mem-
microscopy (IVCM) have been used to examine brane provides an ideal microenvironment for the
cellular morphologic changes at the limbus [30, local stem cells to migrate and proliferate in the
31], while optical coherence tomography (OCT) hope of epithelial and/or limbal restoration.
has been used to measure epithelial thinning Should these initial treatments prove ineffec-
which is common in LSCD [32, 33]. IVCM can tive or if LSCD is at a severe stage of disease
provide high-resolution images which allow pre- progression, such that the resident stem cells are
cise examination of limbal structures and can overtly dysfunctional or too few in number,
identify the loss of these structures or morpho- replacement of the LSCs from an external source
logical abnormalities at the limbus [34, 35]. is required. In severe cases of LSCD where only
OCT is faster and requires less patient co-opera- one eye is affected, autologous limbal transplants
tion, and although it can provide images of lim- from the unaffected eye to the damaged eye were
bal structure, they are not as high resolution as historically the preferred treatment [43]. This has
IVCM [36]. However, neither of these tech- the advantage of utilizing the patient’s own
niques is currently used in standard diagnostic healthy limbal tissue for treatment alleviating any
practice [25]. immunosuppression and infection-related issues
that would be of major concern in cases where a
limbal allograft is necessary [44–46]. The major
21.4 Treatments: Limbal disadvantage of limbal transplant in either case is
Transplants, Limbal Stem that the large amount of limbal tissue required to
Cell Expansion and Stromal be excised from the healthy donor eye creates a
Stem Cells? risk of that eye developing LSCD itself. To cir-
cumvent this major downside, smaller sections of
The basic premise of any treatment for LSCD is limbal tissue can be used to expand the stem cell
the replacement of the lost stem cells and regen- population ex  vivo on supporting matrices such
eration of the limbal structure damaged through as amniotic membrane prior to grafting onto the
injury or disease. affected corneal epithelium [44, 47–49].
In less severe cases or partial LSCD, more The corneal epithelium being the anterior-most
conservative treatment approaches can initially part of the cornea is primarily impaired in LSCD;
be taken. Non-surgical treatments include the use however in most severe cases such as with signifi-
of autologous serum drops, therapeutic contact cant chemical burns, the stroma can be defective,
lenses and eye lubrication which are designed to and the resident keratocytes can also be lost. As
limit any further defects, promote healing of most of the treatments above target epithelial
existing defects and facilitate proliferation and recovery first and foremost, the impact that the
migration of existing stem cells to re-form a loss of keratocytes and their stem and progenitor
healthy limbus [37–40]. Corneal scraping is precursors may have on limbal recovery has not
sometimes used to remove the conjunctival over- been fully elucidated. In LSCD cases where stro-
growth and allow for the remaining healthy cor- mal scarring is evident, it is common for either a
neal epithelial stem cells to repopulate the corneal full-thickness or partial-thickness corneal trans-
surface [41]. This treatment however is depen- plant procedure to be performed immediately
dent on sufficient remaining stems cells to prior to grafting of the ex  vivo expanded LSCs.
actively repopulate the epithelium before the This has the combined desired result of replacing
recurrence of conjunctivalization [41]. An adjunct the defective stroma and potentially facilitating
302 S. Ismail et al.

limbal recovery but is hindered by the age-old characterized and defined cell population for use
problem of donor tissue availability both for in limbal transplantation is desirable. With this
transplant and ex vivo expansion. end goal in mind, we have focused on isolating a
mixed population of cells from the cornea which
in the right circumstances have the potential for
21.5 Regenerative Capabilities: limbal reformation.
Regeneration of the Limbus? The location of the desirable cell types along
the corneal surface is an important consideration.
Although the dysfunction or loss of stem cells is With the limbus as the well-accepted site for lim-
considered the primary cause of LSCD, recent bal stem cells, we assessed the regenerative capa-
findings suggest that significant disturbances to bilities of cells present at the central and
the limbal niche could be responsible for the peripheral cornea. In 2008, we developed a
observed stem cell dysfunction [6] and as such, ‘donut’ excimer laser ablation model with which
LSCD should be considered a limbal niche defi- we showed that corneal epithelial repair occurs
ciency. It is therefore imperative to consider inwards from the peripheral cornea as well as
whether the above treatments for LSCD results in outwards from the central cornea suggesting the
the regeneration of the limbus complete with a presence of regenerative cells at the peripheral
limbal niche. and central cornea [51]. Even when the limbus
Ex vivo expansion on amniotic membrane was completely ablated, epithelial regrowth was
would likely result in cultivation of a mixture of observed, providing evidence that corneal repair
epithelial cells, transient amplifying cells and can occur independent of a functional limbus.
stem cells which collectively have the makings of Despite these findings, we have deemed it likely
the limbal niche, but current practices are unable that limbus-independent corneal regeneration is
to characterize, enumerate or manipulate their essentially a wound healing response for short-­
relative proportions amongst the expanded cells term repair (12 h post-ablation in our study) and
on the amniotic surface prior to grafting onto the that a healthy limbus is essential for the long-­
limbal stem cell deficient eye. In effect, this term maintenance of the cornea.
inability to standardize cell expansion results in With this in mind, we have looked at isolation
unknown numbers of LSCs being grafted, and of cells from the limbus and peripheral cornea
although it has a reported success rate of 70% in and assessed their potential to re-form limbal tis-
improving the patients’ vision [50], this unknown sue. Corneoscleral rims composed of peripheral
entity would certainly contribute to the variabil- cornea, limbus and scleral tissue are often dis-
ity in the success of this procedure. The long-­ carded after the central cornea is removed by the
term success of these treatments is yet to be surgeon for transplantation procedures. This
proven as there is little evidence of where the postsurgical ‘waste’ tissue is an ideal source of
transplanted cells eventually localize and whether limbal and stromal stem and progenitor cells that
the procedure eventually restores the limbus to its could be harvested for use in cell-based therapy.
pre-injury state is unclear given that the criteria The limbal/peripheral corneal region from this
used to determine ‘success’ are often subjective. tissue could be excised from the scleral tissue and
the cells extracted from extracellular tissue com-
ponents. The resultant cell extracts would likely
21.6 Isolation of Spheres represent the individual cellular components of
with Mixed Potential the in  vivo niche. The sphere-forming assay is
one such method that has been used to isolate
Given the challenges associated with the treat- cells with proliferative potential from tissue
ment of LSCD and the ambiguities of the regen- extracts in a non-adherent culture system. Since
erative capabilities of the cells used in current its inception in the early 1990s as ‘neurosphere
treatment practices, the development of a well-­ assay’ [52] in brain tissue and from the early
21  Stem Cell Spheres for Corneal Regeneration 303

2000s in ocular tissue, the sphere-forming assay formation, we have shown that active recruitment
has become a widely used method for isolation may also occur whereby single cells that may not
and enrichment for adult stem and progenitor have been part of a sphere or had previously
cells [53]. Cells cultured in sphere-forming assay migrated out from a sphere are pulled into the
have some advantages to those cultured in tradi- body of the sphere by extended pseudopodia
tional adherent monolayers. The most notable of [55]. Similarly, some proliferating cells at or out-
which is that they retain similar characteristics to side of the periphery of a sphere show rapid
the tissue from which they have been isolated and migration back towards the sphere. Additionally,
can accurately simulate the native behaviours of cells within a sphere are also proliferative with
this tissue [54]. EdU-based proliferation assays that label DNA in
Our laboratory routinely isolates limbal cells dividing cells showing up to 70% of spheres
for sphere-forming assay from corneoscleral rims composed completely of EdU-positive cells at
by mechanical removal of endothelial cells and 7  days in culture [55]. Despite sphere-forming
anterior epithelial cells prior to prolonged tissue culture conditions promoting non-adherent cell
treatment with dispase II, collagenase and hyal- culture, we often observe cells attached to the
uronidase to ensure the complete dissociation of glass surface that form a cell network which
limbal cells from extracellular tissue compo- comes together over time to form semi-attached
nents. The single-cell extracts are subsequently spheres that subsequently become non-adherent
cultured in vitro in a mitogen-driven serum-free mature spheres. These observations combined
culture system [55]. Typically, single cells come indicate that the spheres are not necessarily clo-
together to form cell aggregates with well-defined nogenic as previously thought [56] and are
spherical borders which morphologically differ- formed by a combination of cell aggregation,
entiate true spheres from loose cell aggregates active recruitment and proliferation.
(Fig. 21.1). In our experience, sphere formation The culture conditions used in sphere-forming
can occur within days to several weeks post assay are designed to promote the survival of
in vitro culture. The process of sphere formation stem and progenitor cells and select out the more
is not merely a passive process of cell aggrega- differentiated cell types in extended culture.
tion but rather an active and dynamic one with Given the source of the cell extracts being limbus
multiple facets coming together to form the stem and peripheral cornea and the knowledge that
cell-enriched sphere. Although passive aggrega- sphere formation is non-clonal, spheres are likely
tion of cells that are in close vicinity to form to contain cells of epithelial and stromal origin.
loose clusters may be the initial step in sphere Long-term free-floating culture of spheres for up

Fig. 21.1  Sphere-forming assay: cell isolates from human corneoscleral tissue cultured in sphere-forming conditions
form cell aggregates with well-defined spherical borders
304 S. Ismail et al.

to 4  months without passaging confirms their Collectively, these results suggest that spheres
ability for self-maintenance and ensures stem cell contain mixed cell populations both in terms of
enrichment. Although we have observed consid- cell lineage and stemness and/or differentiation
erable donor variation in both the ability of cell state. This heterogeneity of cell types observed in
extracts to form spheres and the longevity of sphere composition may reflect that observed in
spheres once formed, we have been able to main-
tain healthy spheres in culture for greater than
9 months in some cases. Characterization of the
cells within spheres by assessment of their ability
to respond to stimulus such as a collagen sub-
strate shows an ability to rapidly migrate and pro-
liferate with the greatest proliferation observed at
the periphery of the sphere and amongst migra-
tory cells, while the cells at the core of the sphere
remain relatively quiescent (Fig.  21.2) [55].
Immunocytochemical detection of cell differen-
tiation markers amongst these migrating cells
shows expression of the mesenchymal cell
marker vimentin and the epithelial cell marker
keratin K3/K76 confirming a likely mix of epi-
thelial and stromal origin cells (Fig.  21.3). By
contrast free-floating spheres and non-migratory
cells within collagen-stimulated spheres showed Fig. 21.2  Spheres respond to collagen stimulus with pro-
expression of the limbal basal epithelial marker liferation and migration: fluorescence ubiquitination cell
cycle indicator (Fucci) labelling of spheres placed on a
deltaNp63, the limbal niche marker notch1, the
collagen surface shows greatest proliferation (green/yel-
neural progenitor cell marker nestin and the low) at the periphery of the sphere and amongst migratory
repair cell (myofibroblast) marker α-smooth cells (red), while the cells at the core of the sphere remain
muscle actin (α-SMA) (Fig.  21.4) [55–57]. relatively quiescent

a b

Fig. 21.3  Mixed origin of sphere-derived cells: migra- Cell nuclei labelled with DAPI (blue), proliferative cells
tory cells derived from spheres show expression of the labelled with EdU (pink/red). Panel b: Cell nuclei labelled
mesenchymal cell marker vimentin (a) and the epithelial with propidium iodide (red/orange). (Panel a reproduced
cell marker keratin K3/K76 (b) (green) confirming a in part from reference: [57])
likely mix of epithelial and stromal origin cells. Panel a:
21  Stem Cell Spheres for Corneal Regeneration 305

a b

c d

Fig. 21.4 Sphere cell characteristics: free-floating marker nestin (c, red) and the repair cell (myofibroblast)
spheres and non-migratory cells within collagen-­ marker α-smooth muscle actin (α-SMA) (d, green). Cell
stimulated spheres show expression of the limbal basal nuclei labelled with DAPI (blue). (Panel b reproduced in
epithelial marker deltaNp63 (a, green), the limbal niche part from reference: [57])
marker notch1 (b, green), the neural progenitor cell

the in  vivo limbal niche and facilitate the long-­ Outgrowth of cells would appear to be random
term maintenance of stem and progenitor cells. and in all directions around the circumference of
When exposed to a collagen substrate, spheres the sphere in 75% of spheres (Fig.  21.5a).
show a number of distinct and intriguing behav- However, in 25% of spheres, the response to col-
iours which need to be carefully assessed in order lagen stimulus would appear to be more deliber-
to determine their suitability for use in stem cell ate and targeted with an apparent directional
therapy. Upon placement onto a collagen sub- response (Fig.  21.5b). In all cases, a distinct
strate, spheres cease to be free floating and attach polarity seems to develop with cellular outgrowth
to the surface, and cells rapidly migrate outwards, occurring in opposite directions from two ‘poles’,
and this is often coupled with cell division. while the areas adjacent to the polar outgrowth
306 S. Ismail et al.

a b

c d

Fig. 21.5 Sphere behaviour on collagen substrate: directional response with cell migration along distinct
spheres respond to collagen stimulus in a variety of ways ‘north and south poles’. (b) This directed cell migration
with approximately 75% reacting with random outgrowth creates cell exclusion zones (c, arrow) which over time
from all directions, (a) while approximately 25% show a may be partially or completely repopulated (d)

remain as cell exclusion zones (Fig.  21.5c). In initially occurring, the entire sphere migrates as a
some cases, a few migratory cells venture out whole with cells from each pole displaying dis-
into the apparent cell exclusion zones but do not tinct locomotive behaviour. The cells at the front
continue migrating outwards, instead turning of the motile sphere undergo cycles of
back towards the direction of polar outgrowth or ­lamellipodia extension and traction in a forward
reversing towards the sphere. Over time direction, while the trailing cells at the opposite
(2–10 days), this polar pattern of cell migration is pole show a highly coordinated pattern of retrac-
partially or completely lost with cell migration tion which serves to project the sphere forwards
into the exclusion zone ensuing from both the [55]. The direction of en masse sphere migration
sphere and the directional migratory cells is not linear, with the migratory path able to be
(Fig.  21.5d). Distance from the central sphere changed from one initial direction to another pos-
also appears to be important for cell signaling as sibly in response to microenvironmental signals.
cells furthest from the sphere seem to lose The significance of this en masse migration is
directionality. unclear and must be carefully considered. The
In rare cases (less than 10% of spheres anal- advantage of such behaviour may allow spheres
ysed), a unique phenomenon of en masse migra- to sense its environment and enable it to ‘relo-
tion occurs (Fig. 21.6). With what would appear cate’ to the most favourable location to re-estab-
to be directed cell migration with polar outgrowth lish the limbus. However, on the flip side, this
21  Stem Cell Spheres for Corneal Regeneration 307

Fig. 21.6  Collective sphere migration: en masse migra- begins moving upwards, and by 42–45  h post collagen
tion is observed in less than 10% of spheres where the placement, the sphere begins to migrate out of view. This
entire sphere migrates as a whole with cells from each indicates that sphere migration en masse can take place in
pole displaying distinct locomotive behaviour. Here the multiple directions. Red dot indicates original central
sphere initially migrates slightly downwards and to the position of sphere; white reference line indicates original
right from 5 to 10  h post placement on a collagen sub- horizontal position of centre of sphere
strate. At 30–37  h, the sphere changes direction and

migratory pattern may also result in spheres mov- The dual role of corneal stem cells to replace
ing away from their intended transplanted site lost cells and to respond to injury warranted
and moving to unintended areas of the cornea investigation of this second function amongst our
where undesirable cell migration and prolifera- sphere cultures in order to gauge their wound
tion may occur. healing ability. To this end, we conducted a series
When multiple spheres are seeded in close of direct and remote wounding experiments to
proximity to each other, the migratory cells from evaluate the sphere response to physical injuries
each sphere interact to form both areas of out- and assess their capacity for wound healing
growth and cell exclusion zones which, when (Fig.  21.7). Mature spheres in culture for 3–4
organized in a circular pattern, form a ring-like months post tissue extraction that were most
structure with a large central exclusion zone. likely to be stem cell enriched were placed onto a
These observations, albeit largely qualitative, are collagen substrate and further stimulated 48 hours
not isolated instances and have been observed post attachment by a direct compression injury to
amongst spheres prepared from tissue collected the centre of the sphere using 10  μm glass
from multiple donors over a number of years. ­capillaries with the aid of a micromanipulator.
Collectively, these observations demonstrate our Fluorescence-based staining for live/dead cells
limbal and peripheral corneal spheres to be het- confirmed an increase in cell death upon com-
erogeneous structures that are capable of dynamic pression injury when compared to uninjured
behaviours in response to positive stimuli. It is spheres. This increase in cell death elicited a dif-
our opinion that these behaviours are not innate ferentiation response from the cells within some
predetermined responses but rather determined of the spheres whereby an observable increase in
by the microenvironment that they are placed in the expression of the myofibroblast repair cell
and as such, if correctly manipulated, could be marker α-SMA was detected 24 hours post injury
used to effectively regenerate the limbus. [58]. This increase in α-SMA-positive cells was
308 S. Ismail et al.

also appeared to slow the outward migration of


cells from the sphere at 24  h post injury which
then resumed at 7 days post injury, presumably as
the effect of the injury had diminished. These
results clearly established that quiescent spheres
of 3–4 months in culture are capable of reacting
to injury similar to how one would expect cells at
the limbus to react. However, injuries to the eye
may not always be to the peripheral cornea or
limbal region, and therefore the ability of spheres
to respond to remote injury is also an important
consideration. To test whether our spheres pos-
sessed this ability, we conducted remote scrape
wounding of migrating cells at 50 and 100  μm
distances from each sphere circumference in the
direction that had the greatest level of outward
Fig. 21.7  Sphere response to injury: direct compression cell migration. Having previously established
injury of a sphere results in outward cell migration, that spheres are capable of directed cell migra-
increased proliferation (EdU incorporation, red/pink) and tion on a collagen surface, we measured whether
myofibroblast differentiation (α-smooth muscle actin there was an increase in directed outward cell
expression, green). Radiating concentric circles drawn at
100  μm intervals were used to quantify outward cell migration from the sphere in the direction of the
migration, while a directional response was evaluated by scratch injury. By measuring cell numbers in
enumerating cell migration in quadrants 1–4. Cell nuclei 100  μm concentric circles from the sphere, we
labelled with DAPI (blue). Scale bar = 100 μm showed that scratch injury at the closest distance
to the sphere (50 μm away) resulted in a greater
no longer observed in spheres at days 4, 7 and 14 number of outwardly migrating cells than from
post injury. Migrating cells from injured spheres those spheres that did not receive scratch injury
showed morphological features characteristic of to migrating cells and those that received scratch
myofibroblasts indicating the stromal origin cells injury to cells at a greater distance (100  μm
within the spheres were upregulating α-SMA in away). This may be due to a lack of a strong
what could be considered a wound healing enough signal to provoke a wound healing
response [58]. An important part of any wound response from those cells injured at 100 μm from
healing response is a rapid increase in cell divi- the sphere. When looking at proliferative
sion to replace the dead cells at the injury site response, scratch injury resulted in a similar or
with new healthy cells. In spheres with direct greater number of dividing cells amongst the
compression injury, a significant increase in pro- migratory cells, while the presence of α-SMA-­
liferative cells as detected by EdU-based prolif- positive cells both within the spheres and amongst
eration assay was evident at 4  days post injury migrating cells in remotely injured spheres
when compared to uninjured spheres. This prolif- showed that a stromal cell response had been
erative response was also present at 7 days post triggered. No such response was seen in spheres
injury but had declined somewhat from the day 4 that had not received scratch injury [58].
peak, and by day 14 the proliferative response Collectively, these results show that spheres have
had returned to that of uninjured spheres. Cells the ability to respond to injury (both direct and
that had migrated outwards from the injured remote) by initiating migration, proliferation and
sphere showed no difference in proliferation, differentiation of cells into a repair cell pheno-
indicating that the increase in cell division was a type similar to cells at the limbus. However, the
direct response to the cell death signals elicited spheres did not show a specific directional
by the injured cells. Direct compression injury response to remote scratch injury in  vitro, sug-
21  Stem Cell Spheres for Corneal Regeneration 309

gesting other factors (physical and physiological) organization of cell migration patterns emanating
may play a role in ensuring migratory and repair from the spheres was observed with directed cell
cells reach the site of injury from the limbus. migration, and thirdly, spheres placed in proximity
Having established the properties and behav- to each other showed migration of cells towards
ioural characteristics of spheres in vitro and con- each other (Fig. 21.8b) [57]. In addition, the migra-
firming their ability to respond favourably to tion pattern of cells showed an obvious preference
collagen substrate and to elicit a wound healing towards the corneal direction of the tissue as
response, it was pertinent to assess their ability to opposed to the scleral direction. Proliferation was
repopulate human corneal tissue. Therefore we evident as by day 7 post implantation, the entire
investigated whether spheres could be transplanted corneal surface of the tissue segment had become
into decellularized (repeated freeze-­thaw) corneo- repopulated as evidenced by live cell staining, and
scleral tissue at the position of the limbus and the number of cells present could not have repopu-
whether the implanted spheres would respond in lated the surface by migration alone. The limitation
the same way that was observed in vitro on a colla- to cell migration appeared to only be the extent of
gen-coated surface. To facilitate this, we developed the corneal tissue with cells observed all the way to
a microsurgical technique to create wedge-shaped the furthest edge of the corneal surface (Fig. 21.8c).
incisions at the scleral border of the limbus [59] to This phenomenon was not observed on the scleral
create a shallow trough into which spheres could be side of the tissue where only limited migration had
implanted. The limbal tissue was confirmed to be occurred and most of it had remained in proximity
completely decellularized by a lack of live cell to the limbal region. These observations again
staining prior to sphere implantation and essen- strengthen our earlier thoughts that the spheres do
tially served as a model of LSCD.  Implanted not necessarily have predetermined responses but
spheres from mature sphere cultures could easily rather respond appropriately according to the envi-
be implanted under a dissecting microscope and ronment into which they are placed. Therefore
showed a propensity to attach to the tissue and when placed into the limbal region, they would
remain within the site of surgical incision for the appear to behave in a similar way to how one would
duration of the 241-­hour experiments (Fig. 21.8a). expect limbal cells to behave by migrating towards
In an ex  vivo setting where decellularized tissue the corneal surface. Closer examination of the
implanted with spheres was incubated in serum- migratory cells in situ revealed differing cellular
containing culture medium, spheres were observed organization patterns at the corneal, limbal and
to exhibit a number of similar behaviours to those scleral surfaces with those at the limbal and scleral
observed on a flat expansive collagen-coated sur- region showing thin spindle-like elongated mor-
face. Firstly, spheres were able to respond to the phology, while those at the corneal surface were
collagen fibrils surrounding them by actively send- broader in appearance and at the leading migratory
ing migratory cells outwards. Secondly, the active edge showed branching cellular processes. Cells

a b c

Fig. 21.8  Implantation of spheres into donor corneo- at 3 days post implantation (b) with preferential and com-
scleral rims: surgical implantation of spheres into the lim- plete repopulation of the available corneal surface within
bal region of decellularized corneoscleral rims (a) results 7 days of implantation (c). (Reproduced in part from ref-
in multidirectional live (calcein AM, green) cell migration erence: [57])
310 S. Ismail et al.

appeared more organized at the limbal and corneal spheres elicit upon implantation is that of a wound
surfaces, while those at the sclera did not show healing response with migration and proliferation
such a uniform pattern of alignment. Cross-­ towards the corneal surface. However, the entire
sectional analysis of the sphere-implanted tissue corneal bed was repopulated within 7 days, and the
showed mostly repopulation over the anterior cor- repopulated surface could be maintained for at
neal surface with limited evidence of cell migration least another 7 days which would suggest that cor-
deeper down into the tissue. At 14 days post sphere neal homoeostatic processes may be beginning to
implantation, the stem cell marker deltaNp63 could take place. Although the short time frame for this
be detected along with differentiation markers lam- does not allow us to definitively show this. Longer-
inin and vimentin. Whole-­ mount staining of term sphere implantation experiments that allow us
sphere-implanted tissue also demonstrated the to determine if keratocyte differentiation of stromal
presence of the stem cell marker ABCG2 in some cells and retention of stem cell populations that
cells and the limbal basal epithelial marker facilitate corneal surface maintenance would be
notch1 in others, while the keratocyte marker kera- valuable. However, these results show remarkable
tocan was not detected. Taken together, these promise for the use of our spheres in corneal
results would suggest that stem and progenitor cell regeneration.
repositories as well as limbal components and that Sphere implantation experiments were con-
of transient amplifying cells and differentiated cells ducted using spheres prepared from normal donor
were being retained in implanted tissue. All com- tissue and implanted into decellularized normal
ponents would be present in native cellularized tis- donor tissue. However this does not address the
sue. The lack of keratocan detection may reflect the fact that in order to be an effective cell-based ther-
lack of keratocyte differentiation from stromal cells apy, spheres would be transplanted into patients
at 14 days post implantation and likely due to the with ocular conditions amounting to abnormal tis-
limited cellular infiltration of underlying tissue and sue structures. Therefore to test whether spheres
the serum-containing culture conditions, but the have the equivalent ability to repopulate diseased
presence of the mesenchymal cell marker vimentin tissue, our wedge-shaped incision/sphere implan-
provides evidence that stromal origin cells are tation technique was applied to diseased corneal
being retained in sphere-implanted tissue. Gene tissue. Corneal buttons from keratoconus patients
expression data on sphere-implanted tissue con- and a decompensated cornea from a patient receiv-
firmed a reduction in keratocan and also laminin-1 ing corneal transplant were used to implant spheres
14  days post implantation when compared to the into the central cornea. Again, the tissues were
original sphere cells, while the proliferation marker decellularized by repeated freeze-thaw cycles and
PCNA showed a significant increase at day 4 fol- confirmed to be free of host tissue cells by live cell
lowed by a reduction from the day 4 peak at day 7 staining. This left an acellular diseased tissue
and a return to basal levels at day 14 post sphere matrix which was much weaker and friable than its
implantation. This day 4 proliferation peak is iden- corresponding normal donor equivalent. This yet
tical to the peak in proliferative cells observed in to be published data showed that implanted spheres
the earlier direct compression sphere wounding were able to adhere to the diseased tissue and
experiments. It would seem that day 4 post place- remain in position similar to that of normal tissue.
ment on a collagen stimulus is a critical time point Over the course of 7 days, spheres were also able
for cell proliferation. Gene expression data also to send out migratory cells over the t­issue surface.
confirmed no change in vimentin expression and The extent of migration was variable amongst dif-
stem cell marker (ABCB5, ABCG2 and p63) ferent spheres, and in some cases, initial rapid
expression which may suggest stability in the migration observed at day 3 post implantation had
maintenance of these cell types over time. The evi- reduced by day 7. However this observation was
dence provided by these sphere implantation not unique to the diseased tissue, and a similar
experiments that agree with our in  vitro wound response was observed in concurrently implanted
healing work and suggests that the response that normal corneal buttons suggesting the response
21  Stem Cell Spheres for Corneal Regeneration 311

was a sphere-specific or culture condition-specific the extent of the entire stromal section when mul-
response. Due to the larger surface area of a cor- tiple spheres were placed on the section. In all
neal button compared to a corneoscleral tissue seg- cases, the sphere cell migration showed a propen-
ment, complete surface repopulation was not sity to migrate across the stromal section preferen-
observed by implantation of 3–4 spheres tially to the culture dish surface. These unpublished
(Fig. 21.9). The pattern of repopulation, however, results clearly demonstrated that the abnormal
was similar in both normal and diseased tissue structure of the diseased tissue was not a deterrent
both in terms of cell morphological patterns and to sphere attachment, migration and division
cell orientation patterns. Spheres implanted in although there was no observable increase in any
close vicinity to each other showed circumferen- of these measures in response to diseased tissue
tial migration of cells including towards each other compared to normal tissue either. Of note was the
where they appeared to form cellular bridges fact that keratoconic stromal sections have a much
between spheres when the cells from each sphere looser arrangement of collagen fibrils and this
interacted. The extent of cell migration over 7 days altered structure did have a noticeable effect on the
gave clear evidence of cell proliferation as the cell migration pattern at days 7 and 14 post implan-
sheer numbers of cells present could not have pop- tation with cell orientation taking on a less orga-
ulated the migratory area by outgrowth alone. nized and a more random orientation pattern which
Moreover the spheres maintained themselves in was not observed in normal stromal sections.
terms of their morphological structure in most Qualitative observation of cell morphology of
cases. EdU-based proliferation assays confirmed migratory cells showed limited evidence of a
the presence of proliferative cells at day 10 post keratocytic phenotype in a few cells from spheres
implantation. Furthermore, placement of spheres implanted into diseased tissue. Gene expression
onto en face 10 μm sections of decellularized dis- data showed no apparent differences in stem cell
eased stromal tissue invoked a similar response of markers, epithelial cell markers, keratocyte mark-
attachment, cell migration and proliferation. In ers or extracellular matrix markers between nor-
these experiments, cell migration by day 14 was to mal and diseased sections indicating the sphere

a b

Fig. 21.9  Implantation of spheres into diseased central shows a pattern of cell migration outwards from implanted
corneal tissue: sphere implantation into a decompensated spheres which predominantly increased from day 3 (a) to
central corneal button (recovered from corneal transplant day 10 (b) post implantation
surgery) assessed by calcein AM live cell staining (green)
312 S. Ismail et al.

response to be essentially the same in diseased showed some evidence of collagen expression
tissue as observed in normal tissue. Vimentin from implanted cells indicating they may be
expression showed an increase over time in both capable of laying down new matrix, but the
normal and diseased sections and may reflect a 14-day time period is likely too short to deter-
stromal cell migratory response over the largely mine whether this would be the beginnings of
stromal sections. Whether these stromal cells structural repair to the diseased tissue. A second
would differentiate into keratocytes was not consideration would be the cell-cell interaction
clear, and there was no evidence of this at 14 days between normal donor and diseased recipient
post implantation. Longer-term experiments cells. As all of our in  vitro and ex  vivo experi-
would help to determine if the migrated cells, ments were conducted on decellularized matrix,
once established in their post-migration position the effect of resident cells is largely unknown.
within the stroma, would become resident kerato- Best-case scenario would be that the transplanted
cytes. Culture conditions promoting keratocyte normal cells would be able to use signaling
differentiation would also likely be required to mechanisms to reverse, alter or stop the abnormal
confirm this. Immunohistochemical staining for cellular response of host cells. Any remedial
α-SMA at days 1 and 4 post implantation showed effect of transplanted sphere cells may be depen-
a few cells positive for this repair cell marker dent on the severity of disease progression, and it
indicating a small wound healing response. These may be that there is a critical level of progression
unpublished results confirm the ability of spheres or ‘tipping point’ where any therapeutic effects
to respond to diseased tissue in a similar way to exerted by transplanted cells would not be able to
their response to normal tissue and that the cells reverse the disease progression. All of these ques-
show the correct markers and morphological tions remain to be answered.
(epithelial or stromal) tendencies for the tissue We have recently looked at sphere-forming
structures they were placed in. ability in cells extracted from keratoconic tissue
Further investigations in this area need to and assessed their ability to respond to similar
address whether the implanted spheres from nor- stimuli to that of normal donor spheres
mal cells are capable of altering the structure of (Fig.  21.10). However, as keratoconic tissue is
the diseased tissue. Our initial experiments here usually obtained as central corneal buttons that

Fig. 21.10  Response of a


spheres from
keratoconic cells:
spheres formed from
keratoconic cell isolates
respond more rapidly to
collagen stimulus than
spheres from normal
cells (a, time
point = 25 h post
implantation). The
sphere from keratoconic Normal Keratoconic
cells is unable to b
maintain itself, and
migrating cell numbers
are not retained over an
extended period of time,
while the sphere from
normal cells remains
intact (b, time
point = 266 h)
21  Stem Cell Spheres for Corneal Regeneration 313

have been removed from patients prior to trans- remarkable amount of potential for cell-based
plant, there is a lack of peripheral corneal and therapy as they are enriched with stem and pro-
limbal cells present. In our previous experience genitor cells, they are responsive to stimuli such
looking at sphere-forming ability in normal cen- as collagen and tissue matrices and they possess
tral corneal buttons, we found that sphere forma- the ability to elicit a wound healing response
tion did indeed occur from central corneal cells and appear to show an appropriate level of
but at a diminished rate of 1:2 compared to migration, differentiation and division to repop-
sphere-forming ability in peripheral corneal/lim- ulate decellularized tissue surfaces. Additionally,
bal extracts [56, 60]. This sphere-forming abilitythey do not appear to adversely change their
is further diminished in keratoconic cell extractsbehavioural characteristics when applied to dis-
where we seldom observed sphere-forming cells eased tissue matrix.
from keratoconic tissue. On the rare occasions There are a number of ways in which to
where sphere formation was successful, the apply stem cell-based therapies which in the
spheres did not survive as long as those formed eye currently rely largely on ex vivo expansion
from normal tissue extracts in culture. When on amniotic membrane. In any case, none of the
placed onto collagen stimulus and compared current cell therapy applications address the
with spheres from normal donor tissue, kerato- cellular make-up of the expanded cell popula-
conic spheres appeared to respond much faster tions. Very few details exist of the numbers and
to the substrate with rapid cell migration and types of cells that are being grafted and where
division. This response however was not pro- the grafted cells end up both along the surface
longed, and over time the cells of the normal of the cornea horizontally and within the depth
spheres migrated out further and proliferated of the cornea vertically. Similarly there is little
more than the keratoconic spheres which could long-term evidence of whether these cells
not maintain their migratory and proliferative repopulate the limbus and maintain a fully
response. This may suggest that cells from kera- functional limbus over a prolonged period of
toconic tissue are in a perpetually heightened time. Stem cell-enriched peripheral corneal
state of wound healing response which cannot spheres formed by sphere-forming assay repre-
maintain correct corneal homoeostasis long sent a way to apply a well-defined, known cell
term. Again whether co-culture of keratoconic population for treating a range of corneal dys-
cells with normal cells would revert the diseased trophies. As cells are extracted as single cells
cells to a normal phenotype remains to be from tissue, they have the ability to be sepa-
ascertained. rated and sorted by laboratory methods such as
FACS prior to culture in sphere-forming assay,
and individual sphere cultures can be thor-
21.7 Conclusion: Potential oughly tested in vitro for their ability to respond
of Spheres for Treating LSCD appropriately prior to transplantation. Sphere
and Corneal Dystrophies cell populations could potentially be manipu-
lated for use in the treatment of different condi-
Our work with peripheral corneal- and limbal-­ tions to enhance the therapeutic effect, for
derived cells cultured in sphere-forming assay example, to maximize limbal components for
has shown that spheres cultured in this way are LSCD treatment or to enrich for epithelial stem
stem cell-enriched, discrete populations of cells cells to treat recurrent epithelial erosions.
that are not merely aggregates but dynamic enti- Although further assessment of the potential of
ties that represent a heterogeneous mix of lim- sphere-based cell therapy is required, current
bal niche components and possess unique evidence would suggest that it would be a use-
properties that allow them to react to mechani- ful adjunct to current treatments and represent
cal, environmental and chemical stimuli. To an improved method for long-term sustained
date, we have shown that these spheres hold a corneal regeneration.
314 S. Ismail et al.

Compliance with Ethical Requirements  10. Dhamodaran K, Subramani M, Ponnalagu M, Shetty


R, Das D.  Ocular stem cells: a status update! Stem
Cell Res Ther. 2014;5(2):56.
1. Conflict of Interest 11. Hashmani K, Branch MJ, Sidney LE, Dhillon PS,
Salim Ismail, Charles N. J. McGhee, Jennifer Verma M, McIntosh OD, et  al. Characterization of
J. McGhee, Ye Li, Jeremy John Mathan, Jinny corneal stromal stem cells with the potential for
epithelial transdifferentiation. Stem Cell Res Ther.
Jung Yoon, Himanshu Wadhwa, Stephanie 2013;4(3):75.
U-Shane Huang and Trevor Sherwin declare 12.
Patel DV, McKelvie J, Sherwin T, McGhee
that they have no conflict of interest. C.  Keratocyte progenitor cell transplantation: A
2. Informed Consent/Human Studies novel therapeutic strategy for corneal disease. Med
Hypotheses. 2013;80(2):122–4.
All procedures followed were in accordance 13. Yamagami S, Yokoo S, Mimura T, Takato T, Araie M,
with the ethical standards of the responsible Amano S. Distribution of precursors in human corneal
committee on human experimentation (insti- stromal cells and endothelial cells. Ophthalmology.
tutional and national) and with the Helsinki 2007;114(3):433–9.
14. Amano S, Yamagami S, Mimura T, Uchida S, Yokoo
Declaration of 1975, as revised in 2000. S.  Corneal stromal and endothelial cell precursors.
Informed consent was obtained from all Cornea. 2006;25(10 Suppl 1):S73–7.
patients for being included in the study. 15. Sherwin T. A new niche for the corneal epithelial stem
3. Animal Studies cell. Clin Exp Ophthalmol. 2009;37(7):644–5.
16. Li J, Chen SY, Zhao XY, Zhang MC, Xie HT. Rat lim-
No animal studies were carried out by the bal niche cells prevent epithelial stem/progenitor cells
authors for this article. from differentiation and proliferation by inhibiting
Notch signaling pathway in vitro. Invest Ophthalmol
Vis Sci. 2017;58(7):2968–76.
17. Blanpain C, Horsley V, Fuchs E. Epithelial stem cells:
References turning over new leaves. Cell. 2007;128(3):445–58.
18. Gonzalez G, Sasamoto Y, Ksander BR, Frank MH,
1. Chen Z, de Paiva CS, Luo L, Kretzer FL, Pflugfelder Frank NY. Limbal stem cells: identity, developmen-
SC, Li DQ. Characterization of putative stem cell phe- tal origin, and therapeutic potential. Wiley Interdiscip
notype in human limbal epithelia. Stem Cells (Dayton, Rev Dev Biol. 2018;7(2):e303.
Ohio). 2004;22(3):355–66. 19. Galindo S, Herreras JM, Lopez-Paniagua M, Rey E,
2. Grieve K, Ghoubay D, Georgeon C, Thouvenin O, de la Mata A, Plata-Cordero M, et  al. Therapeutic
Bouheraoua N, Paques M, et  al. Three-dimensional effect of human adipose tissue-derived mesenchy-
structure of the mammalian limbal stem cell niche. mal stem cells in experimental corneal failure due to
Exp Eye Res. 2015;140:75–84. limbal stem cell niche damage. Stem Cells (Dayton,
3. Goldberg MF, Bron AJ.  Limbal palisades of Vogt. Ohio). 2017;35(10):2160–74.
Trans Am Ophthalmol Soc. 1982;80:155–71. 20. Bobba S, Di Girolamo N, Mills R, Daniell M, Chan
4. Townsend WM.  The limbal palisades of Vogt. Trans E, Harkin DG, et al. Nature and incidence of severe
Am Ophthalmol Soc. 1991;89:721–56. limbal stem cell deficiency in Australia and New
5. Dua HS, Shanmuganathan VA, Powell-Richards Zealand. Clin Exp Ophthalmol. 2017;45(2):174–81.
AO, Tighe PJ, Joseph A.  Limbal epithelial crypts: 21. Rossen J, Amram A, Milani B, Park D, Harthan J,
a novel anatomical structure and a putative lim- Joslin C, et al. Contact lens-induced limbal stem cell
bal stem cell niche. Br J Ophthalmol. 2005;89(5): deficiency. Ocul Surf. 2016;14(4):419–34.
529–32. 22. Gonzalez S, Chen L, Deng SX.  Comparative study
6. Yazdanpanah G, Jabbehdari S, Djalilian AR.  Limbal of xenobiotic-free media for the cultivation of human
and corneal epithelial homeostasis. Curr Opin limbal epithelial stem/progenitor cells. Tissue Eng
Ophthalmol. 2017;28(4):348–54. Part C Methods. 2017;23(4):219–27.
7. West-Mays JA, Dwivedi DJ. The keratocyte: corneal 23. Dua HS, Saini JS, Azuara-Blanco A, Gupta P. Limbal
stromal cell with variable repair phenotypes. Int J stem cell deficiency: concept, aetiology, clinical
Biochem Cell Biol. 2006;38(10):1625–31. presentation, diagnosis and management. Indian J
8. Du Y, Funderburgh ML, Mann MM, SundarRaj Ophthalmol. 2000;48(2):83–92.
N, Funderburgh JL.  Multipotent stem cells in 24. Sejpal K, Bakhtiari P, Deng SX.  Presentation, diag-
human corneal stroma. Stem Cells (Dayton, Ohio). nosis and management of limbal stem cell deficiency.
2005;23(9):1266–75. Middle East Afr J Ophthalmol. 2013;20(1):5–10.
9. Pinnamaneni N, Funderburgh JL.  Concise review: 25. Kim KH, Mian SI.  Diagnosis of corneal lim-

Stem cells in the corneal stroma. Stem Cells (Dayton, bal stem cell deficiency. Curr Opin Ophthalmol.
Ohio). 2012;30(6):1059–63. 2017;28(4):355–62.
21  Stem Cell Spheres for Corneal Regeneration 315

26. Chuephanich P, Supiyaphun C, Aravena C, Bozkurt 41. Dua HS. The conjunctiva in corneal epithelial wound
TK, Yu F, Deng SX. Characterization of the corneal healing. Br J Ophthalmol. 1998;82(12):1407–11.
subbasal nerve plexus in limbal stem cell deficiency. 42. Anderson DF, Ellies P, Pires RT, Tseng SC. Amniotic
Cornea. 2017;36(3):347–52. membrane transplantation for partial limbal stem
27. Dua HS, Azuara-Blanco A.  Limbal stem cells
cell deficiency. Br J Ophthalmol. 2001;85(5):
of the corneal epithelium. Surv Ophthalmol. 567–75.
2000;44(5):415–25. 43. Kenyon KR, Tseng SC.  Limbal autograft transplan-
28. Huang AJ, Tseng SC, Kenyon KR. Alteration of epi- tation for ocular surface disorders. Ophthalmology.
thelial paracellular permeability during corneal epi- 1989;96(5):709–22; discussion 22–3.
thelial wound healing. Invest Ophthalmol Vis Sci. 44. Baylis O, Figueiredo F, Henein C, Lako M, Ahmad
1990;31(3):429–35. S. 13 years of cultured limbal epithelial cell ther-
29. Puangsricharern V, Tseng SC.  Cytologic evidence
apy: a review of the outcomes. J Cell Biochem.
of corneal diseases with limbal stem cell deficiency. 2011;112(4):993–1002.
Ophthalmology. 1995;102(10):1476–85. 45. Holland EJ.  Epithelial transplantation for the man-
30. Chidambaranathan GP, Mathews S, Panigrahi AK,
agement of severe ocular surface disease. Trans Am
Mascarenhas J, Prajna NV, Muthukkaruppan V.  In Ophthalmol Soc. 1996;94:677–743.
vivo confocal microscopic analysis of limbal stroma 46. Kolli S, Ahmad S, Lako M, Figueiredo F. Successful
in patients with limbal stem cell deficiency. Cornea. clinical implementation of corneal epithelial stem
2015;34(11):1478–86. cell therapy for treatment of unilateral limbal
31. Jawaheer L, Anijeet D, Ramaesh K.  Diagnostic cri- stem cell deficiency. Stem Cells (Dayton, Ohio).
teria for limbal stem cell deficiency-a systematic 2010;28(3):597–610.
literature review. Surv Ophthalmol. 2017;62(4): 47. Pellegrini G, Traverso CE, Franzi AT, Zingirian M,
522–32. Cancedda R, De Luca M.  Long-term restoration of
32. Le Q, Yang Y, Deng SX, Xu J.  Correlation between damaged corneal surfaces with autologous culti-
the existence of the palisades of Vogt and limbal epi- vated corneal epithelium. Lancet. 1997;349(9057):
thelial thickness in limbal stem cell deficiency. Clin 990–3.
Exp Ophthalmol. 2017;45(3):224–31. 48. Sangwan VS, Basu S, MacNeil S, Balasubramanian
33. Chan EH, Chen L, Yu F, Deng SX.  Epithelial thin- D.  Simple limbal epithelial transplantation (SLET):
ning in limbal stem cell deficiency. Am J Ophthalmol. a novel surgical technique for the treatment of uni-
2015;160(4):669–77. e4 lateral limbal stem cell deficiency. Br J Ophthalmol.
34. Jalbert I, Stapleton F, Papas E, Sweeney DF, Coroneo 2012;96(7):931–4.
M. In vivo confocal microscopy of the human cornea. 49. Shortt AJ, Secker GA, Notara MD, Limb GA,

Br J Ophthalmol. 2003;87(2):225–36. Khaw PT, Tuft SJ, et  al. Transplantation of ex  vivo
35. Shortt AJ, Secker GA, Munro PM, Khaw PT, Tuft SJ, cultured limbal epithelial stem cells: a review of
Daniels JT.  Characterization of the limbal epithelial techniques and clinical results. Surv Ophthalmol.
stem cell niche: novel imaging techniques permit 2007;52(5):483–502.
in  vivo observation and targeted biopsy of limbal 50. Haagdorens M, Van Acker SI, Van Gerwen V, Ni

epithelial stem cells. Stem Cells (Dayton, Ohio). Dhubhghaill S, Koppen C, Tassignon MJ, et  al.
2007;25(6):1402–9. Limbal stem cell deficiency: current treatment
36. Lathrop KL, Gupta D, Kagemann L, Schuman JS, options and emerging therapies. Stem Cells Int.
Sundarraj N.  Optical coherence tomography as a 2016;2016:9798374.
rapid, accurate, noncontact method of visualizing 51. Chang CY, Green CR, McGhee CN, Sherwin

the palisades of Vogt. Invest Ophthalmol Vis Sci. T.  Acute wound healing in the human central cor-
2012;53(3):1381–7. neal epithelium appears to be independent of lim-
37. Fernandes M, Sangwan VS, Rao SK, Basti S,
bal stem cell influence. Invest Ophthalmol Vis Sci.
Sridhar MS, Bansal AK, et  al. Limbal stem cell 2008;49(12):5279–86.
transplantation. Indian J Ophthalmol. 2004;52(1): 52. Reynolds BA, Weiss S.  Generation of neurons

5–22. and astrocytes from isolated cells of the adult
38. Geerling G, Daniels JT, Dart JK, Cree IA, Khaw
mammalian central nervous system. Science.
PT.  Toxicity of natural tear substitutes in a fully 1992;255(5052):1707–10.
defined culture model of human corneal epithe- 53. Yokoo S, Yamagami S, Yanagi Y, Uchida S, Mimura
lial cells. Invest Ophthalmol Vis Sci. 2001;42(5): T, Usui T, et al. Human corneal endothelial cell pre-
948–56. cursors isolated by sphere-forming assay. Invest
39. Mantelli F, Argueso P.  Functions of ocular surface Ophthalmol Vis Sci. 2005;46(5):1626–31.
mucins in health and disease. Curr Opin Allergy Clin 54. Li H, Dai Y, Shu J, Yu R, Guo Y, Chen J. Spheroid cul-
Immunol. 2008;8(5):477–83. tures promote the stemness of corneal stromal cells.
40. Schornack MM.  Limbal stem cell disease: man-
Tissue Cell. 2015;47(1):39–48.
agement with scleral lenses. Clin Exp Optom. 55. Yoon JJ, Wang EF, Ismail S, McGhee JJ, Sherwin
2011;94(6):592–4. T. Sphere-forming cells from peripheral cornea dem-
316 S. Ismail et al.

onstrate polarity and directed cell migration. Cell Biol 58. Huang SU, Yoon JJ, Ismail S, McGhee JJ, Sherwin
Int. 2013;37(9):949–60. T. Sphere-forming cells from peripheral cornea dem-
56.
Chang CY, McGhee JJ, Green CR, Sherwin onstrate a wound-healing response to injury. Cell Biol
T.  Comparison of stem cell properties in cell Int. 2015;39(11):1274–87.
populations isolated from human central and lim- 59.
Mathan J, Ismail S, McGhee J, Sherwin
bal corneal epithelium. Cornea. 2011;30(10): T. Implantation of human peripheral corneal spheres
1155–62. into cadaveric human corneal tissue. Bio-Protocol.
57. Mathan JJ, Ismail S, McGhee JJ, McGhee CN,
2017;7(14):e2412.
Sherwin T. Sphere-forming cells from peripheral cor- 60. Yoon JJ, Ismail S, Sherwin T. Limbal stem cells: cen-
nea demonstrate the ability to repopulate the ocular tral concepts of corneal epithelial homeostasis. World
surface. Stem Cell Res Ther. 2016;7(1):81. J Stem Cells. 2014;6(4):391–403.
Eye Platelet-Rich Plasma (E-PRP)
for Corneal Regeneration
22
Alejandra E. Rodríguez and Jorge L. Alió

22.1 Introduction homeostasis and the regeneration of the ocular


surface [2], as well as growth factors, cytokines
There are different situations, in which the ocu- and elements of the extracellular matrix [3].
lar surface can be altered, from a mild form such These natural physiological processes of res-
as dry eye or environmental exposure to a very toration of damaged tissue are permanent activi-
severe one that can even compromise vision, ties that occur in our body without being aware of
reaching blindness as in the case of ocular cica- it, but there are situations in which by the severity
tricial pemphigoid or Stevens-Johnson syn- of the damage or by the inability of the body
drome [1]. itself to resolve the situation, administration of
Some of the conditions that produce ocular treatments is necessary to solve the problem.
surface disorders are ocular dryness, dry kerato- Treatments range from the administration of
conjunctivitis, recurrent corneal erosions, persis- topical artificial tears to the practice of complex
tent epithelial defects, neurotrophic keratopathies, ocular surface restoration surgeries such as cor-
ocular surface syndrome after LASIK, torpid cor- neal transplants in their various modalities [1, 4–
neal ulcers, graft-versus-host disease and ocular 7]. Within the wide range of therapeutic
cicatricial pemphigoid [1], among others. possibilities available, conventional treatments
After the alteration or injury of the conjunc- are sometimes not effective in solving the patho-
tival or corneal tissue, a biological process of logical condition.
regeneration occurs to restore that damaged In this aspect, the administration of blood
ocular surface. This is a complex process involv- products as a therapeutic option has acquired
ing the epithelial stem cells of the cornea and great prominence, due to its biological composi-
the conjunctiva since they are responsible for tion and its ability to induce the regeneration of
affected tissues [8–10].

A. E. Rodríguez (*)
Laboratory of the Research, Development and 22.2 Platelets
Innovation Department, Vissum Innovation,
Alicante, Spain
Platelets were discovered in 1882 and have tradi-
e-mail: alejandra@vissum.com
tionally been presented as cytoplasmic remains
J. L. Alió
of megakaryocytes [11]. At present, they are con-
Professor and Chairman of Ophthalmology,
University Miguel Hernandez, Vissum-Instituto sidered to be anucleated cells that come from the
Oftalmologico de Alicante, Alicante, Spain megakaryocytes of the bone marrow that, after a

© Springer Nature Switzerland AG 2019 317


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_22
318 A. E. Rodriguez and J. L. Alió

complex mechanism of nuclear and cytoplasmic These proteins perform tissue repair functions
maturation, become free platelets [12]. In a and influence the vascular reactivity of blood
healthy adult, the physiological range of platelets cells, angiogenesis and inflammation [21]. The
is 150–400 × 103 cells/ml, with an average pro- proteins secreted by the alpha granules also inter-
duction of around 1 × 1011 platelets daily [13]. vene in the cellular defence against exogenous
The main function of platelets or thrombo- agents through the production of signalling pro-
cytes is to ensure haemostasis and prevent teins that attract macrophages.
blood loss when blood vessels rupture. In this Platelets are essential for the innate immune
process, they adhere to the exposed sub-endo- response and to fight infections caused by viruses,
thelial collagen, form a platelet plug and then bacteria or other microorganisms. They help to
activate [14]. maintain and modulate inflammation and are an
Platelets also participate in the activation of important source of pro-inflammatory molecules
the coagulation cascade and secondary haemo- since they secrete P-selectin, tissue factor,
stasis [15]. The accidental rupture of platelets CD40L and metalloproteinases.
produces the formation of a certain amount of They also participate in skin diseases, aller-
thrombin, and if the generated thrombin is large gies, rheumatoid arthritis or liver disease, while,
enough, the mechanisms that convert fibrinogen paradoxically, platelet-rich autologous plasma
into fibrin are triggered, forming the thrombus and platelet concentrates are used as an aid to
composed of fibrin and erythrocytes. Thrombosis promote tissue repair and cell growth [22].
causes myocardial infarction if the affected organ Under physiological conditions, platelets are
is the heart [16] and apoplexy if it occurs in the circulating in a non-active form, and on their sur-
brain vessels [17]. face, they express a relatively small number of
Although central role is in the prevention of molecules, which, in their activated state, will
bleeding, platelets contribute to various processes facilitate their interaction with other platelets and
that extend beyond haemostasis and thrombosis, cells in their environment [13].
including the promotion of inflammatory and Platelets change shape with activation due to
immune responses, maintenance of vascular complex modifications in the skeleton and cyto-
integrity and contribution to healing of wounds. skeleton of their membrane. With activation, they
Platelets can recruit leukocytes and progenitor experience the release of granules, dense bodies
cells to sites of vascular injury and thrombosis; and the content of lysosomes.
they store, produce and release pro-inflamma- This activation, also known as degranulation,
tory products, anti-inflammatory and angiogenic can be caused by a physical stimulus, chemical
factors and microparticles in the circulation, in or a combination of both and causes the alpha
addition to stimulating the generation of throm- granules to fuse with the platelet membrane,
bin [18]. where some of the secretory proteins (e.g. cer-
In electron microscopy, the different organ- tain factors of growth) pass to the active state
elles that contain the platelets are observed: mito- and thus are released to the outside where they
chondria, peroxisomes, ribosomes, glycogen and bind to the receptors of the target cells (endothe-
granules; the latter are divided into three types: lial cells, epithelial cells, osteoblasts, among
alpha, delta or dense granules and lambda [19]. others) [23].
The alpha granules of platelets contain more Growth factors released by activated platelets
than 30 bioactive proteins, and despite the initiate and modulate healing because they
absence of the nucleus and DNA, platelets have induce chemotaxis, cell proliferation and differ-
a system to perform protein synthesis and have entiation, neovascularization and deposition of
copies of mRNA for almost a third of known extracellular matrix [24]. Specifically, alpha
proteins in the human genome; they process the granules constitute the largest reservoir of
mRNA and efficiently translate different pro- growth and mitogen factors that are released dur-
teins [20]. ing platelet activation. These factors are involved
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 319

in processes of chemotaxis, cell proliferation, molecules, such as fibroblasts, osteoblasts, endo-


differentiation and angiogenesis, thus promoting thelial cells, leukocytes, monocytes and macro-
healing [22, 25]. phages [26].
We will describe below the most relevant GFs
involved in the ocular regeneration.
22.3 Growth Factors

Growth factors (GF) are a group of substances of 22.3.1 Platelet-Derived Growth


peptide nature whose function is intercellular Factor (PDGF)
communication at the molecular level. These are
capable of modifying cellular biological Platelet-derived growth factor (PDGF) is a known
responses as they intervene by regulating the pro- modulator of fibroblast cell mitosis and chemo-
cesses of migration, proliferation, cell differenti- taxis [28]. It indirectly promotes angiogenesis
ation and metabolism and even apoptosis. through macrophages through a mechanism of
The main function of growth factors is that of chemotaxis [29].
external control of the cell cycle. These promote In studies carried out, the expression of
increased cell size by increasing the protein syn- PDGF-AA, PDGF-BB and PDGF-AB and their
thesis of the cells on which they act [26]. corresponding receptors in the human cornea and
The most important GFs coming from plate- their participation in proliferation and chemo-
lets are transforming growth factor beta (TGF-β), taxis in assays with cultures of human corneal
basic fibroblastic growth factor (FGFb), platelet-­ fibroblasts have been demonstrated, showing that
derived growth factor (PDGF), vascular endothe- the PDGF-BB factor has a significantly higher
lial growth factor (VEGF), connective tissue chemotactic effect compared to PDGF-AA or
growth factor (CTGF), type 1 insulin-like growth PDGF-AB.  Due to the chemotactic effect that
factor (IGF-1) and epidermal growth factor PDGF presents in the healing of corneal wounds,
(EGF) [27]. the hypothesis is hypothesized that said factor
Growth factors act locally. Cell stimulation regulates the chemoattractant of neutrophils,
can be done through an autocrine system, a IL-8 [30, 31].
mechanism by which cells produce and respond Studies conducted with PDGFs and their
to the biological mediator, or by a paracrine sys- receptors (PDGFRs) in animal development and
tem in which the cell that generates the factor is disease have revealed their role as CF in organo-
in the vicinity of the cells on which it acts. genesis, cancer, vascular inflammation and tissue
Generally, GFs are synthesized in the form of fibrosis [32].
precursors, which is why a proteolysis process is PDGF-BB is used to improve wound healing
required for their release in the active form. The in clinical practice for about 25  years.
mechanism of action that they carry out begins Recombinant human PDGF-BB, called becapler-
when they join specific membrane receptors, min, has been introduced into the clinic as a topi-
there being for each FC a receptor or set of spe- cal treatment for chronic lower extremity
cific receptors. The process is mediated by a sys- neuropathic diabetic ulcers [33].
tem of second messengers that activate a cascade It has also been used against pressure ulcers
of signals. Thanks to this mechanism, the action and to accelerate healing in various surgical pro-
of the factors does not stop at the site of the cedures. PDGF-BB-induced tissue repair mecha-
injury, although they have disappeared from the nisms appear to involve fibroblast proliferation,
environment, since they have activated the way of collagen production and neovessel formation
second messengers. [34] and may involve direct effects on mesenchy-
Apart from the alpha granules of the platelets mal cells expressing PDGFR-α [35].
that constitute a large reservoir of growth fac- Its clinical efficacy has been demonstrated in
tors, there are other cell types that produce these several phase III studies [36], and the combined
320 A. E. Rodriguez and J. L. Alió

results suggest that the topical application of towards the signalling pathway of TGF-β [40]
PDGF-BB is safe and well tolerated. A study with can be considered.
more than 900 patients confirmed that topical The TGF-β signalling pathway plays a key
treatment with PDGF improved the healing of role in the function of ocular fibroblasts, their dif-
chronic foot ulcers of full-thickness diabetics [37]. ferentiation and their proliferation. The suppres-
The local administration of PDGF-BB has also sion of the TGF-β pathway has shown a decrease
been tested in patients with severe periodontal in extracellular matrix deposition in ocular fibro-
disease and has been found to increase biomark- blasts in vitro [42].
ers for bone metabolism and rotation [38] and On the other hand, it has been demonstrated
provide improved periodontal regeneration [39]. that cyclosporine A inhibits TGF-β2 induced by
the myofibroblasts from human pterygium fibro-
blasts obtained from primary culture [43], so
22.3.2 Transforming Growth Factor these results could indicate the therapeutic poten-
Beta (TGF-Β) tial of cyclosporine A against progression of the
pterygium.
The transforming growth factor beta (TGF-β) is
multifunctional and one of the most important
factors involved in modulating the behaviour of 22.3.3 Epidermal Growth Factor
ocular tissues. In addition, this ligand participates (EGF)
in cell migration and proliferation, cell death,
protein synthesis during development and repair The epidermal growth factor (EGF) plays a very
of tissues among other physiological or patho- important role in the migration and proliferation
logical processes [40]. of the corneal epithelium to improve the pro-
TGF-β is an intriguing cytokine exhibiting cesses of tissue regeneration, as well as the stim-
dual activities in malignant disease. On the one ulation of the granulation tissue.
hand, it is an important mediator of cancer inva- Studies show that the mitogen effect of EGF
sion, metastasis and angiogenesis, while on the on the proliferation of corneal epithelial cells
other hand, it exhibits antitumour functions [41]. requires the suppression of expression of the
In most cases, TGF-β improves the production Pax-6 gene specific to the eye. Fibroblasts, pro-­
of the extracellular matrix and suppresses cell osteoblasts and pre-chondrocytes express a high
proliferation, for example, that of epithelial cells number of receptors for EGF [44].
in the presence of other factors. On the other On the other hand, it is known that EGF
hand, this GF is also able to induce the produc- increases the concentration of intracellular cal-
tion of other factors such as connective tissue GF cium [Ca++]i and activates the protein kinase ERK
(CTGF), platelet-derived GF (PDGF), fibroblast-­ ½ to stimulate the mucin secretion of conjuncti-
derived GF (FGF) and vascular endothelial GF val goblet cells [45]. It is also known that scarring
(VEGF), as well as stimulating its own produc- of the corneal epithelium is stimulated by EGF
tion. All of them play an important role in restor- and is partially mediated through the 12/15-LOX-­
ing normal tissue after suffering an injury [27]. LXA4 pathway [46].
There are three isoforms of TGF-β, called To achieve an ideal wound-healing process
TGF-β1, TGF-β2 and TGF-β3, whose presence is that promotes healthy growth of tissue with less
found in the tissues of mammals; however, its scarring, a gel-based topical drug delivery system
expression is not uniform. As in other tissues, composed of three different polymers chitosan,
overactivation of TGF-β underlies the pathogen- dextran sulphate and polyvinylpyrrolidone K30
esis of ocular diseases that can lead to impaired (CDP) was prepared. Applied topically, EGF evi-
vision and the alteration of ocular tissue homeo- denced clearly beneficial effects by accelerating
stasis. Therefore, the possibility of developing the wound healing process and producing less
therapeutic strategies for such diseases oriented scarring [47].
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 321

22.3.4 Vascular Endothelial Growth a wide variety of physiological processes such as


Factor (VEGF) phagocytosis, proliferation, adhesion and cell
migration [53, 54].
Vascular endothelial growth factor (VEGF) has Fibronectin is able to bind fibrin, collagen,
been known for a long time as a potent stimulator glycosaminoglycans, heparin, certain proteogly-
of blood vessel proliferation. It is an endothelial cans and cellular plasma membrane proteins such
cell mitogen, angiogenic factor and a potent as integrins. Therefore, it establishes connections
mediator of vascular permeability, in addition to between molecules of the extracellular matrix
being involved in the growth of nerves [48]. and between molecules of the cells with the
In vitro experiments have shown that VEGF extracellular matrix. Fibronectin molecules can
and its receptors are expressed by neurons and appear forming insoluble protein fibres in the
stimulate the proliferation of cortical neurons, connective tissues or soluble in the plasma of
provide protection to central and peripheral neu- body fluids, such as blood [55].
rons against hypoxia-induced death and promote Among its main functions is to intervene in
axonal growth in peripheral neurons [49]. the remodelling of tissues during embryogenesis
On the other hand, overexpression of VEGF is and participate in the healing process of a lesion
a factor that contributes to the development of after the formation of a fibrin clot, generated by
disease. For example, solid tumours require an the activation of the haemostatic system [56].
increase in blood supply if they will continue to
grow beyond a certain size and tumours that
express VEGF are able to continue growing; this 22.5 Properties of Blood
process is called angiogenesis [50]. Derivatives with and Without
Overexpression of VEGF can also lead to vas- Platelets
cular disease in the retina and other parts of the
body. A decrease in the level of VEGF in the pul- The blood products have shown their ability to
monary arteries has also been associated with the improve healing and stimulate the regeneration
condition of pulmonary emphysema [50]. of different tissues, an effect that is due to the
Since cancer growth is stimulated by VEGF, growth factors and bioactive proteins that are
researchers have made numerous attempts to synthesized and present in the blood [21].
decrease its expression to prevent angiogenesis Platelet-rich plasma (PRP) is a blood deriva-
and tumour growth. Two drugs that have been tive obtained by centrifugation of whole blood
successful in slowing the progression of diseases characterized by having a high concentration of
that depend on VEGF are bevacizumab and platelets [29].
ranibizumab [51]. The most important growth factors found in
platelet-rich plasma, from which cells they derive
and which are their main functions, are described
22.4 Adhesion Molecule: in [57].
Fibronectin In addition to platelets, the plasma contains
some cell adhesion molecules such as fibrin, fibro-
Fibronectin is an adhesion protein, which is free nectin and vitronectin which provide structural
in plasma, and is a major constituent of the com- support for cell migration, proliferation and three-
position of the extracellular matrix. Plasma dimensional growth in the tissues in which they
fibronectin levels are approximately 300 ± 100 act [23]. On the other hand, the cytokines PF4 and
μg/mL [52]. CD40L present in the plasma promote tissue repair
This protein has been widely studied for its and influence the vascular reactivity of blood cells,
ability to interact with various cells and macro- angiogenesis and inflammation [21].
molecules, influencing its behaviour and the bio- Different in vitro assays have been performed
chemical reactions of various systems, as well as to study and try to understand the effect that dif-
322 A. E. Rodriguez and J. L. Alió

ferent types of blood preparations based on plate- duction of type I collagen and that the synthesis
lets produce in cells and tissues. of VEGF occurred only in the tendon cells. They
Hartwig et  al. studied the epitheliotrophic concluded that the biological effects of PRGF
capacity of platelet and serum concentrates in may depend on the concentration of platelets
human cells of the corneal epithelium. After applied and the type of cells used [62, 63].
stimulation they observed a significant increase More recent trials have studied the biological
in cell proliferation of the platelet concentrate. effects of PRGF and autologous serum in cul-
The effects of migration and cell differentiation tures of corneal stromal human keratocytes (HK)
were positive with both preparations but slightly and conjunctival fibroblasts (HConF). The results
higher with serum [58]. showed higher significant levels of all the growth
Following this line of work, in 2005 they car- factors analysed in the PRGF when compared
ried out another study also on human epithelial with in the autologous serum. It was also observed
cells in an in vitro cell culture model. This time that PRGF produced a significant increase in the
they compared the epitheliotrophic capacity of biological results of both HK and HFCF cells and
the cells when faced with fresh frozen plasma reduced fibroblast differentiation to myofibro-
(without platelets) and with serum. The results of blasts induced by TGF-β1 in contrast to autolo-
cell proliferation, migration and differentiation gous serum. This latter finding suggests that
were significantly better in serum than in fresh PRGF can improve the treatment of superficial
frozen plasma (without platelets). In this study, wound healing by minimizing scar formation
none of the preparations studied contained plate- compared to AS while promoting cell prolifera-
lets, but it was inferred that the greater epithelio- tion and migration [64, 65].
trophic capacity of the serum could be due to the Freire et al. did a study comparing the effects
higher concentration of proliferation mediators of autologous serum, PRP and PRGF on cultures
such as the EGF and PDGF growth factors and of human corneal cells. All the compounds had
the higher vitamin A content [59]. high levels of growth factors, but the EGF values​​
In a subsequent trial, they compared the cor- were higher with PRGF than with the rest. The
neal epitheliotrophic capacity using three differ- cell proliferation rates were higher with autolo-
ent blood preparations: serum, fresh frozen gous serum and with PRGF [66].
plasma and a platelet concentrate that was acti- Subsequently, the same authors conducted an
vated with thrombin. In this case, the cell prolif- experimental trial in rabbits, in which they
eration results were better with the platelet induced a corneal erosion to be treated with the
concentrate than with the rest, while the serum same three previous blood derivatives. All the
showed better results of migration and cell dif- compounds improved healing, but the corneas
ferentiation [60]. treated with autologous serum and PRGF healed
The plasma rich in growth factors (PRGF) is more quickly [67].
another blood derivative, which as its name indi- Experimental studies were also carried out but
cates has a high concentration of growth factors; with clinical sense performing photorefractive
and due to its preparation, it does not contain keratectomy (PRK) to C57BL/6 mice to assess
platelets [61]. the healing process after the application of
Studies conducted with cultures of human plasma rich in growth factors. The control ani-
fibroblasts obtained from the skin, synovium and mals showed a higher corneal haze index, and the
tendon revealed a significant proliferative epithelium that has been generated was hyper-
response in all these cells when stimulated with plastic, while those treated with PRGF presented
different preparations of PRGF.  The maximum a greater regeneration of the corneal epithelium
proliferation occurred with concentrations of and lower haze [68].
PRGF from twice to four times the initial. They More recent studies have demonstrated the
also observed that PRGF increased the synthesis ability of s-PRGF to promote the proliferation
of hyaluronic acid but that it did not alter the pro- and migration of rabbit epithelial and corneal
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 323

cells and to accelerate corneal re-epithelialization


rich fibrin preparation (PRF) with different blood
in in vivo assays [69]. derivatives. In this study, it determines the con-
There are other studies in which blood deriva- centrations of three pro-inflammatory cytokines
tives rich in platelets have consistently shown to (IL-1β, IL-6 and TNF-α), the anti-inflammatory
enhance the proliferation, migration and differen- cytokine IL-4 and the angiogenesis promoter
tiation of stem cells, with varied applications in VEGF. Their results revealed that the PRF could
tissue engineering and regenerative medicine be an immune regulation node with inflammatory
[70–73]. retro-control capacity. This concept could explain
In addition to the aforementioned blood prod- the reduction of postoperative infections when
ucts, based on plasma with or without platelets, the PRF was used as a surgical additive [78].
we find autologous serum (AS) that is also used Other trials have demonstrated the effects of
in the field of ophthalmology. Autologous serum, PRP on the release of cytokines from monocytes
although it does not contain platelets, presents and the generation of lipoxin A4 (LXA4) which
variable amounts of growth factors, which is the is an endogenous anti-inflammatory. In this case,
reason why it is included in this section [74]. PRP promoted significant changes in the release
In vitro toxicity studies using human corneal of cytokines and pro-inflammatory chemokines
epithelial cells have shown reduced toxicity when mediated by monocytes and increased lipoxin
compared to hypromellose without a preserva- A4. This suggests that PRP can suppress cyto-
tive, used to prepare artificial tears [75]. kine release, limit inflammation and thus pro-
Several experimental studies have also shown mote tissue regeneration [79].
their regenerative properties. Autologous serum Mazzocca et  al. evaluated in  vitro the anti-­
enhances the regenerative effect of mesenchy- inflammatory effects of PRP, ketorolac and meth-
mal stem cells in damaged rat corneas [76]. In ylprednisolone in human umbilical vein
addition, it promotes the migration and differen- endothelial cells. Although PRP and ketorolac
tiation of corneal epithelial cells [58, 59]. AS did not produce such a significant reduction in
also presents high amounts of EGF and induces markers of cellular inflammation as methylpred-
the proliferation of epithelial cells of the human nisolone, they reduced cellular inflammation
cornea [66], promoting corneal reepithelization compared to control. Therefore, they concluded
in rabbits’ eyes with previously debrided epithe- that PRP may have clinical application as an
lium [67]. injectable anti-inflammatory drug [80].
In addition, it has been observed that AS On the other hand, Schar et al. studied the lev-
increases the conjunctival expression of mucin els of various growth factors and the inflamma-
MUC5AC, increasing as a consequence the gob- tory cytokine IL-1β in platelet-rich plasma
let cell’s density in the conjunctiva, mainly in preparations with leukocytes (L-PRP). In this
patients with severe dry eye [77]. trial, the levels of IL-1β and VEGF were signifi-
cantly lower in the L-PRP than in the autologous
serum, demonstrating its lower pro-inflammatory
22.6 Inflammation and Blood and angiogenic power [81].
Derivatives Following this same line of work, Anitua et al.
performed the same assay, comparing a platelet-­
There are many research groups that have studied rich plasma scaffold with leukocytes (L-PRP) with
the markers of inflammation in different blood a PRGF scaffold without leukocytes. The concen-
preparations with platelets with variable results, tration of all the growth factors released was com-
mainly due to the differences in PRP preparation parable between both methods, with no differences
methods [73]. between them. But after 3 days of incubation in the
In fact, there are several who are in favour of L-PRP scaffolds, the VEGF levels decreased dras-
the presence of leukocytes in their formulations, tically. On the other hand, the inclusion of leuko-
such as Dohan et al. who compared their platelet-­ cytes in the L-PRP scaffolds produced a significant
324 A. E. Rodriguez and J. L. Alió

increase in the pro-­inflammatory cytokines, IL-1β On the other hand, other researchers have
and IL-16, with respect to the PRGF scaffolds that applied allogeneic serum in patients with severe
presented very low values. Scaffolds of PRGF dry eye who suffered from graft-versus-host dis-
remained stable during the 8  days of incubation ease, in whom it was not possible to apply autolo-
[82]. The subsequent addition of leukocytes to the gous serum. The patients improved their
PRGF scaffolds stimulated a more pro-inflamma- symptoms and signs of dryness and had no
tory environment, secreting more quantities of adverse effects [88, 89].
pro-­inflammatory cytokines such as IL-1β, IL-6,
IL-8, IL-16 and TNF-α, altering the properties of
fibrin [83]. 22.7 Antimicrobial Activity
The inactivation of plasma by heat has also of the Platelet-Rich Plasma
been studied at 56 °C in order to reduce the activ-
ity of complement and immunoglobulin E and in Searching more properties and biological effects
this way be able to apply it in patients with auto- of platelet-rich plasma, several authors have stud-
immune diseases. The cell proliferation and ied its antimicrobial capacity.
migration assays gave good results with the In addition to the large amount of proteins
plasma heated to 56 °C, compared to the unheated and growth factors, they have discovered that
plasma. On the other hand, most of the proteins platelet-­rich plasma releases immunomodulatory
and morphogens involved in healing were ade- agents with antimicrobial activity such as plate-
quately preserved [84]. let microbicidal proteins (PMPs), thrombocydins,
To evaluate the anti-inflammatory power of RANTES, connective tissue-activating peptide
the plasma rich in growth factors, the blood of and platelet factor 4 (PF-4), among others [90, 91].
patients with graft-versus-host disease was con- Other authors have directly studied the inhibi-
fronted with ocular surface fibroblasts in a pro-­ tory effect of platelet concentrates against differ-
inflammatory environment (IL-1β and ent cultures of microorganisms. Drago et  al.
TNF-alpha). The results concluded that even in demonstrated the inhibitory capacity of PRP
this environment, the plasma rich in growth fac- against strains obtained from the oral cavity of
tors had regenerative power and that it also had Enterococcus faecalis, Candida albicans,
an anti-inflammatory effect [85]. Streptococcus agalactiae and Streptococcus ora-
In a further step towards controlling the anti-­ lis, but not Pseudomonas aeruginosa [92]. In
inflammatory effects of blood derivatives, some general, most authors agree on the inhibitory
authors have studied the effects of allogeneic effects of PRP against Staphylococcus aureus
blood products. and Staphylococcus epidermidis whether they are
Stenwall et al. have investigated the composi- methicillin-sensitive or methicillin-resistant [92–
tion of cytokines and the anti-inflammatory effects 94] and Streptococcus agalactiae [92, 94]; but
of allogeneic serum preparations to assess their they differ in the antimicrobial activity of PRP
best use as eye drops. For this purpose, the serum against Escherichia coli [94, 95] and Enterococcus
of 15 donors was inactivated, and then the regula- faecalis [92, 95]. All agree on the inability of
tory capacity of inducing T cells was studied [86]. PRP to inhibit the growth of Pseudomonas aeru-
Although the vast majority of blood deriva- ginosa [92, 95].
tives used in ophthalmology is of autologous ori-
gin, there are some authors who have treated
patients with serum of allogeneic nature. 22.8 O
 btaining Blood Derivatives
Chiang et  al. applied allogeneic serum from Used in Ophthalmology
relatives to patients with persistent epithelial defect
in whom it was not feasible or appropriate to use The denomination and way of obtaining products
autologous serum, and none of the patients pre- prepared with blood derivatives in the literature is
sented an adverse effect during the treatment [87]. very diverse and leads to constant misunderstand-
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 325

ings in readers, scientists and reviewers [96]; so Briefly, the process to prepare the E-PRP is:
we will try to clarify this aspect as much as
possible. –– Blood collection in ten sterile tubes, each of
Initially, the term “PRP” was described by 10  ml with 1  ml of 3.2% sodium citrate
Marx [23, 97] and, in general, refers to a two-step (Monovette, Stardest) to prevent coagulation.
centrifugation process, whereby a platelet-rich –– Centrifugation in conventional laboratory cen-
plasma is obtained which is subsequently acti- trifuge: 10 min at 1600 rpm [9]. These condi-
vated with calcium chloride or thrombin, imme- tions may vary depending on the centrifuge
diately before use. This PRP has a gel or clot and the size of the tubes used.
consistency or even more solid fibrin mesh and –– Separation of platelet-rich plasma in a laminar
had always been used in surgical procedures as flow hood, in strict conditions of sterility and
an adjunct to healing, in addition to direct appli- with disposable material.
cations on skin ulcers in diabetic patients. –– 4 ml of E-PRP are placed in each of ten sterile
Given that there is a huge variety of systems to glasses or plastic dropper bottles (in case it is
obtain autologous blood derivatives, with vari- prepared as eye drops).
able amounts of platelets and growth factors, we –– If activation is needed: 50  μl 10% calcium
will focus on those used in the field of ophthal- chloride per 1  ml of E-PRP (clot or fibrin
mology, which are those related to the topic of membrane).
this chapter. –– To have a more exhaustive control of the
obtaining process, a haemocytometer or blood
cell counter can be used to know the haemato-
22.8.1 E-PRP (Eye Platelet-Rich logical parameters of the whole blood before
Plasma) the centrifugation (haematocrit, RBC count,
leukocytes, platelets) and after the preparation
The acronym E-PRP comes from its denomina- of the E-PRP (count of red blood cells, leuko-
tion in English as “Eye-PRP” (eye platelet-rich cytes and platelets).
plasma) and refers to the platelet-rich plasma that
is used in ophthalmology. Under these conditions, the resulting E-PRP
The term E-PRP was introduced by Alio et al. has a platelet enrichment index of 1.6–2.5 times
[8] to differentiate it from the term “PRP” that above the baseline values ​​(whole blood). The
had been used in its previous publications related E-PRP eye drops is a 100% autologous com-
to this topic [98–100] and that caused confusion pound since it does not contain any additives.
in the literature. The process of obtaining E-PRP In Fig. 22.1 the different phases of the E-PRP
is different from that used by Marx et al. [23]. preparation can be observed from the centrifuged
The E-PRP is very versatile since it is possible blood.
to use it as eye drops (without activating), as a clot Patients are instructed on how to wash their
(activated) or as an autologous membrane of fibrin hands before using the product, and how to
(activated). The autologous E-PRP in the form of ­perform a correct application avoiding contact
eye drops is used as ocular topical treatment for with the dropper with any part of the eye. These
various pathologies of the ocular surface [98–100] precautions are carried out in order to avoid fur-
and in the form of a clot or autologous fibrin mem- ther contamination.
brane in surgical procedures for the reconstruction A bottle of E-PRP (4 ml) in use can be kept at
of the ocular surface [98, 101, 102]. 4  °C for a week, after which it should be dis-
E-PRP for ophthalmological use is carried out carded. The rest of the E-PRP bottles should be
through a one-step centrifugation process, and kept frozen at −20  °C for a maximum of
the final rate of platelet concentration will depend 3 months.
on whether it will be used as eye drops, as a clot This system of preparing platelet-rich
or as an autologous fibrin membrane. plasma is called “open” since it is necessary to
326 A. E. Rodriguez and J. L. Alió

Conservation:
1 week at 4° C
3 months at –20° C

4 ml

PRP E-PRP eyedrop

Cl2Ca

30
min
37° C
Centrifugated PRP
E-PRP clot
blood
Autologous
thrombin
Cl2Ca

60 min
37° C
PPP
Platelet poor Fibrin membrane
plasma ø: 18–22mm – Thickness: 1mm
Immediate use

Fig. 22.1  Diagram with the steps followed in the preparation of the different types of E-PRP: eye drops, clot and fibrin
membrane

open the tubes with blood for handling, which These systems are commercial kits, consisting
is why they must be performed under sterile of specially designed blood collection tubes, with
conditions. their own consumables. In general, you must pur-
chase the specific centrifuge for that commercial
kit, as it comes with the established ­centrifugation
22.8.2 Closed Systems for Obtaining parameters and cannot be modified by the user.
PRP
22.8.2.1 RegenKit® BCT
There are other methods to obtain platelet-rich It is a closed system for the production of plasma
plasma with a “closed” system; in these cases the rich in platelets. The RegenKit® BCT belongs to
tubes are manipulated to obtain the PRP, but the company RegenLab USA. It is approved by
without the opening of them. They use syringes the FDA (Food and Drug Administration). The
and needles. name of the product is A-PRP, acronym of autol-
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 327

ogous platelet-rich plasma. Each Regen BCT autologous serum lacks platelets and any other
tube with 8  ml of blood generates 4–5  ml of cells. It is considered an “open” method since for
autologous plasma enriched with platelets, at a its preparation, it is necessary to open the tubes to
concentration of 1.6 of the basal levels [96]. manipulate the blood.
Although the methods used to obtain autolo-
22.8.2.2 ANGEL® gous serum do not usually follow standardized
ANGEL is a closed system to obtain platelet- protocols, except very few [105], and the results
rich plasma from the supplier Arthrex. This obtained from the literature are extremely dispa-
device is based on a centrifugal equipment plus rate, the basic steps for obtaining autologous
software that uses a system of tubes and pro- serum would be:
grammed centrifuges, to obtain the PRP.  It
requires user experience and allows selecting the –– Spontaneous coagulation of the blood
degree of platelet concentration in a wide range extracted without anticoagulant
(3× to 18×). The device that acts as a plasma –– Incubation for 1 h at room temperature
fractionator is the centrifuge itself with a com- –– Centrifugation
plex set of fungibles. The content is not exposed –– Collection of supernatant fraction
to the environment [103]. –– Dilution
–– Conservation
22.8.2.3 PRGF Endoret®
The preparation of this plasma rich in growth fac- For its preparation, blood is used without
tors is carried out with the closed system PRGF anticoagulant. The blood is allowed to coagulate
Endoret®, a commercial kit from BTI spontaneously in the tubes for 1 h at room tem-
(Biotechnology Institute, S.L., Miñano, Álava, perature and subsequently centrifuged to release
Spain). It is necessary to acquire the kit and the the serum from the clot with red blood cells and
specific centrifuge (BTI System IV, Vitoria, leukocytes. Some authors prefer to wait up to
Spain). The process is based on a double centrifu- 2 h [105].
gation and activation technique. With respect to centrifugation, there are dis-
Briefly the steps required are the following: crepancies in the literature regarding the amount
of revolutions per minute (rpm) used and the time
–– First centrifugation: to obtain the platelet-rich of centrifugation used for their separation [67,
plasma, using a specific centrifuge (580 g for 105–107]. A standard option would be 3000 rpm
8 min). for 15 min, which does not produce haemolysis
–– Plasma collection and activation (products and separates an adequate amount of serum
provided in the kit). (approximately half of the total blood volume)
–– Incubation of the plasma in glass tubes at while maintaining the unaltered properties [66].
37 °C for 1 h. Most authors perform the preparation of autol-
–– Second centrifugation: 1000 g for 10 min. ogous serum with fresh serum, continuing with
–– Separation of plasma rich in growth factors the procedure immediately after its separation
and storage in mini vials. [59, 105, 108–110]. But here there is no consen-
–– It can be stored up to 3  months frozen at sus, and some authors choose to freeze the serum
−20 °C [104]. at −80 °C to prepare it at another time [107].
In relation to the subsequent dilution that is
made to the AS, as expected, there is also no
22.8.3 Autologous Serum agreement between the authors. Some of them
opt for dilution at 20% [105, 107, 111], and oth-
Autologous serum, unlike plasma, is obtained ers use it at 50% [75, 110, 112–114]. Even some
from blood without anticoagulant. Since all cel- of them also use it 100% when they do not obtain
lular components are trapped in the clot, the the expected results [75]. Most authors use 0.9%
328 A. E. Rodriguez and J. L. Alió

physiological saline for dilution, while others use gic response, without side effects and without
balanced saline solution (BSS) [115] or sodium altering eye homeostasis.
hyaluronate [112]. Blood derivatives for the treatment of ocular
On the other hand, some carry out a filtering surface diseases have been used for a long time
after its preparation to eliminate any possible with very promising results. One of the first
contamination. To do this, they use a syringe with records in the field of ophthalmology belongs to
0.22 μm pore filters that ensure the elimination of Rosenthal et al. [120], who in 1975 used a mix-
bacteria [105, 112, 116]. ture of autologous platelets, fibrinogen and
With respect to conservation, most agree that thrombin as an ocular surface adhesive for the
keeping it for up to 3 months frozen at −20 °C fixation of lamellar keratoplasties in rabbits.
does not alter the properties of autologous serum One of the most popular blood preparations
[105, 109, 110]. Some authors even assert that and used to date in the field of ophthalmology is
storage at −20 °C for 9 months does not alter the autologous serum. Although there are older
concentration of growth factors [74]. records [121], it was when Fox et al. [116] and
Tsubota et  al. [5] published their work for the
treatment of dry eye with Sjögren’s syndrome
22.9 Clinical Applications that the use of autologous serum became more
of Blood Derivatives popular. Since that time, a great variety of studies
have been carried out to demonstrate its clinical
The ocular surface comprises the corneal and efficacy in the treatment of dry eye, epithelial
conjunctival tissue including all its cellular lay- defects, dry eye post LASIK and other alterations
ers. These structures interact with others, such as of the ocular surface [58, 75, 105, 107, 108,
the eyelids with their lacrimal and Meibomian 122–125].
glands to maintain an adequate tear film and con- But it has been recently that several studies
tribute to eye homeostasis. There are many spe- have been published comparing the effects of
cific situations and pathologies that affect this autologous serum with other blood derivatives
complex balance producing a decompensation such as PRP or PRGF [66, 67, 126].
and altering the surface of the eye accordingly. In The first author to use platelet-rich plasma
this sense, the most frequent pathology that (PRP) in the field of ophthalmology was Alio
affects the ocular surface is the dry eye syndrome et al. in 2007 to treat neurotrophic corneal ulcers
from its mild to very severe states in which the [98]. He then continued using E-PRP eye drops
vision of patients is compromised [117]. for the treatment of dry eye [100], in the postocu-
Common treatments include an immense vari- lar surface LASIK syndrome [99] and other ocu-
ety of artificial tears, lubricants with lipid con- lar surface alterations [8, 9].
tent, liposomal spray, inserts, anti-inflammatories, In their last study with moderate to severe
immunosuppressant such as cyclosporine A or cases of dry eye disease, Alio et al. evaluated the
extracellular matrix regenerators, in addition to use of autologous platelet-rich plasma (E-PRP)
other procedures such as occlusion of lacrimal eye drops as monotherapy. Three hundred sixty-­
punctures or the most recent treatments with eight patients with moderate to severe dry eye
intense pulsed light (IPL) [118, 119]. disease (DED) were included. Subjects were
Therapeutic strategies in ocular surface altera- classified as evaporative DED (EDED) or aque-
tions are increasingly focused on the use of ous deficient DED (ADDED). Two hundred
multiple-­action, less artificial treatments without ninety-seven (80.7%) patients were women, and
potential allergens such as preservatives or other 71 (19.3%) were men. Two hundred thirty-two
products that could induce toxicity in a very vul- (63%) patients had EDED, while 136 (37%) had
nerable ocular surface. These therapies must be ADDED.  After 6  weeks of monotherapy treat-
effective in reducing symptomatology and pre- ment with autologous PRP, dry eye symptoms
venting late complications, with little or no aller- improved in 322 (87.5%) cases. A decrease of
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 329

corneal fluorescein staining (CFS) was observed a


in 280 (76.1%) patients. One hundred six (28.8%)
patients improved at least 1 line of BCVA.  The
scores in the Ocular Surface Disease Index
(OSDI) and the Oxford scale of corneal fluores-
cein staining decreased statistically after the
treatment (p  <  0.05). Regarding the number of
rounds of E-PRP treatment, of the 368 patients
treated, 237 (64.4%) received only one round of
PRP (6  weeks consecutively) and 131 (35.6%)
between two and five rounds [127]. b
Recently, Alio et al. published another work
to evaluate the efficacy of autologous platelet-
rich plasma (E-PRP) eye drops for the treat-
ment of chronic ocular surface syndrome (OSS)
following laser in situ keratomileusis (LASIK)
[128]. It was a prospective interventional clini-
cal study (Registration NCT03322917) what
included 156 eyes of 80 patients. 89.6% of the
patients who had positive CFS before treatment
showed a decrease in at least one quadrant
Fig. 22.2 (a) Eye of a patient suffering OSS post LASIK
to total disappearance on CFS.  Conjunctival before the E-PRP treatment. (b) Same patient than A after
hyperaemia improved in 93.3% of the patients receiving E-PRP during 6 weeks
with previous signs of ocular surface inflam-
mation. There was a significant improve-
ment in logMAR CDVA from 0.14  ±  0.19 to Fig. 22.3 shows the corneal ulcers of two
0.06  ±  0.12 (p  =  0.000), and 74 (71.4%) eyes patients post keratoplasty surgery, before and
improved at least 1 line in CDVA.  They con- after the treatment with E-PRP.
clude that monotherapy with autologous E-PRP We can see the eye of a patient before and
is a well-tolerated, safe and effective treatment after the treatment with E-PRP during 6 weeks.
for the management of post-­LASIK ocular sur- In addition to topical applications such as eye
face syndrome. drops, E-PRP can be prepared in the form of clot
In Fig.  22.2 we can see the eye of a patient or fibrin membrane and can be used as an adju-
suffering OSS post-LASIK before and after the vant in surgical processes of ocular surface
treatment with E-PRP during 6 weeks. reconstruction. In some cases of corneal perfora-
Autologous E-PRP has been also used for the tion, they have used the autologous fibrin mem-
treatment of dormant corneal ulcers secondary to brane with E-PRP clots at the site of injury to seal
corneal surgery and unresponsive to conventional the perforation and recover that cornea [101]. For
treatment [129]. A total of 44 eyes of 28 patients this purpose, both the fibrin membrane and the
were included in the study and treated with autol- E-PRP clot were prepared with the patient’s own
ogous platelet-rich plasma during 6 weeks, after blood just before the operation. Nylon 10-0
presenting corneal ulcer post keratoplasty, refrac- stitches were used to fixate the fibrin membrane
tive surgery, cross-linking or chronic postsurgical to the conjunctiva, and then the E-PRP clot was
corneal oedema. In all, 40 (90.9%) experienced placed over the corneal perforation, underneath
an improvement in their symptoms, 28 patients the fibrin membrane. A temporal partial tarsor-
(65.1%) improved their visual acuity and 26 rhaphy was performed at the end of the proce-
(59.09%) had a decrease in the size of the ulcer or dure. The fibrin membrane was present over the
even a total closure. corneal surface for the first 3–5  days and then
330 A. E. Rodriguez and J. L. Alió

a b

c d

Fig. 22.3 (a) and (c) show the corneal dormant ulcers of two patients’ post-keratoplasty surgery. In (b) and (d), the
corneas appear regenerated after 6 weeks of treatment with E-PRP

gradually disappeared. No evidence of infection Kim et al. used PRP to treat epithelial defects
or inflammation was detected, and in all cases, that did not resolve after corneal infections and
the corneal perforation was sealed. compared it with autologous serum. The healing
In others cases, they have used E-PRP clots rates of the corneal epithelium were significantly
together with a collagenous patch derived from higher with PRP than with autologous serum
bovine pericardium (Tutopatch®) to repair perfo- [133]. Lee et al. used the same PRP preparation
rated corneas [102]. Six patients with perforated of Kim et  al. (2012) for recurrent corneal ero-
corneal ulcers were treated with this procedure, sions, being effective in decreasing the recur-
and they were followed up for 3 months with no rence rate [134].
evidence of relapses or perforations in five of In addition, PRP has been injected into the
them. lacrimal glands of patients with severe lacrimal
Before the implementation of the use of dysfunction and severe dry eye, improving tear
Tutopatch® or autologous fibrin membrane, Alio production and decreasing associated keratitis
et al. also used this surgical reconstruction tech- [135]. Other authors have successfully used PRP
nique together with amniotic membrane [98]. as adjuvant during macular hole surgery in
Other collaborators have also used this tech- patients with high myopia [136]. Platelet deriva-
nique of surgical reconstruction of the ocular tives have also been used in subconjunctival infil-
surface with clots and autologous fibrin mem- trations for the treatment of corneal causticity
brane based on E-PRP [130, 131], as well as in [137].
the form of eye drops to see the regeneration The other blood derivative, PRGF, has been
after LASIK [132]. used for multiple purposes by Anitúa et al. since
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 331

1999 [61] and has been employed since then in


several areas of medicine [138–140]. But it was The data shown below are provided by our
not until 2010 that the PRGF began to be used in own work developed in the Laboratory of
the field of ophthalmology. López-Plandolit et al. the Research, Development and Inno­
used PRGF diluted 50% to treat persistent cor- vation Department of Vissum Instituto
neal epithelial defects in 18 patients. The results Oftalmológico de Alicante and Vissum
showed a recovery of the epithelial defects in Innovation (Spain), and they are in process
85% of the cases [141]. Subsequently, the same of publication in a peer-reviewed journal.
authors evaluated the results of the PRGF diluted
to 20% in cases of dry eye. They observed a
reduction in symptoms and an improvement in 22.10.1  Results of E-PRP
squamous metaplasia, although this was not sta- Characterization
tistically significant [141].
On the other hand, Merayo et  al. used PRGF Blood from different healthy volunteers was used
Endoret preparations for the treatment of patients to prepare E-PRP.  The concentration of the
with dry evaporative eye [142] and in patients with growth factors PDGF-BB, TGF-β1, EGF,
ocular surface disorders refractory to conventional VEGF-A and fibronectin was determined in each
treatments with satisfactory results [143]. of the four treatments including fresh E-PRP,
More recently, Sánchez-Ávila et  al. used the fresh spin, frozen E-PRP at −20 °C for 3 months
PRGF to treat neurotrophic keratitis in stages 2 and frozen at −20  °C for 3  months and spin
and 3. The majority of patients resolved the cor- before the ELISA tests.
neal defect, and one patient presented an adverse Platelets in the E-PRP concentrated almost
effect [144]. twice, according to the concentration index
obtained from 1.89. The red blood cells in the
E-PRP obtained were almost insignificant com-
22.10 Characterization of E-PRP pared to the whole blood. The reduction observed
was greater than 99%, and the leukocyte values
Although E-PRP has been used widely in the in E-PRP were very low compared to the whole
clinic with excellent results, there was not enough blood. The reduction was of 82% with respect to
information regarding the determination of the baseline values. The increase of platelets and
concentrations of the most important growth fac- reduction of red and white blood cells in the
tors and its adequate conservation for its later use E-PRP is statistically significant regarding to the
as a treatment for the different affections of the baseline (p ≤ 0.001).
ocular surface. In Table 22.1, we can see the values of the cell
To achieve these objectives, the concentrations count before and after the preparation of the
of platelets, leukocytes and red blood cells in the E-PRP.
whole blood and E-PRP were determined, and the Growth factor concentrations in E-PRP were
main growth factors of E-PRP and also the fibro- analysed on the day of the collection (fresh sam-
nectin adhesion protein were quantified. ELISA ples) and after have been stored frozen at −20 °C
immunoassays were performed for each molecule for 3  months with or without posterior
(eBioscience, Bender MedSystems GmbH, centrifugation.
Vienna, Austria). In parallel, in order to determine On the other hand, it was observed that all the
the biological stability of the autologous E-PRP concentrations of the GFs increased considerably
and demonstrate that it retained its biological in the fresh or frozen E-PRP with respect to the
potential after storage at −20  °C, the concentra- plasmatic levels (Table 22.2). The exception was
tions of the growth factors of fresh E-PRP and fro- the fibronectin whom values increased slightly in
zen E-PRP to −20 °C for 3 months, with or without the fresh or frozen E-PRP, with respect to the
subsequent centrifugation, were compared. basal plasma levels.
332 A. E. Rodriguez and J. L. Alió

Table 22.1  Blood cells count before and after the E-PRP preparation
Platelets × 103/μl Red cells × 106/μl White cells × 103/μl
Whole blood E-PRP Whole blood E-PRP Whole blood E-PRP
Mean 265.17 495.00 4.11 0.04 5.43 0.97
SD 69.75 95.14 0.32 0.01 0.70 0.30
SD standard deviation

Table 22.2  Comparison of levels of growth factors in plasma and in fresh and frozen E-PRP
Molecules Plasmaa Fresh E-PRPb Frozen E-PRPb
PDGF-BB <2000 pg/ml 5065.83 pg/ml 11961.67 pg/ml
TGF-β 1725 pg/ml 84165.71 pg/ml 74710.36 pg/ml
EGF 110–350 pg/ml 338.24 pg/ml 672.25 pg/ml
VEGF 47.3 pg/ml 274.25 pg/ml 494.49 pg/ml
Fibronectin 400 μg/ml 549.02 μg/ml 544.10 μg/ml
Values obtained from the eBioscience Kit protocol
a

Values obtained from our experiments


b

Fig. 22.4 Concentration EGF


* P value £ 0.05
of epidermal growth 900.00
factor (EGF) in E-PRP *
after the application of 800.00 **
four procedures
700.00
***
600.00

500.00 **
mg/ml

*
***
400.00

300.00

200.00

100.00

0.00
Fresh Fresh-spin Frozen Frozen-spin

The concentrations of EGF and PDGF-BB strong and significant positive correlation
obtained in the E-PRP were significantly higher between TGF-β1 and VEGF-A (p < 0.01), which
(p  <  0.05) after freezing it during 3  months at showed a relationship between these two GF.
−20 °C (Figs. 22.4 and 22.5).
The concentration levels reached by the
growth factor TGF-β1 were significantly higher 22.10.2  Discussion of E-PRP
(p < 0.05) in the fresh and frozen at −20 °C, over Characterization
the fresh and spin and frozen at −20 °C and spin
(Fig. 22.6). There are different opinions about which platelet
However, no significant differences were concentration index is appropriate to use platelet-­
observed among the treatments for the growth rich plasma as a therapeutic tool in the field of
factor VEGF-A and fibronectin (Figs.  22.7 tissue regeneration.
and 22.8). On the one hand, there are those who prefer
There was a positive and significant correla- high or very high concentration indices. In this
tion between the growth factor PDGF-BB and case we refer to those who use platelet concen-
platelets (p  <  0.05). In addition, there was also trates more than double the baseline values [145]
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 333

PDGF-BB
* P value £ 0.05
18000.00
**
16000.00 *

14000.00

12000.00
mg/ml

10000.00

8000.00

6000.00 ***

4000.00

2000.00

0.00
Fresh Fresh-spin Frozen Frozen-spin

Fig. 22.5  Concentration of platelet-derived growth factor BB (PDGF-BB) in E-PRP after the application of four
procedures

Fig. 22.6 Concentration TGF-b1


* P value £ 0.05
of transforming growth 100000.00
factor beta (TGF-β) in *
E-PRP after the 90000.00
**
application of four ***
80000.00
procedures
70000.00

60000.00
mg/ml

50000.00 **
40000.00

30000.00

20000.00
*
10000.00 ***

0.00
Fresh Fresh-spin Frozen Frozen-spin

and affirm that more platelets provide a higher And finally there are the authors who prefer
concentration of GF. In these cases, complex sys- that their preparations do not contain platelets,
tems for obtaining PRP with sophisticated equip- either because they use blood without antico-
ment or even several centrifuges are usually agulant and obtain autologous serum (AS),
necessary [135, 146]. leaving the platelets trapped in the blood clot
On the other hand, there are researchers who together with leukocytes and fibrin [112, 147,
concentrate platelets moderately, usually after 148], or those that prepare a plasma rich in
only one centrifugation, and who use the enriched platelets but then activate it with calcium chlo-
plasma as obtained after centrifugation [8, 9, ride or directly with thrombin, to obtain plasma
136] or those who dilute it for clinical application rich in growth factors (PRGF) [104, 141, 142].
[133, 134]. Either in the case of autologous serum or in that
334 A. E. Rodriguez and J. L. Alió

Fig. 22.7 Concentration VEGE-A


of vascular endothelial 1000.00
growth factor A
900.00
(VEGF-A) in E-PRP
after the application of 800.00
four procedures
700.00

600.00

mg/ml 500.00

400.00

300.00

200.00

100.00

0.00
Fresh Fresh-spin Frozen Frozen-spin

Fig. 22.8 Concentration Fibronectin


of the plasma protein 600.00
fibronectin in E-PRP
after the application of 580.00
four procedures
560.00

540.00
mg/ml

520.00

500.00

480.00

460.00
Fresh Fresh-spin Frozen Frozen-spin

of PRGF, both preparations present variable No papers were found in the consulted bibli-
concentrations of GF [74, 104]. ography that expressed the concentration of red
In our study, the results obtained after the cen- blood cells in the blood derivatives used for oph-
trifugation of the blood show that the level of plate- thalmology. Anitua et al. [82] analysed the com-
lets in the E-PRP is almost twice as high as the basal position of PRGF scaffolds and determined that
values in​​ the whole blood, since the concentration the concentration of red cells in the PRGF
index reached was 1.89. Therefore, the centrifuga- Endoret was between 0.02 and 0.04 (×106/μl).
tion speed used for the preparation of the E-PRP Other authors found concentrations of red blood
and the established time are adequate to ensure the cells between 0.01 and 0.03 (×106/μl) in PRP
optimal supply of GF for clinical uses [146, 149]. used for regenerative medicine [150]. All these
Regarding the red blood cell count, the results values are
​​ comparable to those of our study,
have shown that, after centrifugation, the values​​ which turned out to be 0.04 ± 0.01 (×106/μl).
in the E-PRP are almost insignificant compared Regarding leukocytes, the values ​​found in the
to the total blood counts since the percentage E-PRP were very low compared to the whole
found is only 0.9%, obtaining a reduction index blood, showing a reduction of 82% with respect
greater than 99%. to the baseline. The average concentration of
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 335

white blood cells in the E-PRP (0.97  ×  103/μl) variations observed between them. This indicates
was comparable to that found by some authors in that the subsequent centrifugation that removes
PRP [146, 151] or slightly higher than that the platelets from the supernatant does not mod-
described by others for PRGF [64, 82], although ify the EGF concentration, probably because all
none of the latter indicates the basal concentra- of this EGF has been released after freezing.
tion in the whole blood from which they start. Platelet-derived growth factor (PDGF) is a
These small values ​​of leukocytes found in the modulator of the chemotaxis and cellular mitosis
E-PRP do not classify it as a blood derivative of fibroblasts [28]. It is presented in three differ-
“with” leukocytes. This denomination is suitable ent isoforms, PDGF-AA, PDGF-AB and
for those prepared with high concentrations of PDGF-BB, and for our study, we have chosen
white blood cells, close to or equal to those in the PDGF-BB, since it has the highest chemotactic
whole blood. Such is the case of L-PRF and power of the three [30, 31].
L-PRP [81–83], or the PRP prepared in two steps PDGF-BB followed the same pattern as the
traditionally used in other areas of medicine other previous EGF, observing much higher values ​​in
than ophthalmology [29, 152, 153]. Those PRPs frozen E-PRP compared to fresh. This difference
used in ophthalmology after a centrifugation, but was significant for the fresh spin, but not for the
incorporating the buffy coat [133, 134, 146], fresh when compared with the respective frozen
would also fall into this category. E-PRP, although the concentrations found were
As far as possible, leukocytes should be avoided more than double in the latter. Possibly this lack
in PRP preparations due to their potential pro- of statistical significance is due to the variability
inflammatory effect since their presence increases of the data.
the concentration of the pro-­inflammatory cyto- The concentrations obtained for PDGF-BB
kines IL-1β, TNF-α and the metalloproteinases were comparable to those found by other authors
MMP-9 and MMP-13 [154, 155]. [156, 157] and much higher than those of other
On the other hand, it must be considered that researchers [158]. On the other hand, there are
non-leukocyte blood derivatives, such as SA, in interesting studies that compare the GF of sam-
addition to GF, have pro-inflammatory cytokine ples of PRP, PRGF and AS but determine the
values ​​such as IL-1β, significantly higher than PDGF-AB, so the results cannot be compared
those present in platelet-rich plasma with leuko- [64, 74, 104]. Other authors do not specify what
cytes (L-PRP) [81]. This is because at the time of fraction of PDGF they have analysed, so the find-
blood coagulation, the leukocytes are trapped in ings cannot be compared either [66].
the clot releasing their content including the When the different GFs were correlated with
interleukins. the blood cells in the fresh E-PRP, it was observed
The growth factors chosen in this study were that there was a significant and positive correla-
the most representative in the field of ocular tion between the platelets and the PDGF-BB
regeneration. (r  =  0.841 and p  =  0.036). This is because this
The epidermal growth factor (EGF) partici- factor is released during the degranulation of
pates actively in the migration and proliferation platelets when they are in active form [159]. But
of the corneal epithelium in tissue regeneration no correlation was found between the platelet
processes [44]. The concentrations of EGF in the count in the E-PRP and the frozen PDGF-BB
frozen E-PRP samples were significantly higher (p = 0.203).
than the corresponding fresh ones, even doubling Several authors have also found a positive cor-
their values. These values ​​of the frozen samples relation between the platelet count and PDGF-BB
have higher values than
​​ those obtained with the [160–162], while others have found no correla-
PRGF of Anitua et  al. [64, 104] and with the tion between them in platelet-rich plasma prepa-
autologous serum of López-García et al. [74]. rations [156, 157, 163].
The concentrations of EGF in the frozen and The transforming growth factor β1 (TGF-β1)
frozen-spin E-PRP were comparable, with no is a multifunctional GF and one of the most
336 A. E. Rodriguez and J. L. Alió

important that intervene in the modulation of fresh E-PRP (r = 0.408 and p = 0.421). Probably
the behaviour of ocular tissues [40]. It favours this difference is due to methodological issues in
the chemotaxis of fibroblasts, promotes the dif- the realization of the ELISA technique.
ferentiation of myofibroblasts and induces the This positive correlation found between leu-
production of extracellular matrix by stimulat- kocytes and TGF-β1 is most likely because this
ing the production of collagen, fibronectin and GF originates in macrophages, monocytes, neu-
proteoglycans and decreases their degradation trophils and some types of lymphocytes, as well
by inhibiting metalloproteases and other proteo- as in platelets [57].
lytic enzymes. Depending on the context, it may A strong and positive correlation was also
have pro-inflammatory or anti-inflammatory found between TGF-β1 and VEGF-A (r = 0.943
effects [164]. and p  =  0.005). This fact could be explained
The concentrations of TGF-β1 in the fresh and because the presence of VEGF, dose dependent,
frozen E-PRP were high, not observing statistical induces cell migration and cell proliferation,
differences between them. But subsequent cen- accompanied by upregulation of TGF-β1 mRNA,
trifugation drastically decreased the concentra- improving the secretion of this factor and its bio-
tion of this factor. activity [167]. On the other hand, it has also been
Given the results obtained, it can be inferred observed that cytokines derived from leukocytes,
that TGF-β1 does not need to be activated to per- together with the TGF-β1 released by platelets
form its function since there are no differences and plasma-derived factors, are capable of
between the fresh values ​​and after freezing at ­inducing the expression of VEGF-A genes [168].
−20 °C for 3 months. And since the decrease in In this way, VEGF-A and TGF-β1 cooperate
concentration is significant after the centrifuga- because they have a strong relationship, to act in
tion of both preparations, fresh and frozen, it processes of cellular chemotaxis and neovascu-
could be assumed that the TGF-β1 molecules are larization promoting the regeneration of tissues.
somehow adhered to the platelet membrane since Vascular endothelial growth factor (VEGF-A)
when they are eliminated the values descend is a potent angiogenic and endothelial cell mito-
significantly. gen that is present in platelets and in endothelial
When the results obtained were compared cells stimulating the proliferation of blood ves-
with those of the bibliography, it was observed sels [57].
that there was great variability between the dif- With respect to the concentrations obtained of
ferent tests, either by the detection technique, the VEGF-A, there were no significant differences
blood processing method or the individual varia- between any of the treatments, although a slightly
tions [165, 166]. Our TGF-β1 values in ​​ the higher average concentration was found in the
E-PRP were higher than those obtained by other frozen samples. This could be because the expres-
authors in AS [86] and in PRGF [104]. sion of the growth factor VEGF-A is modulated
Some authors have found a positive correla- by other factors such as EGF, TGF-α and β,
tion between TGF-β1 and the initial concentra- PDGF, the GF derived from keratinocytes or the
tion of platelets [161], but this fact has not insulin-like GF type 1, as well as by cytokines. It
occurred in our study (r = 0.676 and p = 0.141), is also affected by stimuli such as hypoxia/isch-
although in view of the results found, it seems emia mainly through the inducible hypoxia factor
that there is some relationship. (HIF-1) [169].
On the other hand, it was observed that there is The concentrations found in the fresh E-PRP
a significant and positive correlation between the were the same as those found by other authors in
leukocytes and the concentration of TGF-β1  in fresh PRGF, while the values o​​ f the frozen E-PRP
the frozen E-PRP (r = 0.932 and p = 0.007). But at −20  °C for 3  months turned out to be twice
this correlation, strangely enough, was not evi- those of the PRGF stored under the same condi-
dent when the white blood cell counts were con- tions [104]. There is a study that compares the
trasted with the TGF-β1 concentrations in the concentrations of GF in samples of AS, PRGF
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 337

and PRP but does not specify the fraction of inferior and the authors assure that their values​​
VEGF analysed, and although the concentrations are normal as in our case [66, 104]. A reasonable
are the same units as our study (pg/ml), the val- explanation for this situation could be the vari-
ues ​​are of another order of magnitude, so they ability in the analytical detection techniques
cannot be matched. used.
In contrast, in another study comparing the Although some authors found a decrease in
levels of VEGF in AS with PRGF, it was found the levels of this protein after storage at −20 °C
that the highest levels of VEGF corresponded to for 3 months [104], this was not the case in our
PRGF [64]. The authors comment that although case since the concentrations obtained were simi-
higher VEGF values ​​were found, they did not lar in fresh and frozen samples. Other authors
detect neovascularization in any of the patients found a fragmentation of the fibronectin mole-
treated with PRGF in previous studies [141, 170]. cule that produced a decline in its activity, but
Unlike the previous ones, we would be inter- this happened after 4–12 months of cryopreserva-
ested that the values of VEGF in the E-PRP are tion [171].
not very high given that a high concentration of Except for PDGF-BB, no correlation was
them contributes to the development of disease, found between the growth factors EGF, TGF-β1
since the overexpression of VEGF can lead to and VEGF-A with the platelet count (p > 0.05).
vascular alterations of the retina [50] and has These results coincide with those reported by
even been associated with the development and other authors who affirm that this fact could be
maintenance of tumours [51]. explained by the high individual variability that
Fibronectin is an adhesion protein, which is exists in the cellular production and storage of
free in plasma, and is one of the main constitu- GF in platelets. The content of GF in individuals
ents of the extracellular matrix. This protein can could be influenced by cells other than platelets
interact with various molecules and cells, estab- such as leukocytes; this means that each individ-
lishing connections between them in the extracel- ual would need different concentrations of plate-
lular matrix. This molecule is one of the first to lets to achieve a comparable biological effect.
arrive when a corneal wound occurs filling the Therefore, the platelet count in the whole blood
bed of the lesion, thus producing cell migration or PRP would not be a predictive factor of HR
during the repair process of the corneal epithe- values in​​ the PRP [157, 163]. Another possible
lium [53–55]. fact is that there are other mechanisms of interac-
In our trial, no significant differences were tion, not dependent on platelets, among the
found in the concentration of fibronectin, since immense number of molecules present in the
the values ​​of this protein remained constant in the blood plasma that modulate the release of growth
four types of procedures performed. The expla- factors.
nation for this fact is because fibronectin is a In this study, different procedures applied to
plasma protein and its presence does not depend E-PRP have been evaluated in order to find out
on platelet activation. how they affected freezing, storage time and cen-
The fibronectin concentrations found in our trifugation in the concentration of the different
study are slightly higher than the blood plasma bioactive molecules of the plasma. According to
reference values ​​ since it is approximately the results obtained, we could say that freezing at
300 ± 100 μg/ml [52]. These variations between −20  °C for 3  months provided higher levels of
healthy individuals, probably physiological, may the growth factors EGF and PDGF-BB and main-
be due to variations in age and sex [52].As it hap- tained the concentrations of TGF-β1, VEGF-A
pened when trying to compare our results with and fibronectin.
the literature to compare the levels of the GF This behaviour is because the process of freez-
found, it is also difficult to do so for fibronectin ing and subsequent thawing of the E-PRP causes
since the values found,
​​ although in the same con- a powerful activation of the platelets it contains
centration (μg/ml), are in an order of magnitude [158]. The PRP must be activated so that the
338 A. E. Rodriguez and J. L. Alió

α-granules release their content, and this activa- ity to provide growth factors involved in the bio-
tion can occur by the addition of collagen, cal- logical responses of cells.
cium and/or thrombin, by contact with the glass E-PRP is very versatile since it can be applied
or by freezing cycles [150, 172]. With this last as an eye drop in various pathologies of the ocu-
procedure, it is possible to arrive until the total lar surface as in the case of corneal ulcers, dry
liberation of its content thus providing great eye and ocular surface dysfunction after LASIK
amount of GF. and as a clot or autologous fibrin membrane for
On the contrary, the centrifugation applied to surgical reconstruction in cases of severe corneal
fresh and frozen E-PRP samples, contrary to the ulcers or corneal perforations.
previous idea that it could activate the platelet But contrary to what might be expected, high
content, did not give the expected results. All the platelet concentrations (1  ×  106/μl), which pro-
concentrations of the molecules studied in the duce very high levels of GF, are not necessary to
“centrifuged/spin” groups, except for TGF-β1, successfully treat corneal lesions [158]. With
were like their homologs without centrifuging, much lower concentrations, the desired effects
and no statistically significant differences were are achieved. In our study, the average concentra-
found. tion of platelets was 495 × 103/μl, which is why
In the particular case of TGF-β1, the concen- this concentration is adequate to obtain sufficient
trations of the centrifuged samples decreased sig- levels of growth factors to treat the various ocular
nificantly with respect to their counterparts surface affectations that require regenerative
without centrifuging, demonstrating that the therapy.
platelets had to be present to obtain values close
​​ The finding of high concentrations of growth
to the initial ones. factors EGF, PDGF-BB and fibronectin in the
The centrifugation parameters used (6000 rpm E-PRP is the key to obtain the desired regenera-
for 5 min) did not produce the activation of the tive effects on the ocular surface.
platelets present in the E-PRP, but their irrevers- This occurs because high levels of fibronectin
ible aggregation as they were agglomerated in the together with high PDGF-BB values ​​stimulate
bottom of the conical tubes, making their dissolu- the migration of corneal epithelial cells in cor-
tion impossible. neal healing processes [173, 174]. On the other
hand, EGF increases proliferation and cell migra-
tion, promoting the reordering of actin filaments
22.11 Conclusions [175], and is also able to stabilize the tear film
and prevent apoptosis [176]. All these molecules,
The interest in the therapeutic application of together with fibroblast GF (FGF), facilitate cor-
platelet-rich plasma has increased in different neal re-epithelialization while being modulated
fields of medicine. It is known that this treatment by several cytokines [164].
has properties that enhance and accelerate the Clinical studies in which platelet-rich plasma
process of tissue regeneration through the stimu- was used as a treatment showed very good results
lation of cell growth and function. Due to its in patients with different pathologies of the ocu-
properties, it has generated great advances in lar surface. All these findings are now reinforced
many areas of regenerative medicine. This bio- by knowing the concentration of the different
logical and autologous product does not generate components of E-PRP in terms of cell elements,
rejection and minimizes the possible adverse growth factors and fibronectin.
effects. Therefore, it is possible to conclude that the
Numerous studies have demonstrated the effi- E-PRP presents the necessary amounts of growth
cacy of E-PRP as a treatment for several ocular factors and adhesion proteins, proving to be an
surface diseases due to their concentration of adequate tool as a regenerative therapy of the cor-
platelets, which remain viable and retain the abil- nea and ocular surface.
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 339

Compliance with Ethical Requirements  Conflict of 12. Rivadeneyra L, Ivani PC, Schattner M, Pozner

Interest RG.  Así comienza la vida plaquetaria: un viaje
Alejandra E. Rodríguez and Jorge L. Alio declare that desde los megacariocitos medulares a las plaqu-
they have no conflict of interest. etas circulantes. Acta Bioquímica Clínica Latinoam.
Informed Consent 2016;50(2):233–45.
All procedures followed were in accordance with the 13. López Farré A, Macaya C. Plaqueta: Fisiologia de la
ethical standards of the responsible committee on human activacion y la inhibicion. Rev Española Cardiol Supl.
experimentation (institutional and national) and with the 2013;13(SUPPL.2):2–7.
Helsinki Declaration of 1975, as revised in 2000. Informed 14. Farndale RW, Sixma JJ, Barnes MJ, De Groot

consent was obtained from all patients for being included PG. The role of collagen in thrombosis and hemosta-
in the study. sis. J Thromb Haemost. 2004;2(4):561–73.
No animal studies were carried out by the authors for 15.
Flores-Rivera OI, Ramírez-Morales K, Meza-­
this chapter. Márquez JM, Nava-López JA. Fisiología de la coagu-
lación. Rev Mex Anestesiol. 2014;37:282–386.
16. Xu W, Wang TY, Becker RC. Enfermedades hemato-
logicas: desde dentro del corazon. Rev Esp Cardiol.
References 2011;64(7):606–13.
17. Hidalgo, Mesa CJ, Cepero Rodríguez I, Berrios

1. Garg A, Sheppard JD, Donnenfeld ED, Meyer D, Águila JE, Ulloa Quintanilla FO, Polanco Rodríguez
Mehta CK. Ojo seco y otros trastornos de la superficie F.  Infarto cerebral: complicaciones y causas de
ocular: diagnóstico y tratamiento en xero-­dacriología. muerte. Rev Cuba Med Mil. 2005;34(1):1–3.
Madrid. España: Editorial Médica Panamericana, 18. Smyth SS, Mcever RP, Weyrich AS, Morrell CN,

S.A.; 2008. 472 p. Hoffman MR, Arepally GM, et  al. Platelet func-
2. Cárdenas Díaz T, Capote Cabrera A, Benítez Merino M tions beyond hemostasis. J Thromb Haemost.
d C, Noriega Martínez JL, Montero Díaz E, Hormigó 2009;7(11):1759–66.
Puertas IF. Medicina regenerativa y superficie ocular. 19. González-Villalva A.  Sangre y Hematopoyesis. In:
Rev Cuba Oftalmol. 2012;25(1):104–18. Kierszenbaum AL, Tres LL, editors. Histología y
3. Pikuła M, Langa P, Kosikowska P, Trzonkowski Biología Celular. Barcelona, España: Elsevier España,
P.  Stem cells and growth factors in wound healing. S.L; 2012. p. 168–201.
Postep Hig Med Dosw. 2015;69:874–85. 20. Spinelli SL, Maggirwar SB, Blumberg N, Phipps

4. Coursey TG, de Paiva CS.  Managing Sjögren’s RP. Nuclear emancipation: a platelet tour de force. Sci
syndrome and non-Sjögren syndrome dry eye Signal. 2010;3(144):pe37.
with anti-inflammatory therapy. Clin Ophthalmol. 21. Anitua E, Andia I, Ardanza B, Nurden P, Nurden
2014;8:1447–58. AT.  Autologous platelets as a source of proteins for
5. Tsubota K, Satake Y, Ohyama M, Toda I, Takano Y, healing and tissue regeneration. Thromb Haemost.
Ono M, et  al. Surgical reconstruction of the ocular 2004;91(1):4–15.
surface in advanced ocular cicatricial pemphigoid 22. Nurden AT. Platelets, inflammation and tissue regener-
and Stevens-Johnson syndrome. Am J Ophthalmol. ation. Thromb Haemost. 2011;105(SUPPL.1):13–33.
1996;122(1):38–52. 23. Marx RE. Platelet-rich plasma: evidence to support its
6. Hick S, Demers PE, Brunette I, La C, Mabon M, use. J Oral Maxillofac Surg. 2004;62:489–96.
Duchesne B.  Amniotic membrane transplantation 24. Nurden AT, Nurden P, Sanchez M, Andia I, Eduardo
and fibrin glue in the management of corneal ulcers A.  Platelets and wound healing. Front Biosci.
and perforations: a review of 33 cases. Cornea. 2008;2(i):3532–48.
2005;24(4):369–77. 25. Gawaz M, Vogel S. Platelets in tissue repair: control
7. Nakamura T, Inatomi T, Sotozono C, Koizumi N, of apoptosis and interactions with regenerative cells.
Kinoshita S.  Ocular surface reconstruction using Blood. 2013;122(15):2550–4.
stem cell and tissue engineering. Prog Retin Eye Res. 26. Diago P, Bielsa JMS. Puesta al día en Factores de cre-
2015;51:187–207. cimiento y proteínas que influyen en el crecimiento
8. Alio J, Arnalich-Montiel F, Rodriguez A.  The role óseo: aplicaciones en implantología oral. Periodoncia.
of ‘eye platelet rich plasma’ (E-PRP) for wound 2001;11(3):205–16.
healing in ophthalmology. Curr Pharm Biotechnol. 27. Barrientos S, Stojadinovic O, Golinko MS, Brem

2012;13(7):1257–65. H, Tomic-Canic M.  Growth factors and cyto-
9. Alio JL, Rodriguez AE, WróbelDudzińska D.  Eye kines in wound healing. Wound Repair Regen.
platelet-rich plasma in the treatment of ocular sur- 2008;16(5):585–601.
face disorders. Curr Opin Ophthalmol. 2015;26(4): 28. Alvarez RH, Kantarjian HM, Cortes JE.  Biology of
325–32. platelet-derived growth factor and its involvement in
10. Nugent RB, Lee GA. Ophthalmic use of blood-derived disease. Mayo Clin Proc. 2006;81(9):1241–57.
products. Surv Ophthalmol. 2015;60(5):406–34. 29. Beca T, Hernandez G, Morante S, Bascones A. Plasma
11. Ribatti D, Crivellato E. Giulio Bizzozero and the dis- rico en plaquetas. Una revisión bibliográfica. Av
covery of platelets. Leuk Res. 2007;31(10):1339–41. Periodoncia. 2007;19(1):39–52.
340 A. E. Rodriguez and J. L. Alió

30. Kim WJ, Mohan RR, Mohan RR, Wilson SE. Effect ium fibroblasts. Biochem Biophys Res Commun.
of PDGF, IL-alpha, and BMP2/4 on corneal fibroblast 2016;482(4):1148–53.
chemotaxis: expression of the platelet-derived growth 44. Wang L, Wu X, Shi T, Lu L. Epidermal growth fac-
factor system in the cornea. Investig Ophthalmol Vis tor (EGF)-induced corneal epithelial wound healing
Sci. 1999;40(7):1364–72. through nuclear factor κB subtype-regulated CCCTC
31. Colciago A, Celotti F, Casati L, Giancola R, Castano binding factor (CTCF) activation. J Biol Chem.
SM, Antonini G, et al. In vitro effects of PDGF iso- 2013;288(34):24363–71.
forms (AA, BB, AB and CC) on migration and pro- 45. Hodges RR, Bair JA, Carozza RB, Li D, Shatos MA,
liferation of SaOS-2 osteoblasts and on migration Dartt DA. Signaling pathways used by EGF to stimu-
of human osteoblasts. Int J Biomed Sci. 2009;5(4): late conjunctival goblet cell secretion. Exp Eye Res.
380–9. 2012;103:99–113.
32. Andrae J, Gallini R, Betsholtz C.  Role of platelet-­ 46. Kenchegowda S, Bazan NG, Bazan HEP. EGF stimu-
derived growth factors in physiology and medicine. lates lipoxin A4 synthesis and modulates repair in cor-
Genes Dev. 2008;22(10):1276–312. neal epithelial cells through ERK and p38 activation.
33. Robson MC, Phillips LG, Thomason A, Robson LE, Investig Ophthalmol Vis Sci. 2011;52(5):2240–9.
Pierce GF. Platelet-derived growth factor BB for the 47. Khanbanha N, Atyabi F, Taheri A, Talaie F, Mahbod
treatment of chronic pressure ulcers. Lancet (London, M, Dinarvand R. Healing efficacy of an EGF impreg-
England). 1992;339(8784):23–5. nated triple gel based wound dressing: in  vitro and
34. Pierce GF, Tarpley JE, Allman RM, Goode PS, Serdar in vivo studies. Biomed Res Int. 2014;2014:1–10.
CM, Morris AB, et al. Tissue repair processes in heal- 48.
Ferrara N, Houck K, Jakeman LYN, Leung
ing chronic pressure ulcers treated with recombinant DW. Molecular and biological properties of the vas-
platelet-derived growth factor BB.  Am J Pathol. cular endothelial growth factor family of proteins.
1994;145(6):1399. Endocr Rev. 1992;13(1):18–32.
35. Pierce GF, Tarpley JE, Tseng J, Bready J, Chang
49. Yu CQ, Zhang M, Matis KI, Kim C, Rosenblatt

D, Kenney WC, et  al. Detection of platelet-derived MI.  Vascular endothelial growth factor mediates
growth factor (PDGF)-AA in actively healing human corneal nerve repair. Invest Ophthalmol Vis Sci.
wounds treated with recombinant PDGF-BB and 2008;49(9):3870–8.
absence of PDGF in chronic nonhealing wounds. J 50. Duffy AM, Bouchier-hayes DJ, Harmey JH. Vascular
Clin Invest. 1995;96(3):1336–50. endothelial growth factor (VEGF) and its role in non-­
36. Smiell JM, Wieman TJ, Steed DL, Perry BH, Sampson endothelial cells: autocrine signalling by VEGF.  In:
AR, Schwab BH. Efficacy and safety of becaplermin VEGF and Cancer. Austin: Landes Bioscience; 2004.
(recombinant human platelet-derived growth factor- p. 133–44.
BB) in patients with nonhealing, lower extremity dia- 51. Stevenson W, Cheng SF, Dastjerdi MH, Ferrari G,
betic ulcers: a combined analysis of four randomized Dana R.  Corneal neovascularization and the utility
studies. Wound Repair Regen. 1999;7(5):335–46. of topical VEGF inhibition: ranibizumab (Lucentis)
37.
Steed DL.  Clinical evaluation of recombinant vs bevacizumab (Avastin). Ocul Surf. 2012;10(2):
human platelet-derived growth factor for the treat- 67–83.
ment of lower extremity ulcers. Plast Reconstr Surg. 52.
Lucena S, Arocha Piñango CL, Guerrero
2006;117(7 Suppl):143S–9S. discussion 150S–151S. B. Fibronectin. Structure and functions associated to
38. Sarment DP, Cooke JW, Miller SE, Jin Q, McGuire hemostasis. Review Invest Clin. 2007;48(2):249–62.
MK, Kao RT, et  al. Effect of rhPDGF-BB on bone 53. Pastar I, Stojadinovic O, Yin NC, Ramirez H,

turnover during periodontal repair. J Clin Periodontol. Nusbaum AG, Sawaya A, et  al. Epithelialization in
2006;33(2):135–40. wound healing: a comprehensive review. Adv Wound
39. Nevins M, Camelo M, Nevins ML, Schenk RK,
Care. 2014;3(7):445–64.
Lynch SE. Periodontal regeneration in humans using 54. Ljubimov AV, Saghizadeh M.  Progress in corneal

recombinant human platelet-derived growth factor- wound healing. Prog Retin Eye Res. 2015;49:17–45.
­BB (rhPDGF-BB) and allogenic bone. J Periodontol. 55. Paniagua R, Manuel N, Pilar S, Manuel A-U, Benito
2003;74(9):1282–92. F, Ramón A, Sáez JF.  Biología Celular. Madrid:
40. Saika S.  TGFbeta pathobiology in the eye. Lab
McGraw-Hill  – Interamericana de España, S.  A. U;
Investig. 2006;86(2):106–15. 2007. 381 p.
41. Costanza B, Umelo I, Bellier J, Castronovo V, Turtoi 56. Makogonenko E, Tsurupa G, Ingham K, Medved

A.  Stromal modulators of TGF-β in Cancer. J Clin L. Interaction of fibrin(ogen) with fibronectin: further
Med. 2017;6(1):7. characterization and localization of the fibronectin-­
42. Stahnke T, Kowtharapu BS, Stachs O, Schmitz K-P, binding site. Biochemistry. 2002;41(25):7907–13.
Wurm J, Wree A, et al. Suppression of TGF-β path- 57. De La Mata J.  Plasma rico en plaquetas: ¿un nuevo
way by pirfenidone decreases extracellular matrix tratamiento para el reumatologo? Reumatol Clin.
deposition in ocular fibroblasts in  vitro. PLoS One. 2013;9(3):166–71.
2017;12(2):e0172592. 58. Hartwig D, Harloff S, Liu L, Schlenke P, Wedel T,
43. Il GS, Kim Y-H, Jung J-C, Kim IG, Lee JS, Lee Geerling G.  Epitheliotrophic capacity of a growth
KW, et  al. Cyclosporine A inhibits TGF-β2-induced factor preparation produced from platelet concen-
myofibroblasts of primary cultured human pteryg- trates on corneal epithelial cells: a potential agent for
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 341

the treatment of ocular surface defects? Transfusion. induce hepatocyte growth factor production by syno-
2004;44(12):1724–31. vial fibroblasts from arthritic patients. Rheumatology.
59. Hartwig D, Herminghaus P, Wedel T, Liu L, Schlenke 2007;46(12):1769–72.
P, Dibbelt L, et al. Topical treatment of ocular surface 71. Assirelli E, Filardo G, Mariani E, Kon E, Roffi A,
defects: comparison of the epitheliotrophic capacity Vaccaro F, et  al. Effect of two different prepara-
of fresh frozen plasma and serum on corneal epithelial tions of platelet-rich plasma on synoviocytes. Knee
cells in an in vitro cell culture model. Transfus Med. Surgery, Sport Traumatol Arthrosc. 2015;23(9):
2005;15(2):107–13. 2690–703.
60. Liu L, Hartwig D, Harloff S, Herminghaus P, Wedel 72. Beitzel K, McCarthy MB, Russell RP, Apostolakos J,
T, Kasper K, et al. Corneal epitheliotrophic capacity Cote MP, Mazzocca AD.  Learning about PRP using
of three different blood-derived preparations. Investig cell-based models. Muscles Ligaments Tendons J.
Ophthalmol Vis Sci. 2006;47(6):2438–44. 2014;4(1):38–45.
61. Anitua E.  Plasma rich in growth factors: prelimi-
73. Masoudi EA, Ribas J, Kaushik G, Leijten J,

nary results of use in the preparation of future Khademhosseini A. Platelet-rich blood derivatives for
sites for implants. Int J Oral Maxillofac Implants. stem cell-based tissue engineering and regeneration.
1999;14(4):529–35. Curr Stem Cell Reports. 2016;2(1):33–42.
62. Anitua E, Andía I, Sanchez M, Azofra J, del Mar 74. López-García JS, García-Lozano I, Rivas L, Ramírez
Zalduendo M, de la Fuente M, et al. Autologous prep- N, Méndez MT, Raposo R.  Stability of growth fac-
arations rich in growth factors promote proliferation tors in autologous serum eyedrops after long-term
and induce VEGF and HGF production by human ten- storage. Curr Eye Res. 2015;3683(November):
don cells in culture. J Orthop Res. 2005;23(2):281–6. 1–7.
63. Anitua E, Sánchez M, Zalduendo MM, De La Fuente 75. Poon AC, Geerling G, Dart JK, Fraenkel GE, Daniels
M, Prado R, Orive G, et al. Fibroblastic response to JT. Autologous serum eyedrops for dry eyes and epi-
treatment with different preparations rich in growth thelial defects: clinical and in vitro toxicity studies. Br
factors. Cell Prolif. 2009;42(2):162–70. J Ophthalmol. 2001;85(10):1188–97.
64. Anitua E, de la Fuente M, Muruzabal F, Riestra A, 76. Pinarli F, Okten G, Beden U, Fışgın T, Kefeli M, Kara
Merayo-Lloves J, Orive G. Plasma rich in growth fac- N, et  al. Keratinocyte growth factor-2 and autolo-
tors (PRGF) eye drops stimulates scarless regeneration gous serum potentiate the regenerative effect of mes-
compared to autologous serum in the ocular surface enchymal stem cells in cornea damage in rats. Int J
stromal fibroblasts. Exp Eye Res. 2015;135:118–26. Ophthalmol. 2014;7(2):211–9.
65. Anitua E, Sanchez M, Merayo-Lloves J, de La Fuente 77. López-García JS, García-Lozano I, Rivas L,
M, Muruzabal F, Orive G. Plasma rich in growth fac- Giménez C, Acera A, Suárez-Cortés T.  Effects of
tors (PRGF-Endoret) stimulates proliferation and autologous serum eye drops on conjunctival expres-
migration of primary keratocytes and conjunctival sion of MUC5AC in patients with ocular surface dis-
fibroblasts and inhibits and reverts TGF-β1-induced orders. Cornea. 2016;35(3):336–41.
myodifferentiation. Investig Ophthalmol Vis Sci. 78. Dohan DM, Choukroun J, Diss A, Dohan SL, Dohan
2011;52(9):6066–73. AJJ, Mouhyi J, et  al. Platelet-rich fibrin (PRF): a
66. Freire V, Andollo N, Etxebarria J, Durán JA, Morales second-generation platelet concentrate. Part III: leu-
M-C.  In vitro effects of three blood derivatives on cocyte activation: a new feature for platelet concen-
human corneal epithelial cells. Invest Ophthalmol Vis trates? Oral Surg Oral Med Oral Pathol Oral Radiol
Sci. 2012;53(9):5571–8. Endod. 2006;101(3):51–5.
67. Freire V, Andollo N, Etxebarria J, Hernáez-Moya
79. El-Sharkawy H, Kantarci A, Deady J, Hasturk H, Liu
R, Durán JA, Morales M-C.  Corneal wound heal- H, Alshahat M, et  al. Platelet-rich plasma: growth
ing promoted by 3 blood derivatives: an in  vitro factors and pro- and anti-inflammatory properties. J
and in  vivo comparative study. Cornea. 2014;33(6): Periodontol. 2007;78(4):661–9.
614–20. 80. Mazzocca AD, Mccarthy BR, Intravia J, Beitzel KA,
68. Anitua E, Muruzabal F, Alcalde I, Merayo-Lloves Apostolakos J, Cote MP, et  al. An in  vitro evalua-
J, Orive G.  Plasma rich in growth factors (PRGF-­ tion of the anti-inflammatory effects of platelet-rich
Endoret) stimulates corneal wound healing and plasma, ketorolac, and methylprednisolone. Arthrosc
reduces haze formation after PRK surgery. Exp Eye J Arthrosc Relat Surg. 2013;29(4):675–83.
Res. 2013;115:153–61. 81. Schär MO, Diaz-Romero J, Kohl S, Zumstein
69. Etxebarria J, Sanz-Lázaro S, Hernáez-Moya R,
MA, Nesic D.  Platelet-rich concentrates dif-
Freire V, Durán JA, Morales MC, et al. Serum from ferentially release growth factors and induce
plasma rich in growth factors regenerates rabbit cell migration in  vitro. Clin Orthop Relat Res.
corneas by promoting cell proliferation, migra- 2015;473(5):1635–43.
tion, differentiation, adhesion and limbal stemness. 82. Anitua E, Zalduendo MM, Prado R, Alkhraisat MH,
Acta Ophthalmol. 2017. https://doi.org/10.1111/aos. Orive G. Morphogen and proinflammatory cytokine
13371. release kinetics from PRGF-Endoret fibrin scaf-
70. Anitua E, Sánchez M, Nurden AT, Zalduendo MM, De folds: evaluation of the effect of leukocyte inclu-
la Fuente M, Azofra J, et al. Platelet-released growth sion. J Biomed Mater Res  - Part A. 2015;103(3):
factors enhance the secretion of hyaluronic acid and 1011–20.
342 A. E. Rodriguez and J. L. Alió

83. Anitua E, Zalduendo M, Troya M, Padilla S, Orive consensus, clinical implications and perspectives.
G.  Leukocyte inclusion within a platelet rich Muscles Ligaments Tendons J. 2014;4(1):3–9.
plasma-­ derived fibrin scaffold stimulates a more 97. Marx RE.  Platelet-rich plasma (PRP): what
pro-­inflammatory environment and alters fibrin is PRP and what is not PRP? Implant Dent.
properties. PLoS One. 2015;10(3):1–19. 2001;10(4):225–8.
84. Anitua E, Muruzabal F, De la Fuente M, Merayo-­ 98. Alio JL, Abad M, Artola A, Rodriguez-Prats JL,
Lloves J, Orive G.  Effects of heat-treatment on Pastor S, Ruiz-Colecha J. Use of autologous platelet-­
plasma rich in growth factors-derived autologous rich plasma in the treatment of dormant corneal
eye drop. Exp Eye Res. 2014;119:27–34. ulcers. Ophthalmology. 2007;114(7):1286–94.
85. Anitua E, Muruzabal F, de la Fuente M, Riestra A, 99. Alio JL, Pastor S, Ruiz-Colecha J, Rodriguez A,
Merayo-Lloves J, Orive G. PRGF exerts more potent Artola A.  Treatment of ocular surface syndrome
proliferative and anti-inflammatory effects than after LASIK with autologous platelet-rich plasma. J
autologous serum on a cell culture inflammatory Refract Surg. 2007;23(6):617–9.
model. Exp Eye Res. 2016;151:115–21. 100. Alio JL, Colecha JR, Pastor S, Rodriguez A,
86. Stenwall PA, Bergström M, Seiron P, Sellberg Artola A.  Symptomatic dry eye treatment with
F, Olsson T, Knutson F, et  al. Improving the anti-­ autologous platelet-rich plasma. Ophthalmic Res.
inflammatory effect of serum eye drops using 2007;39(3):124–9.
allogeneic serum permissive for regulatory T cell 101. Alio JL, Rodriguez AE, Martinez LM, Rio
induction. Acta Ophthalmol. 2015;93(7):654–7. AL.  Autologous fibrin membrane combined with
87. Chiang C-C, Chen W-L, Lin J-M, Tsai solid platelet-rich plasma in the Management of
Y-Y.  Allogeneic serum eye drops for the treat- Perforated Corneal Ulcers. JAMA Ophthalmol.
ment of persistent corneal epithelial defect. Eye. 2013;131(6):745 751.
2009;23(2):290–3. 102. Alio JL, Rodriguez AE, Martinez LM. Bovine peri-
88. Chiang C-C, Lin J-M, Chen W-L, Tsai cardium membrane (Tutopatch) combined with solid
Y-Y. Allogeneic serum eye drops for the treatment of platelet-rich plasma for the management of perfo-
severe dry eye in patients with chronic graft-versus-­ rated corneal ulcers. Cornea. 2013;32(5):619–24.
host disease. Cornea. 2007;26(7):861–3. 103. Moreno R, Carreño MG, María J, Herreros A,
89. Na K-S, Kim MS.  Allogeneic serum eye drops Romero Garrido JA, López-Sánchez P. Plasma rico
for the treatment of dry eye patients with chronic en plaquetas: actualización de los sistemas emplea-
graft-versus-host disease. J Ocul Pharmacol Ther. dos para su obtención. Farm Hosp [Internet]. 2016
2012;28(5):479–83. [cited 2017 May 8];40(5):385–93. Available from:
90. Tang YQ, Yeaman MR, Selsted ME.  Antimicrobial http://scielo.isciii.es/pdf/fh/v40n5/05original05.pdf.
peptides from human platelets. Infect Immun. 104. Anitua E, Muruzabal F, Pino A, Merayo-Lloves
2002;70(12):6524–33. J, Orive G.  Biological stability of plasma rich in
91. Tohidnezhad M, Varoga D, Podschun R, Wruck CJ, growth factors eye drops after storage of 3 months.
Seekamp A, Brandenburg LO, et  al. Thrombocytes Cornea. 2013;32(10):1380–6.
are effectors of the innate immune system releasing 105. Geerling G, Maclennan S, Hartwig D.  Autologous
human beta defensin-3. Injury. 2011;42(7):682–6. serum eye drops for ocular surface disorders. Br J
92. Drago L, Bortolin M, Vassena C, Taschieri S, Del Ophthalmol. 2004;88(11):1467–74.
Fabbro M.  Antimicrobial activity of pure platelet-­ 106. Movahedan H, Ghassemifar V. Treatment of persis-
rich plasma against microorganisms isolated from tent corneal epithelial defect with autologous serum.
oral cavity. BMC Microbiol. 2013;13:47. Asian J Ophthalmol. 2006;8(6):236–41.
93. Anitua E, Muruzabal F, Orive G.  Antimicrobial 107. López-García JS, García-Lozano I, Rivas L,
properties of plasma rich in growth factors (PRGF- Martínez-Garchitorena J.  Use of autologous serum
ENDORET) technology. Sci Microb Pathog in ophthalmic practice. Arch Soc Esp Oftalmol.
Commun Curr Res Technol Adv. 2011(1):414–21. 2007;82(1):9–20.
94. Li H, Hamza T, Tidwell JE, Clovis N, Li B. Unique 108. Tananuvat N, Daniell M, Sullivan LJ, Yi Q,
antimicrobial effects of platelet-rich plasma and McKelvie P, McCarty DJ, et  al. Controlled study
its efficacy as a prophylaxis to prevent implant-­ of the use of autologous serum in dry eye patients.
associated spinal infection. Adv Healthc Mater. Cornea. 2001;20(8):802–6.
2013;2(9):1277–84. 109. Lagnado R, King AJ, Donald F, Dua HSA. Protocol
95. Bielecki T, Gazdzik TS, Szczepanski T.  Benefit for low contamination risk of autologous serum
of percutaneous injection of autologous platelet-­ drops in the management of ocular surface disorders.
leukocyte-­rich gel in patients with delayed union and Br J Ophthalmol. 2004;88(4):464–5.
nonunion. Eur Surg Res. 2008;40(3):289–96. 110. Noble BA, Loh RS, MacLennan S, Pesudovs K,
96. Dohan Ehrenfest DM, Andia I, Zumstein MA, Reynolds A, Bridges LR, et al. Comparison of autol-
Zhang C-Q, Pinto NR, Bielecki T.  Classification ogous serum eye drops with conventional therapy
of platelet concentrates (platelet-rich plasma-PRP, in a randomised controlled crossover trial for ocu-
platelet-rich fibrin-PRF) for topical and infiltra- lar surface disease. Br J Ophthalmol. 2004;88(5):
tive use in orthopedic and sports medicine: current 647–52.
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 343

111. von Hofsten J, Egardt M, Zetterberg M. The use of 125. Pan Q, Angelina A, Marrone M, Stark WJ, Akpek
autologous serum for the treatment of ocular surface EK.  Autologous serum eye drops for dry eye. Pan
disease at a Swedish tertiary referral center. Int Med Q, editor. Cochrane Database Syst Rev 2017;2.
Case Rep J. 2016;9:47–54. https://doi.org/10.1002/14651858. CD009327. pub3.
112. López-García JS, García-Lozano I, Rivas L, Ramírez Review.
N, Raposo R, Méndez MT.  Autologous serum eye 126. Anitua E, Muruzabal F, Tayebba A, Riestra A, Perez
drops diluted with sodium hyaluronate: clinical and VL, Merayo-Lloves J, et al. Autologous serum and
experimental comparative study. Acta Ophthalmol. plasma rich in growth factors in ophthalmology:
2014;92(1):22–9. preclinical and clinical studies. Acta Ophthalmol.
113. Semeraro F, Forbice E, Braga O, Bova A, Di 2015;93(8):e605–14.
Salvatore A, Azzolini C.  Evaluation of the effi- 127. Alio JL, Rodriguez AE, Ferreira-Oliveira R,
cacy of 50% autologous serum eye drops in dif- Wróbel-­Dudzińska D, Abdelghany AA.  Treatment
ferent ocular surface pathologies. Biomed Res Int. of dry eye disease with autologous platelet-rich
2014;2014:826970. plasma: a prospective, interventional, non-ran-
114. Gus PI, Marinho D, Zelanis S, Belló-Klein A, Locatelli domized study. Ophthalmol Ther. 2017;6(2):
C, Nicola F, et  al. A case-control study on the oxi- 285–93.
dative balance of 50% autologous serum eye drops. 128. Alio JL, Rodriguez AE, Abdelghany AA, Oliveira
Oxidative Med Cell Longev. 2016;2016:9780193. RF.  Autologous platelet-rich plasma eye drops
115. Liu L, Hartwig D, Harloff S, Herminghaus P, Wedel for the treatment of post-LASIK chronic ocu-
T, Geerling G.  An optimised protocol for the pro- lar surface syndrome. J Ophthalmol. 2017.
duction of autologous serum eyedrops. Graefes Arch https://doi.org/10.1155/2017/2457620.
Clin Exp Ophthalmol. 2005;243(7):706–14. 129. Alio JL, Rodriguez AE, De Arriba P, Gisbert S,
116. Fox RI, Chan R, Michelson JB, Belmont JB, Abdelghany AA.  Treatment with platelet-rich
Michelson PE.  Beneficial effect of artificial plasma of surgically related dormant corneal ulcers.
tears made with autologous serum in patients Eur J Ophthalmol. 2018;28(5):515–20.
with keratoconjunctivitis sicca. Arthritis Rheum. 130. Ortuño-Prados VJ, Alio JL.  Tratamiento de úlcera
1984;27(4):459–61. corneal neurotrófica con plasma rico en pla-
117. Viso E, Gude F, Rodríguez-Ares MT.  The asso- quetas y Tutopatch®. Arch Soc Esp Oftalmol.
ciation of meibomian gland dysfunction and other 2011;86(4):121–3.
common ocular diseases with dry eye: a population-­ 131. Arnalich F, Rodriguez AE, Luque-Rio A, Alio
based study in Spain. Cornea. 2011;30(1):1–6. JL.  Solid platelet rich plasma in corneal surgery.
118. Al-Saedi Z, Zimmerman A, Bachu RD, Dey S, Shah Ophthalmol Ther. 2016;5(1):31–45.
Z, Baugh R, et  al. Dry eye disease: present chal- 132. Javaloy J, Alió JL, Rodriguez AE, Vega A, Muñoz
lenges in the management and future trends. Curr G.  Effect of platelet-rich plasma in nerve regen-
Pharm Des. 2016;22(28):4470–90. eration after LASIK.  J Refract Surg. 2013;29(3):
119. Muñoz-Hernández AM, Santos-Bueso E, Cuiña-­ 213–9.
Sardiña R, Díaz-Valle D, Gegúndez-Fernández 133. Kim KM, Shin Y-T, Kim HK. Effect of autologous
JA, Benítez-del-Castillo JM.  New therapies for platelet-rich plasma on persistent corneal epithelial
neurotrophic keratitis. Arch Soc Esp Oftalmol. defect after infectious keratitis. Jpn J Ophthalmol.
2016;91(3):105–7. 2012;56(6):544–50.
120. Rosenthal AR, Harbury C, Egbert PR, Rubenstein 134. Lee JH, Kim MJ, Ha SW, Kim HK.  Autologous
E.  Use of a platelet-fibrinogen-thrombin mixture platelet-rich plasma eye drops in the treatment of
as a corneal adhesive: experiments with sutureless recurrent corneal erosions. Korean J Ophthalmol.
lamellar keratoplasty in the rabbit. Inv Ophthalmol. 2016;30(2):101–7.
1975;14(November):872–5. 135. Avila MY.  Restoration of human lacrimal function
121. Ralph R, Doane M, Dohlman C.  Clinical experi- following platelet-rich plasma injection. Cornea.
ence with a mobile ocular perfusion pump. Arch 2014;33(1):18–21.
Ophthalmol. 1975;93(10):1039–43. 136. Figueroa MS, Govetto A, De Arriba-Palomero
122. Semeraro F, Forbice E, Nascimbeni G, Taglietti M, P.  Short-term results of platelet-rich plasma as
Romano V, Guerra G, et  al. Effect of autologous adjuvant to 23-G vitrectomy in the treatment of
serum eye drops in patients with Sjögren syndrome-­ high myopic macular holes. Eur J Ophthalmol.
related dry eye: clinical and in vivo confocal micros- 2015;26(5):491–6.
copy evaluation of the ocular surface. In Vivo 137. Del Cid RMDA, Escoriaza IMDE. Subconjunctival
(Brooklyn). 2016;30(6):931–8. application of regenerative factor-rich plasma for the
123. Soni NG, Jeng BH.  Blood-derived topical ther- treatment of ocular alkali burns. Eur J Ophthalmol.
apy for ocular surface diseases. Br J Ophthalmol. 2009;19(6):909–15.
2016;100(1):22–7. 138. Anitua E, Alkhraisat MH, Orive G.  Perspectives
124. Noda-Tsuruya T, Asano-Kato N, Toda I, Tsubota and challenges in regenerative medicine using
K.  Autologous serum eye drops for dry eye after plasma rich in growth factors. J Control Release.
LASIK. J Refract Surg. 2006;22(1):61–6. 2012;157(1):29–38.
344 A. E. Rodriguez and J. L. Alió

139. Anitua E, Sánchez M, Orive G, Padilla S.  A bio- and author’s perspective. J Cutan Aesthet Surg.
logical therapy to osteoarthritis treatment using 2014;7(4):189–97.
platelet-rich plasma. Expert Opin Biol Ther. 153. Zhong W, Sumita Y, Ohba S, Kawasaki T, Nagai
2013;13(8):1161–72. K, Ma G, et  al. In vivo comparison of the bone
140. Anitua E, Pino A, Martinez N, Orive G, Berridi regeneration capability of human bone marrow
D.  The effect of plasma rich in growth factors on concentrates vs. platelet-rich plasma. PLoS One.
pattern hair loss: a pilot study. Dermatol Surg. 2012;7(7):e40833.
2017;43(5):658–70. 154. McCarrel TM, Minas T, Fortier LA.  Optimization
141. Lopez-Plandolit S, Morales M-C, Freire V, Grau of leukocyte concentration in platelet-rich plasma
AE, Duran JA.  Efficacy of plasma rich in growth for the treatment of tendinopathy. JBJS-American.
factors for the treatment of dry eye. Cornea. 2012;94(19):1–8.
2011;30(12):1312–7. 155. Sundman EA, Cole BJ, Fortier LA.  Growth factor
142. Merayo-Lloves J, Sanchez-Avila RM, Riestra AC, and catabolic cytokine concentrations are influenced
Anitua E, Begoña L, Orive G, et al. Safety and effi- by the cellular composition of platelet-rich plasma.
cacy of autologous plasma rich in growth factors Am J Sports Med. 2011;39(10):2135–40.
eye drops for the treatment of evaporative dry eye. 156. González M, Arteaga-vizcaíno M, Ruiz A, Briceño
Ophthalmic Res. 2016;56(2):68–73. O, Quintero M, Atencio R, et al. Niveles del factor
143. Merayo-Lloves J, Sanchez RM, Riestra AC, Anitua de crecimiento derivado de plaquetas en el plasma
E, Begoña L, Orive G, et  al. Autologous plasma rico en plaquetas antes y despues de antiagregantes
rich in growth factors eyedrops in refractory cases plaquetarios (PDGF levels in platelet-rich plasma
of ocular surface disorders. Ophthalmic Res. before and after anti platelets drugs). Av en Biomed.
2016;55(2):53–61. 2013;2(3):127–36.
144. Sanchez-Avila RM, Merayo-Lloves J, Riestra AC, 157. Weibrich G, Kleis WKG, Hafner G, Hitzler
Fernandez-Vega Cueto L, Anitua E, Begoña L, et al. WE.  Growth factor levels in platelet-rich plasma
Treatment of patients with neurotrophic keratitis and correlations with donor age, sex, and platelet
stages 2 and 3 with plasma rich in growth factors count. J Cranio-Maxillofacial Surg. 2002;30(2):
(PRGF-Endoret) eye-drops. Int Ophthalmol. 2017. 97–102.
https://doi.org/10.1007/s10792-017-0582-7. 158. Ronci C, Ferraro AS, Lanti A, Missiroli F, Sinopoli
145. Tanidir ST, Yuksel N, Altintas O, Yildiz DK, Sener S, Del Proposto G, et  al. Platelet-rich plasma as
E, Caglar Y. The effect of subconjunctival platelet-­ treatment for persistent ocular epithelial defects.
rich plasma on corneal epithelial wound healing. Transfus Apher Sci. 2015;52(3):300–4.
Cornea. 2010;29(6):664–9. 159. González M, Arteaga-Vizcaíno M, Benito M, Benito
146. Mazzocca AD.  Platelet-rich plasma differs accord- M. Application of platelet rich plasma (PRP) and its
ing to preparation method and human variability. J derivatives in dental implantologie and plastic sur-
Bone Jt Surg. 2012;94(4):308. gery. Investig Clin. 2012;53(4):408–18.
147. Urzua C a., Vasquez DH, Huidobro A, Hernandez 160. Kobayashi Y, Saita Y, Nishio H, Ikeda H, Takazawa
H, Alfaro J. Randomized double-blind clinical trial Y, Nagao M, et  al. Leukocyte concentration and
of autologous serum versus artificial tears in dry eye composition in platelet-rich plasma (PRP) influ-
syndrome. Curr Eye Res. 2012;37(8):684–8. ences the growth factor and protease concentrations.
148. Celebi ARC, Ulusoy C, Mirza GE.  The efficacy J Orthop Sci. 2016;21(5):683–9.
of autologous serum eye drops for severe dry 161. Yuan N, Wang C, Wang Y, Yu T, Long Y, Zhang X,
eye syndrome: a randomized double-blind cross- et al. Preparation of autologous platelet-rich gel for
over study. Graefes Arch Clin Exp Ophthalmol. diabetic refractory dermal ulcer and growth factors
2014;252(4):619–26. analysis from it. Chin J Reparative Reconstr Surg.
149. Stellos K, Kopf S, Paul A, Marquardt JU, Gawaz 2008;22(4):468–71.
M, Huard J, et  al. Platelets in regeneration. Semin 162. Zimmermann R, Jakubietz R, Jakubietz M, Strasser
Thromb Hemost. 2010;36(2):175–84. E, Schlegel A, Wiltfang J, et al. Different preparation
150. Amable PR, Carias RBV, Teixeira MVT, da Cruz methods to obtain platelet components as a source
Pacheco I, Corrêa do Amaral RJF, Granjeiro JM, of growth factors for local application. Transfusion.
et  al. Platelet-rich plasma preparation for regen- 2001;41(10):1217–24.
erative medicine: optimization and quantification of 163. Eppley BL, Woodell JE, Higgins J.  Platelet quan-
cytokines and growth factors. Stem Cell Res Ther. tification and growth factor analysis from platelet-­
2013;4(3):67. rich plasma: implications for wound healing. Plast
151. Pochini A de C, Antonioli E, Bucci DZ, Sardinha Reconstr Surg. 2004 Nov;114(6):1502–8.
LR, Andreoli CV, Ferretti M, et  al. Analysis of 164. Ambroziak AM, Szaflik J, Szaflik JP, Ambroziak M,
cytokine profile and growth factors in platelet-rich Witkiewicz J, Skopinski P.  Immunomodulation on
plasma obtained by open systems and commercial the ocular surface: a review. Cent Eur J Immunol.
columns. Einstein (São Paulo). 2016;14(3):391–7. 2016;41(2):195–208.
152. Dhurat R, Sukesh M.  Principles and methods 165. Castillo TN, Pouliot MA, Kim HJ, Dragoo
of preparation of platelet-rich plasma: a review JL. Comparison of growth factor and platelet concen-
22  Platelet-Rich Plasma (E-PRP) for Corneal Regeneration 345

tration from commercial platelet-rich plasma separa- fibronectin-mediated opsonic activity in plasma
tion systems. Am J Sports Med. 2011;39(2):266–71. cryoprecipitate with storage. Role of fibronectin
166. Grainger DJ, Mosedale DE, Metcalfe JC. TGF-beta fragmentation. Vox Sang. 1988;54(3):129–37.
in blood: a complex problem. Cytokine Growth 172. Plöderl K, Strasser C, Hennerbichler S, Peterbauer-­
Factor Rev. 11(1–2):133–45. Scherb A, Gabriel C. Development and validation of
167. Li ZD, Bork JP, Krueger B, Patsenker E, Schulze-­ a production process of platelet lysate for autologous
Krebs A, Hahn EG, et  al. VEGF induces prolif- use. Platelets. 2011;22(3):204–9.
eration, migration, and TGF-beta1 expression in 173. Nishida T.  Translational research in corneal
mouse glomerular endothelial cells via mitogen-­ epithelial wound healing. Eye Contact Lens.
activated protein kinase and phosphatidylinosi- 2010;36(5):300–4.
tol 3-kinase. Biochem Biophys Res Commun. 174. Klenkler B, Sheardown H, Jones L.  Growth fac-
2005;334(4):1049–60. tors in the tear film: role in tissue maintenance,
168. Hoeben A, Landuyt B, Highley MS, Wildiers H, Van wound healing, and ocular pathology. Ocul Surf.
Oosterom AT, De Bruijn EA.  Vascular endothelial 2007;5(3):228–39.
growth factor and angiogenesis. Pharmacol Rev. 175. Shen EP, Hu F-R, Lo S-C, Chen Y-M, Sun Y-C, Lin
2004;56(4):549–80. C-T, et  al. Comparison of corneal epitheliotrophic
169. Shibuya M. Structure and function of VEGF/VEGF-­ capacity among different human blood–derived
receptor system involved in angiogenesis. Cell preparations. Cornea. 2011;30(2):208–14.
Struct Funct. 2001;26(1):25–35. 176. Xiao X, He H, Lin Z, Luo P, He H, Zhou T, et  al.
170. López-Plandolit S, Morales M-C, Freire V, Therapeutic effects of epidermal growth fac-
Etxebarría J, Durán J a. Plasma rich in growth fac- tor on benzalkonium chloride–induced dry eye in
tors as a therapeutic agent for persistent corneal epi- a mouse model. Investig Opthalmology Vis Sci.
thelial defects. Cornea. 2010;29(8):843–8. 2012;53(1):191–7.
171. Blumenstock FA, Valeri CR, Saba TM, Cho E,
Melaragno A, Gray A, et  al. Progressive loss of
Part IV
Regenerative Surgery of the Corneal
Stroma
Applied Anatomy of the Corneal
Stroma
23
Harminder Singh Dua and Dalia G. Said

Key Points layer can strengthen the cornea and


• The corneal stroma is made of arrest or retard progressing of ectasia.
Bowman’s layer anteriorly, the pre-­ • Stromal lamellae are made predomi-
Descemet’s layer (Dua’s layer, also nantly of type I collagen, which are very
termed the Dua-Fine layer) posteriorly compact and intertwined anteriorly,
and the stromal lamellae and kerato- comparatively less compact centrally
cytes, which form the bulk of the cor- and posteriorly and relatively compact
nea, in between. again in the pre-Descemet’s layer.
• The function of Bowman’s layer is not • Organization of stromal collagen
well known as it is absent in some spe- lamellae is complex, being orthogonal
cies and does not seem to affect the cor- centrally and tangentially at the
nea when ablated in laser refractive periphery to accommodate the biome-
surgery. Transplantation of Bowman’s chanical stresses that the cornea is
subjected to.
• The pre-Descemet’s layer also contains
predominantly type I collagen but has
H. S. Dua (*) more elastin relative to other parts of the
Academic Section of Ophthalmology, Division of cornea. It is distinctly connected with
Clinical Neuroscience, University of Nottingham, the trabecular meshwork.
Nottingham, UK
• The intricate arrangement and charac-
Department of Ophthalmology, Queens Medical teristics of the cells and matrix struc-
Centre, University Hospitals NHS Trust,
Nottingham, UK tures collectively contribute to corneal
e-mail: harminder.dua@nottingham.ac.uk transparency.
D. G. Said • Elucidation of the corneal stromal anat-
Academic Section of Ophthalmology, Division of omy, especially of the posterior stroma,
Clinical Neuroscience, University of Nottingham, has enabled development and under-
Nottingham, UK
standing of lamellar corneal surgery and
Department of Ophthalmology, Queens Medical corneal pathology.
Centre, University Hospitals NHS Trust,
Nottingham, UK
Research Institute of Ophthalmology (RIO),
Cairo, Egypt

© Springer Nature Switzerland AG 2019 349


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_23
350 H. S. Dua and D. G. Said

The corneal stroma is one of the most organized Bowman’s layer, upon which rests the basement
connective tissue elements of the body. Lined by membrane of the corneal epithelium. The middle
cells on either side, it constitutes about 90% of component is the corneal stroma, which consti-
the corneal substance. Its structure is designed to tutes 90% (500 microns) of the corneal thickness,
act as the wall of the eyeball to protect the tissues and posteriorly is the pre-­Descemet’s layer (Dua’s
inside and maintain intraocular pressure, to be layer), which is lined by Descemet’s membrane
transparent to allow passage of light, to act as an (DM) and endothelium on its posterior surface.
optical lens to provide two thirds of the focusing
power of the eye and to be resilient to withstand Bowman’s Layer  It is also termed Bowman’s
external and internal forces that deform it yet membrane, Bowman’s zone or anterior limiting
spring back to its basic shape. In the interest of lamina (Fig.  23.1a). It is an 8–10-micron-­ thick
transparency, it is adapted to meet its energy acellular layer of primarily type I and type III col-
requirements without a vascular supply. lagen fibres which are narrower than those found in
the stroma. These are randomly woven into a dense
and compact layer of very uniform thickness for
23.1 Biometrics and Structure most part. It extends from limbus to limbus and
narrows towards its termination, which marks the
The corneal shape is that of an aspheric convex anterior limit of the cornea and the commencement
lens. Both anterior and posterior surfaces of the of the limbus. Like in the stroma, the space between
cornea contribute to its overall refractive power. collagen fibres is filled with proteoglycans and gly-
Externally, the clear cornea is horizontally oval cosaminoglycans (see below). The anterior surface
measuring 11–12 mm in the horizontal axis and of Bowman’s layer is very smooth, while at its pos-
9–11 mm vertically, in adults. The oval shape is terior surface, the fibres extend into anterior stroma
due to the scleralization of the superior and infe- with which they are intricately related. On in vivo
rior approximately 0.5–0.75  mm. When viewed confocal microscopy, it appears as a wavy structure
from the posterior surface, the corneal outline with some prominent crisscrossing lines of colla-
appears circular. The central cornea is between gen in the interface of Bowman’s layer and anterior
500 and 550 microns, increasing gradually to stroma, called K-structures. These are believed to
around 700 microns at the periphery. The radius be the basis of the mosaic pattern seen in the ante-
of curvature differs in the central optical zone rior cornea [3, 4] (Fig. 23.1b). Although the turn-
(3 mm diameter) where it is 7.5–8 mm, from the over of Bowman’s layer is attributed to the
peripheral zone where it is flatter giving the underlying stromal keratocytes, it does not regener-
aspheric optical effect. The total refractive power ate when damaged by injury or photorefractive
of the cornea is 40–44 dioptres. The refractive keratectomy. Its function is not clear, but it is sug-
index of the cornea is 1.376 compared to air, gested that it affords rigidity to the anterior cornea
which is 1.00, and tears and aqueous which are and helps maintain the refractive shape and pro-
both approximately 1.336. The cornea transmits vides a firm, smooth and regular surface for the
about 90% of the incident light [1, 2]. corneal epithelium. However, some species of
mammals do not have a Bowman’s layer [5], and
these functions are still apparently well served.
23.2 Corneal Stroma Transplantation of Bowman’s layer in patients with
keratoconus is being performed to afford additional
In broad terms, the stroma is organized in three strength and rigidity to the cornea in patients with
components based on anatomical, behavioural and keratoconus [6].
functional characteristics determined by the cells
within these components and in part adjacent to Stroma  The stroma is made essentially of extra-
them. The anteriormost component is Bowman’s cellular matrix, which accommodates cells and
membrane or, perhaps technically more correctly, nerves.
23  Applied Anatomy of the Corneal Stroma 351

b c

Fig. 23.1 (a) Resin-embedded semi-thin section of the stain). (b) In vivo confocal microscopy enface image of
anterior corneal stroma and epithelium. The stratified Bowman’s layer with sub-basal nerves and K-lines. (c)
squamous epithelium sits on Bowman’s layer, which Transmission electron microscopy of the corneal stroma
appears as homogenous structure of uniform thickness. showing the transversely (top) and horizontally cut lamel-
The underlying anterior stroma shows collagen fibres of lae (bottom)
the stromal lamellae and keratocytes (toluidine blue

Extracellular Matrix  The extracellular matrix of ment of collagen is predominantly orthogonal,


collagen and glycosaminoglycans accounts for i.e. at right angles to each other (Fig. 23.1c), in
about 70% of the dry weight of the cornea. Type the mid and posterior cornea and the largely iso-
I collagen is the most abundant with lesser tropic, i.e. similar in all directions, in the anterior
amounts of type III, type V and type VI [7, 8]. stroma [10–12]. Furthermore, there is a preferen-
Collagen fibres are bundled together in about tial orientation of collagen along the superior-­
200–250 lamellae of 2 microns thickness that run inferior and nasal-temporal directions which
from limbus to limbus, spanning the entire cornea seems to be an adaptation to absorb biomechani-
[9]. Fibres in the lamellae run parallel to each cal stress exerted by the rectus and orbicularis
other, and the lamellae run parallel to the surface muscles [13]. The arrangement of lamellae in the
of the cornea, but individual lamellae subtend anterior third is more oblique with some branch-
large angles to adjacent lamellae [8]. The arrange- ing [14] and is undulating and interwoven [8, 15].
352 H. S. Dua and D. G. Said

A complex branching pattern of collagen fibres is nent of the stroma, there is also a well-defined
seen in the anterior stroma where fibres extend network of elastin fibres [22, 23]. An annulus of
from Bowman’s layer, intertwine with the stro- an elastic fibre system is also present at the
mal tissue and reinsert in Bowman’s layer. This cornea-­scleral limbus [24, 25]. These structural
pattern diminishes systematically towards the features of the peripheral part of the cornea
posterior stroma [16]. These features confer a explain the circular movement of air at the
greater compactness to the anterior stroma. periphery and the limit of 8.5–9 mm diameter of
the type 1 big bubble in DALK [18] and the
There is also a difference in compactness of greater force required to separate the peripheral
the peripheral and central stroma in the middle stroma [17].
and posterior parts of the stroma. The force
required to separate the stromal tissue is less in Glycosaminoglycans  Glycosaminoglycans also
the centre than at the periphery of the cornea called mucopolysaccharides are made of repeat-
[17]. Air injected in the stroma, as in the proce- ing units of disaccharide that form long
dure of deep anterior lamellar keratoplasty unbranched structures. They are highly polar
(DALK), follows a distinct circular pattern of (having negative and positive charges that do not
movement in the peripheral cornea (Fig. 23.2a), neutralize each other) and attract water. The
which corresponds to the 1.5–2 mm of compact space between collagen fibres is filled with
stroma at the periphery, which also corresponds ground substance made of the glycosaminogly-
to the flatter peripheral part of the prolate cornea. cans keratan sulphate, dermatan sulphate, chon-
Histology of this area after air insufflation sug- droitin and hyaluronan. The latter is most
gests an interweave of collagen that is different abundant in the developing cornea and declines
from that seen in the central cornea [18] thereafter. In the adult cornea, keratan sulphate is
(Fig. 23.2b). There is a ‘circular annulus’ of col- the most abundant and along with dermatan sul-
lagen fibres at the periphery of the cornea. The phate and chondroitin can bind to proteins in the
fibres are oriented tangentially to the limbus stroma to form proteoglycans. The posterior cor-
probably creating an anatomical landmark as nea contains more keratan sulphate, which is
they transition from orthogonal to the tangential more hydrophilic than dermatan sulphate, which
orientation in the annulus adding to the compact- is more in the anterior stroma [26, 27]. This con-
ness and tighter attachment of fibres at the tributes to the differential swelling of the cornea,
periphery [11, 19–21]. Though collagen is by far with greater swelling of the posterior stroma
the most dominant extracellular matrix compo- compared to the anterior stroma [10].

a b

Fig. 23.2  Corneal stromal architecture. (a) Air injected the expanded peripheral cornea with wide interlamellar
in the stroma travels circumferentially along the limbus, spaces (arrows). This arrangement offers less resistance to
in the peripheral 1–1.5 mm of the cornea. This consistent the passage of air compared to the central cornea.
movement of air suggests a special construction of this (Reproduced from the author’s own publication Dua et al.
part of the stroma. (b) Histology of the cornea showing 2018 [18])
23  Applied Anatomy of the Corneal Stroma 353

Keratocytes  Keratocytes constitute most of the By in vivo confocal microscopy (IVCM), it has
cells that reside in the stroma. Macrophages, been estimated that the corneal stroma contains
dendritic cells and transiting bone marrow- about 20,500 cells per cubic millimetre (Fig. 23.3c).
derived cells are also present in very small num- The density of cells in anterior stroma is 765 ± 262
bers. The total cell mass is less than 4% of the cells/mm2, in mid-stroma around 347 ± 64.4 cells/
stromal mass [28, 29]. Keratocytes are akin to mm2 and in posterior stroma about 315 ± 57.2 cells/
the fibroblasts of other connective tissues, hence mm2. The greatest density is in the anterior 10% of
also termed stromal fibroblasts. Keratocytes are the stroma. Keratocyte density is age related and
located between lamellae and have a central flat- declines by about 0.45% per year being about twice
tish spindle-­shaped cell body from which mul- the amount in the anterior stroma. With IVCM,
tiple cytoplasmic processes extend in all keratocytes in the anterior stroma demonstrate
directions, making contact with similar pro- smaller nuclei than those in the posterior stroma
cesses of neighbouring keratocytes (Figs. 23.1a [32, 33]. The limbal stroma acts as a stem cell niche
and 23.3). Cells are capable of communication giving residence to progenitor keratocytes or kera-
with each other by exchange of information via tocyte stem cells. The peripheral and limbal corneal
gap junctions at the tips of keratocyte processes. stromal cells express all features of mesenchymal
This enables them to act as one cell mass and stem cells that have the potential to transdifferenti-
respond collectively to physiologic and patho- ate to an epithelial phenotype [34, 35].
logic stimuli. Keratocytes also express integrins Keratocytes function to produce and maintain
through which they interact with the surround- the extracellular matrix of collagen and glycos-
ing extracellular matrix, and CD34, a trans- aminoglycans. Collagen secretion starts with
membrane phosphoglycoprotein in abundance, intracellular synthesis of pro-α-collagen. Three
which also serves as an adhesion molecule, hydroxylated and glycosylated molecules are
binding to its ligand, L-selectin, in the matrix assembled into procollagen, which is secreted out
[30, 31]. The keratocyte-matrix interaction reg- by the cells. In the extracellular space, procolla-
ulates and modulates keratocyte behaviour such gen is converted to collagen molecules, several of
as mitotic rate and activity. Normally they are which form collagen fibrils, which in turn com-
quiescent and with a very low proliferative rate. bine to form tough collagen fibres with a charac-
Keratocytes have a well-developed cytoskeleton teristic cross striation pattern, with a periodicity
of predominantly actin filaments, which give the of 67  nm. In the corneal stroma, the collagen
cells contractile ability. fibres are very uniform in diameter ranging from
22.5 to 35 nm [8, 36].

a b c

Fig. 23.3  Keratocytes. (a) The anterior stroma shows a of a keratocyte process within the stromal lamellae. An
large number of keratocytes stained for CD34 (an adhe- intracytoplasmic deposit (vacuole) of lipid material is
sion molecule) marker, which is seen on all keratocytes. seen (arrow). (c) In vivo confocal microscopic image of
Bowman’s layer is acellular. The corneal epithelium is keratocytes in the corneal stroma. The nuclei bodies
artefactually detached from Bowman’s layer (immuno- appear white with some showing extensions of keratocyte
peroxidase stain). (b) Transmission electron micrograph processes
354 H. S. Dua and D. G. Said

Keratocytes contain water-soluble structural wounding or disease, keratocytes near the pathol-
proteins called crystallins of which aldehyde ogy break their intercellular connections and then
dehydrogenase is the most prominent. Crystallins behave independently and can transform to myo-
help in maintaining the transparency of the kera- fibroblasts and activated fibroblasts [39]
tocyte and cornea [37, 38]. They also secrete (Fig.  23.4a–c). A common response of kerato-
matrix metalloproteases, which degrade the cytes to injury including laser refractive surgery
extracellular matrix in the physiologic role of (re) and cross-linking is apoptosis, which is believed
modelling the stroma and in disease states, caus- to be a ‘defence’ reaction to limit spread of the
ing ulceration by excessive degradation or lysis infection or injurious agent [40, 41]. In culture,
of the collagen (stromal melting). The function of keratocytes can express cytokeratin 3, suggesting
MMPs is balanced by the presence of tissue possible mesenchymal-epithelial transition [42]
inhibitors of MMPs (TIMPS). In response to (Fig. 23.4d–f).

a b c

d e f

Fig. 23.4  Keratocyte variations in cornea transplanted immunophenotype. (c) The section is stained for alpha
with amniotic membrane on the surface. (a) Transverse smooth muscle actin. All corneal stromal keratocytes are
section of the corneal stroma with the epithelized, trans- negative, but the cells in the amniotic stroma are positive
planted amniotic membrane on the surface (haematoxylin indicating that they are keratocyte-derived myofibro-
and eosin stain). Keratocytes are seen in the stroma of the blasts. (d) Cultured corneal keratocytes stained with anti-
cornea and in the amniotic stroma where they have ­CD34 antibody (red). (e) Cultured corneal keratocytes
migrated from the corneal stroma through breaches in stained with anti-cytokeratin 3 antibody (green). (f)
Bowman’s layer. (b) The corneal stroma shows kerato- Superimposition of the images showing that the same cell
cytes (brown) stained with CD34 antibody. The kerato- expresses both CD34 and cytokeratin 3 – mesenchymal-­
cytes in the amniotic stroma do not show any stain with epithelial transition. (a–c Adapted in part from authors’
CD34 indicating that they have lost the resting keratocyte own publication Said et al. [39])
23  Applied Anatomy of the Corneal Stroma 355

23.3 Stromal Nerves stroma. A close anatomical relationship with


stromal keratocytes and macrophages has also
The cornea is the most sensitive structure in the been described [46, 49]. Stromal nerves show
human body, being 40 times more sensitive than considerable anomalies in corneal pathology, a
dental pulp and over 400 times more than the skin fact that is increasingly being recognized with the
[43]. Corneal innervation is predominantly sen- advent of IVCM [50] (Fig. 23.6).
sory, from the ophthalmic division of the trigemi- Autonomic innervation is from postganglionic
nal (V cranial) nerve. The nasociliary branch of fibres from the SCG that ascend in the plexus
the ophthalmic division of the trigeminal nerve is around the internal carotid artery. Ocular fibres
the main nerve of the ocular surface. It enters the leave in the cavernous sinus and enter the orbit
orbit through the superior orbital fissure. Its through the superior orbital fissure as the sympa-
branches include the anterior and posterior eth- thetic root of the ciliary ganglion. These fibres
moidal nerves, two or three long ciliary nerves either merge directly with the long ciliary nerves
and a communicating branch to the ciliary gan- or pass through the ganglion, without synapse, to
glion. It terminates as the infra-trochlear and emerge in the short ciliary nerves. Human cor-
nasal nerves. Six short ciliary nerves arise from neas have sparse sympathetic innervation [51].
the ciliary ganglion and together with the long
ciliary nerves enter the suprachoroidal space by
penetrating the sclera around the optic nerve. 23.4 P
 re-Descemet’s Layer (Dua’s
They pass anteriorly, supply the iris and ciliary Layer, PDL)
body and terminate in the pericorneal (limbal)
plexus. The limbal plexus thus has both sensory The development and practice of lamellar corneal
and autonomic nerves and is predominantly vaso- surgery especially DALK by the big-bubble tech-
motor in function [44]. nique brought to light a layer at the posteriormost
Eleven nerve bundles per quadrant, carrying a part of the stroma termed pre-Descemet’s layer
mixture of sensory and autonomic fibres, pass or Dua’s layer. The American Association of
through the limbus and enter the cornea in the Ocular Oncologists and Pathologists has recently
middle third of the stroma (Fig. 23.5). Just before termed this the Dua-Fine layer (Fig. 23.7).
or soon after entering the corneal stroma, the The PDL is less than 20 microns
nerve bundles lose their perineurium and con- (10.15 ± 3.6 m) thick and is made up of 5–8 com-
tinue radially and anteriorly between the collagen pact lamellae of predominantly type I collagen
lamellae towards the central area, giving rise to like most of the stroma. The lamellae are orga-
branches that spread through the anterior and nized in longitudinal, transverse and oblique
mid-stroma. The posterior stroma lacks nerves directions. It also has types IV and VI (long-­
though some innervation of the corneal endothe- spacing) collagen, which are more than in the
lium has been described [45]. The axons of the stroma (Fig. 23.7a–e). The width of the lamellae
central stromal nerves are unmyelinated and run is significantly narrower than in the correspond-
in the anterior stroma parallel to collagen bun- ing region of the stroma immediately anterior to
dles. Most of the axons in these bundles are about the PDL, where there are 3–5 layers for the same
0.5 μm in diameter but can be as thick as 2.5 μm width of tissue [52, 53]. The PDL is by and large
[46, 47]. There is a loose sub-Bowman’s plexus acellular though occasional keratocytes or kera-
of nerves from where fibres penetrate Bowman’s tocyte processes can be seen in it [54]. The col-
zone, predominantly in the mid-periphery of the lagen fibres in the PDL are also significantly
cornea, and emerge in the sub-basal plane of the thinner than in the stromal lamellae anterior to it
epithelium where they end in single or multiple [52]. The PDL also has a high elastin content,
bulblike structures which probably represent the more than any other part of the cornea and as
termination and folding of the nerve sheath [48]. much as the trabecular meshwork [24, 25, 55].
A small population of axons terminate in the This explains why, during DALK, the wall of the
356 H. S. Dua and D. G. Said

a b

c d

Fig. 23.5  Corneal stromal nerves. (a) Thick myelinated copy of the thick peripheral (limbal) entrance of a stromal
corneal nerves are seen entering the peripheral cornea nerve. (c) The same nerve narrows towards the mid-­
from the limbus. These become invisible as they lose their periphery and (d) tapers to become thin and fine in the
myelin towards the centre of the cornea. Inset: slit view centre of the corneal stroma on its way to the sub-­
showing the nerves in cross section lying in the middle Bowman’s plexus
third of the stroma (arrows). (b) In vivo confocal micros-

big bubble expands and, on deflation, returns to air after excimer laser ablation of the PDL (in
its original dimensions. The compactness of the experimental eyes) fails to produce a big bubble
layer coupled with the elastin content also prob- indicating that the PDL is unique [56]. The PDL
ably accounts for the fact that the layer is imper- can be visualized in  vivo between DM and the
vious to air. On forceful injection, air in the last row of keratocytes by ultrahigh-resolution
corneal stroma traverses through the stroma optical coherence tomography [57]. There is a
(probably around keratocyte spaces) until it distinct plane of cleavage between the PDL and
reaches the PDL. Here it cannot pass through the the posterior stroma, with the separation occur-
layer and hence cleaves it off the posterior stroma ring along the last row of keratocytes. This plane
to form the type 1 bubble in DALK. Injection of is exploited in big-bubble DALK where the type
23  Applied Anatomy of the Corneal Stroma 357

a d

b e

c f

Fig. 23.6  Altered corneal nerves in bullous keratopathy. stromal nerve is seen. Note the absence/decreased visibil-
In vivo confocal microscopy (IVCM) (a, b, c) and corre- ity of keratocytes. (e) Corresponding image on whole-­
sponding whole-mount acetylcholinesterase-stained (d, e, mount stain showing aberrant hyper-regeneration of fine
f) corneal stromal nerves. (a) Thickened nerve with nerves from the bifurcated trunk that are coiled and tortu-
excrescences (arrows) and involuted excrescence (arrow- ous. These were not visible on IVCM. (c) Loops and coils
head). (d) Corresponding image on whole-mount stain of relatively thicker nerves in the stroma. (f) Corresponding
showing that the excrescences (arrows) could represent images on whole-mount stain showing a similar aberrant
the site of origin of branches, which may be undergoing pattern (Reproduced from the author’s own publication,
involution (arrowhead). (b) A bifurcation of a thickened Al-Aqaba et al. [50])
358 H. S. Dua and D. G. Said

a b c

d e f

Fig. 23.7  The pre-Descemet’s layer (Dua’s layer PDL on the right and the PDL on the left. The PDL shows the
also termed the Dua-Fine layer by the American lamellar structure with some areas of darker long-spacing
Association of Ocular Oncologists and Pathologists). (a) collagen towards the DM interface. (d) Scanning electron
Light photomicrograph of the PDL that was separated micrograph (SEM) of the PDL, separated from the deep
from the underlying stroma by the injection of air. stromal bed anterior to it. The edge of the PDL is clearly
Descemet’s membrane (DM) has been stripped off the visible. (e) SEM of the cornea from which the EL with
PDL. (b) Transmission electron micrograph (TEM) of the DM has been separated. The two layers are scrolled
PDL showing that it is made of multiple thin and com- together with the endothelial cells outside. (f) SEM of
pactly arranged lamellae. The DM has been peeled off. (c) periphery of PDL showing a cluster of fenestrations
Higher-power TEM of the PDL and DM. The dark verti- (arrow). These fenestrations are distributed all along the
cal band in the middle of the image is the banded layer of periphery of PDL and allow air injected in the stroma to
the DM. The non-banded layer and the endothelial cell are pass through PDL to create a type 2 big bubble

1 bubble represents the separation of the PDL during DALK with a type 1 bubble, also termed
together with DM and endothelium from the the DALK triple procedure [61]. Another surgical
­posterior stroma. The plane can also be surgically procedure termed pre-Descemet’s endothelial
accessed when air injection fails to create the big keratoplasty (PDEK) involves use of the PDL
bubble. Such separation is also manifest in cor- together with DM and endothelial cells as the
neal pathology related to infection, offering a transplant tissue to replace host DM and endothe-
plane for spread of fungi [58], and in chronic cor- lium, as an alternative to DMEK with good
neal oedema where spontaneous dehiscence of results [62].
the PDL together with the DM occurs [59]. There At its periphery, all along the circumference,
is evidence to suggest that acute hydrops in kera- the collagen fibres of the PDL start to fan out like
toconus is related to a rupture of the DM and a three-dimensional crisscrossing sieve and con-
PDL [60] and that the PDL is part of descemeto- tinue as the core of the trabecular meshwork
celes [52]. There is also a clear plane of cleavage beams, which too have the same collagen and
between PDL and DM which becomes apparent elastin content as the PDL. Trabecular cells rest-
in DM detachments, DM peeling during ing on basement membrane are seen in the
Descemet’s membrane endothelial keratoplasty peripheral 350 microns of the stroma of the PDL
(DMEK) and in the type 2 bubble of DALK [52, [63]. Clusters of fenestrations are presented in
53]. The PDL is strong and resilient with a burst- the PDL, within 500 microns central to the termi-
ing pressure (of type 1 bubble) of 500–700 mm of nation of DM. Fifteen to twenty fenestrations or
mercury. This allows surgeons to perform phaco- clusters of fenestrations are present along the cir-
emulsification with lens implant under the PDL cumference, on either side of the termination of
23  Applied Anatomy of the Corneal Stroma 359

the DM. These are present within 500 microns of any of these singly or in combination can com-
the termination of DM centrally and between ter- promise transparency and affect vision. Intrinsic
mination of DM and the trabecular meshwork transparency of the cornea is related primarily to
peripherally. The fenestrations measure between the architecture of the stromal collagen. The very
5 and 60 microns with a mean of 20.3 microns uniform diameter of the collagen fibres, the
[63] (Fig. 23.7f). These fenestrations explain the equally uniform and consistent spacing of the
passage of air into the anterior chamber during fibres, the relative state of dehydration (78%
DALK and the occurrence of a type 2 bubble water content), the unmyelinated nerves that pop-
(and mixed bubble) wherein air passes through ulate the stroma, the ‘transparent’ keratocytes
these fenestrations to access the plane between and the absence of blood vessels all make signifi-
PDL and DM. cant contributions to transparency.
There is an embryologic basis for the The mean diameter of the corneal fibres and
PDL. Separation of the lens vesicle from the sur- the interfibre distance is less than half of the
face ectoderm initiates the formation of the cor- wavelength of visible light, which is 400–700 nm.
nea, with the surface ectoderm forming the Thus any scattering of a light ray passing through
two-layered epithelium [64]. The epithelium one fibre is cancelled by the scatter of the adja-
secretes collagen and glycosaminoglycans that cent ray allowing light to pass through the cornea
fill the space between the epithelium and the lens [73]. Any alteration in the finely organized archi-
vesicle as the acellular primary stroma [65–68]. tecture of the collagen fibres, as occurs with
The first wave of migration of neural crest cells oedema and scarring, leads to random scattering
from the rim of the developing optic cup passes and back scatter (diffuse and irregular reflection),
between the posterior layer of the primary stroma limiting the amount and quality (focused) of light
and the lens vesicle to form the embryonic endo- entering the eye.
thelium. The third wave of neural crest cells Despite the water-holding property of the gly-
migrate into the middle of the primary stroma cosaminoglycans, the cornea is maintained in a
and form the stromal keratocytes (second wave state of relative dehydration, in the interest of
forms the iris and pupillary membrane). The ker- transparency. This is largely brought about by the
atocytes lay down the definitive or secondary endothelial pump function of the endothelium.
stroma. The primary stroma anterior to the Surface evaporation of water and the barrier
expanding secondary stroma is compressed and afforded by the epithelium also contribute to loss
forms the basis of Bowman’s layer. It is likely of stromal water. Hence corneal thickness is
that the compressed primary stroma posterior to physiologically slightly more on waking. In
the secondary stroma forms the basis of the patients with early endothelial failure, e.g. Fuchs
PDL. The PDL is also likely to be influenced by endothelial dystrophy, the complaint of blurry
the endothelium because of its proximity to this vision on waking, which can take a few hours to
layer. The endothelium is known to contribute clear, is related to, among other factors, the loss
collagens [69–71] and hyaluronan [72] to the of evaporation during sleep. Epithelial loss (abra-
developing posterior stroma (PDL). sion) is associated with stromal oedema due to
breach of the epithelial barrier function. It is
interesting that localized endothelial dysfunction
23.5 Corneal Transparency or loss is associated with localized oedema of the
overlying stroma only. This suggests that the
Maintaining transparency to allow entry of light endothelial pump function strictly serves the
to the eye is the most important function of the stroma directly anterior to it with little if any lat-
cornea. Several factors both anatomical and eral effect.
physiological contribute to this. The epithelium, The magnitude of the contribution of the endo-
stroma and endothelium all have important roles, thelium to maintaining transparency can be seen
and compromise in structure and/or function of in the fact that an excised cornea swells up to
360 H. S. Dua and D. G. Said

1200 microns thickness with an increase in hydra- 7. McLaughlin JS, Linsenmayer TF, Birk DE.  Type V
collagen synthesis and deposition by chicken embryo
tion from 78% to 98%, related to the swelling corneal fibroblasts in vitro. J Cell Sci. 1989;94:371–9.
pressure of the stroma, i.e. the ability of the polar 8. Komai Y, Ushiki T. The three-dimensional organisa-
glycosaminoglycans to absorb water and separate tion of collagen fibrils in the human cornea and sclera.
the collagen fibres. This pressure is countered by Invest Ophthalmol Vis Sci. 1991;32:2244–58.
9. Rodrigues MM.  Cornea. In: Jakobiec FA, edi-
the endothelial pump with a little help from the tor. Ocular anatomy, embryology and teratology.
intraocular pressure, in vivo. Increased hydration Philadelphia: Harper & Row; 1982.
of the cornea causes the cornea to become hazy, 10. Muller LJ, Pels E, Vrensen GF. The specific architec-
while fibrosis/scarring causes opacity. Clinically ture of the anterior stroma accounts for maintenance of
corneal curvature. Br J Ophthalmol. 2001;85:437–43.
there is an interesting anomaly in that corneas suf- 11. Abahussin M, Hayes S, Knox Cartwright NE,

fering from complete hydration for several months Kamma-Lorger CS, Khan Y, Marshall J, Meek KM.
due to endothelial failure show remarkable clear- 3D collagen orientation study of the human cor-
ance after endothelial keratoplasty. Equally, there nea using X-ray diffraction and femtosecond laser
technology. Invest Ophthalmol Vis Sci. 2009;50:
are some such corneas that retain a degree of 5159–64.
‘haze’ and poor quality of vision post-operatively 12. Meek KM, Knupp C. Corneal structure and transpar-
(authors’ experience), suggesting that prolonged ency. Prog Retinal Eye Res. 2015;49:1–16.
oedema can lead to structural changes in collagen 13. Boote C, Dennis S, Huang Y, Quantock AJ, Meek
KM. Lamellar orientation in human cornea in relation
organization and/or abnormal collagen produced to mechanical properties. J Struct Biol. 2005;149:1–6.
by the stressed keratocytes [1]. 14. Hogan MJ, Alvarado JA, Wedell JE.  In: Hogan MJ,
Alvarado JA, Wedell JE, editors. Histology of the
Declaration of Interest None of the authors have any human eye. Philadelphia: Saunders; 1971.
conflict of interest related to the subject matter and con- 15. Radner W, Zehetmayer M, Aufreiter R, Mallinger

tent of the chapter. HS Dua is Consultant for Dompe, R.  Interlacing and cross-angle distribution of col-
Santen, Thea and Shire. He has shares in NuVision lagen lamellae in the human cornea. Cornea.
BioTherapeutics and GlaxoSmithKline. No human or ani- 1998;17:537–43.
mal studies were carried out by the authors for this 16. Winkler M, Chai D, Kriling S, Nien CJ, Brown DJ,
chapter. Jester B, Juhasz T, Jester JV. Nonlinear optical macro-
scopic assessment of 3-D corneal collagen organiza-
tion and axial biomechanics. Invest Ophthalmol Vis
Sci. 2011;52:8818–27.
References 17. Smolen MK, McCarey BE.  Interlamellar adhe-

sive strength in human eye bank corneas. Invest
1. Nishida T. Basic sciences: cornea, sclera, and ocular Ophthalmol Vis Sci. 1990;31:1087–95.
adnexa anatomy, biochemistry, physiology, and bio- 18. Dua HS, Faraj LA, Kenawy MB, AlTaan S, Elalfy MS,
mechanics. Cornea. In: Krachmer JH, Mannis MJ, Katamish T, Said DG. Dynamics of big bubble forma-
Holland EJ, editors. Cornea, fundamentals of cornea tion in deep anterior lamellar keratoplasty by the big
and external disease. New York: Mosby; 1997. bubble technique: in vitro studies. Acta Ophthalmol.
2. Fares U, Otri AM, Al-Aqaba MA, Dua HS. Correlation 2018;96:69–76.
of central and peripheral corneal thickness in healthy 19. Kokott W.  Ubermechanisch-funktionelle strikturen
corneas. Cont Lens Anterior Eye. 2011;35:39–45. des auges. Albrecht Von Graefes Arch Ophthalmol.
3. Yokogawa H, Kobayashi A, Sugiyama K. Mapping of 1938;138:424–85.
normal corneal K-structures by in vivo laser confocal 20.
Newton RH, Meek KM.  Circumcorneal annu-
microscopy. Cornea. 2008;27:879–83. lus of collagen fibrils in the human limbus. Invest
4. Kobayashi A, Yokogawa H, Sugiyama K. In vivo laser Ophthalmol Vis Sci. 1998;39:1125–34.
confocal microscopy of Bowman’s layer of the cor- 21.
Aghamohammadzadeh H, Newton RH, Meek
nea. Ophthalmology. 2006;113:2203–8. KM. X-ray scattering used to map the preferred col-
5. Gipson IK.  Anatomy of the conjunctiva, cornea and lagen orientation in the human cornea and limbus.
limbus. In: Smolin G, Thoft RA, editors. The cornea, Structure. 2004;12:249–56.
scientific foundations and clinical practice. New York: 22. M'Ilroy JH.  On the presence of elastic fibres in the
Little Brown and company; 1994. cornea. J Anat Physiol. 1906;40:282–91.
6. van Dijk K, Parker JS, Baydoun L, Ilyas A, Dapena 23. Asejczyk-Widlicka M, Srodka DW, Kasprzak H,

I, Groeneveld-van Beek EA, Melles GRJ.  Bowman Pierscionek BK.  Modelling the elastic properties
layer transplantation: 5-year results. Graefes Arch of the anterior eye and their contribution to mainte-
Clin Exp Ophthalmol. 2018. https://doi.org/10.1007/ nance of image quality: the role of the limbus. Eye.
s00417-018-3927-7. 2007;21:1087–94.
23  Applied Anatomy of the Corneal Stroma 361

24. Lewis PN, White TL, Young RD, Bell JS, Winlove CP, system in the modulation of corneal tissue organi-
Meek KM.  Three-dimensional arrangement of elas- zation and wound healing. Exp Eye Res. 1996;62:
tic fibers in the human corneal stroma. Exp Eye Res. 325–57.
2016;146:43–53. 41. Wilson SE, Pedroza L, Beuerman R, Hill JM. Herpes
25. White TL, Lewis PN, Young RD, Kitazawa K, Inatomi simplex virus type-1 infection of corneal epithelial
T, Kinoshita S, Meek KM. Elastic microfibril distri- cells induces apoptosis of the underlying keratocytes.
bution in the cornea: differences between normal and Exp Eye Res. 1997;64:775–9.
keratoconic stroma. Exp Eye Res. 2017;159:40–8. 42. Perrella G, Scott CA, Spelat R, Brusini P, D’Aurizio
26. Bettelheim FA, Plessy B.  The hydration of proteo- F, De P, Dua HS.  Cultured human keratocytes from
glycans of bovine cornea. Biochem Biophys Acta. the limbus and cornea both express epithelial cyto-
1975;383:203–14. keratin 3: possible mesenchymal-epithelial transi-
27.
Castoro JA, Bettelheim FA.  Water gradients tion. Int J Ophthalmic Pathol. 2012;1:2. https://doi.
across bovine cornea. Invest Ophthalmol Vis Sci. org/10.4172/2324-8599.1000101.
1988;29:963–8. 43. Bonini S, Rama P, Olzi D, Lambiase A. Neurotrophic
28. Hassell JR, Birk DE. The molecular basis of corneal keratitis. Eye. 2003;17:989–95.
transparency. Exp Eye Res. 2010;91:326–35. 44. Marfurt CF, Cox J, Deek S, Dvorscak L.  Anatomy
29. Otori T.  Electrolyte content of the rabbit corneal
of the human corneal innervation. Exp Eye Res.
stroma. Exp Eye Res. 1967;6:356–67. 2010;90:478–92.
30. Joseph A, Hossain P, Jham S, Jones RE, Tighe P, 45. Leon-Feliu E, Gomez-Ramos P, Rodriguez-Echandia
McIntosh RS, Dua HS.  Expression of CD34 and EL. Endothelial nerve fibres in the cornea of the frog
L-selectin on human corneal keratocytes. Invest Rana ridibunda. Experientia. 1978;34:1352–3.
Ophthalmol Vis Sci. 2003;44:4689–92. 46. Muller LJ, Pels L, Vrensen GF. Ultrastructural orga-
31. Perrella G, Brusini P, Spelat R, Hossain P, Hopkinson nization of human corneal nerves. Invest Ophthalmol
A, Dua HS.  Expression of haematopoietic stem cell Vis Sci. 1996;37:476–88.
markers, CD133 and CD34 on human corneal kerato- 47. Muller LJ, Vrensen GF, Pels L, Cardozo BN,

cytes. Br J Ophthalmol. 2007;91:94–9. Willekens B.  Architecture of human corneal nerves.
32. Patel S, McLaren J, Hodge D, Bourne W.  Normal Invest Ophthalmol Vis Sci. 1997;38:985–94.
human keratocyte density and corneal thickness 48. Al-Aqaba MA, Fares U, Suleman H, Lowe J, Dua
measurement by using confocal microscopy in vivo. HS.  Architecture and distribution of human corneal
Invest Ophthalmol Vis Sci. 2001;42:333–9. nerves. Br J Ophthalmol. 2010;94:784–9.
33. Niederer RL, Perumal D, Sherwin T, McGhee
49. Seyed-Razavi Y, Chinnery HR, McMenamin PG.  A
CNJ.  Age-related differences in the normal human novel association between resident tissue macrophages
cornea: a laser scanning in vivo confocal microscopy and nerves in the peripheral stroma of the murine cor-
study. Br J Ophthalmol. 2007;91:1165–9. nea. Invest Ophthalmol Vis Sci. 2014;55:1313–20.
34. Branch MJ, Hashmani K, Dhillon P, Jones DR, Dua 50. Al-Aqaba MA, Alomar T, Lowe J, Dua HS. Corneal
HS, Hopkinson A.  Mesenchymal stem cells in the nerve aberrations in bullous keratopathy. Am J
human corneal limbal stroma. Invest Ophthalmol Vis Ophthalmol. 2011;151:840–9.
Sci. 2012;53:5109–16. 51. Toivanen M, Tervo T, Partanen M, Vannas A,

35. Hashmani K, Branch MJ, Sidney LE, Dhillon
Hervonen A. Histochemical demonstration of adren-
PS, Verma M, McIntosh OD, Hopkinson A, Dua ergic nerves in the stroma of human cornea. Invest
HS.  Characterization of corneal stromal stem cells Ophthalmol Vis Sci. 1987;28:398–400.
with the potential for epithelial transdifferentiation. 52. Dua HS, Faraj LA, Said DG, Gray T, Lowe

Stem Cell Res Ther. 2013;4:75. J.  Human corneal anatomy redefined: a novel pre-­
36. Giraud JP, Pouliquen Y, Offret G, Payrau P. Statistical Descemet’s layer (Dua’s layer). Ophthalmology.
morphometric studies in normal human and rabbit 2013;120:1778–85.
corneal stroma. Exp Eye Res. 1975;21:221–9. 53. Dua HS, Faraj L, Said DG.  Dua’s layer: discovery,
37. Jester JV, Moller-Pedersen T, Huag J, Sax CM, Kays characteristics, clinical applications, controversy
WT, Cavangh HD, Petroll WM, Piatigorsky J.  The and potential relevance to glaucoma. Expert Rev
cellular basis of corneal transparency: evidence for Ophthalmol. 2015a;10:531–47.
‘corneal crystallins. J Cell Sci. 1999;112:613–22. 54. Schlötzer-Schrehardt U, Bachmann BO, Tourtas T,
38. Jester JV.  Corneal crystallins and the develop-
Torricelli AA, Singh A, González S, Mei H, Deng SX,
ment of cellular transparency. Semin Cell Dev Biol. Wilson SE, Kruse FE. Ultrastructure of the posterior
2008;19:82–93. corneal stroma. Ophthalmology. 2015;122:693–9.
39. Said DG, Nubile M, Alomar T, Hopkinson A, Gray T, 55. Mohammed I, Ross AR, Britton JO, Said DG, Dua
Lowe J, Dua HS. Histologic features of transplanted HS.  Elastin content and distribution in endothelial
amniotic membrane: implications for corneal wound keratoplasty tissue determines direction of scrolling.
healing. Ophthalmology. 2009;116:1287–95. Am J Ophthalmol. 2018;194:16.
40. Wilson SE, He YG, Weng J, Li Q, McDowall AW, 56. Dua HS, Mastropasqua L, Faraj L, Nubile M, Elalfy
Vital M, Chwang EL. Epithelial injury induces kerato- MS, Lanzini M, Calienno R, Said DG.  Big bubble
cyte apoptosis: hypothesized role for the interleukin-1 deep anterior lamellar keratoplasty: the collagen
362 H. S. Dua and D. G. Said

layer in the wall of the big bubble is unique. Acta the embryonic period proper. Anat Embryol (Berl).
Ophthalmol. 2015b;93:427–30. 1983;168:87–99.
57. Bizheva K, Haines L, Mason E, MacLellan B, Tan B, 65. Hay ED, Revel JP.  Fine structure of the developing
Hileeto D, Sorbara L. Vivo imaging and morphometry avian cornea. Monogr Dev Biol. 1969;1:1–144.
of the human pre-Descemet’s layer and endothelium 66. Bron A, Tripathi R, Tripathi B.  Wolff’s anatomy of
with ultrahigh-resolution optical coherence tomogra- the eye and orbit. 8th ed. London: Chapman & Hall;
phy. Invest Ophthalmol Vis Sci. 2016;57:2782–7. 1998.
58. Liu Z, Zhang P, Liu C, Zhang W, Hong J, Wang 67. Dodson JW, Hay ED. Secretion of collagenous stroma
W. Split of Descemet's membrane and pre-Descemet's by isolated epithelium grown in vitro. Exp Cell Res.
layer in fungal keratitis: new definition of corneal 1971;65:215–20.
anatomy incorporating new knowledge of fungal 68.
Cai CX, Fitch JM, Svoboda KK, Birk DE,
infection. Histopathology. 2015;66:1046–9. Linsenmayer TF. Cellular invasion and collagen type
59. Dua HS, Said DG.  Clinical evidence of the pre-­
IX in the primary corneal stroma in vitro. Dev Dyn.
Descemets layer (Dua’s layer) in corneal pathology. 1994;201:206–15.
Eye. 2016;30:1144–5. 69. Hayashi M, Ninomiya Y, Hayashi K, Linsenmayer TF,
60. Yahia Chérif H, Gueudry J, Afriat M, Delcampe A, Olsen BR, Trelstad RL. Secretion of collagen types I
Attal P, Gross H, Muraine M. Efficacy and safety of and II by epithelial and endothelial cells in the devel-
pre-Descemet’s membrane sutures for the manage- oping chick cornea demonstrated by in situ hybrid-
ment of acute corneal hydrops in keratoconus. Br J ization and immunohistochemistry. Development.
Ophthalmol. 2015;99:773–7. 1988;103:27–36.
61. Zaki AA, Elalfy MS, Said DG, Dua HS. Deep anterior 70. Linsenmayer TF, Gibney E, Gordon MK, Marchant
lamellar keratoplasty–triple procedure: a useful clini- JK, Hayashi M, Fitch JM. Extracellular matrices of the
cal application of the pre-Descemet’s layer (Dua’s developing chick retina and cornea. Localization of
layer). Eye. 2015;29:323–6. mRNAs for collagen types II and IX by in situ hybrid-
62. Agarwal A, Dua HS, Narang P, Kumar DA, Agarwal ization. Invest Ophthalmol Vis Sci. 1990;31:1271–6.
A, Jacob S, et  al. Pre-Descemet’s endothelial kera- 71. Quantock AJ, Young RD. Development of the corneal
toplasty (PDEK). Br J Ophthalmol. 2014;98: stroma, and the collagen–proteoglycan associations
1181–5. that help define its structure and function. Dev Dyn.
63. Dua HS, Faraj LA, Branch MJ, Yeung AM, Elalfy 2008;237:2607–21.
MS, Said DG, Gray T, Lowe J. The collagen matrix 72. Toole BP, Trelstad RL.  Hyaluronate production and
of the human trabecular meshwork is an extension removal during corneal development in the chick. Dev
of the novel pre-Descemet’s layer (Dua’s layer). Br J Biol. 1971;26:28–35.
Ophthalmol. 2014;98:691–7. 73. Maurice DM. The cornea and sclera. In: Davison H,
64. O’Rahilly R.  The timing and sequence of events in editor. The eye, vol. 1B, Vegetative physiology and
the development of the human eye and ear during biochemistry. 3rd ed. Orlando: Academic press; 1984.
Confocal Microscopy of the Cornea
in a Clinical Model of Corneal
24
Stromal Expansion Using Adipose
Stem Cells and Corneal
Decellularized Laminas in Patients
with Keratoconus

Mona El Zarif, Karim Abdul Jawad,


and Jorge L. Alió

24.1 Introduction loss of keratocyte density [5]. Apoptosis of kera-


tocytes [2, 6] or release of enzymes is thought to
Keratoconus is characterized by a progressive be the cause of keratocyte loss and consequently
decrease in corneal strength [1], related to a pro- loss of corneal stroma over time [2, 6]. The pro-
gressive loss of keratocytes of the human eye, the portion of keratocytes in the corneal stroma is
number of keratocytes decrease from anterior to decreased with the progression of the disease [2].
posterior stroma [2], and the progressive thinning In the end stages of keratoconus, the clinical
of the corneal stroma [2, 3]. This definition is aspects of the thin and debilitated cornea are
valid for most patients with keratoconus, although associated with a sharp decrease in the amount of
some variations in the phenotypic expression of keratocytes. There is severe corneal deformation
the disease might be present [4]. Most of kerato- [2], and alteration in the location of the corneal
conus cases have thin corneas and a weak apex [7] consequently causes severe visual loss.
mechanical resistance related to the progressive The treatment of keratoconus should be
directed towards the recovery of the normal char-
acteristics of the corneal stroma. This can be
achieved by substituting the diseased cornea with
M. El Zarif (*) a healthy cornea, by inserting into the cornea dif-
Optica General Sarl, Department of Optometry and ferent materials to increase resistance such as
contactology, Saida, Lebanon
intracorneal rings, or biological tissues, or by
Vissum Instituto Oftalmologico de Alicante, stimulating cells present in the corneal stroma to
Alicante, Spain
produce collagen. Such recovery of the corneal
K. A. Jawad stroma could be induced by either stimulating the
Optica General Sarl, Department of Optometry and
cells that produce collagen or by increasing the
Contactology, Saida, Lebanon
thickness with the use of corneal stroma or any
University of Nicosia, Department of Life and Health
other tissue that could increase the corneal thick-
Sciences, Nicosia, Cyprus
ness and optical performances while improving
J. L. Alió
the corneal thickness with the production of new
Professor and Chairman of Ophthalmology,
University Miguel Hernandez, Vissum-Instituto collagen, substituting the one lost due to the pro-
Oftalmologico de Alicante, Alicante, Spain gression of the disease.

© Springer Nature Switzerland AG 2019 363


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_24
364 M. El Zarif et al.

Based on the authors’ experience in corneal 24.2.2 Methods


regeneration [8–10], we proposed the use of cor-
neal decellularized lamina with or without ADAS 24.2.2.1 D  escription of the Confocal
cells for the treatment of severe thin corneas suf- Microscope Used: (HRT3
fering from keratoconus [8]. The extremely thin RCM Device
cornea that exists in severe cases could at the from Heidelberg)
same time further increase in thickness with the With confocal microscopy using optical section-
use of a corneal lamina [9], cut by a femtosecond ing, the microstructure of the cornea can be
laser from corneas of eye bank donors, that could observed under better conditions than with previ-
be implemented in its biological value with the ously available devices. For instance, the utiliza-
associated use of the above-described stem cells tion of coherent light has allowed the improvement
with the purpose of increasing the corneal thick- of contrast and the quality of the images of the
ness even further [8]. Along this chapter, we are stroma that is observed under the confocal micro-
going to describe the evolution of confocal scope [2].
microscopy following a clinical experiment, in Considered as a noninvasive technique, confo-
which autologous stem cells have been injected cal microscopy allows in vivo observation of the
in the corneal stroma [8]. Corneal human donors microstructure of the cornea and a more accurate
unsuitable for corneal graft with or without the quantification of keratocyte density (Fig.  24.1
impregnation of autologous stem cells, for the and Table 24.1).
purpose of treating of patients with advanced
stages of keratoconus [9, 10]. 24.2.2.2 P  reparation of the Confocal
We are going to report herein the results of Microscope
this experiment with the use of confocal micros- While the device is turned off, the Rostock
copy as an investigational tool to ascertain the Cornea Module (RCM) objective must be set at
fate of the stem cells used along the experiment +12 D (Fig. 24.2). Then, a drop of contact gel is
and the impact that this cellular therapy might applied on the front surface of the microscope’s
have in the future treatment of keratoconus [8]. lens of the RCM. The gel must not contain air
bubbles (Fig. 24.3). The TomoCap is then placed
on top of the RCM objective (Fig.  24.4). The
24.2 Patients and Methods camera is placed on the side of the examined
eye [11].
24.2.1 Patients After getting the consent of the patient, mak-
ing sure there is no allergy towards the gel and
Fourteen consecutive keratoconus patients were anesthetic applied, and informing how the exami-
enrolled in the study and randomly distributed nation will be performed, the patient is prepared
into three study groups: by applying a drop of topical anesthetic in the eye
to be examined. Then, a gel tear substitute is
Group 1: Autologous ADASC implantation (5 applied in both eyes. The gel is also applied on
patients) the outer surface of the TomoCap in order to
Group 2: Decellularized human corneal stroma improve the contact surface between the cap and
transplantation (5 patients) the eye of the patient. Heidelberg Engineering
Group 3: Autologous ADASC + decellularized recommends using the same high viscosity gel
human corneal stroma transplantation (4 for both applications on the eye of the patient and
patients) the TomoCap in order to prevent any confusion
Thirteen patients completed the study, only one between the gels, since two different gels can be
abandoned it due to independent personal used in each application. We recommend pre-
reasons. serving the gel at 4 °C for a period of time before
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 365

Fig. 24.1  The HRT3 1


RCM device from
Heidelberg [11] 7
2 8
3

4
9

10
11
5

6 12

Table 24.1  Illustration of the confocal microscope with


detailed description of its different parts [11]
1 Forehead rest 7 CCD camera
2 Headrest column with red 8 RCM objective
marks
3 External fixation light 9 Chin rest
4 Camera head 10 Adjustment screw
for camera position
5 Adjustment screw for 11 Adjustment screw
vertical and horizontal for chin rest
camera position
6 CCD camera cable 12 RCM objective
cable

its use and maintaining the environment of the


examination room at a relatively low
temperature.
Before proceeding with the examination, the
patient must remain stable, comfortable, and
aligned with the lens of the confocal microscope. Fig. 24.2  RCM objective set at +12 D [11]
Afterwards, the patient is instructed not to move
the eyes, to keep on looking at the light inside the TomoCap, and then the focal plane position is set
lens, not to blink, and not to look outside the lens to “0” by clicking on “reset” the focal section
in order not to hinder the observations made with (Fig. 24.5). Resetting the focal position to “0” at
the microscope [11]. the superficial epithelial cells of the examined
Afterwards, the CCD camera is adjusted per- eye is crucial to measure the depth of the corneal
pendicularly to the optical axis of the scanning cell layer relative to the corneal surface.
laser camera. The RCM objective and the focal Following this, the scanning laser camera is
plane are adjusted to the ocular surface of the slowly moved towards the patient’s eye
366 M. El Zarif et al.

Fig. 24.3  Front surface of the contact lens of the micro-


scope where the gel is applied [11] 4

Fig. 24.5  Image seen after adjustment of the focal plane


to the ocular surface of the TomoCap [11]

Fig. 24.4  TomoCap attached to the objective [11]

(Fig. 24.6). It is advised to apply gel on the exte- 1


rior face of the TomoCap to improve the contact
between the cap and the patient’s cornea. At this Fig. 24.6 Reflection of the laser from the confocal
point, the patient is asked to widely open the eye, microscope in the center of the pupil indicating the right
and the scanning camera is slowly moved towards alignment between the microscope and the eye of the
patient [11]
the eye until the TomoCap touches the patient’s
cornea. The RCM is turned clockwise or counter-
clockwise, and the focal plane is adjusted to the
desired cell layer once a contact between the 24.2.2.3 D  ifficulties in Handling
patient’s cornea and the TomoCap has been the Confocal Microscope
made. The headrest must not be moved transver- The use of the confocal microscope for the study
sally. Finally, an image is obtained by pressing of the cornea is always somehow cumbersome
the foot switch [11]. for both the patient and the examiner.
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 367

The patient must be able to collaborate with Unlike in normal epithelial tissues, it can be
the examiner in maintaining a stable posture, a ­difficult to distinguish the borders between the
good fixation, and a wide palpebral opening of epithelial cells, and the basal epithelium shows
the eye in order to allow contact between the distortions and irregularities (Figs.  24.9, 24.10,
TomoCap and the eye [11]. and 24.11) [13].
The feeling of contact of the TomoCap with
the eye of the patient may cause tearing and pho-
tophobia in some patients, which increases the
difficulty of the examination.

24.2.2.4 Confocal Sampling


of the Corneal Stromal
in Advanced Keratoconus
In confocal microscopic investigations of kera-
toconic patients, superficial epithelial cells that
are elongated in oblique direction, are consis-
tently found alongside highly reflective
(Fig. 24.7), pathological alterations are found in
principle at the fourth stage of the disease in the
central area of the cornea, as reported by various
authors [12].
In our experience, we also found (Figs.  24.7
and 24.8) that in the epithelial layer, superficial
cells are elongated and distorted in an oblique Fig. 24.8  Superficial epithelium cells at −1 μm from a
34-year-old keratoconic female patient show different
direction instead of them being presented as a sizes of superficial epithelial cells and an elongated one in
mosaic form of regular polygonal surface [13]. oblique direction

Fig. 24.7  Superficial epithelium cells at −3 μm from a Fig. 24.9  Corneal epithelium at 32  μm from a kerato-
34-year-old keratoconic female patient. It shows elon- conic female patient
gated cells in oblique direction, highly reflective with evi-
dent area of missing cells
368 M. El Zarif et al.

Fig. 24.10  Corneal epithelium at 50  μm from a kerato- Fig. 24.12  Bowman’s membrane of a 34-year-old female
conic female patient showing the nerve plexus, irregular with advanced keratoconus shows a visible dark striae at
and distorted epithelial cells and small area of missing cells 61 μm

Fig. 24.11  Basal epithelium cells at 61  μm from a Fig. 24.13  Basal epithelium, Bowman’s membrane, and
32-year-old keratoconic female patient. It shows highly subepithelial plexus nerve of a 31-year-old male patient;
reflective nodules related to scar tissue at the level of the fibers are disposed in oblique direction and display
Bowman’s membrane, with areas of missing cells fine irregularity

Extending from the basal membrane to the (Fig. 24.12) has also been described in other stud-
immediate stroma beneath it, ruptures in ies that found that there is a reflection of bright
Bowman’s membrane are represented by a highly background illumination underneath Bowman’s
reflective scar tissue [13]. Presence of striae membrane due to a disarrangement of collagen
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 369

Fig. 24.14 Preoperative corneal anterior stroma at Fig. 24.15  Deep corneal stroma at 396  μm from a
158 μm of a 35-year-old keratoconic female patient show- 50-year-old keratoconic female patient showing oblique,
ing oblique dark striae with a moderate number of reflec- vertical, and horizontal dark striae with a moderate num-
tive keratocytes ber of hyperreflective keratocytes

fibers [12]. Sometimes, the thin nerve fibers pres-


ent in the area of corneal protrusion, disposed in
an oblique horizontal pattern that is coherent with
the direction of the stretched epithelial cells
(Fig. 24.13) [13]. In addition, an ­elevation in the
anterior stroma and Vogt’s striae in the posterior
stroma’s reflectivity are consistent with pathologi-
cal processes of the corneal stroma [14]. One of
the main abnormalities of the stroma is the pres-
ence of microstriae that are evident in the whole
stroma of the cornea, which also appear as multi-
ple dark and thin lines in comparison with the
brighter stromal reflectivity (Fig.  24.14).
These microstriae could possibly result from the
degenerative process that leads to changes in the
extracellular collagen lamellae of the cornea [13].
Fig. 24.16  Anterior corneal stroma at 63  μm from a
These microstriae can appear as horizontal, verti-
33-year-old keratoconic female patient, with a reduced
cal, oblique, or reticular lines (Fig.  24.15), and number of keratocytes and thin nuclei content
they appear to extend from the most anterior
stroma to Descemet’s membrane [13].
Morphology of keratocytes changes from the shape ( Fig.  24.17), and more consistent and
anterior to the posterior stroma. They have thin bigger nuclei size in the mid of stroma
­
and consistent nuclei in the anterior stroma (Fig. 24.18). The morphology of the keratocytes
(Fig. 24.16), some stromal cells have a dendritic in the posterior stroma is similar to the one in the
370 M. El Zarif et al.

Fig. 24.19  Deep corneal stroma at 439  μm from a


Fig. 24.17  Preoperative mid-stroma at 217  μm from a
33-year-old keratoconic female patient, with a significant
31-year-old keratoconic male patient. It shows  kerato-
decrease in the number of keratocytes, and less reflective
cytes in dendritic form
and dense nuclei

Fig. 24.18  Mid-stroma at 220  μm from a 33-year-old


keratoconic female patient, with a reduced number of Fig. 24.20  Descemet’s membrane from a 31-year-old
keratocytes with more dense nuclei keratoconic male patient, showing an acellular layer with
a highly reflective zone

middle stroma, but they are smaller and less cell layer might show changes such as an increase
bright (Fig. 24.19). in cell area, a decrease in cell density, polymega-
Finally, the descemet membrane appears com- thism, and pleomorphism depending on the pro-
pletely acellular (Fig. 24.20) and the endothelial gression of the disease (Fig. 24.21) [13].
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 371

24.2.2.5 Methods of Cell Counting Method for Manual Cell Counting


of Keratocytes
Automated Cell Counting of Keratocytes First, an image of the corneal stroma of a patient
The automatic count that is calculated by confo- is selected. The image has to have a good quality
cal corneal microscopy devices usually is not of contrast and illumination, and must ­contain
precise, which leads to an increase in errors when keratocytes. An area known as the region of inter-
calculating the number of cells. Therefore, for the est (ROI) is determined. This region has plenty of
purpose of this investigation, we preferred to per- cells. It is advised to choose the same approxima-
form a manual cell count of the keratocytes of the tion of ROI (e.g., 0.1000 mm2) for all cell counts
corneal stroma [11]. in all layers of the corneal stroma [2, 11].
The cell count begins with mid brightness and
contrast. At this point, the cells that are more illu-
minated and more refringent [5, 6] are selected
with the marker of the device, which usually is
either white or light gray and have clear borders
[2]. The dark gray cells should not be taken into
consideration, because these cells do not belong
to the chosen plane. This high brightness is usu-
ally attributed to the nuclear content in these
cells, and they have irregular oval-shaped bodies
(Fig.  24.22) [5], while the difference in bright-
ness is due to the metabolic activation and the
direction of incidence of the laser light [5].
After the first count is done, the contrast is
increased to the maximum, and the brightness is
decreased to the minimum. Keratocytes that dis-
Fig. 24.21  Endothelium at 531  μm from a 24-year-old appear from the chosen plane are eliminated, and
keratoconic male patient with mild pleomorphism only the ones that remain are used for the final

Fig. 24.22  Cell count performed with mid brightness and contrast; the keratocytes are marked in blue
372 M. El Zarif et al.

Fig. 24.23  Elimination of keratocytes that do not belong to the plane under observation using low brightness and high
contrast settings

Fig. 24.24  Refining of the cell count using three-quarter high brightness and a quarter low contrast at which the image
appears in grayscale

count of cells even if they are faintly identified in become highly illuminated, which allows a final
the image (Fig. 24.23). refinement of the cell count.
Finally, the contrast is decreased to a quarter, When the images are very dark, it is recom-
and the brightness is increased to three quarters mended to increase the contrast to the maxi-
(Fig.  24.24). In this newly adjusted image, the mum and the brightness to three quarters
plane becomes grayscale, and the keratocytes (Fig.  24.25). This facilitates the count of the
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 373

Fig. 24.25  Maximum contrast and three-quarter brightness is used in the case of very dark images

Fig. 24.26 Count of ADAS cells and keratocytes, Fig. 24.27  Count of cells for the same patient 1  year
1  month following surgery of a 33-year-old keratoconic after surgery
patient. The cells are marked in blue
in the first 6 months after the surgery, the ADAS
keratocytes, and the cell count can be resumed cells have a round shape and are more luminous
as mentioned above. and refringent than normal keratocytes
(Fig. 24.26) [8]. After these 6 months, the ADAS
Cell Counting of ADAS Cells cells assimilate a shape similar to normal kera-
The counting method for transplanted ADAS tocytes [8], thus  reducing any morphological
cells is done similarly to the way normal kerato- differences between them (Fig.  24.27).
cytes of the corneal stroma are counted. The Therefore, the cell count becomes the same for
only difference between the two methods is that both kinds of cells.
374 M. El Zarif et al.

Cell Count on a Decellularized Lamina ­transplanted (Fig.  24.  29). Thus, the images of
The images provided by the confocal micro- these cells lack the depth seen in the normal cor-
scope shows that the lamina appears totally acel- neal keratocytes of the patient. Three months
lular in the first month (Fig. 24.28) but the newly after the surgery, the morphology of patients’
colonized transplanted lamina by the keratocytes cells colonized on the lamina started to show
of the patient’s corneal stroma show that these some  new keratocyte appearance (Fig.  24.29).
cells differ in morphology from the native kerato- The shape of these cells at 6 months became
cytes of the cornea. This is caused by the reduced more similar to the keratocytes (Fig. 24.30) [9],
transparency of the lamina right after it is before fully developing into cells that have the
same morphology of keratocytes 1 year post-
operation (Fig.  24.31) [10]. Therefore, it is

Fig. 24.28  Anterior aspect of a decellularized lamina of


a female, which appears to be without cells 1 month
Fig. 24.30  Cell count in the anterior aspect of the lamina
post-operation
in a 29-year-old keratoconic patient, 6  months
post-operation

Fig. 24.29 Cell count of a 24-year-old keratoconic


patient 3 months post-operation on the posterior aspect of Fig. 24.31  Cell count on the anterior aspect of the lam-
the lamina ina of a 31-year-old patient, 1 year after the surgery
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 375

important to identify the change in morphology point, it is advisable to lower the illumination
of these cells throughout the months, which helps to 0% and increase the contrast to 100% as
in achieving a more accurate cell count. indicated in (Fig. 24.23).
• If the images have low resolution, the contrast
Cell Count on a Recellularized Lamina must be modified towards 25%, and the illumi-
The cell count on a recellularized lamina is rela- nation is raised to approximately 75%. By
tively similar to the one made on a decellularized doing so, this image is seen in the clear back-
one. In this case, a few ADAS cells can be seen ground, and the keratocytes appear with distin-
within the lamina at 1 month after the surgery in guished border and are clearly illuminated.
two of the four patients. In addition, the recellular- Therefore, they can be marked more easily. We
ized lamina consistently contained a higher num- call this regulation the grayscale (Fig. 24.24).
ber of keratocytes than in the decellularized one, • If the confocal images are very dark, it is
when observed 3–12 months after the surgery [10]. advisable to set the illumination to 75% and
the contrast to 100% (Fig. 24.25).
24.2.2.6 Corneal Density Calculation • If a line of separation between both structures
To obtain cell density, the examiner must use the of attached cells is not clear, it is advisable to
cellular counting software in the confocal micro- interchange between the contrast and illumi-
scope [11] and delimit a fixed area depending on nation until a discrimination of the orientation
the images measured by the specialist, with a of the nuclear content of both cells can be
periphery that contains abundant keratocytes. made. If the content appears parallel, it is
Since it is difficult to achieve great corneal con- more likely that they belong to the same
tact between the TomoCap and the patient, the nucleus. However, if the nuclear contents of
specialist is required to select an area that con- the two cells are in different directions, then
tains abundant cells. they belong to two different cells.
The following data appear on the software • If the cells at the edge of the selected plane are
screen: not seen in their complete shape, every two
halves of cells are counted as one single cell.
• Number of cells: Which are the number of
cells counted by the observer When observing the transplanted laminas, it
• ROI (mm2): The area chosen by the same was noticed that their morphology is either decel-
observer lularized or recellularized; has an extracellular
matrix characterized by the presence of abundant
Density +/− std (cell in mm2): Density calcu- streaks, which are larger in size and are more pro-
lated by mm2 by the same device with an error of found; and has changed during the next months
more or less the number of cells in the area [6, 11]. after the transplant. This changed morphology
For example: remained constant afterwards.

Density ( cells in mm 2 ) = 200 ± 12



24.3 Results

24.2.2.7 Difficulties in Cell Counting 24.3.1 Confocal Microscopy


During our experience in counting keratocytes, of the Corneal Stroma
we encountered difficulties until we established Following the Stromal
specific criteria and methodology for the purpose Enhancement Intervention
of cell counting. We recommend the following:
24.3.1.1 Patients with  Transplanted
• The cells that appear in the captured images as ADAS
if they belonged to the same plane of 50 One to three months following the implantation
micron of depth cannot be counted. These surgery, the transplanted ADAS cells were round in
cells appear with a faded gray color. At this shape and more refringent [8], in comparison to the
376 M. El Zarif et al.

Fig. 24.32  Confocal microscopy of corneal stroma from Fig. 24.34  Morphological change in the shape of ADAS
a 33-year-old keratoconic female patient at 265  μm cells, from round to fusiform in the same 273  μm, of a
1 month after transplantation of ADAS cells 33-year-old female patient 6 months after the operation

Fig. 24.35  Keratocytes in the surgical plane at 250 μm,


Fig. 24.33  Confocal microscopy at 273  μm from a 1 year after the transplant
33-year-old keratoconic female patient 3  months after
transplantation of ADAS cells. The blue dots show the
transplanted cells in the plane of the lamellar pocket. They of the original keratocytes of the patient’s cornea
have a round morphology and are more refringent than the (Fig. 24.34). The round morphology of these cells
original keratocytes of the patient’s stroma
is a strong indicator of their ability to survive, espe-
cially during the early postoperative period [8]. We
fusiform morphology of the keratocytes [5], pres- also found that the number of keratocytes increased
ent in the recipient cornea (Figs. 24.32 and 24.33). 6 months after the operation [9]. One year after the
However, the shape of ADAS cells changed from surgery, it is noticeable that the number of kerato-
round to fusiform 6 months after the operation [8], cytes increased even more (Fig.  24.35) [10].
and they assimilated a morphology similar to that A gradual increase (P value = 0.07) in the density
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 377

Table 24.2  Evolution of the density of keratocytes of the patients over 1 year
Preop 6 months 12 months
Group
Group 1 Group 2 Group 3 Group 1 Group 2 Group 3 1 Group 2 Group 3
Anterior stroma 236.45 222.6 187.6 254.15 215.3 286.6 288.5 441.2a 467.4
(cells/mm2) (220, (163.3, (176.6, (210, (190, (210, (210, (392.5, (333.3,
280) 286.6) 198) 306.6) 273.3) 356.6) 340) 474) 594)
Posterior stroma 202.5 181.8 192.8 255.8 239.16 237.4 322.5 459.7a 416.5
(cells/mm2) (190, (136.6, (174, (190, (213.3, (163.3, (290, (376.6, (388,
225) 216) 203.3) 320) 265) 293.3) 340) 540) 447.5)
Keratocyte cellular density (mean value and range) for G 1 (autologous ADASC implantation; n:4), G 2 (decellularized
human corneal stroma implantation; n: 5) and G 3 (autologous ADASC recellularized human corneal stroma implanta-
tion; n: 4) before, 6 and 12 months after surgery [10]
a
Shows statistically significant (P ≤ 0.05) differences between the preoperative and 12 months postoperative values for
each parameter and for each study group separately (for mid-stroma, this comparison is between 6 and 12  months
post-operation)

Fig. 24.36  Posterior face of a decellularized lamina of a


Fig. 24.37  Anterior face of the decellularized lamina of
31-year-old male, 1  month after the operation, seen at
the same patient, 3  months after the operation, seen at
289 μm. The lamina appears to be almost acellular
207 μm, showing early colonization of the keratocytes to
the lamina
of cells was evident in the anterior, middle, and
posterior stroma of the patients 1 year after the the surgery (Fig.  24.37). On the other hand, the
operation, in comparison to the density level preop- posterior aspect of the lamina appeared to be
erative (Table 24.2). clearly colonized by some keratocytes at the same
time of examination in three of the five patients
24.3.1.2 P  atients with Decellularized (Fig. 24.38). Six months after the transplant, the
Lamina anterior aspect of the lamina was clearly colo-
One month after the surgery, the transplanted lam- nized by keratocytes in some patients (Fig. 24.39),
ina generally remained acellular in most of the while the posterior aspect was more abundant in
patients (Fig. 24.36) [9]. However, two out of five these cells (Fig. 24.40). Moreover, 1 year after the
patients showed in the anterior aspect of the lam- operation, both the anterior and posterior aspects
ina the beginning of keratocyte colonization, became fully colonized by keratocytes (Figs. 24.41
which is a change normally seen 3 months after and 24.42) [10]. The density of cells on the lamina
378 M. El Zarif et al.

Fig. 24.38  Posterior face of the lamina at 271  μm, Fig. 24.40  Posterior face of the lamina that is highly
3 months after the operation clearly colonized by the sur- colonized by keratocytes from the cornea 6 months after
rounding keratocytes the operation

Fig. 24.39  Colonized anterior face of the lamina at Fig. 24.41  Anterior face of the lamina showing high
176 μm at 6 months quantity of keratocytes 12 months after the surgery

was similar to what is expected in a normal cor- of the patient’s stroma. It shows clear and abun-
nea. We were able to demonstrate that there was a dant dark striae, some of which have consider-
significant increase in keratocytes within the lam- able thickness and have vertical, horizontal, and
ina and keratocyte density in both the anterior and oblique directions (Figs. 24.43 and 24.44).
posterior stroma was statistically significant (P We observed at the periphery and border of
value  =  0.008) 1 year after the operations were the lamina highly reflective structures that are
performed (Table 24.2). caused by the mechanical trauma during the
The collagen matrix morphology of the decel- surgery (Figs. 24.45 and 24.46). The transition
lularized lamina differs from the collagen matrix zone between the lamina and the corneal stroma
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 379

Fig. 24.42  Posterior face of the lamina colonized by


high numbers of keratocytes 12  months after the
operation Fig. 24.44  Anterior face of the decellularized lamina of
a 35-year-old keratoconic female patient showing dark
and abundant striae. The matrix is clearer than the normal
stroma and more reflective

Fig. 24.43  Anterior face of the decellularized lamina of


a 31-year-old male patient, 3 months postoperatively, seen
at 154 μm, showing striae that are darker and larger than
the ones present in a normal stroma
Fig. 24.45  Anterior face of the decellularized lamina of
a 34-year-old female, 1  month post-operation, shows
highly reflective peripheral structures on the lamina due to
can be  easily differentiated (Fig.  24.47); also,
mechanical trauma
the ­lamina’s border has, in addition, a reflective
structure (Fig.  24.47). We observe as well a
clear keratocyte migration from the corneal Twelve months after the operation, both anat-
stroma to the decellularized lamina (Fig. 24.48). omy of the corneal anterior and posterior stroma
Also, the lamina’s light reflectivity is higher and the keratocyte density did not differ in patients
than the matrix’ corneal stroma (Figs.  24.44 to the normal corneal stroma at the confocal
and 24.48). microscopy level (Figs. 24.49 and 24.50) [10].
380 M. El Zarif et al.

Fig. 24.46  Decellularized lamina present at 272 μm, of a


31-year-old male, 3 months after surgery, showing reflec- Fig. 24.48  Transition zone at 306 μm between the poste-
tive peripheral structures, due to the mechanical surgical rior face of the decellularized lamina and the recipient
trauma, needle-shaped bodies, and some keratocytes with stroma with a number of migrating keratocytes with den-
morphological change dritical shapes and the corneal stroma with a number of
keratocytes, 12  months post-operation of a 35-year-old
female patient

Fig. 24.47  Transition zone at 190 μm between the border


of the anterior face of the decellularized lamina and the
Fig. 24.49  Anterior stroma at 103 μm of a 24-year-old
corneal stroma, 3  months post-operation, for a 35-year-­
patient, 12 months after the surgery
old female showing highly reflective structures, with a
few dendritical keratocytes

a few number of ADAS on their surface, which


24.3.1.3 P atients with Recellularized are smaller than normal keratocytes in the ante-
Lamina rior face of the lamina (Figs.  24.51 and 24.52),
One month after the surgery, both the anterior but they are more similar to them in the posterior
and posterior faces of the patients’ laminas face, unlike the decellularized laminas, which
showed in two of the four patients the presence of were completely acellular. At 3  months, the
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 381

Fig. 24.50  Posterior stroma at 410  μm of the same Fig. 24.52  Aspect of the posterior face of the lamina at
24-year-old patient, 12 months after the surgery 275 μm, 1 month post-operation

Fig. 24.53  Anterior face of the recellularized lamin at


Fig. 24.51  Anterior face of a recellularized lamina at 195 μm, 6 months after the surgery
174  μm of a 50-year-old patient, 1  month after the
operation

n­ umber of cells increased slightly, and all the keratocytes [10], similar to that in a normal cor-
patients showed a few number of cells similar to nea (Figs. 24.57 and 24.58). Results derived from
keratocytes, some of them presenting with den- the anterior stroma (P  =  0.008) and posterior
dritical shapes. Six and twelve months after the stroma (P  =  0.008) showed a statistically
­operation, the number of cells that colonized the ­significant increase in the keratocytes’ density of
laminas was higher in the recellularized laminas the patients from the G2  +  G3 (Table  24.2). In
than in the decellularized ones (Figs.  24.53, addition, we could demonstrate that there was a
24.54, 24.55, and 24.56) [9, 10]. significant increase in keratocytes within the
The histology of the anterior and posterior recellularized lamina more higher in the group 3
corneal stroma showed a very high number of than the group 2.
382 M. El Zarif et al.

Fig. 24.54  Posterior face of the lamina 6 months post- Fig. 24.56  Posterior face of the lamina at 299  μm,
operation with a high number of keratocytes 12 months post-operation. It shows a very high number of
keratocytes

Fig. 24.55  Anterior face of the recellularized lamina at Fig. 24.57  Anterior corneal stroma with abundant kera-
215 μm 12 months after the surgery tocytes in a 33-year-old female patient, 12  months after
the operation, seen at 106 μm

24.4 Discussion density of corneal keratocytes. This favored the


production of new collagen, which resulted in an
These sets of experiments represent the first improved corneal thickness [9]. In  addition, this
attempt at using stem cell therapy for human study showed that the implantation of a decellular-
­corneal stroma in patients affected by advanced ized lamina, either colonized or not by ADASC,
keratoconus. also produced changes in the corneal topography,
We have previously reported that after the thickness, and corneal r­ egularity that may be con-
injection of autologous stem cells (ADASC) in the sidered as therapeutic with a potential application
stroma of the cornea, there was an increase in the to the treatment of keratoconus [9, 10].
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 383

lamina were also observed. The decellularized


lamina was colonized by the patient’s own kera-
tocytes from the first month. After 12 months, the
lamina was completely recellularized.
Confocal microscopy of the cornea during the
study has been able to show us from the first post-
operative month the following findings: accumu-
lations of the mesenchymal cells implanted on
the lamina, cell morphological changes have
been clearly seen first on the posterior face fol-
lowed by the anterior face, and the number of
cells has increased statistically and significantly
between immediately after the surgery and a
12-month period.
Confocal microscopy is considered a noninva-
sive technique [6]. But difficulties can arise when
Fig. 24.58  Posterior corneal stroma at 507 μm with high
patients are being examined, especially when the
number of keratocytes for the same 33-year-old patient contact between the TomoCap and the cornea of
the patient must not cause any harm for the
patient. The fragility of the operated cornea is
In this innovative study, we have reported also a concern when examining a patient with the
the  outcomes of the use of corneal confocal confocal microscope. Before performing the test
microscopy for the observation of ADASC in the with the confocal microscope, it is advisable to
selected surgical plane in vivo, which allowed a introduce a silicone hydrogel disposable contact
qualitative and quantitative assessment of them lens, which improves the contact between the eye
along the experience. Moreover, it allowed moni- and the TomoCap. This lens prevents the gel from
toring of the progressive morphological changes leaking. In the absence of this lens, the gel dries
that occurred in the decellularized and recellular- out with time or dissipates outside the eye. In
ized laminas throughout the observation and such case, the examination is interrupted, and
assisted in determining the change in the cell additional gel must be reapplied in the eye of the
densities in the grafted tissue as well as in all the patient. Also, the addition of this lens reduces
corneal stroma. corneal sensitivity, which allows the patient to
Confocal corneal microscopy has shown to be keep the eye open for a longer time, which ulti-
an essential tool in the assessment and “in vivo” mately facilitates the examination process. In
follow-up of the corneas implanted with mesen- addition, sterilization must be maintained
chymal cells for corneal regeneration purposes. throughout the process in order to prevent any
Corneal bioconfocal microscopy demonstrates contamination in the patient’s eye, to reduce error
an increase in the cellularity of the anterior and that may hinder the results obtained from the
posterior stroma following the implantation of observation, and to preserve the quality of the
ADAS cells into the corneal pockets created images obtained by the confocal microscope
using a femtosecond laser. (Figs. 24.59 and 24.60).
Relevant changes in the morphology were One of the limitations of this study is the
observed in the implanted cells from the moment method used for counting cells as it was
of their implantation into the corneal pocket, ­performed with a manual method. Although it
showing changes from clusters around the was performed by two independent observers in
­individual cells to their development into adult all cases following the same guidelines, this
keratocytes. These findings occurred in the ante- method may introduce a certain amount of sub-
rior, mid, and posterior stroma of the implanted jectivity and interpretation bias. However, this
corneas. Moreover, changes inside the acellular method yields at present moment more accurate
384 M. El Zarif et al.

Fig. 24.59 Anterior face of decellularized lamina at


155 μm in a 33-year-old patient, 12 months after the sur- Fig. 24.61  Confocal microscopy automatic cell counting
gery; image taken with silicone hydrogel contact lenses of corneal stroma, from a 33-year-old female with kerato-
conus at 151 μm seen 12 months after transplantation of
ADAS cells

Fig. 24.60  Posterior face of decellularized lamina at Fig. 24.62  Anterior face of the recellularized lamina at
280 μm in a 33-year-old patient, seen 12 months after the 195 μm of a 35-year-old female, 6 months after the sur-
surgery; image taken with silicone hydrogel contact lenses gery, demonstrates automatic cell counts with high and
abnormal number of cells

results than an automated cell counting micro- should have been counted within the corneal
scope which are affected by a large variability stroma were not accounted for by the device
and misinterpretations. We performed a refined (Fig.  24.61) on both sides of the transplanted
automatic cell count, and we observed a bias lamina (Fig.  24.62). An artifacted result on the
towards high and abnormal number of cells that lamina’s border in the transition zone between
were counted for the normal structure of the cor- the lamina and stroma was seen with this device
nea, and a considerable number of cells that as well (Fig. 24.63).
24  Confocal Microscopy of the Cornea in a Clinical Model of Corneal Stromal Expansion Using Adipose… 385

d­ ecellularized laminas should be made, along-


side the growth of keratocytes in the corneal
stroma in both types of transplants, and relate it
to the improvement of the visual acuity of the
operated patients.
Furthermore, comparisons among different
morphologies of keratocytes should be estab-
lished, comparing non-dendritic cells, with oth-
ers that have a high dendritic shape and investigate
the biological implications that such morphologi-
cal changes may have.
A particular interest has been for the observa-
tion of hyper-reflective area, acellular or with very
few cells within the corneal stroma studied by the
confocal microscope. Further investigation should
verify if this fact is due to secondary reactions
caused by mechanical manipulation after surgery,
or to the formation of new collagen related to the
Fig. 24.63  Transition zone at 154 μm between the border ADAS cells implanted in the surgical plan.
of the anterior face of the decellularized lamina and the Comparisons of studies between different age,
corneal stroma, 6  months post-operation, for a 31-year-­
old male, showing highly reflective structures, the blue sex, and ethnic groups can yield many explana-
dots are generated by automatic count cells tions about prevalence and frequencies in the
population. We believe that it is necessary to
establish a more objective method of cell count-
24.5 Future Perspectives ing based on image processing programs, where
one can edit the contrast and illumination and use
In our study, confocal microscopy was very different specialized filters to optimize the visu-
­useful for keratocyte cell counting and determin- alization of keratocyte cells and to better calcu-
ing the evolution of the implanted cells in terms late their size.
of morphology and densities along the follow-up In addition, the use of digital cameras captur-
time of the experience. ing at high frames per second and high resolution
Although only 14 patients participated in this in the future can reduce the examination time and
study, sufficient statistical significance levels of provide a more reliable measurement of the
confidence were obtained that allow us to ­confirm thickness evolution of the cornea and the newly
differences among the clinical groups, which implanted corneal layers, with fewer errors
demonstrates the importance of the study estab- caused by the movement of the patient’s eye.
lished by the confocal microscope. However, it is Also, the development of a new TomoCap that
necessary to increase in future experiences the has a biconcave or planoconcave silicone hydro-
clinical number of participants in order to gel lens fixated to its top, with the lens made of a
strengthen the statistical significances found in specific thickness containing a viscous gel, may
the present study. shorten and provide more comfort in the confocal
Further research should be carried out to exam of the implanted patients. Also, it would be
answer questions like if all morphological very interesting to have a new  RCM objective
changes begin in the posterior stroma and then in that can be applied on the eye, while the patient is
the anterior, just as it was observed in our study, lying down on his back. This would restrict the
which seems to be a specific finding of this patient’s body and head movements. In this case,
investigation. a better fixation of the laser that is projected on
Comparisons between the recellularized the cornea can be made, which can provide better
laminas containing Bowman’s membrane and images of the different layers of the cornea and
­
386 M. El Zarif et al.

less error while measuring the pachymetry of the 5. Mastropasqua L, Nubile M.  Normal corneal mor-
phology. In: Confocal microscopy of the cornea.
cornea. This innovative biomicroscope objective Thorofare: SLACK; 2002.
would allow more comfortable and probably pre- 6. Javadi M, Kanavi M, Mahdavi M, Yaseri M, Rabiei
cise study of the confocal cornea microscopy H, Javadi A, et al. Comparison of keratocyte density
evolution of the experience, as it may allow to between keratoconus, post-laser in situ keratomileu-
sis keratectasia, and uncomplicated post-laser in situ
study the collagen fibers the implanted laminas keratomileusis cases. A Confocal Scan Study. Cornea.
as they are added to the collagen matrix structure 2009;28(7):774–9.
of the patient’s cornea. 7. Edmund C.  Assessment of an elastic model in the
pathogenesis of keratoconus. Acta Ophthalmol.
2009;65(5):545–50.
Acknowledgments The authors would like to thank
8. Alió del Barrio J, El Zarif M, de Miguel M, Azaar
Jorge L.  Alió del Barrio, MD, PhD; Ziad Abdul Jawad,
A, Makdissy N, Harb W, et al. Cellular therapy with
Od; Peggy Saba, Od; María P. de Miguel, PhD; Nehman
human autologous adipose-derived adult stem cells for
Makdissy, PhD; Francisco Arnalich, MD, PhD; and
advanced keratoconus. Cornea. 2017;36(8):952–60.
Ibrahim Achkar, MD, for their essential contribution to
9. Alió del Barrio J, El Zarif M, Azaar A, Makdissy N,
the clinical and cell biology part of this study.
Khalil C, Harb W, et al. Corneal stroma enhancement
with decellularized stromal laminas with or without
stem cell recellularization for advanced keratoconus.
A J Ophthal. 2018;186:47–58.
References 10. Alió J, Alió del Barrio J, El Zarif M, Azaar A,

Makdissy N, Khalil C, et  al. Regenerative medicine
1. Piñero D, Alio J, Barraquer R, Michael R, Jiménez of the corneal stroma for advanced keratoconus: one
R.  Corneal biomechanics, refraction, and cor- year outcomes. (Not published).
neal aberrometry in keratoconus: an integrated 11.
Guthoff R.  Rostock cornea modul. Klinische
study. Invest Opthalmol Visual Sci. 2010;51(4): Monatsblätter für Augenheilkunde. 2005;222(S 3).
1948. 12. Wygledowska-Promienska D, Rokita-Wala I, Gierek-­
2. Ku J, Niederer R, Patel D, Sherwin T, McGhee Ciaciura S, Piatek-Koronowska G. The alterations in
C.  Laser scanning in  vivo confocal analysis of ker- corneal structure at III/IV stage of keratoconus by
atocyte density in keratoconus. Ophthalmology. means of confocal microscopy and ultrasound biomi-
2008;115(5):845–50. croscopy before penetrating keratoplasty. Klin Ocz.
3. Alió J. Keratoconus: recent advances in diagnosis and 1999;101(6):427–32.
treatment. In: Keratoconus pathology to pathogenesis: 13. Mastropasqua L, Nubile M. Confocal microscopy in
Springer; 2017. keratoconus. In: Confocal microscopy of the cornea.
4. Alió J, Piñero D, Alesón A, Teus M, Barraquer R, Thorofare: SLACK; 2002.
Murta J, et  al. Keratoconus-integrated characteriza- 14. Somdi S, Hahnel C, Slowik C, Richter A, Weiss DG,
tion considering anterior corneal aberrations, internal Guthoff R. Confocal in vivo microscopy and confocal
astigmatism, and corneal biomechanics. J Cataract laser-scanning fluorescence microscopy in keratoco-
Refract Surg. 2011;37(3):552–68. nus. Ger J Ophthalmol. 1996;5(6):518–25.
Limbal Stromal Stem Cells
in Corneal Wound Healing:
25
Current Perspectives
and Future Applications

Noopur Mitragotri, Mukesh Damala, Vivek Singh,


and Sayan Basu

25.1 Introduction: Cornea is the mechanism involved in focusing an image


and Vision of an object. However, change in both the s­ ystems
is observed on the basis of changes in the power
The eye is a complex organ comprised of many of the lens by varying its curvature in the human
cell types. Each region has a specific optical eye, i.e. the size of the lens can be adjusted,
function to aid the transmission of light onto the whereas, in the cameras, there are a number of
retina. The cornea is one of the most important lenses that change their relative position [1].
organs which help to achieve this purpose. The Figure  25.1 shows the high-­ resolution normal
human eye can be considered optically equiva- eye image.
lent to a photographic camera, where the cornea The sclera, limbus, ciliary body and vitreous
and human crystalline lens focus the incident of the eye are also important for wound healing,
rays of light on the retina just like the camera lens accommodation, movement, nourishment, main-
focus the light on film. The iris works like the taining the shape of the eye, etc. Many extrinsic
shutter and the pupil as an aperture; the retina and intrinsic factors affect the transparency of the
acts like a film. In spite of having many similari- cornea, like the stromal scars or wound due to
ties, a major difference between the two systems infections, physical damage, chemical burns, etc.
Earlier, there was no other clinical therapy to
treat these stromal scars or wound, other than
N. Mitragotri · V. Singh (*) photo keratectomy. However, many groups have
Center to Ocular Regeneration (CORE); and Brien
Holden Eye Research Center, LV Prasad Eye recently reported the use of stem cell- or cell-­
Institute, Hyderabad, Telangana, India based therapy for treating these wounds in animal
e-mail: viveksingh@lvpei.org models. This chapter will cover various modali-
M. Damala ties along with our own method of using the lim-
Center to Ocular Regeneration (CORE); and Brien bal stem cells in these cases.
Holden Eye Research Center, LV Prasad Eye
Institute, Hyderabad, Telangana, India
School of Life Sciences, University of Hyderabad, 25.1.1 Cornea
Hyderabad, Telangana, India
S. Basu (*) The cornea is present on the external surface of
Center for Ocular Regeneration (CORE), LV Prasad
Eye Institute, Banajara Hills, Hyderabad, the eye, and its optical clarity is important for
Telangana, India vision as it helps to transmit light onto the retina
e-mail: sayanbasu@lvpei.org as mentioned earlier. It is a transparent layer, and
© Springer Nature Switzerland AG 2019 387
J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_25
388 N. Mitragotri et al.

Fig. 25.1  Anatomy of


a
the human eye: (a)
Normal human eye:
Image showing the
external appearance of a
normal human eye. (b)
Cornea: Normal
transparent cornea
giving the clear view of
intraocular iris and
limbal region. (c)
Limbus: High-resolution
image of normal limbus
located between the
conjunctiva and sclera,
the source of epithelial
and stromal stem cells b c

the transparency is maintained by the regular so it helps to maintain the structural frame-
arrangement of collagen. The corneal transpar- work. It also helps to maintain the avascularity,
ency is maintained by the three main and two transparency and immunity of the cornea.
auxiliary layers. Figure  25.2 shows different Stroma occupies more than 90% of the corneal
­layers of the cornea. Out of these five layers, thickness. It consists of an extracellular matrix,
­epithelium, stroma and endothelium are the most keratocytes (corneal fibroblasts) and nerve
important layers. Structure of cells in each layer fibres. This is how stroma is responsible for
plays an important role in corneal functioning. corneal strength and transparency [6]. When
there is a defect in the corneal stroma, it leads
(a) Corneal epithelium: It is the outermost layer to oedema, rupturing of the Descemet’s mem-
of the cornea which is approximately 50 μm brane and collagen arrangements which lead to
in thickness and has four to six layers of non-­ a loss of corneal transparency as shown in
keratinized, stratified squamous epithelial Fig. 25.3. Any problem in the stroma leads to
cells [2]. Epithelium with tear film maintains the obstruction of clear vision. Scarring of the
the optically smooth surface of the cornea corneal stroma because of surgery, trauma and
and acts as a barrier to external biological infections affects the vision of millions world-
and chemical insults [3, 4]. Epithelial cells wide [7]. Physical or chemical injuries also
can be identified by the keratin markers like result in forming scars which eventually leads
CK3 and CK12 as shown in Fig. 25.2. to vision loss or blindness.
(b) Bowman’s layer: The transparent film of tissue (d) Descemet’s membrane: It is the basement
layer behind the corneal epithelium is Bowman’s membrane of the corneal endothelium. It is
layer. It is 8–12 μm thick. It is an acellular, trans- 3–10 μm thick and made of collagen fibrils.
parent and non-regenerating layer which is It acts as a protective barrier against infec-
composed of strong layered collagen fibres. It tions and injuries and helps in maintaining
protects the subepithelial nerve plexus [5]. the corneal dehydration [5].
(c) Stroma: The major portion of the cornea is (e) Endothelium: The endothelium is the thin,
occupied by stroma. It is mainly composed of innermost layer of the cornea composed of
collagen and water. It is 450–500 μm thick and flat polygonal cells. They keep the cornea
25  Limbal Stromal Stem Cells in Corneal Wound Healing: Current Perspectives and Future Applications 389

a b

c d

Fig. 25.2  Cornea through histopathological view: (a) terstained with DAPI (blue). (c) Cross-section of the cor-
Histopathological section of a normal cornea showing the nea with epithelial hyperplasia and stromal fibrosis. (d)
epithelium, stroma and endothelium layers. (b) Section of Abnormal cornea with hypertrophy, stromal oedema and
a normal cornea showing the distribution of keratocytes ruptured Descemet’s membrane. Scale: 200 μM
stained with the CK3 marker (red) and their nuclei coun-

Fig. 25.3 
a
Vascularization of the
cornea leading to vision
loss. (a) Image of an
inflamed eye with
corneal scarring and
vascularization. (b)
Cornea with partial
limbal stem cell
deficiency with 6 clock
hour normal limbus and
6 clock hour
vascularization. (c)
Inflamed limbus with
neovascularization and
dysfunctional limbus
with early limbal stem b c
cell deficiency
390 N. Mitragotri et al.

dehydrated and clear. Generally, fluid gets cause of blindness. The epidemiology of corneal
leaked slowly from inside the eye into the blindness is complicated and encompasses a wide
stroma. The function of endothelium is to variety of infectious and inflammatory eye dis-
pump excess water out of stroma to avoid eases that cause corneal scarring, which ultimately
swelling and opacity [8]. The endothelial cell leads to blindness. To address the huge burden of
density decreases according to age from blindness in the developing world, there is a need
approximately 3000–4000 to 2500 cells/ to pay some extra attention in developing new and
mm2. Unlike epithelium, endothelial cells do promising therapies like stem cell-based therapy.
not demonstrate proliferative ability. The increasing interest in cell-based therapies is
Endothelial cells do not undergo division. because of increasing approvals and promising
Thus, if they get damaged by diseases or and better outcomes leading to less clinical follow-
trauma, they cannot be regenerated in vivo. ­up. Photo keratectomy has its own limitations and
complications mainly with the suturing, clinical
The corneal surface is more vulnerable to visits, long-term steroid usage, glaucoma, etc.
infections and injuries as it is continuously Diseases affecting the cornea are a major cause of
exposed to the environment. However, stroma blindness. The epidemiology of corneal blindness
and endothelium also get affected. This leads to is complicated and encompasses a wide variety of
optical damage or visual dysfunction or a combi- infectious and inflammatory eye diseases that
nation of both. cause corneal scarring, which ultimately leads to
functional blindness which can be understood
from Fig.  25.4. Therefore, treatment of corneal
25.2 The Necessity diseases is the necessity of the world. Corneal
of Regenerative Medicine transplantation is the standard of care as per cur-
in Corneal Blindness rent clinical practice which is used to regain the
vision. But the supply of donor tissue is poor com-
As the cornea plays the primary role in vision, any pared to the demand. However, it is necessary to
damage to it hampers the vision. Though cataract look for the new alternative treatments/therapies
is responsible for nearly 20 million of the 45 mil- other than the commonly practised methods.
lion blind people in the world, the next major
cause is trachoma, corneal scarring and vascular-
ization, blinding 4.9 million individuals [4]. 25.3 Limbus and the Stem Cells
Ocular trauma and corneal ulceration are underre-
ported mostly but are observed to be the signifi- The eye is the most sensitive organ of the body
cant causes of corneal blindness responsible for and it is continuously functional. The layers of
1.5–2.0 million new cases of monocular blindness corneal epithelium and stroma are continuously
every year. In India, approximately 6.8 million under stress because of pollution, environmental
people are estimated to have vision less than factors and microbes. Because of continuous
6/60 in at least one eye due to corneal diseases; of wear and tear, epithelial cells of the cornea are
these, about a million have bilateral involvement. consistently regenerated by the stem cells present
The number is expected to increase by 2020 to in the limbus.
10.6 million [9]. According to the National Between the optically clear cornea and opaque
Programme for Control of Blindness (NPCB) esti- sclera, a transitional region of the stem cells
mates, currently, 120,000 people are blind because exists. This transitional zone is called limbus
of corneal disorders who are estimated to increase (Fig.  25.1), which continuously regenerates the
by 25,000–30,000 every year [10]. In developing corneal epithelial cells [11, 12] and stromal stem
countries, stromal dysfunction, infections, scar- cells [13–15]. It contains radially arranged ridge-­
ring, injuries, etc. are the common causes of blind- like structures called the palisades of Vogt
ness. Diseases affecting the cornea are a major (Fig.  25.1c). The corneal epithelial stem cells
25  Limbal Stromal Stem Cells in Corneal Wound Healing: Current Perspectives and Future Applications 391

a b c

d e

Fig. 25.4  Representative images of the different clinical (d) RAFT oedema with vascularization and scarring. (e)
causes for loss of corneal transparency. (a) Scarring: cor- Graft infection
neal fibrosis. (b) Corneal oedema. (c) Corneal infection.

generate from the progenitor cells residing in thickest layer of the cornea is stroma; it contains
structures called limbal crypts, which extend a population of adult stem cells having mesen-
from the interpalisade ridges [16]. The palisades chymal stem cell (MSC) characteristics. These
protect from ultraviolet light-induced damage cells are quite in a proximity to the human limbal
[17] and a rich supply of blood vessels [18] to the epithelial stem cells. These stem cells not only
corneal stem cells. The palisades of Vogt provide support epithelial cells but also suppress inflam-
the niche for limbal epithelial stem cells (LESCs). mation by remodelling the affected stromal tissue
These stem cells give rise to transient-amplifying and help the cornea restore the transparency [20].
cells (TACs) that have the potential to divide only Limbal stem cells can be expanded to large quan-
a limited number of times. They migrate centrip- tities and thus can be used for bioengineering the
etally and are found in the limbal and peripheral epithelial-stromal tissue equivalents. This cell-­
corneal basal epithelium. Initially, the TACs dif- based therapy can be used for treating ocular sur-
ferentiate into postmitotic cells of suprabasal cor- face defects [21].
neal epithelium and later to terminal differentiated
cells. This is how epithelium is regenerated and
preventing the overgrowth of conjunctival epithe- 25.4 Corneal Wound Healing
lium onto the corneal epithelium [19]. The main
function of limbal epithelial stem cells (LESCs) The human body has a natural mechanism to fight
is to divide, differentiate and migrate centripe- against the external insults and injuries and also,
tally towards the cornea which essentially helps to repair the damaged parts of tissues and organs.
to maintain the transparency of the cornea. Epithelium, stroma and nerves maintain the
Recently, the superficial population of limbal homeostasis of the cornea. A cascade mediated by
stromal stem cells has been recognized. The paracrine and autocrine interactions of cytokines,
392 N. Mitragotri et al.

growth factors and chemokines produced by the are also important in the corneal haze formation
epithelial cells, stromal cells, immune cells and and regression due to stromal remodelling. Our
lacrimal gland along with the corneal nerves previous work in mice model has shown that
involving interactions between the epithelial cells, TGF-β and PDGF are the main factors which
stromal keratocytes corneal nerves, lacrimal convert corneal keratocyte into myofibroblast
glands, tear film and cells of the immune system [23, 24]. Our study also has shown that the bone
is responsible for corneal wound healing [22]. marrow cells may migrate to corneal stroma
after wound healing for the formation of
myofibroblasts.
25.4.1 Epithelial Injury

Cytokines are released following epithelial 25.4.3 Stromal Remodelling


injury. It includes interleukin-1 (IL-1), TNF-α,
epidermal growth factor (EGF) and platelet-­ In stromal wound healing, fibroblast growth fac-
derived growth factor (PDGF), while the stromal tor-­2 (FGF-2) induces the expression of PAX-6
keratocyte response includes the release of IL-1-­ and ABCG2 progenitor markers, which leads to
mediated Fas ligands to induce apoptosis [23, the differentiation of corneal keratocytes in vitro.
24]. IL-1 is the modulator which regulates the These cells fulfil the criteria for MSC by express-
cascade of all the factors involved in healing [25]. ing markers of MSCs. Remodelling is an impor-
It regulates hepatocyte growth factor and kerati- tant mechanism contributing to the morphogenesis
nocyte growth factors which are mediators of of repair. Corneal epithelial cells express the
stromal and epithelial interactions produced by PDGF receptors [28]. PDGF is found in the epi-
keratocytes and myofibroblasts to regulate prolif- thelial basement membrane, at very high levels,
eration and motility. IL-1 also controls the and modulates corneal fibroblast proliferation,
expression of collagenase, metalloproteases and chemotaxis and possibly their differentiation
other enzymes which leads to collagen remodel- [29]. TNF-α might also have a function in initiat-
ling and healing [25, 26]. ing the early wound healing responses.
Factors that are involved in corneal wound
healing are shown in Table 25.1.
25.4.2 Myofibroblast Migration Wound healing by limbal stem cells is similar
and Differentiation to that of the general wound healing mechanism,
and these cells suppress the inflammatory
Myofibroblasts are the cells which play a crucial responses. Epithelial and stromal damages trig-
role in wound healing process. They have con- ger a healing process mediated by progenitor
tractile pseudopodia and alpha-smooth muscle cells which are located in the limbus [30–33].
actin (α-SMA) which are present in the anterior Limbal stem cells exert therapeutic effects
stroma below areas of the epithelial layer. They which induce the repairing by secreting soluble
can be seen after disruption by stains against factors. These factors suppress the inflamma-
these components after 1–2 weeks following tion and angiogenesis which is why expression
injury. They respond to the transforming growth level of interleukin-2 and metalloproteinase is
factor beta (TGF-β) and so are considered to be observed to be decreased. In injured corneas,
the derivatives of keratocytes. They show the upregulation of IL-2, interferon γ (IF γ) and
reduced transparency because of altered corneal vascular endothelial growth factor (VEGF) has
crystalline production and participate mainly in been reported. For healing of the wound, limbal
collagen and extracellular matrix remodelling stem cells secrete certain growth factors to facili-
through the production of collagen, glycosami- tate the survival of injured cells and fasten the tis-
noglycans, collagenases, gelatinases and matrix sue regeneration procedure. Higher levels of
metalloproteinases (MMPs) [27]. Myofibroblasts growth factors, including VEGF, EGF and TGF-β,
25  Limbal Stromal Stem Cells in Corneal Wound Healing: Current Perspectives and Future Applications 393

Table 25.1  List of the factors involved in corneal wound healing, describing their source of secretion and their
functions
Sl
# Factor Secreted by Function
1 Basic fibroblast growth factor Endothelial cells Mitosis/chemotaxis of epithelial cells
(bFGF)
2 Connective tissue growth TGF-β induced fibroblast cells Mediator for TGF-β, fibroblast,
factor collagen synthesis
3 Epidermal growth factor Epithelial cells, lacrimal gland Mitosis of epithelial cells, ECM
(EGF) production
4 Platelet-derived growth factor Corneal epithelial cells Myofibroblast proliferation and
(PDGF) migration [23, 24]
5 Transforming growth factor Corneal epithelial cells Stimulation of epithelial cells
alpha (TGF-α)
6 Transforming growth factor Corneal epithelial cells, lacrimal gland, Regulation of fibroblast proliferation
beta (TGF-β) endothelial cells, inflammatory cells [23, 24]
7 Keratinocyte growth factor Keratocytes, lacrimal gland Epithelial proliferation, mitosis,
(KGF) migration
8 Interleukin-1 (IL-1) Corneal epithelial cells, keratocytes, Stimulation of Fas and Fas L
macrophages, endothelial cells, production, fibroblast and MMP
fibroblasts, eosinophils production
9 Fas Keratocytes Binds with Fas L, involved with
autocrine suicide pathway
10 Tumour necrosis factor alpha Macrophages Protease synthesis, inducing
(TNF-α) apoptosis
11 Interleukin-4 (IL-4), T-lymphocytes Anti-inflammatory, inhibits
interleukin-10 (IL-10) production of TNF α, IL-1, IL-6 [57]

can be observed after the application of limbal into a variety of progenitor cells of connec-
stem cells [34]. tive tissues which can be demonstrated by
observing the expression of CD90, CD44,
CD73, CD105, CD45 and CD34 markers. It
25.5 V
 arious MSc Sources has also been reported that BM-derived stem
for Stem Cell-Based cells/keratocytes are important for wound
Therapies for Cornea Wound healing after corneal haze. Several studies
Healing have also demonstrated that MSCs suppress
the inflammatory cells [23, 24, 35]. It is
MSCs are non-haematopoietic, multipotent stem observed that severity of injury lessens dur-
cells having multi-lineage differentiation poten- ing the acute phase of injury because of
tial. They home to the injured cornea. Many MSCs [36].
groups have been looking for the alternative ther- 2 . Adipose tissue: MSCs are easily isolated from
apies using stem cells from adipose tissue, bone the stromal vascular fraction and are more
marrow (BM), limbus, etc. accessible with higher yield. Adipose-derived
stem cells (ADSCs) have potential to differen-
1. Bone marrow: It is the most commonly used tiate into fat, cartilage, bone and muscle under
and explored source of MSCs. They can be specific culture conditions [37]. ADSCs can
differentiated into mesoderm, neuroectoderm be isolated by the lipoaspiration method. They
and endodermal cells in  vitro. Two major have potential to synthesize and secrete
stem cell populations can be obtained from keratocyte-­ specific proteins. ADSCs have
bone marrow, viz. HSC, haematopoietic stem been shown to be helping in wound healing by
cells, and MSCs. MSCs can be differentiated replacing and regenerating the new cells [38].
394 N. Mitragotri et al.

3. Dental pulp: Dental pulp stem cells (DPSCs) 25.5.1 Limitations of These Sources
are similar to the BM-MSCs and are derived of MSCs
from the pulp of extracted teeth (generally the
third molar exfoliated deciduous). They 1. Though natural body cells are used in the cell-­
express some of the markers for BM-MSCs based therapies, there are few limitations to it.
and have similar differential potential [39]. 2. The role in corneal regeneration is very limited.
DPSCs have potential to differentiate into 3. Long-term preservation of these cells.
bone, fat, dental, cartilage, nerve and muscle 4. Possible induction of an immunogenic

cells and are very friendly with most of the response.
biomaterials. DPSCs have been reported to 5. Risky, painful and invasive procedures for the
regenerate epithelial cells in animal models isolation of BM-MSCs which always have a
with LSCD [40]. possibility of infections.
4. Umbilical cord: Umbilical cord MSCs 6. BM-MSCs show reduced plasticity, growth with
(U-MSCs) are very naive, proliferative and the increase in donor age and passage number
immunosuppressive than the adult or BM-­ and the number of cells acquired being low.
MSCs [41]. U-MSCs show the homing prop- 7. The exact role of U-MSCs is still unknown.
erty because of which they differentiate into 8. Graft survival rate is not up to the mark and
the corneal endothelial cells when practised induction of inflammation and neovascular-
on the corneal injury. ization [47].
5. Embryonic stem cells (ESCs): Preliminary 9. Neural stem cell therapies are still under clini-
results have been reported by the company cal trial phase I/II.  A lot of information and
Ocata Therapeutics (formerly ACT) on clini- confirmation for treating patients is yet to be
cal safety trials for the use of retinal pig- acquired.
mented epithelial (RPE) cells derived from
hESCs for dry macular degeneration and Therefore, limbal stem cells are the preferred
Stargardt’s macular dystrophy [42, 43]. source of stem cells for patients with corneal dis-
6. Induced pluripotent stem cells (iPSCs): The orders, burns and chemical injuries. These trans-
iPSC inference of the limbal genealogy may, plants have the highest success ratio because of
in the long run, give a restorative wellspring renewing and restoring the capacity of the cor-
of limbal cells. Late advances in the recog- neal stroma and epithelium. This keeps the cor-
nizable proof of key administrative qualities nea transparent. Even a tiny part of the limbus
in limbal improvement, separation and devel- can be expanded in  vitro. In our recently pub-
opment are probably going to fasten this lished data of 125 patients, using the simple lim-
restorative open door [44]. There is also a bal epithelial transplant surgery has shown a
report of one Japanese patient who received a success rate of more than 75% with 4  years of
transplant of a sheet of iPSC-derived RPE follow-up. Similarly, cultivated limbal epithelial
[45]. This is the first of its kind in a human transplantation by Pellegrini group and by our
study that may be the leader in many other group showed good success ratio in patients with
applications of iPSC derivatives. While it limbal stem cell deficiency.
might be expected that iPSCs will be used as
autologous therapies, there is a strong move-
ment for their use as allogeneic transplants or 25.6 Results of the Preclinical
as partially compatible HLA haplotyped Trials
derivatives [46].
7. Neural stem cells: Many studies have shown In our previous preclinical studies, in the mouse
that human neural brain stem cells can be used model of corneal opacity, stromal stem cells effec-
for the treatment of retinal degeneration tively regenerated extracellular matrix and collagen
caused by dry macular degradation and for the [48]. Mesenchymal stem cells have shown to be
maintenance of photoreceptor cells. restoring corneal transparency in lumican k­ nockout
25  Limbal Stromal Stem Cells in Corneal Wound Healing: Current Perspectives and Future Applications 395

mice. Mesenchymal stem cells obtained from lim- limbal stem cells did not show increased scatter-
bal biopsy also show immune-­suppressing prop- ing compared to the normal preoperative corneas.
erty and thus help to prevent inflammation, scarring Neovascularization is another threat to the vision.
and immune rejection of transplanted tissue. It is Corneal scarring is healed when treated with lim-
published that cells obtained from expanding lim- bal stem cells and the healed corneas do not
bal biopsy cultures can be used as the cell-based express CD31 marker (which is the hallmark
therapy. This has been proven by applying the lim- for  vascular endothelial cells/ blood vessels).
bal stroma stem cells embedded in fibrin gel to the Basu et al. have shown that the cells isolated from
surface of healing murine debridement wound biopsy procedure can be used for autologous
[49]. Our team has proved that stromal cells derived transplants and these cells suppress the accumu-
from limbal biopsies differentiate into functional lation of fibrotic scar tissue [53]. The study
keratocyte and regenerate the damaged stromal tis- proves that the MSCs obtained from the limbal
sue. Isolation of cells using collagenase resulted in source are functionally similar to corneal stromal
faster cell growth and proliferation. This led to the cells with the evidence of sphere-forming capac-
upregulation of ABCG2, Nestin, OCT4, PAX6, ity, gene expression pattern and organization of
NGFR and SOX2 gene expression which is associ- ECM.  This is how limbal stem cells induce
ated with adult and pluripotent stem cells. repairing of ablated tissue with transparent ECM
Downregulated expression of genes ABCG2 and containing well-organized collagen.
Nestin was observed in the limbal stem cells cul- This preclinical study was carried out to deter-
tured on collagen substratum in serum-free kerato- mine the mesenchymal stem cells present in the
cyte differentiation medium (KDM) [50–52]. human cornea have the ability to differentiate into
When the ability of cells cultured in KDM keratocyte which helps to prevent corneal scar-
was examined on mouse corneal debridement ring. This was confirmed by various experiments
model, limbal stem cells induced fibrotic matrix like clonogenic assay, sphere formation, gene
deposition and thick ECM structures. Fifty thou- expression, immunostaining and optical coher-
sand limbal stem cells were applied to the wound ence tomography (OCT), described below in
which was labelled with 3–3′-dioctadecyloxa- Figs. 25.5 and 25.6. Human LMSCs were shown
carbocyanine. These labelled limbal stem cells to suppress the immune response, viz. when
were found to be distributed throughout the human stromal cells are injected in the mouse
wounded area. Inflammation or rejection was not model, T-cell-mediated response was not induced.
observed during this time, and after 4  weeks, This immune-suppressing quality of hLMSCs
limbal stem cells were observed to contain makes them favourable and the first choice for
human keratocan and type I collagen compo- transplantation. They provide cell-­based therapy
nents of the normal transparent stromal extracel- for corneal scarring. A Lum−/− mouse, which had
lular matrix (ECM) [53]. a thin cornea, haze and abnormal collagen distri-
As shown in Fig.  25.6, fibrotic markers hyal- bution, was treated with limbal stem cells. Cell
uronan, fibronectin, tenascin C, biglycan and rejection was not observed and the stromal cells
decorin were abundant showing scar formation in got distributed on the mouse stroma [48].
the controls (without limbal stem cells).
Proteoglycan, decorin and the stromal matrix were
observed in the wounded corneas treated with lim- 25.7 Human Limbal-Derived
bal stem cells. Additionally, mRNA for type III Stromal/Mesenchymal Stem
collagen and SPARC (secreted protein acidic and Cells (hLMSCs): Our
rich in cysteine) was upregulated after 2  days in Approach for Treating
debrided corneas. Visible scarring was observed in Corneal Scar and Other
all the eyes which did not receive the limbal stem Corneal Pathologies
cells through the low magnification photos.
Corneal scars scatter the light leading to Therapeutically accepted and serologically
reduced visual acuity, but the eyes treated with tested cadaveric donor corneas were obtained
396 N. Mitragotri et al.

Fig. 25.5  Repairing mechanism of the limbal stem cells keratocan and type I collagen. The nuclei in all sections
in a mouse model. (a) Whole-mount immunostaining of were counterstained with DAPI (blue). Ctrl, control;
the murine cornea that received hLMSCs, showing the scale: 50 μM, 500 μM. (*Note – This figure is reproduced
persistence of hLMSCs, DiO-labelled (green), 1 week with permission (copyright licence number
after the transplantation. (b) Histological sections of the 4307021062699) from Basu et  al. [53], Science
murine cornea showing positive expression of human Translational Medicine journal)

from the Ramayamma International Eye Bank, bus from them. In aseptic conditions, a 360°
L V Prasad Eye Institute, approved for process- limbal ring was excised using optical magnifica-
ing, storing and distributing human ocular tis- tion and surgical instruments. Then it is washed
sue. Corneas were washed with antibiotic with buffered saline and minced into smaller
fortified buffered saline before extracting lim- fragments. The minced tissue fragments were
25  Limbal Stromal Stem Cells in Corneal Wound Healing: Current Perspectives and Future Applications 397

Biglycan No cells LBSC


Decorin

100 µm
Fibronectin
Tenascin C

100 µm
binding protein
Hyaluronan

100 µm

Fig. 25.6 Image showing the blockage of fibrotic nuclei in all sections were counterstained with DAPI
matrix deposition by hLMSCs in murine corneas. (blue). Scale: 100  μM. (*Note  – This figure is repro-
Collage of the immunostaining analysis, showing the duced with permission (copyright licence number
negative expression of the fibrotic markers, decorin, 4307021062699) from Basu et  al. [53], Science
tenascin C and hyaluronan, in the murine corneas Translational Medicine journal)
treated with the limbal-derived stromal stem cells. The

collected into incomplete media (plain DMEM/ complete media fortified with 2% serum. The
F12 media, Lonza, BE04-­687F/U1) and were digested tissues were then spun down at
subjected to collagenase digestion by adding 1000  rpm for 3  min at room temperature, in
20  μL of reconstituted collagenase IV saline. Subsequently, the addition of complete
(17104019, Thermofisher) at a ­concentration of media was done to dissolve the pellet, and the
10  IU/μL, to the tissue suspension. Incubation isolated hLMSC suspension was then placed in
of the limbal fragments with the collagenase culture flasks. Media were replaced after every
was carried out for 16  h at 37  ° C in 5% CO2 2 days. Primary cultures yield mostly epithelial
chamber. cells which were sub-cultured for three genera-
After 16 h of incubation, enzymatic digestion tions to obtain a pure population of the
of collagenase was stopped by adding 2 mL of hLMSCs.
398 N. Mitragotri et al.

The Application/Delivery of the hLMSCs  The 25.8 Challenges in a Clinical Trial


technique of delivering the hLMSCs established
by Funderburgh et al. involves (1) debridement 1. The use of hLMSCs is still under trial.

of the corneal epithelium using dry sponge, (2) Obtaining an adequate number of donor cor-
half-­
a-­million cells mixed in 0.05  mL of the neas is still a big challenge.
thrombin component of the fibrin glue and 2. The number of cells obtained per square mil-
allowing the two components to gel, and (3) a limetre reduces with age of the donor.
bandage contact lens placed on the eye. Patients 3. Transportation is a major issue as the cells
in this study group were prescribed prophylactic have to be preserved or transported at a very
topical antibiotics without any corticosteroids. low temperature that requires definite logistics
Patients in the control group were given the stan- and equipment.
dard medical therapy, including topical cortico- 4. As the cells will be used for patients, sterility
steroids and debridement with fibrin glue but has to be maintained which requires cGMP
without any cells. (Current Good Manufacturing Practices) labs
which are very expensive.
The Funderburgh technique of using autolo-
gous and allogenic limbal stromal stem cells has The demonstration of safety and efficacy of
shown to enhance vision, promote epithelializa- the drug product intended for use in humans is
tion in the cornea, improve the corneal clarity and essential before it is approved by the Central
reduce corneal scarring, thus obviating the need Drugs Standard Control Organization (CDSCO),
for corneal transplantation in the eyes with cor- for import or manufacturing. The information
neal burns, ulcers and scars (Fig. 25.7). required for approval of an application to import

a b c

d e f

g h

Fig. 25.7  Events of tissue processing and application of cells population after 14 days of culture. (e) Image of the
the hLMSCs during transplantation. (a) Debris and pure stromal cell population which can be obtained after
unwanted layers of corneal tissues removed. (b) A 360° few passages. (f) Trypsinized cell pellet sent for transplan-
limbal rim dissected out. (c) Collagenase-digested limbal tation. (g) Mixing of the hLMSCs with fibrin glue. (h) Cells
explants were cultured in T25 flasks. (d) Image of epithelial with fibrin glue being applied onto the debrided cornea
25  Limbal Stromal Stem Cells in Corneal Wound Healing: Current Perspectives and Future Applications 399

or manufacture a new drug for marketing regula- the production of pharmaceutical or cell products
tions under Appendix I, IA and VI of Schedule Y [56]. It incorporates the assembling space, the
is described well. In India, for the approval of capacity stockroom for crude and completed
new drug, all the necessary requirements along items and other research facility territories.
with the NDA to FDA should be provided [54], Manufacturing space, the storage warehouse for
which were already obtained by our institute, for raw and finished product and support laboratory
this particular clinical trial. Before starting the areas are included in the cGMP facilities. The
clinical trials (http://www.clinicaltrials.gov, Ref# cGMP facilities are organized and designed
NCT02948023) using the hLMSCs, various doc- according to quality system for collection, pro-
uments have been produced and submitted to the cessing and storage of cell therapy products. For
DCGI.  All procedures followed were in accor- cell-based products, terminal sterilization of the
dance with the ethical standards of the responsi- product or removal/inactivation of microbial con-
ble committee on human experimentation taminants cannot be done. Thus, tested and certi-
(institutional and national) and with the Helsinki fied starting materials and validated processes
Declaration of 1975, as revised in 2000. Informed have been followed, and sterility testing for the
consent was obtained from all patients for being product has been done [55].
included in the study. The facilities must be planned and designed as
per Good Manufacturing Practice for
1. Ethical committee approval/IRB approval
Pharmaceutical Manufacturers including quality
(Ref # LEC 01-14-015) control and quality affirmation programmes,
2. Data of the preclinical studies which set up a quality system way to deal with
3. ICSCR (Institutional Committee for Stem
control gathering, preparing, stockpiling and the
Cell Research) approval arrival of cell treatment items and furthermore
4. Form 44 of the CDSCO that includes the
address the accompanying components like:
study protocol, CV of the principal investiga-
tor, informed consent form, case report form, 1 . Facility (outline, access and maintenance)
master formula record, etc. 2. Equipment (buy, utilization and support)
3. Raw material specifications (buy, stockpiling
and utilization)
Requirements of cGMP Facility  In order to 4. Quality assurance (quality control, approval,
conduct human clinical trials, there is a require- capability and documentation)
ment of few sophisticated facilities and the neces-
sity to have standard SOPs. These are the main The quality assurance is based on these listed
regulatory criteria for ensuring the pharmaceuti- elements: quality programme, organization and
cal quality of a product which will be used for the personnel, standard operating procedures, envi-
patients. Good manufacturing production facili- ronmental control, equipment monitoring, sup-
ties are always a prerequisite for cell- and tissue-­ plies and reagents, process controls, process
based therapies. The facilities are designed as per changes, process validation, labelling design and
the purpose and are accredited by the national control, storage requirements, records, tracking,
regulatory body. The scientists and co-workers non-conformances and complaints management,
work as per the standard protocols of quality risk assessment, reporting and reviewing. In
assurance and ethics committees to ensure the areas where quality assurance interacts with hos-
safety of cellular products [55]. pital departments and infrastructures, such as
housekeeping, engineering, information technol-
Cell-based medicinal products (CBMPs) are ogy and supply departments, the aim is to estab-
handled with lots of precautions. Any method lish relevant service-level agreements in order
associated with CBMPs requires a strict control to  ensure an adequate level of regulatory
in cGMP facility. A cGMP facility is meant for compliance.
400 N. Mitragotri et al.

The quality system approach is chance based. optimized and have a very good success rate.
For the facilities, it is of principal significance to Yet, these corneal transplants need lifelong fol-
forestall potential defilement of both microbio- low-up along with standard medical care. With
logical and endotoxin contaminations because of the emergence of cell-based therapies, mainly
the imperfections in natural conditions, handlers, the stem cells, they have made a paradigm shift
culture holders, crude materials or cross tainting from the conventional therapies to regenerative
with different items arranged in a similar genera- medicine. Our previous work in mice has shown
tion plant. Care has to be taken with techniques that stromal stem cells obtained by culturing in
for compartment disinfection and control of low serum media can prevent corneal scarring.
crude materials and other reagents. Utilization of We have proven that these mesenchymal stem
high-efficiency particulate absorbing (HEPA) cells of limbal origin, when given in 50,000
channel to anticipate airborne cross tainting, iso- cells along with fibrin sealant, can heal the
late treatment of materials from various patients mouse corneal scars in 4  weeks. We now have
and so on has to be done. Consequently, the most taken this technology to the clinic, with the fully
critical rooms of these facilities incorporate the equipped GMP facility at our institute. Our pilot
alleged clean rooms. These clean rooms are study has shown safety and efficacy in treating
grouped into four classes relying upon air and various corneal pathologies mainly the corneal
dust particles, in view of the number of particles scars.
of two sizes (≥0.5 μm, ≥5 μm). Different param-
eters, for example, temperature, moisture and Acknowledgements  We acknowledge Mr. Abhinav
pressure, have been considered and checked on Reddy Kethiri, Brien Holden Eye Research Center, and
Mr. Sridhar Rao Boyinpally, Department of Pathology, LV
account of their potential effect on microorgan- Prasad Eye Institute, for helping in generating the images
ism expansion. The flow of materials and the per- for this chapter. We thank Champalimaud Foundation and
sonnel is kept seperate and unidirectional to limit Tej Kohli Cornea Institute and LV Prasad Eye Institute for
cross-contamination. All the activities are docu- providing the financial assistance and manpower for this
chapter.
mented in real time. The specialized staff trained
in essential cleanliness measures required for
Conflict of Interest  Noopur Mitragotri, Mukesh Damala,
control in clean rooms. The specialized staff has Vivek Singh and Sayan Basu  declare that they have
sufficient capability for both the lead and recon- no conflict of interest.
naissance of all exercises. Generation and dis- Informed consent was obtained along with the consent
semination of CBMPs or ATPs are controlled by for publication.
the applicable local and national authorities.
Note  No animal studies were carried out by the authors
for this article.

25.9 Summary

Corneal clarity is essential for clear vision. This References


is often affected by various factors like mechan- 1. Oommen V, Kanthakumar P.  A simple model of the
ical injuries, burns and scars, which turn out to accommodating lens of the human eye. Adv Physiol
be the major causatives of visual impairment. Educ. 2014;38(2):183–4.
The underlying and natural mechanism of cor- 2. Hanna C, Bicknell DS, O’Brien JE.  Cell turn-
over in the adult human eye. Arch Ophthalmol.
neal repair involves the layering of fibrotic tis- 1961;65:695–8.
sue that makes the cornea cloudy and opaque. 3. Torricelli AA, Singh V, Santhiago MR, Wilson
The standard and current care of therapy for any SE. The corneal epithelial basement membrane: struc-
such corneal stromal defects are penetrating ture, function, and disease. Invest Ophthalmol Vis Sci.
2013;54(9):6390–400.
keratoplasty or lamellar keratoplasty, replacing 4. Torricelli AA, Santhanam A, Wu J, Singh V, Wilson
the whole cornea or the affected stromal region, SE. The corneal fibrosis response to epithelial-stromal
respectively. These clinical treatments are fully injury. Exp Eye Res. 2016;142:110–8.
25  Limbal Stromal Stem Cells in Corneal Wound Healing: Current Perspectives and Future Applications 401

5. Jacobsen IE, Jensen OA, Prause JU.  Structure and 21. Kureshi AK, Dziasko M, Funderburgh JL, Daniels
composition of Bowman’s membrane. Study by frozen JT. Human corneal stromal stem cells support limbal
resin cracking. Acta Ophthalmol. 1984;62(1):39–53. epithelial cells cultured on RAFT tissue equivalents.
6. Nishida K, Kawasaki S, Kinoshita S.  Clusterinmay Sci Rep. 2015;5:16186.
be essential for maintaining ocular surface epithelium 22. Eraslan M, Toker E.  Mechanisms of corneal wound
as non-keratinizing epithelium. Adv Exp Med Biol. healing and its modulation following refractive sur-
1998;438:629–35. gery. Marmara Med J. 2009;22(2):169–78.
7. Funderburgh JL, Funderburgh ML, Du Y. Stem cells 23. Singh V, Barbosa FL, Torricelli AA, Santhiago MR,
in the limbal stroma. Ocul Surf. 2016;14(2):113–20. Wilson SE.  Transforming growth factor beta and
8. Murphy C, Alvarado J, Juster R.  Prenatal and post- platelet-derived growth factor modulation of myo-
natal growth of the human Descemet’s membrane. fibroblast development from corneal fibroblasts
Invest Ophthalmol Vis Sci. 1984;25(12):1402–15. in vitro. Exp Eye Res. 2014;120:152–60.
9. Dandona R, Dandona L. Corneal blindness in a south- 24. Singh V, Jaini R, Torricelli AA, Santhiago MR,

ern Indian population: need for health promotion Singh N, Ambati BK, Wilson SE.  TGFbeta and
strategies. Br J Ophthalmol. 2003;87(2):133–41. PDGF-B signaling blockade inhibits myofibroblast
10. Gupta N, Tandon R, Gupta SK, Sreenivas V, Vashist development from both bone marrow-derived and
P.  Burden of corneal blindness in India. Indian J keratocyte-­derived precursor cells in  vivo. Exp Eye
Community Med. 2013;38(4):198–206. Res. 2014;121:35–40.
11. Dua HS, Azuara-Blanco A.  Limbal stem cells
25. Wilson SE, Schultz GS, Chegini N, Weng J, He

of the corneal epithelium. Surv Ophthalmol. YG.  Epidermal growth factor, transforming growth
2000;44(5):415–25. factor alpha, transforming growth factor beta, acidic
12. Kulkarni BB, Tighe PJ, Mohammed I, Yeung AM, fibroblast growth factor, basic fibroblast growth fac-
Powe DG, Hopkinson A, Shanmuganathan VA, Dua tor, and interleukin-1 proteins in the cornea. Exp Eye
HS. Comparative transcriptional profiling of the lim- Res. 1994;59(1):63–72.
bal epithelial crypt demonstrates its putative stem cell 26. Weng J, Mohan RR, Li Q, Wilson SE.  IL-1 up-­
niche characteristics. BMC Genomics. 2010;11:526. regulates keratinocyte growth factor and hepato-
13. Branch MJ, Hashmani K, Dhillon P, Jones DR, Dua cyte growth factor mRNA and protein production
HS, Hopkinson A.  Mesenchymal stem cells in the by cultured stromal fibroblast cells: interleukin-1
human corneal limbal stroma. Invest Ophthalmol Vis beta expression in the cornea. Cornea. 1997;
Sci. 2012;53(9):5109–16. 16(4):465–71.
14. Funderburgh ML, Du Y, Mann MM, SundarRaj N, 27. Jester JV, Petroll WM, Cavanagh HD. Corneal stromal
Funderburgh JL.  PAX6 expression identifies pro- wound healing in refractive surgery: the role of myofi-
genitor cells for corneal keratocytes. FASEB J. broblasts. Prog Retin Eye Res. 1999;18(3):311–56.
2005;19(10):1371–3. 28. Kamiyama K, Iguchi I, Wang X, Imanishi J. Effects
15. Damala M, Kethiri R, Tavakkoli F, Raju E, Singh of PDGF on the migration of rabbit corneal fibroblasts
V.  The basics of stem cells and their role in vision. and epithelial cells. Cornea. 1998;17(3):315–25.
In: Trends in life science research. Hauppauge, 29.
Kim WJ, Helena MC, Mohan RR, Wilson
New York: Nova Science Publishers, Inc; 2018. SE. Changes in corneal morphology associated with
16. Yoshida S, Shimmura S, Nagoshi N, Fukuda K,
chronic epithelial injury. Invest Ophthalmol Vis Sci.
Matsuzaki Y, Okano H, Tsubota K. Isolation of multi-­ 1999;40(1):35–42.
potent neural crest-derived stem cells from the adult 30. Frank MH, Frank NY. Restoring the cornea from lim-
mouse cornea. Stem Cells. 2006;24(12):2714–22. bal stem cells. Regen Med. 2015;10(1):1–4.
17. Shortt AJ, Secker GA, Munro PM, Khaw PT, Tuft SJ, 31. Kruse FE, Volcker HE.  Stem cells, wound healing,
Daniels JT.  Characterization of the limbal epithelial growth factors, and angiogenesis in the cornea. Curr
stem cell niche: novel imaging techniques permit Opin Ophthalmol. 1997;8(4):46–54.
in vivo observation and targeted biopsy of limbal epi- 32. Secker GA, Daniels JT. Limbal epithelial stem cells of
thelial stem cells. Stem Cells. 2007;25(6):1402–9. the cornea. In: StemBook. Cambridge, MA: Harvard
18.
Mathews S, Chidambaram JD, Lanjewar S, Stem Cell Institute; 2008.
Mascarenhas J, Prajna NV, Muthukkaruppan V, 33. Oie Y, Nishida K. Regenerative medicine for the cor-
Chidambaranathan GP. In vivo confocal microscopic nea. Biomed Res Int. 2013;2013:428247.
analysis of normal human anterior limbal stroma. 34. Gao J, Dennis JE, Muzic RF, Lundberg M, Caplan
Cornea. 2015;34(4):464–70. AI. The dynamic in vivo distribution of bone marrow-­
19. Cotsarelis G, Cheng SZ, Dong G, Sun TT, Lavker derived mesenchymal stem cells after infusion. Cells
RM. Existence of slow-cycling limbal epithelial basal Tissues Organs. 2001;169(1):12–20.
cells that can be preferentially stimulated to pro- 35. Liu H, Zhang J, Liu CY, Hayashi Y, Kao WW. Bone
liferate: implications on epithelial stem cells. Cell. marrow mesenchymal stem cells can differentiate
1989;57:201–9. and assume corneal keratocyte phenotype. J Cell Mol
20. Pinnamaneni N, Funderburgh JL.  Concise review:
Med. 2012;16(5):1114–24.
stem cells in the corneal stroma. Stem Cells. 36. Lan Y, Kodati S, Lee HS, Omoto M, Jin Y, Chauhan
2012;30(6):1059–63. SK.  Kinetics and function of mesenchymal stem
402 N. Mitragotri et al.

cells in corneal injury. Invest Ophthalmol Vis Sci. istration does not improve corneal graft survival out-
2012;53(7):3638–44. come. PLoS One. 2015;10(3):e0117945.
37. Tholpady SS, Llull R, Ogle RC, Rubin JP, Futrell JW, 48. Du Y, Carlson EC, Funderburgh ML, Birk DE,

Katz AJ. Adipose tissue: stem cells and beyond. Clin Pearlman E, Guo N, Kao WW, Funderburgh JL. Stem
Plast Surg. 2006;33(1):55–62. cell therapy restores transparency to defective murine
38. Du Y, Roh DS, Funderburgh ML, Mann MM, Marra corneas. Stem Cells. 2009;27(7):1635–42.
KG, Rubin JP, Li X, Funderburgh JL.  Adipose-­ 49. Basu S, Fernandez MM, Das S, Gaddipati S,

derived stem cells differentiate to keratocytes in vitro. Vemuganti GK, Sangwan VS.  Clinical outcomes of
Mol Vis. 2010;16:2680–9. xeno-free allogeneic cultivated limbal epithelial trans-
39. Ranganathan K, Lakshminarayanan V. Stem cells of plantation for bilateral limbal stem cell deficiency. Br
the dental pulp. Indian J Dent Res. 2012;23(4):558. J Ophthalmol. 2012;96(12):1504–9.
40. Syed-Picard FN, Du Y, Lathrop KL, Mann MM,
50. Du Y, Sundarraj N, Funderburgh ML, Harvey SA,
Funderburgh ML, Funderburgh JL.  Dental pulp Birk DE, Funderburgh JL.  Secretion and organiza-
stem cells: a new cellular resource for corneal stro- tion of a cornea-like tissue in vitro by stem cells from
mal regeneration. Stem Cells Transl Med. 2015;4(3): human corneal stroma. Invest Ophthalmol Vis Sci.
276–85. 2007;48(11):5038–45.
41. El Omar R, Beroud J, Stoltz JF, Menu P, Velot E, 51. Karamichos D, Funderburgh ML, Hutcheon AE,

Decot V. Umbilical cord mesenchymal stem cells: the Zieske JD, Du Y, Wu J, Funderburgh JL.  A role for
new gold standard for mesenchymal stem cell-based topographic cues in the organization of collagenous
therapies? Tissue Eng Part B Rev. 2014;20(5):523–44. matrix by corneal fibroblasts and stem cells. PLoS
42.
Schwartz SD, Hubschman JP, Heilwell G, One. 2014;9(1):e86260.
Franco-­ Cardenas V, Pan CK, Ostrick RM, Lanza 52. Chan AA, Hertsenberg AJ, Funderburgh ML, Mann
R.  Embryonic stem cell trials for macular degen- MM, Du Y, Davoli KA, Mich-Basso JD, Yang L,
eration: a preliminary report. Lancet. 2012; Funderburgh JL. Differentiation of human embryonic
379(9817):713–20. stem cells into cells with corneal keratocyte pheno-
43. Schwartz SD, Regillo CD, Lam BL, Eliott D, Rosenfeld type. PLoS One. 2013;8(2):e56831.
PJ, Gregori NZ, Hubschman JP, Davis JL, Heilwell 53. Basu S, Hertsenberg AL, Funderburgh MK, Burro
G, Spirn M, Maguire J, et al. Human embryonic stem MM, Man M, Du Y, Lathrop K, Syed-Picard FM,
cell-derived retinal pigment epithelium in patients Adam SE, Bir D, Funderburgh J.  Human limbal
with age-related macular degeneration and Stargardt’s biopsy-derived stromal stem cells prevent corneal
macular dystrophy: follow-up of two open-label scarring. Sci Transl Med. 2014;6(266):266ra172.
phase 1/2 studies. Lancet. 2015;385(9967):509–16. 54. Ravinder RB, Suresh N.  Regulatory stages for

44. Pellegrini G, De Luca M.  Eyes on the prize: limbal new drug approvals. 2011. Available at www.
stem cells and corneal restoration. Cell Stem Cell. observerindia.com/cms/export/orfonline/modules/
2014;15(2):121–2. occasionalpaper/attachments/Drug_Discovery_
45. Reardon S, Cyranoski D.  Japan stem-cell trial stirs Book_1260179432814.pdf
envy: researchers elsewhere can't wait to test iPS cells 55. Giancola R, Bonfini T, Iacone A. Cell therapy: cGMP
in humans. Nature. 2014;513(7518):287–9. facilities and manufacturing. Muscles Ligaments
46. Turner M, Leslie S, Martin NG, Peschanski M, Rao Tendons J. 2012;2(3):243–7.
M, Taylor CJ, Yamanaka S, Wilmut I.  Toward the 56. Dietz AB, Padley DJ, Gastineau DA.  Infrastructure
development of a global induced pluripotent stem cell development for human cell therapy translation. Clin
library. Cell Stem Cell. 2013;13(4):382–4. Pharmacol Ther. 2007;82(3):320–4.
47. Fuentes-Julián S, Arnalich-Montiel F, Jaumandreu
57. Lim M, Goldstein MH, Tuli S, Schultz GS.  Growth
L, Leal M, Casado A, García-Tuñon I, De Miguel factor, cytokine and protease interactions during cor-
MP. Adipose-derived mesenchymal stem cell admin- neal wound healing. Ocul Surf. 2003;1(2):53–65.
Cell Therapy of the Corneal Stroma
Using Ex Vivo Cultured Extraocular
26
Cells

Jorge L. Alió del Barrio

26.1 Introduction been demonstrated by several authors, including


reports from our research group [3–5], the capa-
The stroma constitutes more than 90% of the cor- bility of these stem cells to not only survive and
neal thickness. Many features of the cornea, differentiate into adult human keratocytes in
including its strength, morphology and transpar- xenogeneic scenarios without inducing any
ency, are attributable to the anatomy and biome- inflammatory reaction but also to produce new
chanical properties of the corneal stroma [1]. collagen within the host stroma [3, 6], to modu-
Many diseases as corneal dystrophies, scars or late preexisting scars by corneal stroma remodel-
ectatic disorders induce a distortion of its anat- ling [7, 8] and to improve corneal transparency in
omy or physiology leading to a transparency loss animal models for corneal dystrophies by colla-
and subsequent loss of vision. Despite the great gen reorganization as well as in animal models
efforts in the last decade to try to replicate the for metabolopathies by the catabolism of the
corneal stroma in the laboratory in order to look accumulated proteins [9–12]. Mesenchymal stem
for an alternative to classical corneal transplanta- cells have also shown immunomodulatory prop-
tion, this has not been accomplished yet due to erties in syngeneic, allogeneic and even xenoge-
the extreme difficulty to mimic the highly com- neic scenarios [12, 13]. Actually, the first clinical
plex ultrastructure of the corneal stroma, obtain- data regarding the safety and preliminary efficacy
ing substitutes that either don’t achieve enough of the cellular therapy of the corneal stroma from
transparency or strength properties [2]. phase 1 human clinical trials is now available [14,
In this scenario, in the last few years has 15], leading to a new and promising field for
gained a lot of interest the cellular therapy of the future research that may end up providing a real
corneal stroma by the use of mesenchymal stem alternative treatment option for corneal diseases
cells from either ocular or extraocular sources, in the near future.
capable to differentiate into adult keratocytes Considering existing scientific evidence, it
in vitro and in vivo in animal models [1]. It has seems that all types of mesenchymal stem cells
(MSCs) have similar behaviour in  vivo
(Table 26.1) and thus are able to achieve a kerato-
cyte differentiation and modulate the corneal
stroma with immunomodulatory properties [16].
It has also been recently reported that MSC
J. L. Alió del Barrio (*)
University Miguel Hernandez, Vissum-Instituto secretes paracrine factors such as vascular endo-
Oftalmologico de Alicante, Alicante, Spain thelial growth factor (VEGF), platelet-derived

© Springer Nature Switzerland AG 2019 403


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_26
404 J. L. Alió del Barrio

Table 26.1  Stem cells used for the regeneration of the corneal stroma
CSSC BM-MSC ADASC UMSC ESC iPS
Keratocyte differentiation in vitro demonstrated Yes Yes Yes Yes Yes Yes
Keratocyte differentiation in vivo demonstrated Yes Yes Yes Yes No No
Possible autologous use Yes/no Yes Yes Yes/no No Yes
Evidence of keratocyte differentiation and their possible autologous applications

growth factor (PDGF), hepatocyte growth factor profile as ADASCs, but their extraction by bone
(HGF), transforming growth factor beta 1 marrow puncture is a more complicated and pain-
(TGFβ1), etc. that promote cell migration and ful procedure that requires general anaesthesia.
keratocyte survival by apoptosis inhibition and Umbilical MSCs (UMSCs) present an attractive
upregulate the expression of extracellular matrix alternative, but their autologous use would be
(ECM) component genes in keratocytes, subse- expensive and is currently impossible as umbili-
quently enhancing corneal re-epithelialization cal cord is not generally stored after birth.
and stromal wound healing [17]. MSC can be Embryonic stem cells have great potential but
obtained from many human tissues, including also present important ethical issues. On the
adipose tissue, bone marrow, umbilical cord, other hand, the use of iPS technology [21] could
dental pulp, gingiva, hair follicle, cornea and pla- solve such problems, and their capability to gen-
centa [18, 19]. erate adult keratocytes has already been proven
Corneal stroma stem cells (CSSCs) are a in vitro [22].
promising source for cellular therapy as far as In this chapter we will review the different
their isolation technique and culture methods are types of extraocular stem cells (mesenchymal or
being optimized and refined [20], and presum- not) that have been suggested for the regenera-
ably they should have a more efficient keratocyte tion of the corneal stroma, as well as the current
differentiation as they are already corneal cells. existing evidence either in  vitro or in  vivo.
On the other hand, CSSC isolation technique is Finally, we will review the different surgical
more technically demanding considering the approaches that have been suggested (in vivo) for
small amount of tissue that is usually processed the application of this stem cell therapy for the
and still requires a contralateral healthy eye (for corneal stroma regeneration.
autologous use), which is not always available
(bilateral disease) or accepted by patients. An
allogeneic use would still require living or cadav- 26.2 E
 xtraocular Stem Cells Used
eric donor corneal tissue. These drawbacks may for Corneal Stroma
limit their use in clinical practice (mainly for Regeneration
autologous use). On the other hand, extraocular
stem cells don’t present these limitations. Human 26.2.1 Bone Marrow Mesenchymal
adult adipose tissue has shown to be an ideal Stem Cells (BM-MSCs)
source of autologous stem cells, as it satisfies all
the requirements, easy accessibility to the tissue, Park et  al. reported that human BM-MSCs dif-
high cell retrieval efficiency and the ability of its ferentiate in vitro into keratocyte-like cells when
stem cells (h-ADASCs), to differentiate into mul- they are grown in specific keratocyte differentia-
tiple cell types (keratocytes, osteoblasts, chon- tion conditions [23]. They demonstrated a strong
droblasts, myoblasts, hepatocytes, neurons, etc.) expression of keratocyte markers as lumican and
[3]. This cellular differentiation occurs due to the ALDH, together with the loss of expression of
effects of very specific stimulating factors or stem cell markers such as α-smooth muscle actin.
environments for each cell type, avoiding the mix However, they couldn’t demonstrate an evident
of multiple kinds of cells in different niches. expression of keratocan on these differentiated
Bone marrow MSCs (BM-MSCs) have the same cells [23]. Trosan et  al. showed that mice
26  Cell Therapy of the Corneal Stroma Using Ex Vivo Cultured Extraocular Cells 405

BM-MSC cultured in the presence of corneal inflammatory response. These findings were later
extracts and insulin-like growth factor-I (IGF-I) reproduced and confirmed by other authors in
efficiently differentiate into corneal-like cells several research papers [16].
with expression of corneal-specific markers, such The relatively easier extraction and isolation
as cytokeratin 12, keratocan and lumican [24]. techniques of these cells with respect to bone
The survival and differentiation of human marrow sources make these cells so far the best
BM-MSCs into keratocytes have also been dem- extraocular candidates for clinical use for corneal
onstrated in vivo when these cells are transplanted stroma regeneration.
inside the corneal stroma of an animal model.
Keratocan expression was observed without any
sign of immune or inflammatory response [25]. 26.2.3 Umbilical Cord Mesenchymal
Stem Cells (UCMSCs)

26.2.2 Adipose-Derived Adult Human mesenchymal stem cells isolated from


Mesenchymal Stem Cells neonatal umbilical cords have shown similar
(ADASCs) behaviour to other types of mesenchymal stem
cells when transplanted inside the corneal stroma
Human ADASCs (h-ADASCs) cultured in  vitro in  vivo, expressing keratocyte-specific markers
(Fig. 26.1) under keratocyte differentiation con- such as keratocan without inducing immune or
ditions express collagens and other corneal-­ rejection responses [27]. Liu et al. reported that
specific matrix components. This expression is the injection of these cells inside the corneal
quantitatively similar to that achieved by the dif- stroma of lumican-null mice improved corneal
ferentiated h-CSSCs [26]. transparency and increased the stromal thickness
The differentiation of h-ADASCs in func- with a reorganized collagen lamellae, also
tional human keratocytes has also been demon- improving the host keratocyte function through
strated in  vivo, for the first time, in a previous enhanced expression of keratocan and ALDH in
study of our group using a rabbit animal model these mice [10]. These data is encouraging,
[3]. These cells, once implanted intrastromally, although up to date, the autologous use of
express not only collagens type I and VI (the UCMSCs is not possible as far as the umbilical
main components of corneal extracellular matrix) cord from new births is not generally stored.
but also keratocyte-specific markers such as kera-
tocan or ALDH, without inducing an immune or
26.2.4 Embryonic Stem Cells (ESCs)

Current experience with these human pluripotent


stem cells for the corneal stroma regeneration is
much more limited. Chan et al. reported that dif-
ferentiation of these cells into a keratocyte lin-
eage can be induced in  vitro, demonstrating an
upregulation of keratocyte markers including
keratocan [28].
To the best of our knowledge, no studies
in vivo with these cells have been performed in
the field of regenerative medicine for the corneal
stroma. The use of these cells also raises many
ethical issues and, together with the lack of
Fig. 26.1  Microscopic appearance (phase-contrast pho- in vivo data so far, discourages for their use in the
tograph) of human ADASCs (×10 magnification) clinical setting at this time.
406 J. L. Alió del Barrio

26.2.5 Induced Pluripotent Stem and corneal epithelium/limbal stem cell niche
Cells (iPSCs) regeneration, topic not treated here as it has been
already discussed in a previous chapter of this
As already discussed, embryonic stem cells have book. However, surface implantation of MSC
fallen into a certain abandonment partly because would still have a place for the prevention or
of their ethical problems but especially because modulation of anterior stromal scars after an ocu-
of the appearance of the iPSC [21], which do not lar surface injury as a chemical burn. As already
present these problems as they derive from adult discussed before, MSC secretes paracrine factors
cells. In 2012, the Japanese Shinya Yamanaka that enhance corneal re-epithelialization and stro-
and the British John B.  Gurdon received the mal wound healing [17]. Thus, MSC benefits
Medicine Nobel Prize for the discovery that onto the ocular surface may be justified more by
mature, specialized cells can be reprogrammed to these paracrine effects rather than by a differen-
go back to an immature or stem cell state, with a tiation of the MSC into epithelial cells, being the
differentiation capacity to almost any cell line of evidence for the latest controversy. On this
the organism. The iPSCs promise to be the future regard, Di et  al. assayed subconjunctival injec-
of tissue and cellular engineering. tions of BM-MSC in diabetic mice, reporting an
Regarding their application into the corneal increased corneal epithelial cell proliferation and
stroma regeneration, human iPSC has shown the an attenuated inflammatory response mediated
capability to differentiate into neural crest cells by tumour necrosis factor-α-stimulated gene 6
(the embryonic precursor to keratocytes) and, by (TSG6) [30].
culturing them on cadaveric corneal tissue, to Holan et  al. suggested the MSC application
promote their keratocyte differentiation by onto the ocular surface by the use of nanofibre
acquiring a keratocyte-like morphology and scaffolds. They reported that BM-MSC growth
expressing markers similar to corneal keratocytes onto these scaffolds can enhance re-­
[22]. It has been also shown that iPSC-derived epithelialization and supress neovascularization
MSCs exert immunomodulatory properties in the and local inflammatory reaction when applied
cornea similar to those observed with BM-MSCs onto an alkali-injured eye in a rabbit model, and
[29]. To the best of our knowledge, no studies these results were comparable to the ones
have been published reporting the capability of obtained with limbal epithelial stem cells and
the iPSC to differentiate into adult keratocytes both of them better than with ADASC [31]. The
in vivo in the animal model. same group suggested that these results can be
improved when these nanofibre scaffolds seeded
with rabbit BM-MSC are covered with cyclospo-
26.3 Corneal Stroma rine A (CsA)-loaded nanofibre scaffolds, observ-
Regeneration Techniques ing an even greater scar suppression and healing
results with the combination of both nanofibres
All these types of stem cells have been used in (MSC and CsA) [32].
various ways in several research projects in order Topical application of a suspension of autolo-
to find the optimal procedure to regenerate the gous ADASC has been reported in an isolated
human corneal stroma. These approaches could clinical case report where authors describe the
be classified into seven basic techniques: healing of a neurotrophic ulcer nonresponsive to
conventional treatment [33]. The lack of further
scientific evidence for this delivery method since
26.3.1 Ocular Surface Implantation 2012 raises questions about its real efficacy.
of Stem Cells Finally, Basu et  al. suggested the delivery of
MSC by the use of fibrin glue [20]. They resus-
Surface implantation of MSC would be the opti- pended CSSC in a solution of human fibrinogen,
mal approach for ocular surface reconstruction and this was added onto a wounded ocular s­ urface
26  Cell Therapy of the Corneal Stroma Using Ex Vivo Cultured Extraocular Cells 407

with thrombin onto the wound bed. Subsequently keratocytes to catabolize accumulated GAG
the fibrinogen gelifies. By this application, they products [11].
demonstrated the prevention of corneal scarring Recently, our group has published the first
in the mice model, together with the generation of clinical trial where the preliminary safety and
new stroma with a collagen organization indistin- efficacy of the cellular therapy of the human cor-
guishable from that of native tissue. Currently, neal stroma are reported [14, 35]. In this pilot
this group is enrolled in a clinical trial to validate clinical trial, we implanted autologous ADASC
these findings, using autologous and heterologous (obtained by elective liposuction) within a mid-­
CSSC from limbal biopsies for cases of chemical stroma femtosecond laser-assisted lamellar
burns, neurotrophic ulcers and stablished scars, pocket in patients with advanced keratoconus
preliminary reporting and improvement in visual (and already candidates for corneal transplanta-
parameters, corneal epithelization, corneal neo- tion). No signs of inflammation or rejection were
vascularization and corneal clarity [34]. observed, confirming all previous evidence
reported in the animal model, and we could also
demonstrate, by anterior segment optical coher-
26.3.2 Intrastromal Implantation ence tomography (AS-OCT), the production of
of Stem Cells Alone new collagen in the area of the MSC implantation
(Fig.  26.2), with a mean central pachymetric
Direct injection of stem cells inside the corneal improvement of 15 μm. This neo-collagen band
stroma of an animal model has been assayed did not induce any clinical haze (Fig. 26.2), and
in  vivo in some studies, demonstrating the dif- all patients moderately improved their visual
ferentiation of the stem cells into adult kerato- function (mean improvement of two lines in all
cytes without signs of immune rejection. In our visual parameters), while keratometric values
study, we also demonstrated by immunohisto- remained stable (Fig.  26.2) [14, 35]. Further
chemistry the production of human extracellular studies with larger samples are required in order
matrix when h-ADASCs were transplanted inside to confirm this preliminary data, but encouraging
the rabbit cornea [3]. As expected, collagens type results were observed, opening a new and excit-
I and type VI were founded to be expressed in the ing line of therapy for research. As we have dis-
rabbits’ corneal stroma as well as in the trans- cussed, the production of new extracellular
planted h-ADASCs, and collagens III and IV, not matrix by the implanted mesenchymal stem cells
expressed normally in the corneal stroma, were occurs, but is not quantitatively enough to be able
detected neither in the host corneal stroma nor in to restore the thickness of a severely diseased
the transplanted h-ADASCs. Du et al. [9] reported human cornea (as in extreme keratoconic cor-
a restoration of the corneal transparency and neas). However, the direct injection of stem cells
thickness in lumican-null mice (thin corneas, may provide a promising treatment modality for
haze and disruption of normal stromal organiza- corneal dystrophies, for corneal stroma progres-
tion) 3 months after an intrastromal transplant of sive opacification in the context of systemic met-
human CSSCs. They also confirmed that human abolic disorders and for the modulation of corneal
keratan sulphate was deposited in the mouse scarring.
stroma and the host collagen lamellae were reor-
ganized, concluding that delivery of h-CSSCs to
scarred human stroma may alleviate corneal scars 26.3.3 Intrastromal Implantation
without requiring surgery [9]. Very similar find- of Stem Cells Together
ings were reported by Liu et  al., using human with a Biodegradable Scaffold
UMSCs in the same animal model [10]. Coulson-­
Thomas et  al. found that, in a mice model for To enhance the growth and development of the
mucopolysaccharidosis, transplanted human stem cells injected into the corneal stroma, trans-
UMSCs participate both in extracellular glycos- plantation together with biodegradable synthetic
aminoglycan (GAG) turnover and enable host extracellular matrixes has been performed.
408 J. L. Alió del Barrio

Axial / Sagittal Curvature (Front)


90º
0º 60
8 9mm 12 º

+1.3
+1.3 +1.5

30
15
4 +1.2 +1.2 +0.9

º
+0.5 +0.2
+0.8 -0.9 -0.9 +0.7

180º
0 -1.4 0.0


-0.3
+0.2 -0.4 -1.5 +0.2
+0.1 -0.9
4 +0.1 -0.7 -0.6


21

33
-0.3 -0.6


-0.5

8 24 0º
0º 30

a b 270º
c
8 4 0 4 8

Fig. 26.2  Autologous h-ADASC corneal stroma implan- plane in the first postoperative month. Stem cell survival
tation for advanced keratoconus. (a) Slit lamp picture 1 is confirmed by the presence of cells showing a more
year after the procedure and (b) topographic changes rounded morphology (white arrows); (d) AS-OCT picture
(Pentacam) between pre-op and 12 months after surgery. 1 year postoperatively. Note the patched hyper-reflective
Observe the stability of the keratometric parameters; (c) areas (red arrows) at the level of the stromal pocket com-
corneal confocal biomicroscopy pictures at the surgical patible with areas of new collagen production

Espandar et al. injected h-ADASCs with a semi- engineering methods. The major obstacle to the
solid hyaluronic acid hydrogel into rabbit corneal production of a successfully engineered cornea is
stroma, reporting better survival and keratocyte the difficulty with reproducing (or at least simu-
differentiation of the h-ADASCs when compared lating) the stromal architecture. The majority of
to their injection alone [6]. Ma et al. used rabbit stromal analogs for tissue-engineered corneas
ADSCs with a poly(lactic-co-glycolic) (PLGA) have been created by seeding human corneal
biodegradable scaffold in a rabbit model of stro- stromal cells into collagen-based scaffoldings,
mal injury, observing newly formed tissue with which are apparently designed to be remodelled
successful collagen remodelling and less stromal [2]. In addition, new and improved biomaterials
scarring [36]. Initial data shows that these scaf- compatible with human corneas have been devel-
folds could enhance stem cell effects over cor- oped leading to advanced scaffolds that can be
neal stroma, although more research is required. used to engineer an artificial cornea (keratopros-
thesis), such as poly(hydroxyethyl methacrylate)
hydrogels, collagen chondroitin sulphate hydro-
26.3.4 Intrastromal Implantation gels and polyurethanes [37]. The combination of
of Stem Cells with a Non-­ these scaffolds with cells can generate promising
biodegradable Scaffold corneal equivalents, and some studies have
already been published that use mainly corneal
At the present time, no clinically viable human cell lines, including stromal keratocytes, pro-
corneal equivalents have been produced by tissue viding positive results regarding adhesion and
26  Cell Therapy of the Corneal Stroma Using Ex Vivo Cultured Extraocular Cells 409

cellular survival in  vitro [38]. However, limited 26.3.5 Intrastromal Implantation
research has been done in  vivo and using stem of Stem Cells
cells in combination with this kind of synthetic with a Decellularized Corneal
scaffolds. Mimura et  al. used corneal fibroblast Stroma Scaffold
precursors together with porous gelatin hydro-
gels in  vivo in a rabbit model, detecting an The complex structure of the corneal stroma has
increased expression of type I collagen, but the not been yet replicated, and there are well-known
authors later stated that these scaffolds were drawbacks to the use of synthetic scaffold-based
weak and unstable [39]. In a previous study by designs: strong inflammatory responses induced
our group, we analysed the survival and biointe- upon their biodegradation, and nearly all polymer
gration of scaffolds composed by macroporous materials cause a nonspecific inflammatory
membranes of poly(ethyl acrylate) (PEA) colo- response.
nized by h-ADASCs, inside the rabbit corneal Recently, several corneal decellularization
stroma in  vivo [4]. Despite demonstrating stem techniques have been described, which provide an
cell survival in vivo after 12 weeks, h-ADASCs acellular corneal extracellular matrix (ECM) [40].
did not improve the extrusion rate of the biomate- These scaffolds have gained attention in the last
rial when compared with PEA membranes few years as they provide a more natural environ-
implanted without stem cells. Our opinion is that ment for the growth and differentiation of cells
stem cells do not differentiate properly into kera- when compared with synthetic scaffolds. In addi-
tocytes in the presence of these synthetic bioma- tion, components of the ECM are generally con-
terials. They lose the potential benefits served among species and are tolerated well even
demonstrated with biodegradable scaffolds or by by xenogeneic recipients. Moreover, keratocytes
isolated injection and fail to encroach on the bio- are essential for remodelling the corneal stroma
material despite a macroporous structure and for the normal epithelial physiology [41].
(Fig. 26.3). This highlights the importance of transplanting a

a b c

d e

Fig. 26.3 h-ADASC colonized poly(ethyl acrylate) of PEA graft within the corneal stroma 3 months after sur-
(PEA) membrane grafted within rabbit’s corneal stroma gery; (d) histological sections stained with Masson’s tri-
in vivo; (a) SEM image of a poly(ethyl acrylate) (PEA) chrome. Observe the absence of a real biointegration of
macroporous membrane (scale bar 200 mm); (b) fluores- the implant that detaches from the surrounding tissue
cence of Vybrant CM-DiI labelled h-ADASC (white along the histological processing; (e) PEA graft with
arrows, magnification 400×); (c) anterior segment picture extrusion 3 months after surgery
410 J. L. Alió del Barrio

cellular substitute together with the structural sup- (Fig.  26.4a), without observing any rejection
port (acellular ECM) to undertake these critical response despite the graft being xenogeneic [5].
functions in corneal homeostasis. To the best of We also demonstrated the differentiation of
our knowledge, all attempts to repopulate decel- h-ADASCs into functional keratocytes inside
lularized corneal scaffolds have used corneal cells these implants in vivo, which then achieved their
[42–44], but as already discussed, these cells have proper biofunctionalization (Fig. 26.4b, c). Other
significant drawbacks that limit their autologous authors have later assayed the integration of
use in clinical practice (damage of the donor tis- decellularized pig articular cartilage ECM colo-
sue, lack of cells and more difficult cell subcul- nized with mice BM-MSC into the rabbit’s cor-
tures), thus the efforts to find an extraocular neal stroma in  vivo, reporting similar findings,
source of autologous cells. In a previous study by although the transparency of this decellularized
our group, we showed the perfect biointegration scaffolds was not clearly reported [45].
of human decellularized corneal stromal sheets In the author’s opinion, the implantation of
(100 μm thickness) with and without h-ADASC MSC together with decellularized corneal ECM
colonization inside the rabbit cornea in  vivo would be the best technique to effectively restore

b c

Fig. 26.4  Corneal stroma enhancement with decellular- labelled with CM-DiI, around and inside the implant con-
ized human corneal stroma with h-ADASC recellulariza- firming the presence of living human cells inside the rab-
tion in the rabbit animal model. (a) Haematoxylin-eosin bit corneal stroma; and (c) same section showing human
staining of a rabbit cornea with an implanted graft of keratocan and thus their differentiation into human kera-
decellularized human corneal stroma with h-ADASC col- tocytes (arrows) (magnification 400×). Abbreviations:
onization (magnification 200×); (b) human cells (arrows), Epi, epithelium; Str, stroma
26  Cell Therapy of the Corneal Stroma Using Ex Vivo Cultured Extraocular Cells 411

the thickness of a diseased and severely weak- anatomically restoring these advanced kerato-
ened human cornea, as the implantation of MSC conic corneas (Fig. 26.5e). Although no clinical
alone only achieves a limited new ECM forma- differences could be demonstrated between
tion and thickness restoration [3, 14]. Moreover, decellularized and recellularized implants, the
through this technique, and by using autologous host keratocyte recellularization of the implants
MSC from a given patient, it is theoretically pos- and increase in the keratocyte density within the
sible to transform allogenic grafts into functional anterior and posterior stroma were more intense
autologous grafts, thus avoiding any risk of rejec- in those patients where autologous ADASCs
tion. Following this research line, we have were also implanted (Fig. 26.5c, d) [15, 35].
recently published the first clinical trial using
these decellularized human corneal stroma scaf-
folds (120  μm thickness and 9.0  mm diameter 26.3.6 Anterior Chamber Injection
laminas), with or without autologous ADASC of Stem Cells
recellularization, in patients with advanced kera-
toconus [15, 35]. We could demonstrate a moder- Demirayak et  al. reported that BM-MSC and
ate but significant improvement in all visual ADASC, suspended in phosphate-buffered solu-
parameters (about two lines of improvement), tion (PBS) and injected into the anterior cham-
together with a refractive sphere reduction, sig- ber after a penetrating corneal injury in a mice
nificant anterior keratometric flattening model, are able to colonize the corneal stroma
(Fig.  26.5f) and improvement of corneal aber- and increase the expression of keratocyte-spe-
rometry, especially the spherical aberration [15, cific markers as keratocan, with an increase in
35]. A complete corneal transparency without the keratocyte density by confocal microscopy
significant haze was obtained 3 months after the [8]. However, it is very questionable the possible
procedure (Fig.  26.5a, b), and all pachymetric side effects of this MSC injection into the ante-
parameters improved over 100  μm as expected, rior chamber for the lens epithelium and trabecu-

a b c d

e
Axial / Sagittal Curvature (Front)

f 0º
90º
60
8 9mm 12 º

-0.4
-0.8 +2.2

30
15

4 +2.7 -2.5 +2.7


º

-0.7 +1.2
+3.2 -3.7 -1.1 +1.7
180º

0 +1.5 -3.9 -0.1


+1.6 -1.5 -4.0 +0.2


-1.2 -3.1
4 -0.2 -5.0 -0.4

21

-2.5 -2.2
33

-2.8

8 24 0º
0º 30
270º
8 4 0 4 8

Fig. 26.5  Corneal stroma enhancement with decellular- showed a complete acellular pattern 3  months after sur-
ized human corneal stroma with or without autologous gery; (d) intense recellularization signs of the implanted
h-ADASC recellularization. (a, b) Slit lamp pictures lamina 12 months after surgery; (e) AS-OCT image 1 year
1  week (a) and 3  months (b) after surgery. Observe the after corneal stroma enhancement; and (f) topographic
complete corneal transparency restoration; (c) corneal changes between pre-op and 12  months after surgery.
confocal biomicroscopy picture. The implanted lamina Observe the significant flattening of the keratometry
412 J. L. Alió del Barrio

lar meshwork, where it may induce scarring and stroma, despite being a xenogeneic transplant,
a subsequent glaucoma. Considering this, the no signs of rejection or inflammation have never
potential clinical use of this approach, in the been reported [3–12]. This is congruent with the
author’s opinion, might be limited. strong evidence regarding the immunomodula-
tory and immunosuppressive properties of MSC
that aid them in evading and surviving from host
26.3.7 Intravenous Injection of  immune rejection by inhibiting adhesion and
Stem Cells invasion, and inducing cell death of inflamma-
tory cells, partially thanks to a rich extracellular
The systemic use, by intravenous injection, of glycocalyx which contains tumour necrosis
MSC has also been tested. Intravenous mice factor-α-­ stimulated gene 6 (TSG6) [13, 48].
BM-MSC injected after an allograft corneal TSG6 has been demonstrated to play a critical
transplant has been shown to be able to colonize role in the immunosuppressive properties of
the transplanted cornea and conjunctiva (inflamed MSC [12, 29, 30]. Considering all this, the use
ocular tissues), but not the contralateral ungrated of heterologous MSC would simplify tremen-
cornea, simultaneously decreasing immunity and dously the clinical application of MSC, as clini-
significantly improving allograft survival rate cal application centres would not need any
[46]. Yun et al. recently reported similar findings specific equipment as potential MSC banks
with the intravenous injection of iPSC-derived could store and supply the stem cells for their
MSC and BM-MSC after a surface chemical use in patients. Actually, there are already avail-
injury, where they could observe that the corneal able low-cost systems capable of enhancing the
opacity, inflammatory infiltration and inflamma- preservation of MSC at hypothermic tempera-
tory markers in the cornea were markedly tures while maintaining their normal function,
decreased in those treated mice, without signifi- allowing to extend the time window for distribu-
cant differences between both MSC types [29]. tion between the sites of manufacture and the
In contrast, our group could not find any benefit clinic and reducing the wastage associated with
in corneal allograft survival and rejection rates the limited shelf life of cells stored in their liq-
after systemic injection of rabbit ADASC prior to uid state [49]. In addition, Funderburgh et  al.
surgery, during surgery and at various time points recently reported that MSC from different
after surgery in rabbits with vascularized corneas donors may have different immunosuppressive
(model more similar to human corneal trans- properties and, in consequence, different abili-
plants than those reported in mice). Actually, a ties to regenerate and relieve stromal scars [50].
shorter graft survival compared with the non-­ Considering this important finding, best donors
treated corneal grafts was observed [47]. could be selected in MSC banks in order to
expand and supply only those MSC with the
highest immunosuppressive and regenerative
26.4 Autologous Versus capacity, being then not necessary to use autolo-
Heterologous MSC gous cells. We should also consider that adult
keratocytes obtained from autologous MSC
A critical question for future clinical trials to may still carry the same genetic defect that leads
further assess the feasibility of the cellular ther- to the corneal disease in case of corneal dystro-
apy of the corneal stroma is whether the use of phies. In this scenario, the use of heterologous
autologous MSC is really necessary or, on the instead of autologous MSC would be interest-
other hand, heterologous MSC could achieve ing. Actually, a recent study observed gene
the same benefit without still adding any risk of expression differences between the iPSC-
inflammation or rejection. If we consider all derived keratocytes generated from fibroblasts
published evidence in the animal model where of both keratoconic and normal human cor-
human MSC was implanted into the corneal neal stromata, influencing cellular growth and
26  Cell Therapy of the Corneal Stroma Using Ex Vivo Cultured Extraocular Cells 413

proliferation, confirming that, at least in kerato- centres for their use are the future for this promis-
conus cases, adult cells obtained from MSC ing treatment modality, although before we get
may still not be functionally normal [51]. there, there is still a lot of research work to do.

Compliance with Ethical Requirements Conflict of


26.5 MSC Exosomes interest: The authors, their families, their employers and
their business associates have no financial or proprietary
interest in any product or company associated with any
Exosomes are nanosized extracellular vesicles device, instrument or drug mentioned in this article. The
which originate from the fusion of intracellular authors have not received any payment as consultants,
multivesicular bodies with cell membrane and reviewers or evaluators of any of the devices, instruments
or drugs mentioned in this article. All procedures followed
are released into the extracellular spaces [48]. were in accordance with the ethical standards of the
They have been implicated in the ability of MSC responsible committee on human experimentation (insti-
to repair damaged tissue. Funderburgh et  al. tutional and national) and with the Helsinki Declaration of
have recently shown that exosomes isolated 1975, as revised in 2000. Informed consent was obtained
from all patients for being included in the study. No ani-
from the culture media of human CSSC had mal studies were carried out by the author for this
similar immunosuppressive properties and also chapter.
significantly reduced stromal scarring in
wounded corneas in  vivo in the animal model
[52]. This finding suggests that for some dis- References
eases, as prevention or reduction of cornel scars,
MSC exosomes may allow a non-cell-based 1. De Miguel MP, Casaroli-Marano RP, Nieto-Nicolau
therapy [50]. Zhang et  al. suggested the exo- N, et  al. Frontiers in regenerative medicine for cor-
nea and ocular surface. In: Rahman A, Anjum S, edi-
somes released by transplanted UCMSC in the tors. Frontiers in stem cell and regenerative medicine
diseased cornea are able to enter into host cor- research, vol. 1. 1st ed. Cambridge, PA: Bentham
neal keratocytes and endothelial cells and e-Books; 2015. p. 92–138.
enhance their functions [48]. Actually, in their 2. Ruberti JW, Zieske JD. Prelude to corneal tissue engi-
neering—gaining control of collagen organization.
in  vitro experiment in mucopolysaccharidosis Prog Retin Eye Res. 2008;27:549–77.
VII mice, they discovered that UCMSC secreted 3. Arnalich-Montiel F, Pastor S, Blazquez-Martinez
exosomes assisted in the recycling process of A, et  al. Adipose-derived stem cells are a source
accumulated GAGs in the lysosomes in diseased for cell therapy of the corneal stroma. Stem Cells.
2008;26:570–9.
cells [11]. These findings open as well an excit- 4. Alió del Barrio JL, Chiesa M, Gallego Ferrer G,
ing new field for research as the use of exosomes et  al. Biointegration of corneal macroporous mem-
per se could overcome some of the limitations branes based on poly(ethyl acrylate) copolymers in an
and risks associated with the intrastromal cellu- experimental animal model. J Biomed Mater Res A.
2015;103:1106–18.
lar injection, as exosomes can potentially be 5. Alió del Barrio JL, Chiesa M, Garagorri N, et  al.
applied topically [27]. Acellular human corneal matrix sheets seeded with
In conclusion, cellular therapy of the corneal human adipose-derived mesenchymal stem cells inte-
stroma is a novel treatment modality for stromal grate functionally in an experimental animal model.
Exp Eye Res. 2015;132:91–100.
diseases that still requires further studies in form 6. Espandar L, Bunnell B, Wang GY, et  al. Adipose-­
of clinical trials with larger samples in order to derived stem cells on hyaluronic acid-derived
definitely stablish its safety and efficacy for the ­scaffold: a new horizon in bioengineered cornea. Arch
different stromal diseases; the initial data Ophthalmol. 2012;130:202–8.
7. Mittal SK, Omoto M, Amouzegar A, et al. Restoration
obtained from the first pilot clinical trials is very of corneal transparency by mesenchymal stem cells.
encouraging, showing very exciting data with Stem Cell Rep. 2016;7:583–90.
great potential. In the author’s opinion, the cre- 8. Demirayak B, Yüksel N, Çelik OS, et  al. Effect
ation of cellular banks storing and expanding het- of bone marrow and adipose tissue-derived mes-
enchymal stem cells on the natural course of cor-
erologous cells or their exosomes and their neal scarring after penetrating injury. Exp Eye Res.
shipping and delivery to the different clinical 2016;151:227–35.
414 J. L. Alió del Barrio

9. Du Y, Carlson EC, Funderburgh ML, et al. Stem cell of murine mesenchymal stem cells into corneal-like
therapy restores transparency to defective murine cor- cells. Stem Cells Dev. 2016;25(11):874–81.
neas. Stem Cells. 2009;27:1635–42. 25. Liu H, Zhang J, Liu CY, Hayashi Y, Kao WW. Bone
10. Liu H, Zhang J, Liu CY, et al. Cell therapy of congeni- marrow mesenchymal stem cells can differentiate
tal corneal diseases with umbilical mesenchymal stem and assume corneal keratocyte phenotype. J Cell Mol
cells: lumican null mice. PLoS One. 2010;5:e10707. Med. 2012;16:1114–24.
11.
Coulson-Thomas VJ, Caterson B, Kao 26. Du Y, Roh DS, Funderburgh ML, et  al. Adipose-­

WW. Transplantation of human umbilical mesenchy- derived stem cells differentiate to keratocytes in vitro.
mal stem cells cures the corneal defects of mucopoly- Mol Vis. 2010;16:2680–9.
saccharidosis VII mice. Stem Cells. 2013;31:2116–26. 27. Ziaei M, Zhang J, Patel DV, McGhee CNJ. Umbilical
12. Kao WW, Coulson-Thomas VJ. Cell therapy of cor- cord stem cells in the treatment of corneal disease.
neal diseases. Cornea. 2016;35(Suppl 1):S9–S19. Surv Ophthalmol. 2017;62(6):803–15.
13. De Miguel MP, Fuentes-Julián S, Blázquez-Martínez 28. Chan AA, Hertsenberg AJ, Funderburgh ML, et  al.
A, et al. Immunosuppressive properties of mesenchy- Differentiation of human embryonic stem cells into
mal stem cells: advances and applications. Curr Mol cells with corneal keratocyte phenotype. PLoS One.
Med. 2012;12:574–91. 2013;8:e56831.
14. Alió del Barrio JL, El Zarif M, de Miguel MP, et al. 29. Yun YI, Park SY, Lee HJ, Ko JH, Kim MK, Wee
Cellular therapy with human autologous adipose-­ WR, et  al. Comparison of the anti-inflammatory
derived adult stem cells for advanced keratoconus. effects of induced pluripotent stem cell-derived and
Cornea. 2017;36(8):952–60. bone marrow-derived mesenchymal stromal cells
15. Alió Del Barrio JL, El Zarif M, Azaar A, et  al.
in a murine model of corneal injury. Cytotherapy.
Corneal stroma enhancement with decellularized 2017;19(1):28–35.
stromal laminas with or without stem cell recellular- 30. Di G, Du X, Qi X, Zhao X, Duan H, Li S, et  al.
ization for advanced keratoconus. Am J Ophthalmol. Mesenchymal stem cells promote diabetic corneal
2018;186:47–58. epithelial wound healing through TSG-6-dependent
16. Harkin DG, Foyn L, Bray LJ, Sutherland AJ, Li
stem cell activation and macrophage switch. Invest
FJ, Cronin BG.  Concise reviews: can mesenchy- Ophthalmol Vis Sci. 2017;58(10):4344–54.
mal stromal cells differentiate into corneal cells? 31. Holan V, Trosan P, Cejka C, Javorkova E, Zajicova
A systematic review of published data. Stem Cells. A, Hermankova B, et al. A comparative study of the
2015;33(3):785–91. therapeutic potential of mesenchymal stem cells and
17. Jiang Z, Liu G, Meng F, Wang W, Hao P, Xiang Y, limbal epithelial stem cells for ocular surface recon-
et  al. Paracrine effects of mesenchymal stem cells struction. Stem Cells Transl Med. 2015;4(9):1052–63.
on the activation of keratocytes. Br J Ophthalmol. 32. Cejka C, Cejkova J, Trosan P, Zajicova A, Sykova
2017;101(11):1583–90. E, Holan V. Transfer of mesenchymal stem cells and
18. Hendijani F. Explant culture: an advantageous method cyclosporine A on alkali-injured rabbit cornea using
for isolation of mesenchymal stem cells from human nanofiber scaffolds strongly reduces corneal neovas-
tissues. Cell Prolif. 2017;50(2):e12334. cularization and scar formation. Histol Histopathol.
19. Górski B. Gingiva as a new and the most accessible 2016;31(9):969–80.
source of mesenchymal stem cells from the oral cavity 33. Agorogiannis GI, Alexaki VI, Castana O, Kymionis
to be used in regenerative therapies. Postepy Hig Med GD.  Topical application of autologous adipose-­
Dosw. 2016;70(0):858–71. derived mesenchymal stem cells (MSCs) for persis-
20. Basu S, Hertsenberg AJ, Funderburgh ML, Burrow tent sterile corneal epithelial defect. Graefes Arch
MK, Mann MM, Du Y, et al. Human limbal biopsy-­ Clin Exp Ophthalmol. 2012;250(3):455–7.
derived stromal stem cells prevent corneal scarring. 34. Basu S.  Limbal stromal stem cell therapy for acute
Sci Transl Med. 2014;6(266):266ra172. and chronic superficial corneal pathologies: early
21. Takahashi K, Yamanaka S.  Induction of pluripotent clinical outcomes with the Funderburgh technique.
stem cells from mouse embryonic and adult fibroblast Oral presentation at The Association for Research in
cultures by defined factors. Cell. 2006;126:663–76. Vision and Ophthalmology (ARVO) annual meeting,
22. Naylor RW, McGhee CN, Cowan CA, Davidson
Baltimore; 2017.
AJ, Holm TM, Sherwin T.  Derivation of corneal 35. Alió JL, Alió Del Barrio JL, Azaar A, et al.

keratocyte-­like cells from human induced pluripotent Regenerative Medicine of the Corneal Stroma for
stem cells. PLoS One. 2016;11(10):e0165464. Advanced Keratoconus: One Year Outcomes. Am J
23. Park SH, Kim KW, Chun YS, Kim JC. Human mes- Ophthalmol 2018 [In peer review process].
enchymal stem cells differentiate into keratocyte-like 36. Ma XY, Bao HJ, Cui L, Zou J. The graft of autologous
cells in keratocyte-conditioned medium. Exp Eye adipose-derived stem cells in the corneal stromal after
Res. 2012;101:16–26. mechanic damage. PLoS One. 2013;8:e76103.
24. Trosan P, Javorkova E, Zajicova A, Hajkova M,
37. De Miguel MP, Alio JL, Arnalich-Montiel F, et  al.
Hermankova B, Kossl J, et  al. The supportive role Cornea and ocular surface treatment. Curr Stem Cell
of insulin-like growth factor-I in the differentiation Res Ther. 2010;5:195–204.
26  Cell Therapy of the Corneal Stroma Using Ex Vivo Cultured Extraocular Cells 415

38. Hu X, Lui W, Cui L, Wang M, Cao Y. Tissue engineer- 46. Omoto M, Katikireddy KR, Rezazadeh A, Dohlman
ing of nearly transparent corneal stroma. Tissue Eng. TH, Chauhan SK. Mesenchymal stem cells home to
2005;11:1710–7. inflamed ocular surface and suppress allosensitization
39. Mimura T, Amano S, Yokoo S, et al. Tissue engineer- in corneal transplantation. Invest Ophthalmol Vis Sci.
ing of corneal stroma with rabbit fibroblast precursors 2014;55(10):6631–8.
and gelatin hydrogels. Mol Vis. 2008;14:1819–28. 47. Fuentes-Julián S, Arnalich-Montiel F, Jaumandreu L,
40. Lynch AP, Ahearne M.  Strategies for develop-
Leal M, Casado A, García-Tuñon I, et  al. Adipose-­
ing decellularized corneal scaffolds. Exp Eye Res. derived mesenchymal stem cell administration does
2013;108:42–7. not improve corneal graft survival outcome. PLoS
41. Wilson SE, Liu JJ, Mohan RR.  Stromal-epithelial
One. 2015;10(3):e0117945.
interactions in the cornea. Prog Ret eye Res. 48. Zhang L, Coulson-Thomas VJ, Ferreira TG, Kao

1999;18:293–309. WW.  Mesenchymal stem cells for treating ocular
42. Choi JS, Williams JK, Greven M, et al. Bioengineering surface diseases. BMC Ophthalmol. 2015;15 Suppl
endothelialized neo-corneas using donor-derived 1:155.
corneal endothelial cells and decellularized corneal 49. Swioklo S, Constantinescu A, Connon CJ. Alginate-­
stroma. Biomaterials. 2010;31:6738–45. encapsulation for the improved hypothermic preserva-
43. Shafiq MA, Gemeinhart RA, Yue BY, Djalilian
tion of human adipose-derived stem cells. Stem Cells
AR. Decellularized human cornea for reconstructing Transl Med. 2016;5(3):339–49.
the corneal epithelium and anterior stroma. Tissue 50. Funderburgh JL.  Assessing the potential of stem

Eng Part C Methods. 2012;18:340–8. cells to regenerate stromal tissue. Oral presenta-
44. Gonzalez-Andrades M, de la Cruz CJ, Ionescu AM, tion at The Association for Research in Vision and
Campos A, Del Mar PM, Alaminos M. Generation of Ophthalmology (ARVO) annual meeting, Baltimore;
bioengineered corneas with decellularized xenografts 2017.
and human keratocytes. Investig Ophthalmol Vis Sci. 51. Joseph R, Srivastava OP, Pfister RR. Modeling kera-
2011;52:215–22. toconus using induced pluripotent stem cells. Invest
45. Bai H, Wang LL, Huang YF, Huang JX.  An experi- Ophthalmol Vis Sci. 2016;57:3685–97.
mental study of mesenchymal stem cells in tissue 52. Shojaati G.  Regenerative Potential of Stem cell-­

engineering scaffolds implanted in rabbit corneal Derived Exosomes. Oral presentation at The
lamellae to increase keratoprosthesis biointegration. Association for Research in Vision and Ophthalmology
Zhonghua Yan Ke Za Zhi. 2016;52(3):192–7. (ARVO) annual meeting, Baltimore; 2017.
Part V
Regenerative Surgery of the Corneal
Endothelium
Corneal Endothelium: Applied
Anatomy
27
Francisco Arnalich-Montiel

27.1 Introduction gives rise to hematopoietic stem cells, and to


angioblasts, the precursors of endothelial cells. It
Current knowledge of the development, struc- would be, therefore, more correctly referred to as
ture, and function of corneal endothelium allows the posterior cell layer of the cornea or the poste-
understanding of the effects of external trauma or rior epithelium [2].
ocular and systemic disease on this apparently The cornea is formed as a result of the last
simple monolayer of cells. We will review the series of major inductive events during eye devel-
anatomy, physiology, and immunohistochemistry opment at ∼5–6 weeks of human gestation, when
of human corneal endothelium in this chapter. the surface ectoderm interacts with the lens vesicle
[3]. When the surface ectoderm and the lens vesi-
cle are completely separated, the space between
27.2 Anatomy of the Corneal them is filled with perinuclear neural crest cells
Endothelium (NCCs). These NCCs originate during gastrula-
tion at the neural plate border and migrate from
27.2.1 Embryology folds of the neural ectoderm as the neuroepithe-
lium closes to form the neural tube [4]. In humans,
The major structures of the eye have different the NCC migrates in three waves into the eye,
embryological origins. In particular, the cornea is while in mice and chick, there appears to be only
derived from two distinct embryologic tissues: two waves [5]. In humans, the first wave of NCCs
the corneal epithelium from epidermal ectoderm migrates into the space between the anterior sur-
and the rest from neural crest cells [1]. face of the lens and the surface ectoderm destined
Developmentally, the human corneal endothe- to form the corneal endothelium [5]. A second
lium cannot be considered true endothelium, as wave of cells migrates between the corneal epithe-
its origin is quite distinct from that of vascular lium and endothelium to become the keratinocytes
endothelium. The latter derives from the heman- of the corneal stroma. The corneal epithelium syn-
gioblast, a mesoderm-derived progenitor cell that thesizes components of the extracellular matrix for
the formation of primary stroma when the lens
vesicle detaches from the surface ectoderm, while
a third wave of NCCs migrates to the angle
between the posterior cornea (endothelium) and
F. Arnalich-Montiel (*) the anterior edge of the optic cup, eventually con-
Vissum Corporation, Madrid, Spain tributing to the ciliary body and iris stroma [5].

© Springer Nature Switzerland AG 2019 419


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_27
420 F. Arnalich-Montiel

The first wave of NCCs forms a loosely endothelial cells starting at 8  weeks. Secondly
arranged sheet of overlapping cells, and by the there is a prenatal differentiation into a striated
8th week of human gestation, it thins to a mono- basement membrane through the gradual addi-
layer of orderly arranged cuboidal cells [1]. tion of short and thin cross-linking bridges sepa-
Around the 78-mm stage, these cells flatten and rated by 110-nm intervals that are disposed in a
become tightly connected to one another by tight plane perpendicular to the lamellae. This struc-
junctions. By the 120- and 165-mm stages, the ture has a distinctive banded appearance when
endothelial monolayer is of uniform thickness, viewed by electron microscopy and is approxi-
spans the entire posterior corneal surface, and mately 3 μm thick at the time of birth. Finally in
fuses peripherally with the trabecular beams. The the postnatal period, the membrane continues to
endothelial cells stay arrested in the G1-phase of grow in thickness by deposition of a nonstriated,
mitosis due to different factors that include lack nonlamellar material posterior to the striated pre-
of response to positive mitogenic stimulation natal layer, which has an amorphous texture
in  vivo, suppression of the cell cycle entry by when viewed by electron microscopy, reaching
TGF-beta2 in aqueous humor, and contact inhibi- up to 10 μm thick.
tion induced by the formation of mature cell–cell
junctions [6]. At birth, the endothelial monolayer
is approximately 10 μm thick [7]. 27.2.2 Morphology
Immediately anterior to the flattened layer is a
discontinuous homogeneous acellular layer, Human corneal endothelium consists on a mono-
which in time becomes Descemet’s membrane layer of flat cells on an amorphous collagenous
[3]. Three major processes can be distinguished membrane (Descemet’s membrane). These cells
during the formation of Descemet’s membrane adopt a honeycomb-like mosaic when viewed
[8]. First, there is a growth in thickness by lamel- from the aqueous side, showing a hexagonal pat-
lar deposition during the prenatal period by the tern in healthy corneas (see Fig. 27.1). At birth,

Auto
Number 61 SD 230 um2
CV 59 %
CD 2558 /mm2
Max 1420 um2
AVG 391 um2 Min 117 um2
CCT 577 um

Area (Polymegathism) Apex (Pleomorphism)


000-100 um2 0% 3 0%
100-200 um2 20 % 4 6%
200-300 um2 20 % 5 29 %
300-400 um2 20 % 6A 24 %
400-500 um2 21 % 7 24 %
500-600 um2 5% 8 12 %
600-700 um2 8% 9 6%
700-800 um2 2% 10- 0%
800-900 um2 2%
900- um2 3%

Fig. 27.1  Normal human corneal endothelium as seen with specular microscopy. CD cell count density, Avg average
cell area in microns, CCT central corneal thickness, CV coefficient of variation
27  Corneal Endothelium: Applied Anatomy 421

the endothelial monolayer is approximately normal corneas. A minimal numerical density of


10  μm thick and tends to flatten over time and 400–500 cells/mm2 is required to sustain the
stabilize at 4 μm [1]. The cellular morphological pumping activity of the endothelium [10].
features evolve with aging, as some cells grow
and others disappear; the mosaic shows poly-
megathism (different cell size) and polymor- 27.3 Physiology of Corneal
phism (different geometric cell morphology other Endothelium
than a hexagon).
They are highly polarized cells, with an apical The corneal endothelium has three major roles; it
surface devoided of surface villi, unlike the epi- functions as a barrier and as a water pump, and
thelium. The interaction between tight junctions also it is implicated in the production of
and a submembranous network of actomyosin is Descemet’s membrane.
responsible for the hexagonality of the apical sur- The morphology and arrangement of the cor-
face. The lateral membranes, which support neal endothelium are optimized to fulfill its main
enzymatic pumps such as the Na+/K+ -adenosi- physiological function that of acting as a regula-
netriphosphatase (ATPase) pump, presented tor of the entrance and exit of fluids to or from the
extensive lateral interdigitations between adja- cornea. In order to optimize the barrier function,
cent cells [9]. The basal side of the endothelial the hexagonal shape of the cells is the most effi-
membrane contains numerous hemidesmosomes cient to cover the surface of the posterior surface
that promote adhesion to Descemet’s membrane. of the cornea without leaving gaps between them
During life, endothelial cell density and hex- [7]. A hexagonal shape also minimizes the area
agonality decline (Fig. 27.2). At birth endothelial that the individual cell must occupy, allowing the
cell count is over 4000 cells/mm2, with a rapid maximum number of cells per mm2 and thus
decline in the first 2 years of an 8% likely related maximizing pump site density.
to growth in corneal diameter and hence surface
area [10]. From the second to eighth decades of
life, the cell density declines to around 2600 27.3.1 Barrier Function
cells/mm2 [11]. At the same period of time, the
percentage of hexagonal cells declines from The endothelial monolayer functions as a per-
approximately 75% to approximately 60% [12]. meability barrier that restricts the movement of
In this period, the central endothelial cell density water and solutes into the hydrophilic stroma
decreases at an average rate of 0.6% per year in due to the presence of intercellular gap junc-

3000-4000 cell/mm2 2500 cell/mm2 <2000 cell/mm2

18µ > 40 µm
Fig. 27.2  Changes in
corneal endothelial cell From 2° to 8th
density and size At birth decade Old age
throughout life
422 F. Arnalich-Montiel

tions (GP) and tight junctions (TJ) at the apical 2. Intracellular carbonic anhydrase pathway
membrane of endothelial cells. However, unlike [14]. Carbon dioxide also diffuses into the
the corneal epithelium, which forms a tight bar- cytoplasm of the endothelial cells, and in
rier, these junctions do not provide a complete combination with water, the carbonic anhy-
seal around the cell, so there is a constant leak drase produces bicarbonate ions that diffuse
of aqueous humor into the stroma at a slow rate. or are transported into the aqueous humor
This constant leak also provides the major coupled with an efflux of water.
source of nutrients for cells of the nonvascular
cornea [1].
27.3.3 Secretory Function

27.3.2 Pump Function Lastly, endothelial cells also have a secretory func-
tion and are responsible for Descemet’s membrane
The endothelial monolayer has a pump activity to formation and postnatal development. They
generate a state of relative stromal deturgescence secrete collagen type VIII which is an important
(78% water content). This state of stromal dehy- component of Descemet’s membrane [15].
dration allows the necessary arrangement of the
collagen fibers and glycosaminoglycans for tis-
sue transparency [13]. 27.4 Immunohistochemistry
Therefore, endothelial activity consists on a of Corneal Endothelium
pump-leak mechanism in which there is dynamic
equilibrium between two opposite forces: (1) a Although the endothelial layer is only 5 μm thick,
tendency of the stroma to swell due to the imbibi- there is a great three-dimensional complexity due
tion pressure (60 mm Hg) produced by the pro- to the high polarization of the cell [16]. The CEC
teoglycan matrix that surrounds each collagen when seen from the apical surface is usually seen
fiber of the stroma [1] and (2) a net outflow of as flat cells with a polygonal pattern resembling
fluid from the corneal stroma into the aqueous terra-cotta tiles. On the basal side, however, there
humor driven by ionic gradients located in the are numerous interdigitations between these cells
basolateral side of the membrane. Although the resembling more like an amoebic pattern.
net outflow of fluid from the corneal stroma into The classic molecular markers are used to
the aqueous humor is passive and it does not characterize CEC such as Zonula occludens-1
require energy consumption, it does depend on (ZO-1), aquaporin 1, Na+/K+ pump, or neuron-­
aerobic metabolism to maintain the favorable specific enolase [17]. More recently CD 200 (a
osmotic gradient that allows this by means of member of the immunoglobulin superfamily of
intracellular and membrane ion transport proteins), glypican-4, CD 166 (an activated leu-
systems. kocyte cell adhesion protein), and the antioxidant
The two main ion transport systems responsi- enzyme peroxiredoxin-6 have been also used to
ble for the osmotic gradient which allows move- characterize corneal endothelium [18, 19].
ment of fluid from a relatively hypoosmotic However most of them have been found to be
corneal stroma toward a relatively hypertonic nonspecific as they can be expressed in other cor-
aqueous humor are: neal cells, as well as other ocular and non-ocular
tissues. In a recent study, N-cadherin was the
1. Membrane-bound Na+/K+ -ATPase sites only one of all the targets that was found exclu-
[14]. An osmotic gradient of sodium is present sively in CECs and not in other corneal cells [16].
between the aqueous humor and the stroma Nevertheless the subcellular distribution features
and results in the influx of sodium ions from of some of these proteins can be used as specific
the aqueous humor and in an efflux of potas- markers to distinguish these cells from the other
sium ions to the aqueous humor. corneal cell types (see Table 27.1).
27  Corneal Endothelium: Applied Anatomy 423

Table 27.1  Endothelial specific 3D mapping proteins


Fluorescence pattern
Target protein Function Endothelium Epithelium Stroma
Zonula occludens (ZO-1) Tight junction Zigzag line at the apical Continuous Cytoplasm
complex surface interrupted at the Y regular straight
junction line
ATPase, Na+/K+ transporting Active pump Flower shaped at the lateral Cell to cell Cytoplasm
beta 1 polypeptide membrane junction
Actin Microfilaments Submembranous network at Cytoplasm Cytoplasm
the apical surface exclusively
Myosin IIa Microfilaments Submembranous network at Cytoplasm Cytoplasm
the apical surface exclusively
Integrin α3β1 Cell adhesion Polarization at the basal Cell to cell Absent
surface junction
N-cadherin Cell adhesion Flower shaped at the lateral Absent Absent
membrane
Activated leukocyte adhesion Cell adhesion Flower shaped at the lateral Cell to cell Faint in the
molecule (ALCAM or CD membrane junction cytoplasm
166)
Neural cell adhesion Cell adhesion Flower shaped at the lateral Absent Cytoplasm
molecule 1 (NCAM or CD56) membrane
Peroxiredoxin-6 (Prdx-6/ Antioxidant Flower shaped at the lateral Absent Absent
antibody Tag2A12) enzyme membrane

Three differentiated staining patterns can be


identified:

1. The hexagonal apical surface pattern of the


cell can be well characterized by ZO-1 (see
Fig. 27.3), which is distributed along a jigsaw
pattern interrupted at the Y junction between
three cells, and by a submembraous network
of actin and myosin IIa, exclusively located at
the apical surface.
2. The lateral membrane begins immediately
beneath the apical surface with fingerlike pro-
cesses progressively expanding toward
Fig. 27.3  ZO-1 staining for rabbit corneal endothelial
Descemet’s membrane. This side of the cell cells (red) highlighting the hexagonal pattern with a zig-
can be revealed with similar staining patterns zag distribution. Nuclei were counterstained by Hoechst
by targeting N+/k + ATPase, CD166, Prdx-6, 33342 (in blue)
NCAM, and N-cadherin.
3. The basal surface shows a polarization of or by fibroblastic conversion, also known as
integrin α3ß1. endothelial-to-mesenchymal transition (EnMT)
that becomes evident in early passages. In a
Finding a specific marker for endothelial recent study, the markers proposed [20] were
cells is so important and is becoming highly CD56, or neural cell adhesion molecule
important in clinical research. It allows the (NCAM), a surface glycoprotein; CD248, or
identification of high-quality human corneal endosialin which belongs to the family of C-type
endothelial cells in culture, to distinguish this lectin transmembrane receptors; coxsackie-­
subpopulation from other subpopulations that adenovirus receptor (CAR) is part of the junc-
might be contaminated by stromal keratocytes tional adhesion molecules (JAMs); and CD109,
424 F. Arnalich-Montiel

a cell surface antigen involved in the TGF-β 11. Yee RW, Matsuda M, Schultz RO, Edelhauser

HF.  Changes in the normal corneal endothelial cel-
pathway. Another study identified five genes, lular pattern as a function of age. Curr Eye Res.
CLRN1, MRGPRX3, HTR1D, GRIP1, and 1985;4(6):671–8.
ZP4, as novel markers of corneal endothelial 12. Rao SK, Ranjan Sen P, Fogla R, Gangadharan S,
cells, and the specificities of these genes were Padmanabhan P, Badrinath SS.  Corneal endothelial
cell density and morphology in normal Indian eyes.
successfully confirmed by independent experi- Cornea. 2000;19(6):820–3.
ments at both the RNA and protein levels [21]. 13. Geroski DH, Matsuda M, Yee RW, Edelhauser

HF. Pump function of the human corneal endothelium.
Compliance with Ethical Requirements Francisco Effects of age and cornea guttata. Ophthalmology.
Arnalich-Montiel declares no conflict of interest. No 1985;92(6):759–63.
human or animal studies were carried out by the author for 14. Bonanno JA. Identity and regulation of ion transport
this article. mechanisms in the corneal endothelium. Prog Retin
Eye Res. 2003;22(1):69–94.
15. Puk O, Dalke C, Calzada-Wack J, Ahmad N,

Klaften M, Wagner S, et  al. Reduced corneal thick-
References ness and enlarged anterior chamber in a novel
ColVIIIa2G257D mutant mouse. Invest Ophthalmol
1. Tuft SJ, Coster DJ.  The corneal endothelium. Eye Vis Sci. 2009;50(12):5653–61.
(Lond). 1990;4(Pt 3):389–424. 16. He Z, Forest F, Gain P, Rageade D, Bernard A, Acquart
2. Meier S.  Initiation of corneal differentiation S, et al. 3D map of the human corneal endothelial cell.
prior to cornea-lens association. Cell Tissue Res. Sci Rep. 2016;6(1):29047.
1977;184(2):255–67. 17. Okumura N, Hirano H, Numata R, Nakahara M, Ueno
3. Zavala J, López Jaime GR, Rodríguez Barrientos M, Hamuro J, et  al. Cell surface markers of func-
CA, Valdez-Garcia J.  Corneal endothelium: devel- tional phenotypic corneal endothelial cells. Invest
opmental strategies for regeneration. Eye (Lond). Opthalmol Vis Sci. 2014;55(11):7610.
2013;27(5):579–88. 18. Cheong YK, Ngoh ZX, Peh GSL, Ang H-P, Seah X-Y,
4. Beebe DC, Coats JM. The lens organizes the anterior Chng Z, et  al. Identification of cell surface markers
segment: specification of neural crest cell differentia- glypican-4 and CD200 that differentiate human cor-
tion in the avian eye. Dev Biol. 2000;220(2):424–31. neal endothelium from stromal fibroblasts. Invest
5. Gage PJ, Rhoades W, Prucka SK, Hjalt T. Fate maps Opthalmol Vis Sci. 2013;54(7):4538–47.
of neural crest and mesoderm in the mammalian eye. 19. Ding V, Chin A, Peh G, Mehta JS, Choo A. Generation
Invest Ophthalmol Vis Sci. 2005;46(11):4200–8. of novel monoclonal antibodies for the enrichment
6. Joyce NC. Proliferative capacity of the corneal endo- and characterization of human corneal endothelial
thelium. Prog Retin Eye Res. 2003;22(3):359–89. cells (hCENC) necessary for the treatment of corneal
7. DelMonte DW, Kim T. Anatomy and physiology of the endothelial blindness. MAbs. 2014;6(6):1439–52.
cornea. J Cataract Refract Surg. 2011;37(3):588–98. 20. Bartakova A, Alvarez-Delfin K, Weisman AD, Salero
8. Murphy C, Alvarado J, Juster R.  Prenatal and post- E, Raffa GA, Merkhofer RM, et  al. Novel identity
natal growth of the human Descemet’s membrane. and functional markers for human corneal endothelial
Invest Ophthalmol Vis Sci. 1984;25(12):1402–15. cells. Invest Ophthalmol Vis Sci. 2016;57(6):2749–62.
9. Stiemke MM, Edelhauser HF, Geroski DH.  The 21. Yoshihara M, Ohmiya H, Hara S, Kawasaki S,

developing corneal endothelium: correlation of mor- Hayashizaki Y, Itoh M, et  al. Discovery of molecu-
phology, hydration and Na/K ATPase pump site den- lar markers to discriminate corneal endothelial cells
sity. Curr Eye Res. 1991;10(2):145–56. in the human body. PLoS One. 2015;10(3):e0117581.
10. Sobottka Ventura AC, Wälti R, Böhnke M.  Corneal Kerkis I, editor.
thickness and endothelial density before and after
cataract surgery. Br J Ophthalmol. 2001;85(1):18–20.
Corneal Endothelium: Isolation
and Cultivation Methods
28
David Mingo-Botín, Marie Joan Therese D. Balgos,
and Francisco Arnalich-Montiel

28.1 Introduction Human corneal endothelial cells (HCECs) are


derived from the cranial neural crest. They have a
The corneal endothelium is the innermost layer limited capacity to proliferate in vivo, with majority
of the human cornea, composed of a single layer of cells arrested in the G1 phase of the cell cycle and
of hexagonal cells that are organized in a regular cannot go further in the cell cycle phase in spite of
mosaic on their basement membrane  – the endothelial cell injury or loss [3]. This has been
Descemet membrane. Corneal endothelial cells attributed to the following factors as reviewed by
maintain corneal transparency by controlling Cui et al. – the presence of negative regulator mole-
deturgescence in the corneal stroma [1]. They act cules (such p21, p27, p16, p15, p19, and p53), trans-
as a partial or leaky barrier to nutrients and other forming growth factor β (TGF-β) in aqueous humor,
molecules in the aqueous humor in a mechanism contact inhibition of densely packed cells, or stress-
explained by the pump-leak hypothesis. The cor- induced premature senescence [4]. Residual endo-
neal endothelial cells are able to accomplish this thelial cells thus enlarge and migrate to cover the
with the use of tight junctions, simple diffusion, defects left by lost cells [1, 5]. To achieve corneal
facilitated diffusion, and active transport mecha- deturgescence and transparency, a minimal func-
nisms. In the absence of this mechanism, the tional capacity is required, which depends on both
strict organization of the collagen fibrils in the endothelial cell density and quality. There is an
stroma, critical for corneal transparency, would inverse relationship between age and corneal endo-
be compromised [1, 2]. thelial cell density [1, 6]. The high endothelial cell
densities found in infants, up to 5000 cells/mm2,
decrease during early childhood until gradual physi-
ologic decline of approximately 0.6% per year is
reached in adulthood. This can occur without
impairing corneal transparency. However, certain
D. Mingo-Botín endothelial dystrophies and traumatic endothelial
Cornea Unit, Department of Ophthalmology,
injury can dramatically accelerate this process, mak-
Hospital Universitario Ramón y Cajal, Madrid, Spain
ing endothelial cell density fall below the threshold
M. J. T. D. Balgos
of 300–500 cells/mm2 [2]. At this stage, irreversible
Research and Development Department,
Cornea, Cataract and Refractive Surgery Department, corneal edema occurs, leading to vision loss.
Vissum Alicante, Alicante, Spain Corneal disease is a leading cause of visual
F. Arnalich-Montiel impairment worldwide. Corneal opacities make
Vissum Corporation, Madrid, Spain up 4% of the leading causes of blindness, as per

© Springer Nature Switzerland AG 2019 425


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_28
426 D. Mingo-Botín et al.

the 2010 Global Estimates of the World Health nea could be divided into three further parts – the
Organization [7, 8]. While nearly 80% of causes peripheral corneoscleral rim with epithelium and
of corneal blindness are avoidable or preventable, stroma, an anterior lamellar stromal disc, and a
when these measures have been unavailable or posterior lamellar disc with Descemet and endo-
unsuccessful, the only recourse is for replace- thelium [13].
ment of the diseased corneal tissue [9–11]. Another strategy being evaluated is to maxi-
Indications for corneal transplantation vary mize the remaining corneal endothelial cells by a
among countries, but the most common are topical treatment that would decrease endothelial
Fuchs’ dystrophy, keratoconus, bullous keratopa- cell loss and promote the function of the remain-
thy, and infections. Improvement in surgical ing cells. Koizumi et al. and Okumura et al. have
technique and available technology and the reported that the topical use of Rho-kinase
establishment of eye banks all over the world (ROCK) inhibitor Y-27632 in rabbit and primate
have contributed greatly to the establishment of corneal injury models slowed the progression of
corneal transplantation as the standard treatment endothelial cell degeneration and led to restora-
for visual impairment due to corneal pathologies tion of normal cell counts [14–17]. It has also
[10]. A longer recovery time, the risk of develop- been partially successful in four human patients
ing high or irregular astigmatism, glaucoma, and with corneal endothelial dysfunction [17]. The
immune-mediated graft rejection are the main exact mechanism by which Rho-kinase inhibition
concerns that have been associated with penetrat- increases endothelial cell count is still unknown,
ing keratoplasty (PK). These are limited with the and other studies have presented conflicting data
use of endothelial and anterior lamellar kerato- [18]. Hence, more studies are needed on the
plasty, although both techniques require a higher safety and efficacy of these molecules as possible
learning curve than PK. Endothelial keratoplasty topical treatment for corneal endothelial disease.
is also associated with a higher degree of postop-
erative endothelial cell loss than PK [4].
Furthermore, long-term graft survival after PK is 28.2 H
 uman Corneal Endothelial
limited – with 5- and 10-year corneal graft sur- Cell Engineering
vival estimates at 74% and 64%, respectively  –
and repeat grafting is likely necessary [12]. There is now considerable interest in the develop-
Corneal transplantation rates are hampered by ment of suitable alternatives for donor endothe-
the number of available suitable cadaver corneas lial graft material through tissue engineering and
for transplantation. A generally low cornea dona- regenerative medicine. Intensive collaborative
tion rate, coupled with strict criteria for screening and multidisciplinary research has been done on
and exclusion of donor corneas (which include the topic to realize the expansion of functional
serology tests, the presence of notable systemic HCECs in  vitro and to transplant them onto a
medical conditions, and the presence of structural recipient cornea to recover normal or near-­normal
defects in the corneal tissue), has led to an enor- corneal endothelial function.
mous gap between corneal supply and demand. Although HCECs have limited proliferative
Global demand for cornea is at 12.7 million, yet capacity in  vivo, it has been demonstrated that
the range of corneal transplants that have been after the release of cell-cell junctions, and in the
performed is only at 100,000–185,000 [9]. This presence of appropriate growth factors, they are
imbalance will probably be aggravated by able to proliferate ex vivo [3]. Thus, tissue engi-
increasing global population growth and a con- neering gives us the opportunity to replace dys-
comitant aging of the general population. To cope functional corneal endothelium without relying
with the shortage of donor cornea, steps have on the limited available cadaveric donor cornea
been proposed to maximize one donor cornea for supply.
use with multiple patients. Using anterior or pos- Prior to transplantation, tissue-engineered
terior lamellar grafts, a good-quality donor cor- grafts would need to fulfill similar requirements
28  Corneal Endothelium: Isolation and Cultivation Methods 427

to those that endothelial grafts from an eye bank also divide, young donors have a larger cell
must meet [19]. These grafts should be functional population that retain the capacity to prolifer-
and highly cellular to compensate for the cell loss ate in response to mitogenic stimuli. Likewise,
associated with surgery. They should be deliv- peripheral HCECs exhibit a higher replication
ered on a biocompatible carrier that does not alter competence than cells from the central cornea,
the optical and biomechanical properties of the regardless of donor age [28, 29].
cornea and allow a convenient surgical technique. (b) Stem cells are the ideal potential source for
In addition to its availability and high cellularity, tissue engineering, and they may be derived
other potential advantages of corneal endothe- from presumed endothelial progenitors or
lium culture would be the production of autolo- from adult stem cells from other locations.
gous tissue (although it would be limited to Progenitor cells in the corneal endothelium
secondary endothelial diseases), eliminating were first located in a monkey model. Several
immunological rejection, and its ability to with- studies indicate the presence of corneal endo-
stand cryopreservation [20]. thelial progenitors, known as Schwalbe’s line
Since the successful expansion of HCECs cells, at the transition zone between the ante-
reported by Mannagh and Irving in 1965 [21], rior extension of the trabecular meshwork
research in this field has focused especially on and endothelial periphery [30–33].
optimizing the methods of culture as well as the An examination of the microanatomy of
biomaterials used as carriers for transplantation. the endothelium showed that cells at the
As in other engineered tissues, there are a num- extreme periphery were organized in small
ber of key variables that determine their prolifer- clusters around the Hassall-Henle bodies,
ative and functional capacity. These include the arranged in centripetal rows. These cells
viability of the donor cells, the procedure for iso- exhibited lesser differentiation than endothe-
lation, the substrate, and the composition of the lial cells in the cornea, with preferential
culture medium. expression of stem cell markers (nestin,
telomerase, and occasionally breast cancer
resistance protein). This suggested a continu-
28.2.1 Source of HCECs ous slow centripetal migration of endothelial
cells from these niches [31]. Similar results
There are three main sources of HCECs: primary were reported by another study. Manifest
HCECs, established cell lines, and stem cells, which telomerase reaction and BrdU-linked alkaline
may originate from presumed endothelial progeni- phosphatase activity staining, which demon-
tors or from adult stem cells from other locations. strates cellular division, were detected in
peripheral corneal endothelium and enhanced
(a) Primary HCECs are isolated from donor cor- after mechanical wounding of the cornea.
neoscleral rims remaining after corneal These point to a wound-induced cell response
transplantation or human cadaver corneas which suggests that cells in the corneal endo-
that are ineligible for surgery. The viability thelium may be renewed by stem-­like cells
of these cells may be influenced by cell den- originating from the peripheral endothelium
sity, death to preservation time, preservation and migrating toward the center, in order to
medium and period, age, overall health of the minimize endothelial cell loss [32]. This was
donor, and specific cause of death [22, 23]. clinically observed in an in vivo model in an
In particular, the effects of donor age and observational case series of three human
topographical source on the proliferative patients with chemical eye injuries, as well as
capacity of primary HCECs have been studied in an in  vivo endothelial chemical injury
extensively. Using ex vivo cornea [24] and tis- model in rabbit eyes. Corneal endothelial cell
sue culture methods [23, 25–27], it has been density was found to increase at 6  months
shown that, although older donor cells can post-injury in human eyes, compared to
428 D. Mingo-Botín et al.

immediately after and 1 month after. This was therefore it is essential to prevent them from con-
accompanied by an improvement in corneal taminating the HCEC culture. This was achieved
edema and reduced corneal thickness. In the with the use of explant or scraping techniques
rabbit eyes, corneal edema was noted to [45], although even and fibroblast growth-inhib-
decrease over time (although corneal endo- iting media [46] or magnetic cell separation [47]
thelium of rabbits has been known to regener- have been proposed. However the current refine-
ate) albeit without improvement in central ment of Descemet membrane (DM) extraction
corneal opacity. The peripheral cornea was techniques, similar to those used in DM endothe-
clear. Vital staining with alizarin S red and lial keratoplasty (DMEK) [48], allows for a much
trypan blue showed cell damage centrally and cleaner extraction. Briefly, the donor corneo-
peripherally immediately after injury. Within scleral button is positioned on a support, usually
7 days of the injury, corneal endothelial cells with suction vacuum; peripheral and very shal-
were already present in the periphery. Corneal low marking is performed by means of a needle
endothelial cells were only detected in the or a trephine, and, always under balanced salt
center 14 days after injury [34]. solution, the edges of the DM are raised to obtain
Sphere-forming cells derived from HCECs a plane of cleavage from which to pull gently
may also represent a potential source of pro- until the dissection is complete. After peeling the
genitor cells [35, 36] – they have been dem- DM with the endothelium, the cells are detached
onstrated to be highly proliferative with a and dissociated by enzymatic or nonenzymatic
propensity to later differentiate into HCEC-­ methods. Currently this is the most widespread
like cells. It appears that this sphere-forming strategy, and it is known as “peel and digest”
assay results in longer telomeres and higher [22].
telomerase activity, and they are better formed Enzymatic digestion methods mainly use
from endothelial peripheral cells [37]. trypsin, collagenase, or dispase [19, 49]. Its main
(c) Immortalized cell lines are easily available drawback is that enzymatic digestion requires a
and have the advantage of allowing the large prolonged incubation time to separate the cells
quantities of homogeneous cells required for from the matrix, which is accompanied by fur-
many in vitro studies. The first line of HCECs ther cell degradation and increases the possibil-
was developed in 2000 [38] by transfection ity of contamination by stromal cells.
with simian virus 40 T antigen. Two clonal Nonenzymatic methods use EDTA to release
cell lines were later separated from this cell intercellular junctions, which also may promote
population [39]: B4G12 [40], which is con- cell proliferation [50], but can induce some cell
sidered a good model of differentiated HCEC, degradation too. It is also possible to combine
and H9C1, with developing HCEC character- methods, such as that described by Li et  al.,
istics. Other immortalized HCEC line used where collagenase A is first used to detach cells
transduction of human papilloma virus 16 E6/ from DM and aggregates of HCECs are obtained
E7 oncogenes [41]. To avoid the potential that could be preserved in serum-free medium
carcinogenesis induced by viral oncogenes, and maintain HCEC characteristics for at least 3
other attempts have been made by transduc- weeks. Trypsin and EDTA are then added to sep-
tion with human telomerase reverse transcrip- arate the cells, resulting in higher proliferative
tase [42], Cdk4R24C/Cyclin D1 [43], and response [51, 52]. To optimize the isolation of
immortalizing untransfected HCECs [44]. viable cells, the use of a stabilization medium
(serum-supplemented with no added growth fac-
tors) for up to 6 days prior to isolation has also
28.2.2 Isolation been proposed. This could enable the selection
of viable cells and remove the negative impact of
Stromal keratocytes have much greater prolifera- the apoptotic neighboring cells [53], increasing
tive capacity in vitro than endothelial cells, and the yield of subsequent culture [54].
28  Corneal Endothelium: Isolation and Cultivation Methods 429

28.2.3 Substrate or Extracellular lular morphology. Na+/K+ATPase expression was


Matrices also significantly higher in collagen type I-,
fibronectin-, and laminin-coated culture plates.
At the initiation of primary cell culture, adher- Yamaguchi et al. investigated also the expression
ence to the substrate is important to initiate cell of laminin-­5 and its receptor-α3β1 integrin in
growth, whereas detachment of a confluent cell HCECs cultures and concluded that it promoted
layer in later phases is necessary for transplanta- adhesion, migration, and moderate proliferation
tion purposes [55]. While it has been suggested [60].
that matrix coating might not be mandatory for Extracellular matrices from cultured bovine
culturing HCECs [56], it is known that proteins endothelial cells have been extensively used for
in the basement membranes are implicated in the HCEC cultures [61, 62]; however the use of ani-
processes of cell adhesion, proliferation, migra- mal components gives rise to the possibility of
tion, morphogenesis, and differentiation [57, 58]. xenogenic contamination and immunogenicity.
For this purpose, the ideal substrates would be Numata et al. used pericellular matrix of decidua-­
those similar to DM in their physiological and derived mesenchymal cells (PCM-DM) as sub-
mechanical characteristics. They should provide strate in order to avoid animal-derived ECM
a favorable microenvironment for endothelial sources. They demonstrated that this xeno-free
activity, be transparent and nontoxic, and be easy coating PCM-DM substrate enhanced cell adhe-
to handle during transplantation [55]. Multiple sion via integrin, promoted cell proliferation, and
biological, synthetic, and biosynthetic substrates suppressed apoptosis [63].
have been used and described. In the same way that collagen type I sheets have
Extracellular matrix (ECM) proteins found in been used both as ECM and as carrier [61], other
DM include fibronectin, laminin, collagen types tissues have been tested with the same goal.
I, IV, VI, and VIII, and proteoglycans [57, 59], Decellularized human corneal stroma (seeded with
which can be employed as culture substrates. The feline [64] or human [59, 65] corneal endothelial
coatings based on these natural polymers can be cells) and bovine [66] corneal stroma (seeded with
composed of single proteins such as collagen, HCECs) were shown to be possible scaffolds for
fibronectin, laminin, or gelatin) or combinations cultivation and transplantation of HCECs.
of different proteins (such as ECM from cultured Other human-derived sources of tissue sub-
bovine corneal endothelial cells, and FNC strate and carrier are amniotic membrane (AM)
Coating Mix®, containing a mixture of fibronec- and anterior lens capsule. The anti-inflammatory
tin, collagen, and albumin, (Athena Environmental and non-immunogenic properties of AM could be
Sciences Inc.). favorable properties for this purpose [67, 68]. Its
Choi et al. evaluated adhesion, proliferation, main problems are that it is translucent (although
and phenotypic maintenance between uncoated using only basement membrane could improve
and culture plates coated with different ECM – this issue [68]), it is heterologous tissue, and
including collagen type I, collagen type IV, there might be high inter-donor variability which
fibronectin, laminin, and FNC Coating Mix [57]. could influence the outcome. AM can be used
All the coated cultured plates had increased inte- cryopreserved or lyophilized [69]. The potential
grin expression as adhesion markers (especially use of de-epithelized human anterior lens capsule
integrin ɑ2) compared to the uncoated dishes, coming from cataract surgeries has been explored
and centrifugation assays showed improved by Yoeruek et  al. [70], who found optimal cell
adhesion strength in all except for the laminin- density and phenotype after using this scaffold
coated plate. HCECs formed a compact mono- for culture. This has been later confirmed by
layer after a week, without significant differences other investigators [71, 72], although its biologi-
between coated and uncoated plates; but cells cal origins still confer inter- and intra-donor vari-
grown on collagen IV-, fibronectin-, and FNC- ability with the theoretical risk of disease
coated plates showed more typical compact cel- transmission.
430 D. Mingo-Botín et al.

Synthetic and biosynthetic substrates have which can also vary depending on the donor
been developed to function as carriers, but samples. In order to reduce serum dependence,
their discussion is not included in the scope of additional supplementation with serum com-
this chapter. Within however, it is worth men- ponents is required. These serum components
tioning thermoresponsive plates grafted with include insulin, selenium, transferrin, lipids,
poly(N-isopropylacrylamide) (PIPAAm), essential amino acids, or ascorbic acid [77].
which can be hydrophobic or hydrophilic Mitogenic growth factors used include epider-
depending on the temperature and allow the mal growth factor (EGF), vascular endothelial
detachment of endothelial cell layer for trans- growth factor (VEGF), insulin-­ like growth
plantation [73, 74]. factor (IGF), and fibroblast growth factor
Table 28.1 gives examples of natural culture (FGF)  – which could also play a role as dif-
substrates or extracellular matrices. ferentiation factor [49] or promote fibroblastic
transformation [78]. Recently, leukemia inhib-
itory factor (LIF) has been shown to stimulate
28.2.4 Culture Media HCEC growth and preserve cell phenotype in
the presence of bFGF in a media without serum
Many attempts to optimize culture media have [79].
been made, using different basal media sup- Some of the more commonly cited formula-
plemented with several growth factors, serums, tions are summarized in Table 28.2.
and nutrients  – all aimed to stimulate HCEC Although a variety of protocols have proven to
expansion and help maintain cell phenotype. support continuous cell proliferation, the cells
Published protocols vary greatly, and each one often undergo morphological transformation
involves serum as an essential component. toward a fibroblast-like morphology after few
This causes major uncertainty because the use passes. This is known as endothelial-to-­
of serum introduces a complex mixture of mesenchymal transition (EMT) [78, 86]. It is
ingredients, at poorly defined concentrations believed that this phenomenon is promoted by
the use of trypsin-EDTA to break the cell-cell
junctions and bFGF in the culture medium via
Table 28.1  Examples of natural culture substrates or activation of the canonical Wnt signaling, espe-
extracellular matrices cially when TGF-­β1 signaling is also activated
Substrate Reference [86–90]. This highlights the importance of char-
Biological substrates acterization of the obtained endothelium with
 Decellularized Proulx et al. [64] each model  – and such an assessment should
corneal stroma Honda et al. [65]
Choi et al. [59]
evaluate morphology, protein expression (typi-
Bayyoud et al. [66] cally Na/K-ATPase and zonula occludens pro-
Yoeruek et al. [75] teins), and function in  vitro or in  vivo [19].
 Amniotic membrane Ishino et al. [67] Recently, a panel of novel markers correlated
Wencan et al. [68] with function in culture has been described [91].
 Lens capsule Yoeruek et al. [70]
Kopsachilis et al. [71]
Using RNA sequencing, it is also possible to
Bogerd et al. [72] compare transcriptomic profile between ex vivo
Natural polymers HCEC, primary HCEC cultures, and immortal-
 Collagen types I and Choi et al. [59], Choi et al. ized HCEC lines. With this technique, Frausto
IV [57], Numata et al. [63] et  al. found that, at a molecular level, primary
 Laminin Yamaguchi et al. [60]
HCECs resembled ex vivo HCECs more closely
 Bovine endothelial Miyata et al. [26]
ECM Mimura et al. [61] than immortalized lines [92]. Other specific
Hitani et al. [62]. HCEC markers, specially CD166, have been
 FNC Coating Mix ® Joyce et al. [76] identified through the use of monoclonal antibod-
Choi et al. [57] ies [93, 94].
28  Corneal Endothelium: Isolation and Cultivation Methods 431

Table 28.2  Basal media, supplements, and sera for HCEC culture
Basal medium Supplements Serum Reference
F99 (Ham’s F12 and M199, 1:1 ratio) 20 μg/ml ascorbic acid 2–5% Engelmann et al. [80]
20 μg/ml bovine insulin
2.5 μg/ml transferrin
0.6 ng/ml sodium selenite
10 ng/ml bFGF
MEM 5 μg/ml insulin 10% Blake et al. [81]
5 μg/ml transferrin
5 ng/ml sodium selenite
50 μg/ml gentamicin
100 U/ml penicillin
0.1 mg/ml streptomycin
0.25 μg/ml amphotericin B
150 μg/ml ECGS
DMEM 30 mg/l L-glutamine 15% Miyata et al. [26]
2.5 mg/l Fungizone
2.5 mg/l doxycycline
2 ng/ml bFGF
DMEM 2 ng/ml bFGF 10% Ishino et al. [67]
50 U/ml penicillin
50 mg/ml streptomycin
Opti-MEM-I 5 ng/ml EGF 8% Zhu et al. [23]
20 ng/ml NGF
100 μg/ml pituitary extract
20 μg/ml ascorbic acid
200 mg/l calcium chloride
0.08% chondroitin sulfate
50 μg/ml gentamicin
1/100 antibiotic/antimycotic
SHEM (Ham’s F12 and DMEM; 1:1 ratio) 0.5% DMSO 5% Li et al. [51]
2 ng/ml mouse EGF
5 μg/ml insulin
5 μg/ml transferrin
5 ng/ml selenium
0.5 μg/ml hydrocortisone
1 nM cholera toxin
50 μg/ml gentamicin
1.25 μg/ml amphotericin B
EGM-2® VEGF 10% Choi et al. [59]
(Lonza Group Ltd, Basel, Switzerland) EGF
IGF
bFGF
Ascorbic acid
Hydrocortisone
Gentamicin
Amphotericin B
Endothelial SFM® 10 ng/ml FGF-2 – Jäckel et al. [82]
(Life Technologies/Thermo Fisher Scientific, USA) Antibiotics
DMEM:F12 (1:1 ratio) 20% HAF – Feizi et al. [83]
MESCM 10% knockout serum, – Zhu et al. [84]
(DMEM and F12; 1:1 ratio) 5 μg/ml insulin
5 μg/ml transferrin
5 ng/ml sodium selenite
4 ng/ml bFGF
10 ng/ml hLIF
50 μg/ml gentamicin
1.25 μg/ml amphotericin
F99 (Ham’s F12 and M199, 1:1 ratio) 0.1 mg/ml phPL – Thieme et al. [85]
DMEM Dulbecco’s modified Eagle’s medium, DMSO dimethyl sulfoxide, ECGS endothelial cell growth supplement,
EGM endothelial growth medium, HAF human amniotic fluid, MESCM modified embryonic stem cell medium, SFM
serum-free medium, SHEM supplemented hormonal epithelial medium, phPL processed human platelet lysate
432 D. Mingo-Botín et al.

28.2.4.1 A  pproaches to Avoid 28.2.4.2 Animal-Free Approaches


Endothelial-to-Mesenchymal Another important line of research is dedicated to
Transition eliminating animal by-products from the HCEC
Peh et al. evaluated four different culture media: culture process. Given the therapeutic target of
Dulbecco’s modified Eagle’s medium (DMEM) these cultures and regulatory issues, it is neces-
[67], Opti-MEM-I [23], DMEM/F12 [51], and sary to achieve HCEC propagation under xeno-­
Ham’s F12/M199 [80] in FNC-coated plates using free conditions in order to eliminate the risk of
HCECs from paired donor corneas in order to animal disease transmission to humans. In this
avoid donor variability (Peh 2011) [52]. They regard, several laboratories have developed
found that isolated hCECs could be established in serum-free and xeno-free media.
all four media for a short period of time, but only Jackel et  al. compared MEM with fetal calf
Opti-MEM-I and Ham’s F12/M199 [23, 80] were serum and endothelial serum-free medium (SFM)
able to support the continual propagation of and found that cells cultured in MEM were more
hCECs and expression of Na+ K+/ATPase and susceptible to apoptosis in the absence of exoge-
ZO-1. In all of the media, cells lost their character- nous noxious stimuli, while HCEC cultured in
istic morphology beyond the third passage. In SFM seemed to be protected from cell death even
order to overcome EMT transformation – due to when the apoptosis-inducer staurosporine was
the growth factors included in the culture medium added [82]. Another serum-free medium, previ-
and loss of contact inhibition by trypsin-EDTA ously used in limbal stem cell cultures, is modi-
treatment – some authors suggest the use of a dual fied embryonic stem cell medium (MESCM). It
media approach [78]. This dual dual media contains LIF and bFGF and allowed dramatic
approach involves the combined use of a prolifera- expansion of the size of HCEC monolayers [84].
tive medium [80] with greater expansion proper- A different approach is to replace FBS with
ties, with a stabilization medium [23] with better human-derived sera. A promising candidate as
capacity to preserve the cellular morphology of the source of nutrients and growth factors in basal
primary HCECs in vitro. HCECs propagated using growth media is human platelet lysate [96].
this dual media approach were homogeneous, Thieme et  al. compared FBS and processed
retained their characteristic cellular morphology, human platelet lysate (phPL) in immortalized
and expressed higher levels of corneal endothe- and primary HCEC cultures and found superior
lium-associated markers than in those cultured in viability in cultures with phPL [85]. Another
proliferative medium alone, which were heteroge- study compared the use of human amniotic fluid
neous and fibroblastic in appearance. obtained from diagnostic amniocentesis and
The addition of ROCK inhibitor Y27632, FBS, and the first exhibited greater stimulatory
extensively studied by the laboratory of Kinoshita, effect on HCEC growth [83]. Recently, the group
seems to preserve proliferation and maintain of Peh et  al. have described a regulatory-­
HCEC phenotype in vivo [17, 95]. The use of this compliant protocol based on their dual media
molecule is believed to inhibit EMT that occurs approach, collagen IV substrate, and EquaFETAL
after intercellular junctions are disrupted as in serum (derived from specially controlled ani-
wounding processes [88], maintaining the physi- mals) and applied it to a rabbit model of bullous
ologic suppression of RhoA activation with con- keratopathy, demonstrating that it is possible to
tact inhibition. A different technique was translate tissue engineering into the clinic [97].
developed by Zhu et al. using interference RNA
technology to knockdown p120 catenin and
Kaiso transcription suppressor. This unlocks 28.3 Conclusion
mitotic contact inhibition but avoids disruption of
intercellular junctions or activating of canonical Great progress has been made in the isolation and
Wnt signaling [84]. The resultant expanded cultivation of HCEC, especially thanks to the
HCECs have a phenotype suggestive of a neural efforts of several laboratories to improve sub-
crest progenitor reprogramming. strates and culture media. However, a standard
28  Corneal Endothelium: Isolation and Cultivation Methods 433

culture method that satisfies all the requisites for come of corneal transplantation. Ophthalmology.
2009;116(12):2354–60.
human use has not yet been achieved. The 13. Vajpayee RB, Sharma N, Jhanji V, Titiyal JS, Tandon
advances in the knowledge of the HCEC cell R. One donor cornea for 3 recipients: a new concept
cycle along with better identification of endothe- for corneal transplantation surgery. Arch Ophthalmol
lial progenitors, as well as the use of xeno-free (Chicago Ill 1960). 2007;125(4):552–4.
14. Okumura N, Koizumi N, Ueno M, Sakamoto Y,

culture systems, may allow in the not-so-distant Takahashi H, Tsuchiya H, et al. ROCK inhibitor con-
future the application of tissue engineering in verts corneal endothelial cells into a phenotype capa-
patients with corneal endothelial disease, thus ble of regenerating in  vivo endothelial tissue. Am J
meeting the great world demand for this tissue. Pathol. 2012;181(1):268–77.
15. Koizumi N, Sakamoto Y, Okumura N, Tsuchiya H,
Torii R, Cooper LJ, et al. Cultivated corneal e­ ndothelial
Compliance with Ethical Requirements  David Mingo- transplantation in a primate: possible future clinical
Botín, Marie Joan Therese Dr. Balgos, and Francisco application in corneal endothelial regenerative medi-
Arnalich-Montiel declare that they have no conflict of cine. Cornea. 2008;27 Suppl 1:S48–55.
interest. No human or animal studies were carried out by 16. Okumura N, Koizumi N, Kay EP, Ueno M, Sakamoto
the authors for this article. Y, Nakamura S, et al. The ROCK inhibitor eye drop
accelerates corneal endothelium wound healing.
Invest Ophthalmol Vis Sci. 2013;54(4):2493–502.
17. Okumura N, Kinoshita S, Koizumi N. Application
References of rho kinase inhibitors for the treatment of
corneal endothelial diseases. J Ophthalmol.
1. Edelhauser HF. The balance between corneal transpar- 2017;2017:1–8.
ency and edema the proctor lecture. Invest Opthalmol 18. Pipparelli A, Arsenijevic Y, Thuret G, Gain P, Nicolas
Vis Sci. 2006;47(5):1755. M, Majo F.  ROCK inhibitor enhances adhesion and
2. Srinivas SP.  Dynamic regulation of barrier integ- wound healing of human corneal endothelial cells.
rity of the corneal endothelium. Optom Vis Sci. Connon CJ, editor. PLoS One. 2013;8(4):e62095.
2010;87(4):1. 19. Proulx S, Brunette I.  Methods being developed for
3. Joyce NC. Proliferative capacity of corneal endothe- preparation, delivery and transplantation of a tissue-­
lial cells. Exp Eye Res. 2012;95(1):16–23. engineered corneal endothelium. Exp Eye Res.
4. Cui Y-B, Wu J.  Research progress on the negative 2012;95(1):68–75.
factors of corneal endothelial cells proliferation. Int J 20. De Miguel MP, Alio JL, Arnalich-Montiel F, Fuentes-­
Ophthalmol. 2012;5(5):614–9. Julian S, de Benito-Llopis L, Amparo F, et al. Cornea
5. Yee RW, Geroski DH, Matsuda M, Champeau EJ, and ocular surface treatment. Curr Stem Cell Res
Meyer LA, Edelhauser HF.  Correlation of corneal Ther. 2010;5(2):195–204.
endothelial pump site density, barrier function, and 21. Mannagh J, Irving AR. Human corneal endothelium:
morphology in wound repair. Invest Ophthalmol Vis growth in tissue cultures. Arch Ophthalmol (Chicago
Sci. 1985;26(9):1191–201. Ill 1960). 1965;74(6):847–9.
6. Yee RW, Matsuda M, Schultz RO, Edelhauser 22. Zavala J, López Jaime GR, Rodríguez Barrientos CA,
HF.  Changes in the normal corneal endothelial cel- Valdez-Garcia J. Corneal endothelium: developmental
lular pattern as a function of age. Curr Eye Res. strategies for regeneration. Eye. 2013;27(5):579–88.
1985;4(6):671–8. 23. Zhu C, Joyce NC. Proliferative Response of Corneal
7. Wong KH, Kam KW, Chen LJ, Young AL.  Corneal Endothelial Cells from Young and Older Donors.
blindness and current major treatment concern-graft Invest Opthalmol Vis Sci. 2004;45(6):1743.
scarcity. Int J Ophthalmol. 2017;10(7):1154–62. 24. Senoo T, Joyce NC.  Cell cycle kinetics in corneal
8. Pascolini D, Mariotti SP.  Global estimates of visual endothelium from old and young donors. Invest
impairment: 2010. Br J Ophthalmol. 2012;96(5):614–8. Ophthalmol Vis Sci. 2000;41(3):660–7.
9. Gain P, Jullienne R, He Z, Aldossary M, Acquart S, 25. Joyce NC. Cell cycle status in human corneal endo-
Cognasse F, et  al. Global survey of corneal trans- thelium. Exp Eye Res. 2005;81(6):629–38.
plantation and eye banking. JAMA Ophthalmol. 26. Miyata K, Drake J, Osakabe Y, Hosokawa Y, Hwang
2016;134(2):167. D, Soya K, et al. Effect of donor age on morphologic
10. Güell JL, El Husseiny MA, Manero F, Gris O, Elies variation of cultured human corneal endothelial cells.
D. Historical review and update of surgical treatment Cornea. 2001;20(1):59–63.
for corneal endothelial diseases. Ophthalmol Therapy. 27. Enomoto K, Mimura T, Harris DL, Joyce NC.  Age
2014;3(1–2):1–15. differences in cyclin-dependent kinase inhibitor
11. Lamm V, Hara H, Mammen A, Dhaliwal D, Cooper expression and rb hyperphosphorylation in human
DKC.  Corneal blindness and xenotransplantation. corneal endothelial cells. Invest Ophthalmol Vis Sci.
Xenotransplantation. 2014;21(2):99–114. 2006;47(10):4330–40.
12. Borderie VM, Boëlle P-Y, Touzeau O, Allouch C, 28. Bednarz J, Rodokanaki-von Schrenck A, Engelmann
Boutboul S, Laroche L.  Predicted long-term out- K. Different characteristics of endothelial cells from
434 D. Mingo-Botín et al.

central and peripheral human cornea in primary cul- human corneal endothelial cells yields functional
ture and after subculture. In Vitro Cell Dev Biol Anim. hexagonal monolayers. Lewin A, editor. PLoS One.
1998;34(2):149–53. 2012;7(12):e51427.
29. Mimura T, Joyce NC.  Replication competence
43. Yokoi T, Seko Y, Yokoi T, Makino H, Hatou S, Yamada
and senescence in central and peripheral human M, et  al. Establishment of functioning human cor-
corneal endothelium. Invest Ophthalmol Vis Sci. neal endothelial cell line with high growth potential.
2006;47(4):1387–96. Kerkis I, editor. PLoS One. 2012;7(1):e29677.
30. Raviola G.  Schwalbe line’s cells: a new cell type in 44. Fan T, Zhao J, Ma X, Xu X, Zhao W, Xu

the trabecular meshwork of Macaca mulatta. Invest B.  Establishment of a continuous untransfected
Ophthalmol Vis Sci. 1982;22(1):45–56. human corneal endothelial cell line and its biocom-
31. He Z, Campolmi N, Gain P, Ha Thi BM, Dumollard patibility to denuded amniotic membrane. Mol Vis.
J-M, Duband S, et al. Revisited microanatomy of the 2011;17:469–80.
corneal endothelial periphery: new evidence for con- 45. Peh GSL, Beuerman RW, Colman A, Tan DT, Mehta
tinuous centripetal migration of endothelial cells in JS.  Human corneal endothelial cell expansion for
humans. Stem Cells. 2012;30(11):2523–34. corneal endothelium transplantation: an overview.
32. Whikehart DR, Parikh CH, Vaughn AV, Mishler K, Transplantation. 2011;91(8):811–9.
Edelhauser HF. Evidence suggesting the existence of 46. Engelmann K, Böhnke M, Friedl P. Isolation and long-­
stem cells for the human corneal endothelium. Mol term cultivation of human corneal endothelial cells.
Vis. 2005;11:816–24. Invest Ophthalmol Vis Sci. 1988;29(11):1656–62.
33. McGowan SL, Edelhauser HF, Pfister RR, Whikehart 47. Peh GSL, Lee M-X, Wu F-Y, Toh K-P, Balehosur D,
DR. Stem cell markers in the human posterior limbus Mehta JS. Optimization of human corneal endothelial
and corneal endothelium of unwounded and wounded cells for culture: the removal of corneal stromal fibro-
corneas. Mol Vis. 2007;13:1984–2000. blast contamination using magnetic cell separation.
34. Choi SO, Jeon HS, Hyon JY, Oh Y-J, Wee WR,
Int J Biomater. 2012;2012:601302.
Chung T, et al. Recovery of corneal endothe- 48. Parekh M, Baruzzo M, Favaro E, Borroni D, Ferrari
lial cells from periphery after injury. PLoS One. S, Ponzin D, et al. Standardizing descemet membrane
2015;10(9):e0138076. endothelial keratoplasty graft preparation method
35. Yokoo S, Yamagami S, Yanagi Y, Uchida S, Mimura in the eye Bank  — experience of 527 Descemet
T, Usui T, et al. Human corneal endothelial cell pre- membrane endothelial keratoplasty tissues. Cornea.
cursors isolated by sphere-forming assay. Invest 2017;36(12):1458–1466.
Opthalmol Vis Sci. 2005;46(5):1626. 49. Engelmann K, Bednarz J, Valtink M.  Prospects

36. Mimura T, Yamagami S, Yokoo S, Usui T, Amano for endothelial transplantation. Exp Eye Res.
S. Selective isolation of young cells from human cor- 2004;78(3):573–8.
neal endothelium by the sphere-forming assay. Tissue 50. Senoo T, Obara Y, Joyce NC.  EDTA: a promoter of
Eng Part C Methods. 2010;16(4):803–12. proliferation in human corneal endothelium. Invest
37. Yamagami S, Yokoo S, Mimura T, Takato T, Araie M, Ophthalmol Vis Sci. 2000;41(10):2930–5.
Amano S. Distribution of precursors in human corneal 51. Li W, Sabater AL, Chen Y-T, Hayashida Y, Chen S-Y,
stromal cells and endothelial cells. Ophthalmology. He H, et al. A novel method of isolation, preservation,
2007;114(3):433–9. and expansion of human corneal endothelial cells.
38.
Bednarz J, Teifel M, Friedl P, Engelmann Invest Opthalmol Vis Sci. 2007;48(2):614.
K. Immortalization of human corneal endothelial cells 52. Peh GSL, Toh K-P, Wu F-Y, Tan DT, Mehta

using electroporation protocol optimized for human JS. Cultivation of human corneal endothelial cells iso-
corneal endothelial and human retinal pigment epithe- lated from paired donor corneas. Mohan RR, editor.
lial cells. Acta Ophthalmol Scand. 2000;78(2):130–6. PLoS One. 2011;6(12):e28310.
39. Valtink M, Gruschwitz R, Funk RHW, Engelmann 53. Gregory CD, Pound JD.  Microenvironmental influ-
K. Two clonal cell lines of immortalized human cor- ences of apoptosis in  vivo and in  vitro. Apoptosis.
neal endothelial cells show either differentiated or 2010;15(9):1029–49.
precursor cell characteristics. Cells Tissues Organs. 54. Spinozzi D, Miron A, Bruinsma M, Lie JT, Dapena
2008;187(4):286–94. I, Oellerich S, et  al. Improving the success rate of
40. HCEC-B4G12 | Creative bioarray [Internet]. [cited human corneal endothelial cell cultures from single
2018 Mar 13]. Available from: https://www.creative- donor corneas with stabilization medium. Cell Tissue
bioarray.com/HCEC-B4G12-CSC-C3457-item-1468. Bank. 2017;19(1):9.
htm. 5 5. Navaratnam J, Utheim T, Rajasekhar V, Shahdadfar
41. Kim H-J, Ryu Y-H, Ahn J-I, Park J-K, Kim
A.  Substrates for expansion of corneal endothe-
J-C.  Characterization of immortalized human cor- lial cells towards bioengineering of human cor-
neal endothelial cell line using HPV 16 E6/E7 on neal endothelium. J Funct Biomater. 2015;6(3):
lyophilized human amniotic membrane. Korean J 917–45.
Ophthalmol. 2006;20(1):47–54. 56. Chen KH, Azar D, Joyce NC. Transplantation of adult
42. Schmedt T, Chen Y, Nguyen TT, Li S, Bonanno
human corneal endothelium ex  vivo: a morphologic
JA, Jurkunas UV.  Telomerase immortalization of study. Cornea. 2001;20(7):731–7.
28  Corneal Endothelium: Isolation and Cultivation Methods 435

57. Choi JS, Kim EY, Kim MJ, Giegengack M, Khan as carrier matrix for cultivated human corneal endo-
FA, Khang G, et  al. In vitro evaluation of the inter- thelial cells. Cornea. 2009;28(4):416–20.
actions between human corneal endothelial cells 71. Kopsachilis N, Tsinopoulos I, Tourtas T, Kruse

and extracellular matrix proteins. Biomed Mater. FE, Luessen UW.  Descemet’s membrane substrate
2013;8(1):14108. from human donor lens anterior capsule. Clin Exp
58. Engler C, Kelliher C, Speck CL, Jun AS. Assessment Ophthalmol. 2012;40(2):187–94.
of attachment factors for primary cultured human cor- 72. Van den Bogerd B, Dhubhghaill SN, Zakaria

neal endothelial cells. Cornea. 2009;28(9):1050–4. N. Characterizing Human Decellularized Crystalline
59. Choi JS, Williams JK, Greven M, Walter KA, Laber Lens Capsules as a Scaffold for Corneal Endothelial
PW, Khang G, et  al. Bioengineering endothelialized Tissue Engineering. J Tissue Eng Regen Med.
neo-corneas using donor-derived corneal endothelial 2018:1–9.
cells and decellularized corneal stroma. Biomaterials. 73. Hsiue G-H, Lai J-Y, Chen K-H, Hsu W-M.  A novel
2010;31(26):6738–45. strategy for corneal endothelial reconstruction with
60. Yamaguchi M, Ebihara N, Shima N, Kimoto M,
a bioengineered cell sheet. Transplantation. J Tissue
Funaki T, Yokoo S, et  al. Adhesion, migration, and Eng Regen Med. 2018;12(4):e2020–8.
proliferation of cultured human corneal endothe- 74. Lai J-Y, Chen K-H, Hsiue G-H.  Tissue-engineered
lial cells by laminin-5. Invest Ophthalmol Vis Sci. human corneal endothelial cell sheet transplanta-
2011;52(2):679–84. tion in a rabbit model using functional biomaterials.
61. Mimura T, Yamagami S, Yokoo S, Usui T, Tanaka K, Transplantation. 2007;84(10):1222–32.
Hattori S, et  al. Cultured human corneal endothelial 75. Yoeruek E, Bayyoud T, Maurus C, Hofmann J, Spitzer
cell transplantation with a collagen sheet in a rabbit MS, Bartz-Schmidt K-U, et  al. Decellularization of
model. Invest Ophthalmol Vis Sci. 2004;45(9):2992–7. porcine corneas and repopulation with human cor-
62. Hitani K, Yokoo S, Honda N, Usui T, Yamagami S, neal cells for tissue-engineered xenografts. Acta
Amano S.  Transplantation of a sheet of human cor- Ophthalmol. 2012;90(2):e125–31.
neal endothelial cell in a rabbit model. Mol Vis. 76. Joyce NC, Zhu CC.  Human corneal endothelial cell
2008;14:1–9. proliferation: potential for use in regenerative medi-
63. Numata R, Okumura N, Nakahara M, Ueno M,
cine. Cornea. 2004;23(8 Suppl):S8–19.
Kinoshita S, Kanematsu D, et  al. Cultivation of 77. Okumura N, Inoue R, Kakutani K, Nakahara

corneal endothelial cells on a pericellular matrix M.  Corneal endothelial cells have an absolute
prepared from human decidua-derived mesen- requirement for cysteine for survival. Cornea.
chymal cells. Connon CJ, editor. PLoS One. 2017;36(8):988–94.
2014;9(2):e88169. 78. Peh GSL, Chng Z, Ang H-P, Cheng TYD, Adnan
64. Proulx S, Audet C, d’arc UJ, Deschambeault A, Carrier K, Seah X-Y, et  al. Propagation of human corneal
P, Giasson CJ, et al. Tissue engineering of feline cor- endothelial cells: a novel dual media approach. Cell
neal endothelium using a devitalized human cornea as Transplant. 2015;24(2):287–304.
carrier. Tissue Eng Part A. 2009;15(7):1709–18. 79. Liu X, Tseng SC, Zhang M-C, Chen S-Y, Tighe S, Lu
65. Honda N. Descemet stripping automated endothelial W-J, et al. LIF-JAK1-STAT3 signaling delays contact
keratoplasty using cultured corneal endothelial cells in inhibition of human corneal endothelial cells. Cell
a rabbit model. Arch Ophthalmol. 2009;127(10):1321. Cycle. 2015;14(8):1197–206.
66. Bayyoud T, Thaler S, Hofmann J, Maurus C, Spitzer 80. Engelmann K, Friedl P.  Growth of human corneal
MS, Bartz-Schmidt K-U, et al. Decellularized bovine endothelial cells in a serum-reduced medium. Cornea.
corneal posterior lamellae as carrier matrix for culti- 1995;14(1):62–70.
vated human corneal endothelial cells. Curr Eye Res. 81. Blake DA, Yu H, Young DL, Caldwell DR.  Matrix
2012;37(3):179–86. stimulates the proliferation of human corneal endo-
67. Ishino Y, Sano Y, Nakamura T, Connon CJ, Rigby H, thelial cells in culture. Invest Ophthalmol Vis Sci.
Fullwood NJ, et al. Amniotic membrane as a carrier for 1997;38(6):1119–29.
cultivated human corneal endothelial cell transplanta- 82. Jäckel T, Knels L, Valtink M, Funk RHW, Engelmann
tion. Invest Ophthalmol Vis Sci. 2004;45(3):800–6. K. Serum-free corneal organ culture medium (SFM)
68. Wencan W, Mao Y, Wentao Y, Fan L, Jia Q, Qinmei but not conventional minimal essential organ culture
W, et al. Using basement membrane of human amni- medium (MEM) protects human corneal endothelial
otic membrane as a cell carrier for cultivated cat cor- cells from apoptotic and necrotic cell death. Br J
neal endothelial cell transplantation. Curr Eye Res. Ophthalmol. 2011;95(1):123–30.
2007;32(3):199–215. 83. Feizi S, Soheili Z-S, Bagheri A, Balagholi S,

69. Niknejad H, Deihim T, Solati-Hashjin M, Peirovi
Mohammadian A, Rezaei-Kanavi M, et al. Effect of
H.  The effects of preservation procedures on amni- amniotic fluid on the in vitro culture of human corneal
otic membrane’s ability to serve as a substrate endothelial cells. Exp Eye Res. 2014;122:132–40.
for cultivation of endothelial cells. Cryobiology. 84. Zhu Y-T, Li F, Han B, Tighe S, Zhang S, Chen S-Y,
2011;63(3):145–51. et  al. Activation of RhoA-ROCK-BMP signaling
70. Yoeruek E, Saygili O, Spitzer MS, Tatar O, Bartz-­ reprograms adult human corneal endothelial cells. J
Schmidt KU, Szurman P. Human anterior lens capsule Cell Biol. 2014;206(6):799–811.
436 D. Mingo-Botín et al.

85. Thieme D, Reuland L, Lindl T, Kruse F, Fuchsluger 92. Frausto RF, Le DJ, Aldave AJ. Transcriptomic analy-
T. Optimized human platelet lysate as novel basis for sis of cultured corneal endothelial cells as a valida-
a serum-, xeno-, and additive-free corneal endothe- tion for their use in cell replacement therapy. Cell
lial cell and tissue culture. J Tissue Eng Regen Med. Transplant. 2016;25(6):1159–76.
2018;12(2):557–64. 93. Ding V, Chin A, Peh G, Mehta JS, Choo A. Generation
86. Lee JG, Kay EP.  FGF-2-mediated signal transduc- of novel monoclonal antibodies for the enrichment
tion during endothelial mesenchymal transforma- and characterization of human corneal endothelial
tion in corneal endothelial cells. Exp Eye Res. cells (hCENC) necessary for the treatment of corneal
2006;83(6):1309–16. endothelial blindness. MAbs. 2014;6(6):1439–52.
87. Zhu Y-T, Chen H-C, Chen S-Y, Tseng SCG. Nuclear 94. Dorfmueller S, Tan HC, Ngoh ZX, Toh KY, Peh G,
p120 catenin unlocks mitotic block of contact-­ Ang H-P, et  al. Isolation of a recombinant antibody
inhibited human corneal endothelial monolayers specific for a surface marker of the corneal endothe-
without disrupting adherent junctions. J Cell Sci. lium by phage display. Sci Rep. 2016;6(1):21661.
2012;125(Pt 15):3636–48. 95. Koizumi N, Kinoshita S, Okumura N. The role of rho
88. Zhu Y-T, Tighe S, Chen S-L, John T, Kao WY, Tseng kinase inhibitors in corneal endothelial dysfunction.
SCG.  Engineering of human corneal endothelial Curr Pharm Des. 2017;23(4):660–6.
grafts. Curr Ophthalmol Rep. 2015;3(3):207–17. 96. Wang TJ, Chen MS, Chou ML, Lin HC, Seghatchian
89. Lee JG, Ko MK, Kay EP.  Endothelial mesenchymal J, Burnouf T.  Comparison of three human platelet
transformation mediated by IL-1β-induced FGF-2  in lysates used as supplements for in  vitro expansion
corneal endothelial cells. Exp Eye Res. 2012;95(1):35–9. of corneal endothelium cells. Transfus Apher Sci.
90. Liu Y, Sun H, Hu M, Zhu M, Tighe S, Chen S, et al. 2017;56(5):769–73.
Human corneal endothelial cells expanded in vitro are 97. Peh GSL, Ang H-P, Lwin CN, Adnan K, George BL,
a powerful resource for tissue engineering. Int J Med Seah X-Y, et al. Regulatory compliant tissue-engi-
Sci. 2017;14(2):128–35. neered human corneal endothelial grafts restore cor-
91. Bartakova A, Alvarez-Delfin K, Weisman AD, Salero neal function of rabbits with bullous keratopathy. Sci
E, Raffa GA, Merkhofer RM, et  al. Novel identity Rep. 2017;7(1):14149.
and functional markers for human corneal endothelial
cells. Invest Opthalmol Vis Sci. 2016;57(6):2749.
Corneal Endothelial Cell
Transplantation: Animal Models
29
Brad P. Barnett and Albert S. Jun

29.1 Introduction Historically, animal models have played a


vital role in the study of penetrating keratoplasty
Over the past decade, endothelial transplantation and more recently in the study of endothelial
has replaced full-thickness corneal transplanta- transplantation. Notably, Samuel Bigger was the
tion as the definitive treatment for endotheliopa- first to report on corneal transplant in an animal
thies. In addition to partial-thickness model in 1837 in which he performed a penetrat-
transplantation, emerging alternative therapies ing keratoplasty on a domesticated gazelle [25,
include the use of cultured cells either to form 26]. During this same period, Eduard Zirm and
grafts or as injectable cellular therapy, as opposed others experimented on penetrating keratoplasty
to relying on donor tissues as the sole means of in chickens and rabbits [27, 28]. Since that time,
cellular replacement. In this chapter we will animal models have been recognized as essential
review the history of animal models used in to the understanding of the process of endothelial
endothelial transplantation studies as well as gen- graft rejection, in addition to aiding in the cre-
eral insights from full-thickness penetrating kera- ation of possible therapeutics.
toplasty in various animal models. Care will be The majority of initial research efforts in this
taken to discuss how research groups have area were tailored to help improve transplantation
matched experimental aims to the animal model success in human penetrating keratoplasty by bet-
best suited to achieve those aims. To assist future ter understanding immunological graft rejection.
researchers in determining the best animal model The rabbit eye is a popular choice in the study of
for their purposes, the relative pros and cons of corneal grafting due its size, making it easily trans-
each animal model will also be reviewed. A sum- latable to human grafting. Moreover, the rabbit
mary of all animal models used to date for endo- displays various signs of rejection also seen in
thelial transplantation can be found in Table 29.1. humans such as retrocorneal membranes and epi-
thelial decompensation [27]. Despite its popular-
B. P. Barnett ity, various shortcomings of the rabbit model then
Wilmer Eye Institute at Johns Hopkins, Department led investigators to rodent animal models for
of Ophthalmology, Baltimore, MD, USA
endothelial transplantation [29, 30]. The rodent
A. S. Jun (*) model now serves as the staple in immunological
Wilmer Eye Institute at Johns Hopkins, Department
studies, whereas large animal models are primarily
of Ophthalmology, Baltimore, MD, USA
dedicated to studying graft rejection in [31, 32].
Division of Cornea, Cataract and External Eye
In addition to full-thickness grafts, the appli-
Diseases, Baltimore, MD, USA
e-mail: aljun@jhmi.edu cation of carriers for creation of human corneal

© Springer Nature Switzerland AG 2019 437


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_29
438 B. P. Barnett and A. S. Jun

Table 29.1  Summary of endothelial transplant animal models


Host Endothelial graft Procedure Author/reference
Mouse Human spheres with magnetic nanoparticles AC injection Mimura [1]
Mouse Mouse single cells AC injection Yamada [2]
Mouse Autologous mouse on Descemet’s membrane DSAEK Joo [3]
Rat Human spheres with magnetic nanoparticles AC injection Mimura [1]
Rat Human on rat devitalized cornea PKP Tchah [4]
Rabbit Human spheres with magnetic nanoparticles AC injection Mimura [1, 5, 6]
Rabbit Human on collagen carrier DSAEK Mimura [7]
Rabbit Rabbit on Descemet’s membrane DSAEK Lange [8]
Rabbit Human on collagen carrier DSAEK Lai [9]
Rabbit Human on poly(N-isopropylacrylamide) DSAEK Lai [10]
Rabbit Human on human stroma DSAEK Honda [11]
Rabbit Rabbit on silk fibroin carrier DMEK Vazquez [12]
Rabbit Human on amniotic membrane PKP Ishino [13]
Rabbit Bovine on homografts PKP Gospodarowicz [14, 15]
Rabbit Rabbit on homografts PKP Jumblatt [16]
Cat Autologous cat on human devitalized cornea PKP Proulx [17, 18]
Cat Bovine on homografts PKP Bahn [19]
Monkey Monkey on collagen carrier DSAEK Koizumi [20, 21]
Monkey Human neonatal on denuded human cornea PKP Insler [22–24]
AC anterior chamber, PKP penetrating keratoplasty, DSAEK Descemet stripping automated endothelial keratoplasty,
DMEK Descemet’s membrane endothelial keratoplasty

endothelial cell sheets has been reported by many sis. Starting with in  vivo analysis, the slit-lamp
laboratories during previous decades (Fig. 29.1). exam is useful for serially monitoring the trans-
In early animal transplantation studies of cul- parency of a graft, in addition to looking for
tured corneal endothelial cells, endothelial cells inflammation and possible immune rejection.
from humans were implanted on Descemet’s Not unlike clinical ophthalmology, using
membranes from rabbits [14–16], cats [15], mice pachymetry to monitor corneal thickness is also a
[3], and cattle [8, 14, 16] or on gelatin mem- useful way to monitor for corneal edema as a
branes [7] to generate full-thickness corneal marker of endothelial cell function. Anterior-­
grafts utilizing denuded homo-, allo-, or xeno- segment optical coherence tomography also pro-
grafts that were then transplanted with penetrat- vides a means of measuring corneal thickness
ing keratoplasty. Infant and human neonatal and provides precise cross-sectional images.
corneal endothelial cells have been employed Specular microscopy can also be employed to
with success to create such grafts [22–24]. More provide endothelial cell count and morphology.
recent efforts have focused upon partial-­thickness Postmortem techniques allow for histological
endothelial transplantation in animal models as examination which provides a whole host of
well as cellular therapeutics designed to be information not readily available with in  vivo
injected into the anterior chamber [1, 5–7, 11]. assessment modalities. For example, histological
analysis enables direct verification of the pres-
ence of inflammatory cells, fibrosis, and
29.2 In Vivo and Postmortem ­neovascularization [32, 34, 35]. Moreover, it
Assessment allows the researcher to assess for repopulation
of the graft with the host’s stromal keratocytes
Methods of assessing the viability of a graft and and to assess for the presence of nerve regrowth
the presence of rejection are essential to any ani- into the graft [32, 34, 35].
mal model. Efforts to ascertain the above can be In this chapter we will discuss both transplan-
achieved by both in vivo and postmortem analy- tation of graft tissue and the burgeoning field of
29  Corneal Endothelial Cell Transplantation: Animal Models 439

a b

c d

e f

Fig. 29.1  Morphology of human corneal endothelial were seeded onto various carriers to generate grafts for
cells (HCECs) cultured in monolayer (a) and HCECs penetrating keratoplasty (b, c) or Descemet stripping with
seeded onto a carrier to generate a bioengineered endothe- automated endothelial keratoplasty (d–f). Scale
lial graft (b–f). (a) 62-year-old donor HCECs demonstrate bars = 100 μm. (Used with permission [33])
the typical hexagonal morphology. (b–f) Cultured HCECs
440 B. P. Barnett and A. S. Jun

cell therapy in which cells or clusters of cells are also be employed [44]. Unlike indirect methods
transplanted into the anterior chamber to repopu- of cell labeling, direct methods do not provide
late the endothelium. Broadly, within any field of any information about cell viability, as the fluo-
experimental transplantation, a means of tracking rescent agent will still produce signal in a dead
the fate of transplanted cells is necessary. cell. In addition, in the setting of dividing cells,
Historically, the primary means of determining direct methods suffer from the dilution of con-
the fate of transplanted cells was through histo- trast as the total amount of contrast agent remains
logical assessment. Histologically Y chromo- constant, while the number of cells increases.
somes could be detected when grafting male cells Moreover, in certain instances contrast agent is
in a female host [14, 15, 36], and in the case of unequally distributed between the parent and
xenogeneic transplantation, native proteins only daughter cell, with the parent cell retaining the
present in the graft species could be detected in majority of contrast [41].
the detection of a human protein when a human Indirect cell labeling methods require genetic
graft is transplanted in swine [37]. Transgenically alteration of the cell to introduce a reporter gene
expressed reporter genes have also been employed [45]. The reporter gene may code for a fluorescent
for histological analysis such as beta-­or bioluminescent protein, an enzyme, or a recep-
galactosidase and alkaline phosphatase [38, 39]. tor. In the event of the latter, it is the interaction
The use of histological analysis to track cell between the substrate and the encoded reporter
fate has a number of shortcomings. First and protein that provides an optical signal that enables
foremost it prevents real-time serial monitoring in  vivo detection. As reporter genes will be
of graft function and correlation of graft viability expressed by parent and daughter cell alike, no
with other measures of corneal health. With an dilution effect exists. Moreover, as the genes will
acknowledgment of these limitations, various only be expressed by live cells, a means of moni-
groups have developed in vivo cell imaging tech- toring in  vivo viability is also achieved. Various
niques. Due to the optical clarity of the cornea, reporters utilized for assessing the health of cor-
fluorescent, bioluminescent, and other cell track- neal grafts include the use of bioluminescent
ing techniques that are detected optically are imaging of firefly luciferase and fluorescent imag-
ideal for monitoring corneal endothelial trans- ing of green fluorescent protein [3, 13, 46].
plants. In order to create a cell that is fluorescent
or bioluminescent, direct or indirect cell methods
can be employed. 29.3 Clinical Criteria of Corneal
Compared to indirect methods, direct cell Graft Rejection
labeling methods are more straightforward and
do not involve genetic modification of the cell Rejection criteria differ and may not be translat-
[40]. In direct cell labeling, an agent is introduced able from tissue to tissue. For example, skin graft
into or attached to cells prior to transplantation. rejection is an obvious event with marked tissue
Such labeling can be achieved with transfection sloughing and discoloration. Conversely, corneal
agents [40, 41], electroporation [42], and anti- rejection is commonly more subtle and character-
body labeling [43]. Following transplantation, ized largely by opacification of tissue without
the labeling agent can be imaged as a surrogate sloughing, although sloughing has been wit-
marker of the grafted cells. For optical imaging, nessed in small xenografts [32] . Short-term cor-
quantum dots and other fluorescent semiconduc- neal opacification has been observed in both mice
tor nanocrystals can provide a robust fluorescent and rats. Applying clinical criteria for graft rejec-
signal when either loaded into cells or attached to tion, this temporary opacification can ultimately
the surface with antibodies. Fluorochromes can be significant enough to qualify as complete
also be employed in a similar manner [40]. The rejection [47]. However, reversal of this presenta-
use of contrast agents visible on optical coher- tion has been observed after anywhere from
ence tomography, such as gold nano-rods, could many days to weeks later [32, 47].
29  Corneal Endothelial Cell Transplantation: Animal Models 441

Some researchers feel that a mild degree of mounts has been the primary means of evalua-
corneal opacification does not necessarily imply tion. In the future, in vivo specular and confocal
a terminal event. In a seminal study, grafts that microscopy will likely be widely available for
were undergoing rejection as measured by murine models. Indeed, researchers have already
opaqueness did not necessarily lead to terminal begun using such techniques in smaller order ani-
graft failure. For these experiments, grafts were mals such as a recent study utilizing specular
obtained from C57Bl6 mice that transgenically microscopy in the Chinese tree shrew [53], an
expressed green fluorescent protein. These grafts animal comparable in size to the rat. Moreover,
were then transplanted into a Balb/c mouse host. new techniques designed to adequately image
In grafts that developed permanent opacification, corneal endothelial cells in rats such as micro-­
all donor endothelium were lost as no green fluo- optical coherence tomography are currently
rescent protein was seen with confocal micros- being employed shortly after sacrifice in other-
copy [48]. Interestingly, some grafts displayed wise intact rat eyes [54]. In the not too distant
transient opacification between 10 and 21  days future, it is likely similar techniques will be
but then ultimately recovered clarity. These grafts employed for serial in vivo assessment of murine
retained donor endothelium despite a transient endothelial cells.
opacification episode [48]. In this study and a
series of follow-up studies, it was found that a
minimum of 48 percent endothelial cell coverage 29.4 I nsights in Corneal Graft
was present at all times in the mouse corneas that Rejection from Animal
eventually cleared [48, 49]. Taken together, in the Models
murine model, although corneal opacity is cer-
tainly a strong predictor of graft failure, it is The brain, testes, and the eye are unique in terms
important to assess both the grade of corneal of transplant sites as they are immune-privileged.
opacity and wait a sufficient time after transplan- Due to this fact, the anterior chamber has proven a
tation (>8 weeks in the mouse) before diagnosing hospitable location for host-incompatible grafts
irreversible graft rejection [48, 49]. that would otherwise be rejected in other graft loca-
In addition to frank opacification of corneal tions [55, 56]. This well-accepted immune privi-
transplants, ophthalmologists regard the presence lege has been coined anterior c­ hamber-­associated
of epithelial rejection lines and the even more immune deviation (ACAID). ACAID has been
ominous endothelial rejection, or Khodadoust, exploited not only in the case of endothelial cell
lines as indicative of rejection [50]. Comparable transplantation, but has also been explored for allo-
rejection episodes are seen in animal experiments. geneic pancreatic islet cell transplantation in
In these studies, the development of rejection baboons [57] with early clinical studies being per-
lines was predictive of total graft rejection, espe- formed in humans by Perez’s group in an ongoing
cially when not adequately treated with immuno- study.
suppression [51, 52]. To emphasize this deleterious In addition to allotransplantation, the eye has
outcome, in the presence of Khodadoust lines, all also proven as a potential site for xenogeneic
sheep eventually developed completely scleral- grafts. For example, when human endothelial
ized and opaque white cornea [32]. cells on a collagen carrier were transplanted into
Although presence of corneal rejection lines rabbits, slit-lamp microscopy revealed no overt
or corneal opacification is useful for assessing signs of rejection such as fibrin deposition or
graft rejection, for endothelial transplantation the keratic precipitates for the observed period of 1
most reliable diagnostic tests involve direct eval- month (Fig. 29.2) [33]. It remains unclear to what
uation of the endothelial monolayer. To this end, degree using collagen carriers as opposed to
specular microscopy and confocal ophthalmos- human stroma helped prevent rejection in this
copy can be utilized in the large animal models. instance. Endothelial transplantation in this set-
In mice and rats, microscopy of corneal flat ting is unique as the xenogeneic transplanted
442 B. P. Barnett and A. S. Jun

a
(µm)
1400

1200
Corneal thickness

1000

800 *

600
† † † †
400

200

0
preop 0.5 1 3 7 14 21 28
Time (days)
b c

Fig. 29.2  Change in central corneal thickness (a) and cor- or carrier alone (open circles) after stripping of Descemet’s
neal appearance through slit lamp (b, c) after transplantation membrane. Twenty-eight days after transplantation, slit-
of HCEC seeded on collagen carrier DSAEK grafts. (a) lamp photography of the DSAEK group revealed a thin cor-
There was a significant difference in rabbit corneal thickness nea without stromal edema (b), while rabbits that received
between groups (*p  <  0.05 and †p  <  0.01) that received transplantation of collagen carrier alone had severe corneal
DSAEK grafts of HCECs on collagen carrier (filled circles) edema (c). (Used with permission [33])

cells do not directly contact host cells, but instead cultured allogeneic corneal endothelial cells
face the anterior chamber. showed little in the way of rejection [58].
Some evidence exists in mice that partial-­ Specifically, no mixed lymphocyte reaction or
thickness grafts are better tolerated than full-­ delayed hypersensitivity to assess rejection was
thickness grafts in terms of rejection. Hayashi detected. In contrast, full-thickness corneal
et  al. demonstrated in the mouse that chimeric allografts exhibited a high rate of rejection.
corneal endothelial cell allografts produced from Conversely, when allogeneic epithelium and
29  Corneal Endothelial Cell Transplantation: Animal Models 443

stroma from BALB/c mice were reconstituted plant approaches including cellular therapeutics
with immortalized C3H/HeN mouse corneal and bioengineered corneal grafts. In the latter
endothelial cells and transplanted into C3H/HeN case, selection of the animal model that most
recipients, a similar rate of rejection occurred as closely approximates human anatomy and physi-
compared to full-thickness allografts. Taken ology is desired to easily facilitate the transfer of
together, Hayashi’s group suggested these results developments from animal model to humans
demonstrate that endothelial cells alone neither when creating corneal endothelium. The primary
help promote tolerance nor induce allograft rejec- goal of using animal models is to assess the pro-
tion. Instead, they argued it is the non-endothelial cess and potential therapeutics for graft rejection,
component of the full-thickness allografts that in addition to exploring and evaluating cellular
induce rejection [58]. therapeutics as well as bioengineered corneal
Numerous groups have proposed mechanisms grafts. Particular attention is necessary to evalu-
of corneal endothelial allograft acceptance. This ate the multiple factors impacting the rejection of
tolerance has been attributed to immunologic corneal grafts in humans. Despite not being HLA
ignorance of draining lymph node T-inducer matched, most human corneal grafts demonstrate
priming cells and active immunosuppression a substantial acceptance rate1 year out from
with T-cell regulatory induction and T-cell anergy grafting [65]. Principles to optimize corneal
[59–63]. Mice exhibiting long-term acceptance transplantation in cat, rabbit, and the outbreed
of full-thickness corneal transplantation reveal sheep model have proven translatable to optimiz-
how active immunosuppression is induced [2, 47, ing transplant in humans [32, 65, 66].
59]. However, active immunosuppression failed The potential for expanding focused immuno-
to be induced in corneal endothelial cell logical studies related to rejection of corneal grafts
allografts, as demonstrated by adoptive transfer has been attainable with the use of the rat model
of splenocytes from mice that had accepted chi- [29] and particularly the use of the mouse model
meric corneal endothelial cell allografts and [32]. Due to their extensive use in immunological
regraft challenge in corneal endothelial cell research, much is known about these processes in
allograft-accepted mice [2, 47, 59]. rodents. Moreover, numerous strains are available
ACAID induction was demonstrated by including both inbred and congenic strains. In
Niederkorn’s group and Amano’s group after the recent years, numerous gene knockout strains
injection of corneal endothelial cells into the including conditionally deleted strains as well as
anterior chamber [58, 64]. ACAID induction was mice that transgenically express cell reporters
not demonstrated however after corneal endothe- have been developed. Using these various models,
lial cell allograft transplantation [58]. These researchers can tailor the rodent strain chosen to
results indicate that corneal endothelial cell graft the specific research question at hand.
survival without rejection was not associated However, clinically both the mouse model and
with immunosuppression or T-cell anergy. the rat model exhibit important anatomic vari-
Immunologic ignorance appears to be the pri- ances when compared to human grafts. Due to
mary reason for rejection-free acceptance of chi- the small size of rodents, advanced surgical skill
meric corneal endothelial cell allografts [58]. is necessary to avoid the mouse or rat being sub-
jected to procedural trauma that thereby induces
the innate immune response. Regardless, when
29.5 Choice of Animal Model looking at the models, particularly the mouse
model, both have provided a wealth of under-
In general ethics dictates the use of the animal of standing of the pathogenesis of immune-related
lowest order that will achieve the research aims. rejection and specifically the value of indirect
Primary applications of animal models are to allorecognition [60, 67].
assess basic processes of and potential therapeu- For ethical and cost considerations, primate
tics for graft rejection and to develop novel trans- transplantation is generally reserved for treat-
444 B. P. Barnett and A. S. Jun

ments that have been vetted in animal models of The mouse model, like other species, demon-
lower order. Larger animals provide not only ease strates rejection through corneal opacification.
of surgical intervention but also the ability to On the other hand, the rat model presents addi-
employ clinical techniques familiar to the oph- tional signs like corneal edema, corneal vascular-
thalmologist, such as slit-lamp biomicroscopy ization, and inflammatory cell invasion.
and confocal microscopy, in order to assess for Comparing a grading system incorporating these
graft rejection and investigate changes in other signs of rejection against the use of the
endothelium. degree of opacification alone has indicated that
the degree of opacification is correlated with the
extent of rejection [30, 47, 69–75]. Corneal
29.6 Types of Animal Models opacification alone is inadequate to identify non-
reversible rejection. Temporary corneal opacifi-
29.6.1 Rodent cation secondary to the extent of procedural
trauma often occurs in the both the rat and mouse
Primarily due to the size of the eyes, small model within days of graft placement. Instead the
rodent animals such as the mouse and rat are degree of opacification and the length of time
not the preferred models for preclinical studies, opacification present are necessary in order to
even though they have been utilized to study determine nonreversible rejection.
corneal endothelium tissue engineering [3, 4, A vast amount of information on the process
68]. The rabbit model has been effective for of immune graft rejection has been obtained by
functional in vivo assessments due to the acces- studying the mouse model. Nevertheless, the dis-
sibility, cost, small size, and ease of handling. similarity of the mouse immune system to that of
However, controlled preclinical studies on the human, as well as the inherent size discrep-
endothelial tissue engineering are impeded by ancy between the species, has made the use of
the capacity of the rabbit corneal endothelium larger animal models essential.
to proliferate extensively in vivo. For this rea-
son, rodent models are increasingly being
employed. 29.6.2 Rabbit
Corneal transplantation in rodents has proven
invaluable in the understanding of xenorejection The rabbit model was the earliest model to be
and allorejection. In a study by Amano’s team, used in a standardized study of corneal grafts and
human corneal endothelial cells were seeded secondary lymphoid tissue [76]. Beginning in the
onto a carrier to make a full-thickness graft, and 1930s, corneal grafts were effectively placed in
penetrating keratoplasty was performed in immu- rabbits [66, 77] with subsequent reports presented
nodeficient nude rats [68]. In the setting, the in literature [78, 79]. In addition to penetrating
xenogeneic graft was well tolerated with few keratoplasty, rabbits were also employed for pos-
signs of rejection. As opposed to nude rats, in the terior lamellar keratoplasty [80], limbal epithelial
work by She et  al. of murine corneal grafts, a grafts [81], and limbal stem cell grafts [82, 83].
robust immune response was seen. In this model, Toward the end of the 1970s, the first tissue-­
a 2  mm donor graft is transferred to a 1.5  mm engineered cornea was constructed by
host bed and secured with 8–12 sutures [30]. Gospodarowicz and Greenburg [14, 15]. They
Several groups have adopted this model with an seeded cultured bovine corneal endothelial cells
aim of understanding the immune mechanisms at onto a cell carrier and transplanted into rabbits
play during graft rejection [30, 47, 69–75]. The after endothelium removal [14, 15]. In the fol-
variety of genetically manipulated strains, as well lowing years, Amano’s team pushed the field fur-
as the plethora of scientific reagents available ther toward clinical application when they
specifically for mice, have proven quite helpful in utilized cultured human corneal endothelial cells
this aim. on a carrier to make full-thickness grafts [84, 85].
29  Corneal Endothelial Cell Transplantation: Animal Models 445

In 2004, using collagen carriers for human cor- to prevent the endothelial, stromal, or epithelial
neal endothelial cells, Mimura et  al. [33] per- cells of the carrier from contaminating the graft.
formed DSAEK in the rabbit (Fig. 29.3a–c, e–g). This process of native cell destruction is antici-
They followed this study by seeding cultured pated to decrease carrier immunogenicity [86,
human corneal endothelial cell suspensions onto 87]. Preservation of corneal transparency was
human stroma and again performed DSAEK in recognized by Amano et al. when liquid nitrogen
the rabbit [11]. gas was employed to decellularize the corneal
Additional groups have also explored cultured stroma [88]. Engelman et  al. discovered that
corneal endothelial cells to create grafts that were denuding Descemet’s membrane without com-
then transplanted with DSAEK in a rabbit model. promising cell adhesion could be accomplished
Using human endothelial cells seeded onto a in a single freeze-thaw cycle [89]. Utilizing
poly(N-isopropylacrylamide) and gelatin-based ammonium hydroxide chemical debridement
carrier, Lai et  al. demonstrated uniform corneal yielded comparable results [17, 18] where well-­
transparency within 2  weeks of DSAEK in the organized native collagen fibers, essential for
rabbit [10]. Similarly, Vasquez et  al. employed corneal transparency, were preserved. These
silk fibroin films to culture both rabbit and human examples of devitalized stromal carriers provided
corneal endothelial cells. Utilizing a DMEK-like desirable tissue-engineered endothelium adher-
procedure, artificial silk fibroin endothelial grafts ence and were favorable for the recipient after
were transplanted in rabbits [12]. At the 6-week full-thickness transplantation in the rabbit.
mark following transplantation of the grafts in In lieu of full-thickness or even partial-­
the rabbit model, both normal thickness and thickness transplantation, some groups have
transparency of the cornea returned. Optical championed the use of cellular therapeutics con-
coherence tomography of the cornea with the silk sisting of single cells or clusters of cells injected
fibroin endothelial graft compared to the unaf- into the anterior chamber (Fig. 29.3d, h). Prior to
fected contralateral eye demonstrated compara- anterior chamber injection, various techniques
ble thickness of the cornea. The silk fibroin graft have been utilized to remove the diseased
was found to be well incorporated into the host endothelium from the host. Cryoablation has
­
tissue and displayed proper intercellular junction been explored, but this procedure has demon-
and cellular pump function [12]. Moreover, his- strated irreversible damage to keratocytes [90].
tological analysis revealed the silk fibroin graft Alternative techniques include Descemet’s mem-
was attached and well integrated with host tissue brane stripping, enabling cells injected into the
with little inflammatory reaction. Additionally, anterior chamber to directly contact the host
rabbit corneal endothelial cells displayed appro- stroma [68]. With this technique, proper head
priate monolayer morphology and further positioning for a sufficient time is required in
expressed ZO-1 and Na+-K+ ATPase as deter- order to allow the transplanted cells to settle and
mined by immunohistochemistry [12]. adhere on the posterior corneal surface and to
Another study undertaken in the rabbit minimize adherence to the trabecular meshwork,
involved making a 7 mm full-thickness trephina- iris, or lens.
tion in the rabbit host [13]. To the excised button, To help ensure proper localization of trans-
Ishino et  al. then performed a DSAEK-like planted cells to the posterior cornea, Mimura
lamellar procedure with a graft of human corneal et al. [91] and Patel et al. [92] have recommended
endothelial cells on an amniotic membrane car- loading iron particles or superparamagnetic
rier. The corneal button was then sewn back into microspheres into cells prior to transplantation.
the host. With this technique, graft transparency These techniques have been previously explored
was maintained for 1 week. for different applications, namely, labeling cells
Devitalization is a novel approach as an option with an iron-based MRI contrast agent [41, 42].
other than use of fresh native tissue. Devitalization Following injection into the anterior chamber,
functions to remove living cells from the carrier these iron-loaded cells are attracted to the poste-
446 B. P. Barnett and A. S. Jun

a b

c d

e f

g h

Fig. 29.3  Slit-lamp (a–d) and histologic (e–h) photo- graft comprised of HCECs seeded onto rat corneal stroma
graphs of corneas posttransplantation of bioengineered denuded of endothelium. (c, g) Slit-lamp photography and
HCEC grafts or injected into the anterior chamber. (a, e) histology of rabbit cornea post DSAEK with a graft com-
Slit-lamp photography and histology of rabbit corneas prised of cultured HCECs seeded onto human amniotic
after full-­thickness transplantation of PK grafts generated membrane. (d, h) Slit-­lamp photography and histology of
from cultured HCECs on rabbit corneal stroma denuded rabbit cornea after in  vivo corneal endothelial stripping
of endothelium. (b, f) Slit-lamp photography and histol- followed by cultured HCECs injected into the anterior
ogy of a rat cornea after penetrating keratoplasty with a chamber. Scale bars = 50 μm. (Used with permission [33])
29  Corneal Endothelial Cell Transplantation: Animal Models 447

rior corneal surface by an external magnet. Using the graft. A similar exclusion of immune ele-
this approach, complete resurfacing of the poste- ments is also achieved in cell microencapsulation
rior cornea, both in  vitro [92] and in  vivo, has for allotransplantation and xenotransplantation
been demonstrated in the rabbit model [92]. [37]. Due to its biodegradable nature, with time
However, safety factors of long-term iron require the collagen disc was broken down thereby pro-
further study. viding direct contact of the transplanted endothe-
As opposed to single cell suspensions, in a lial cells to the anterior chamber in the proper
serious of work from Amano et  al., endothelial apical-basal orientation [10].
cells were coaxed to form spheroid bodies in cul- Various factors have been postulated to con-
ture [5, 6, 93]. These spheroid bodies were then tribute to graft rejection in the rabbit. Not unlike
injected into the anterior chamber [93]. Prone human transplants, the presence of vasculariza-
position postsurgery was necessary for these tion at the host-graft junction at the time of trans-
spheroid bodies to attach to the posterior corneal plant is associated with increased rates of
surface. Subsequent restoration of endothelial rejection [95]. The rate of graft rejection has also
function as noted by resolution of corneal edema been shown to increase when a skin graft from
was observed [93]. the donor was transplanted to the host following
The rabbit model has also been explored for prior corneal transplantation from the same donor
transplantation of human corneal endothelial [52].
cells on electron beam-grafted poly(N-­ Rabbits are known to have rapid and extensive
isopropylacrylamide)-based cell culture carriers. clotting of the aqueous when the anterior cham-
Following stripping of endothelium and trans- ber is entered [52]. Although possibly under-­
plant of these grafts, corneal thickness signifi- identified, a similar occurrence has not been
cantly decreased, as compared to control groups reported in other species. Blocking this anterior
[10]. In this study, the researchers encountered chamber fibrin response with intracameral hepa-
frequent tearing of the graft during transfer due to rin during surgery and topical heparin postsur-
its fragile nature [10]. In order to overcome this gery has proven effective in decreasing the rate of
fragility, a circular poly(vinylidene fluoride) rabbit allograft rejection [76].
membrane was utilized to stabilize the cell sheet In addition to identifying the now eponymous
during transfer. The use of this poly(vinylidene endothelial rejection lines, Khodadoust uncov-
fluoride) membrane protected the graft during ered a variety of factors that influence graft rejec-
transfer by enabling adherence to the posterior tion in the rabbit. He demonstrated that the size
corneal stroma in the proper orientation [10]. of the graft and the location of the graft contrib-
Use of a gelatin disc to facilitate transfer of an uted to rejection. For example, with 7 mm grafts
endothelial graft into a rabbit after Descemet in the rabbit, widespread rejection has been
stripping has been described [10]. Importantly observed [76]. However, smaller grafts of
collagen was not used as a carrier containing 4–5 mm did not appear to induce rejection with
attached cells. Instead these collagen discs faced grafts maintaining continued transparency [52].
the anterior chamber and sandwiched the endo- In this same study, he also found that grafts
thelial graft to the recipient stroma. The porous placed closer to the limbus demonstrated higher
nature of the gelatin disc enabled transfer of rejection rates, especially when vascularization
nutrients from the aqueous humor to the graft and occurred [52].
allowed cellular waste products from the endo- In the rabbit model, Khodadoust also demon-
thelial cell sheet to escape into the anterior cham- strated the use of heparin to prevent fibrin forma-
ber [9, 94]. tion and early suture removal to prevent
Although not directly investigated, it is also neovascularization impacted rejection rates.
possible that this semi-porous collagen disc also Although lamellar grafts had a better prognosis
prevented interaction of immune cells circulating than penetrating grafts, rejection was commonly
in the anterior chamber to directly interact with seen in both when sutures were left in place long
448 B. P. Barnett and A. S. Jun

enough to induce vascularization [34]. With care- 29.6.3 Pig


ful surgical technique to minimize trauma, appro-
priate suture removal, and the use of topical The pig model has been of subject of interest in
heparin to prevent clotting during the procedure, ophthalmic research due to its anatomic and
Khodadoust achieved successful grafts in 95% of physiologic parallels to the human eye.
7 mm grafts and 100% of 5 mm grafts [66]. Additionally, the affordable cost and easy acces-
Although not fully understood, it has been sibility of pig tissue further support interest in
proposed that neovascularization of the graft this model. The cornea of an adult pig has a hori-
leads to rejection as the efferent arm of the zontal diameter that typically measures from 14.5
immune response can be carried directly into the to 16.5 mm [100, 101] with a central pachymetry
graft [50]. An alternate or potentially simultane- of approximately 665 um [102, 103]. The endo-
ous mechanism of graft rejection involves the thelial cells of pigs exhibit minor proliferative
increased afferent cell traffic to the draining ability [104] with endothelial cell counts averag-
lymph node when graft neovascularization occurs ing approximately 4000 cell/mm2 [105]. Various
thereby further driving vasculogenesis and lym- stem cell [106] and other corneal endothelial cell
phangiogenesis [96]. Regardless of the mecha- culture studies [46] have utilized the pig cornea,
nism, early suture removal has proven to help and the majority of the development of refractive
prevent neovascularization. In regard to suture corneal procedures have been developed in the
choice, various groups have used interrupted and pig model [107–109]. Despite the wide interest in
continuous sutures of varying sizes (8-0 to 10-0) and popularity of the pig model, this model is not
[97]. Certain groups have suggested that continu- commonly utilized for full-thickness or partial-­
ous sutures pose less of a nidus for neovascular- thickness corneal transplantation studies.
ization [97]. Most studies have employed
polyglactin or another dissolvable suture.
Nevertheless, as many researchers have seen vas- 29.6.4 Cat
cularization crossing the host-graft junction in as
little as 9–10  days [97], the use of nylon or Although less common than previously described
another non-dissolvable suture with early suture animal models, corneal transplantation has been
removal may be advantageous. explored in the cat. Various groups have utilized
In the rabbit, typically rejection is first marked cat corneal endothelium to create tissue-­
by neovascularization of the graft [50]. In a series engineered grafts [17, 18, 110]. Cats have also
or early experiments, double exchange of corneal been used as the recipient of both xenogeneic and
grafts between pairs of rabbits with selective allogeneic endothelial transplants [14, 15, 17–19,
removal of one or more layers of cornea was 100, 110, 111]. The average adult cat has a cor-
undertaken [51]. This enabled the regrafting of neal diameter that ranges from 15.5 to 18  mm
donor-host chimeric grafts to the original donor. [100] with central pachymetries ranging from
In these experiments, it was the epithelium that 545 to 650  μm [112–116]. Not unlike humans,
was rejected first [51]. Accordingly, the follow- cat endothelial cells generally do not replicate
ing vascularization epithelial rejection lines are [112, 117, 118]. On average cats have endothelial
commonly seen during rejection. Subsequent to cell counts approximating 2300–2900 cells/mm2
the development of an epithelial rejection line, an [112–116]. In contrast to the pig model, posterior
endothelial rejection line develops in the rabbit. lamellar transplant [119] and penetrating kerato-
This Khodadoust line consists of a combination plasty [19, 31, 120–123] have been successfully
of mononuclear cells and dying endothelial cells achieved.
[51]. A terminal event in rabbit allograft rejection Brunette et  al. [124] favored the cat model
is the development of a retrocorneal membrane. over the pig model for full-thickness corneal
This membrane is formed when the failing endo- transplantation, as the desired surgical technique
thelium is engulfed by inflammatory cells and was easier to accomplish. Additionally, Brunette’s
spindle-shaped fibroblastoid cells [98, 99]. groups have performed a series of studies to
29  Corneal Endothelial Cell Transplantation: Animal Models 449

determine the minimal endothelial biopsy size heparin to prevent anterior chamber clotting [32].
necessary to construct a full corneal endothelium. Allograft rejection occurred on average at 20 days
In the cat model, a 5 mm biopsy (representative of survival time with related significant vascular-
of about 8% of the cat’s endothelial eye surface) ization [32]. The presence of fibrin was the first
was necessary to reconstruct a full corneal endo- reported sign of clinical rejection followed by
thelium [125]. Brunette’s group also revealed keratic precipitates [67].
that the functional endothelium of grafted cor-
neas was minimally impacted by injected corneal
endothelial cells in the cat model. With or with- 29.6.6 Monkey
out central endothelial scraping, feline corneal
endothelial cells supplied with Y-27632 and For obvious reasons, the ideal research model for
marked with 3,30-dioctadecyl-5,50(4-­ human application is the primate. There have
sulfophenyl)oxacarbocyanine were injected in been reports of engineered corneal endothelium
cats. Results revealed that the healthiest endothe- in monkeys [20, 21, 126] as well as endothelial
lium was achieved with injection of Y-27632 transplants of both human and pig xenografts
without corneal endothelial cell scraping [124]. [22–24] and monkey allografts and homografts
Use of the cat model has been instrumental in [20, 21]. The corneal diameter in the full-grown
assessing endothelial cell counts to evaluate graft rhesus monkey approximates 10.6  mm with a
survival and graft rejection. Evaluation of the pachymetry of 470 μm at the center of the cornea
allograft 1 month post procedure revealed an [127]. Although described endothelial cell counts
average of 30% cell loss over the entire donor approximating 4000 cells/mm2 [128, 129] exceed
area. Conversely, homograft endothelial cell loss those found in humans, replication in animal
was 15% [31]. Other cat model experiments con- models has not been demonstrated [24, 117, 130].
centrated on removal of the endothelium from the The availability and cost of primates in both
donor before grafting, and no endothelium regen- North America and Europe are limiting factors in
eration was observed in the host [31]. Further sup- the use of this model.
porting these outcomes, corneal homograft cell In Wuzhishan pigs, the survival of pig-rhesus
counts decreased by about one-half over a course corneal xenografts was explored after donor bone
of 6 weeks to 9 months after the graft was placed marrow transplantation. The pigs were utilized as
[98] with a notable polymegathism increase. A donors, while the rhesus monkeys received corneal
rationale for this progressive cell loss following xenotransplantation. The study group was com-
homografting has not been identified [50]. prised of 12 rhesus monkeys separated into two dif-
ferent groups. The treatment group was given
cyclophosphamide intravenously with subsequent
29.6.5 Sheep autologous bone marrow transplantation. The con-
trol group received cyclophosphamide intrave-
Although sheep models of endothelial cell trans- nously alone. Slit-lamp microscopy was used to
plantation are lacking, sheep are the latest large evaluate the xenografts. The recipient’s immuno-
animal model to be utilized for research focusing logical status was analyzed after transplantation by
on corneal graft rejection [32]. In a Morino out- assessing for a mixed lymphocyte reaction, the
bred strain, 12  mm grafts were transferred presence of serum immunoglobulin and comple-
between animals. Fixation of the graft was ment, as well as chimerism formation [131].
obtained with 9-0 polyglactin sutures. Of note, The graft survival time for the bone marrow
the use of smaller 10-0 polyglactin sutures fre- transplant group was 36.0 ± 4.7 days and 17.1 ±
quently broke in the sheep model and thus was 3.2 days for the control group. The bone mar-
abandoned. In the sheep model, a weaker inflam- row transplant arm revealed decreased serum
matory response was observed compared to other immunoglobulin and complement levels as well
animal models thereby negating the need for as decreased mixed lymphocyte reaction.
450 B. P. Barnett and A. S. Jun

Additionally, the corneal xenografts in the bone aims. To date, the various animal models have
marrow transplant arm demonstrated negligible proven complementary in providing researchers
inflammatory cell infiltration and no eosinophil and clinicians with a means of developing new
infiltration. These results indicate that autolo- surgical techniques as well as assessing the func-
gous bone marrow transplantation may sup- tion of various corneal grafts and novel cellular
press immune reaction thereby prolonging therapeutics. Moreover, the various animal mod-
corneal xenograft survival and chimerism for- els have proven invaluable in providing insight
mation [131]. into the immunological processes that take place
Another study in primates undertaken by during corneal graft rejection. The future remains
Koizumi et al. involved cultured corneal endothe- bright for endothelial transplantation. We cer-
lial cells seeded onto a type I collagen carrier. tainly are indebted to the clinical and scientific
They transplanted these grafts through DSAEK pioneers and various animals for their contribu-
in monkeys [20, 21]. Although the transplanted tions to the field of corneal endothelial cell
sheets detached from the host cornea within 1 transplantation.
week post grafting, the corneas recovered trans-
parency within 6 months posttransplantation. Compliance Statements Brad P.  Barnett and Albert
Although not directly addressed by Koizumi’s S. Jun declare that they have no conflict of interest.
No human studies were carried out by the authors for
group, these results have three potential explana- this article.
tions. The first possibility is endothelial cells No animal studies were carried out by the authors for
separated from the detached graft and repopu- this article.
lated the cornea. Another more likely possibility
is the native host peripheral endothelial cells
repopulated the corneal endothelium and the References
graft itself had no impact on endothelial recovery.
A final possibility, ripe for exploration, is that the 1. Mimura T, et  al. Magnetic attraction of iron-­
endocytosed corneal endothelial cells to Descemet’s
transplanted graft provided cytokines or other membrane. Exp Eye Res. 2003;76(6):745–51.
factors that coaxed the native endothelial cells 2. Yamada J, et  al. Allogeneic corneal toler-
into division, thereby repopulating the cornea. ance in rodents with long-term graft survival.
Transplantation. 2005;79(10):1362–9.
3. Joo CK, et  al. Repopulation of denuded murine
Descemet’s membrane with life-extended murine
29.7 Conclusion corneal endothelial cells as a model for corneal cell
transplantation. Graefes Arch Clin Exp Ophthalmol.
Endothelial transplantation and more broadly 2000;238(2):174–80.
4. Tchah H.  Heterologous corneal endothelial cell
corneal transplantation have been explored in a transplantation  – human corneal endothelial cell
variety of animal models. In general, the various transplantation in Lewis rats. J Korean Med Sci.
animal models can be divided into small inbred 1992;7(4):337–42.
animal models and large outbred models. Smaller 5. Mimura T, et al. Sphere therapy for corneal endothe-
lium deficiency in a rabbit model. Invest Ophthalmol
animal models have the benefit of being rela- Vis Sci. 2005;46(9):3128–35.
tively inexpensive and easier to maintain. In addi- 6. Mimura T, et al. Necessary prone position time for
tion, numerous inbred strains are available with human corneal endothelial precursor transplantation
various mutations and novel reporters transgeni- in a rabbit endothelial deficiency model. Curr Eye
Res. 2007;32(7–8):617–23.
cally expressed. Larger animal models enable the 7. Mimura T, et al. Treatment of rabbit bullous keratop-
use of surgical techniques employed clinically athy with precursors derived from cultured human
with humans and moreover enable the use of corneal endothelium. Invest Ophthalmol Vis Sci.
grafts that are closer in size to human grafts. The 2005;46(10):3637–44.
8. Lange TM, Wood TO, McLaughlin BJ.  Corneal
choice of animal model is largely guided by the endothelial cell transplantation using Descemet's
experimental aims, and ethics dictates the use of membrane as a carrier. J Cataract Refract Surg.
the animal of lowest order that will achieve these 1993;19(2):232–5.
29  Corneal Endothelial Cell Transplantation: Animal Models 451

9. Lai JY, et al. Characterization of cross-linked porous tary on Eduard Zirm’s landmark paper of 1906. Br J
gelatin carriers and their interaction with corneal Ophthalmol. 2006;90(10):1222–3.
endothelium: biopolymer concentration effect. PLoS 26. Crawford AZ, Patel DV, McGhee C. A brief history
One. 2013;8(1):e54058. of corneal transplantation: from ancient to modern.
10. Lai JY, Chen KH, Hsiue GH.  Tissue-engineered Oman J Ophthalmol. 2013;6(Suppl 1):S12–7.
human corneal endothelial cell sheet transplanta- 27. Rycroft B. The corneal graft – past, present and future.
tion in a rabbit model using functional biomaterials. Trans Ophthalmol Soc U K. 1965;85:459–517.
Transplantation. 2007;84(10):1222–32. 28. Rycroft PV.  Corneal graft membranes. Trans
11. Honda N, et al. Descemet stripping automated endo- Ophthalmol Soc U K. 1965;85:317–26.
thelial keratoplasty using cultured corneal endo- 29. Williams KA, Coster DJ. Penetrating corneal trans-
thelial cells in a rabbit model. Arch Ophthalmol. plantation in the inbred rat: a new model. Invest
2009;127(10):1321–6. Ophthalmol Vis Sci. 1985;26(1):23–30.
12. Vazquez N, et al. Silk fibroin films for corneal endo- 30. She SC, Steahly LP, Moticka EJ.  A method for
thelial regeneration: transplant in a rabbit Descemet performing full-thickness, orthotopic, penetrat-
membrane endothelial Keratoplasty. Invest ing keratoplasty in the mouse. Ophthalmic Surg.
Ophthalmol Vis Sci. 2017;58(9):3357–65. 1990;21(11):781–5.
13. Ishino Y, et al. Amniotic membrane as a carrier for 31. Cohen KL, et  al. Cat endothelial morphol-
cultivated human corneal endothelial cell transplan- ogy after corneal transplant. Curr Eye Res.
tation. Invest Ophthalmol Vis Sci. 2004;45(3):800–6. 1990;9(5):445–50.
14. Gospodarowicz D, Greenburg G.  The coating of 32. Williams KA, et al. A new model of orthotopic pen-
bovine and rabbit corneas denuded of their endothe- etrating corneal transplantation in the sheep: graft
lium with bovine corneal endothelial cells. Exp Eye survival, phenotypes of graft-infiltrating cells and
Res. 1979;28(3):249–65. local cytokine production. Aust N Z J Ophthalmol.
15. Gospodarowicz D, Greenburg G, Alvarado 1999;27(2):127–35.
J. Transplantation of cultured bovine corneal endo- 33. Mimura T, et  al. Corneal endothelial regenera-
thelial cells to rabbit cornea: clinical implications tion and tissue engineering. Prog Retin Eye Res.
for human studies. Proc Natl Acad Sci U S A. 2013;35:1–17.
1979;76(1):464–8. 34. Williams KA, et  al. A comparison of the effects
16. Jumblatt MM, Maurice DM, McCulley of topical cyclosporine and topical steroid on rab-
JP.  Transplantation of tissue-cultured cor- bit corneal allograft rejection. Transplantation.
neal endothelium. Invest Ophthalmol Vis Sci. 1985;39(3):242–4.
1978;17(12):1135–41. 35. Tuft SJ, Williams KA, Coster DJ.  Endothelial
17. Proulx S, et al. Tissue engineering of feline corneal repair in the rat cornea. Invest Ophthalmol Vis Sci.
endothelium using a devitalized human cornea as 1986;27(8):1199–204.
carrier. Tissue Eng Part A. 2009;15(7):1709–18. 36. Lagali N, et al. Donor and recipient endothelial cell
18. Proulx S, et al. Transplantation of a tissue-­engineered population of the transplanted human cornea: a two-­
corneal endothelium reconstructed on a devitalized dimensional imaging study. Invest Ophthalmol Vis
carrier in the feline model. Invest Ophthalmol Vis Sci. 2010;51(4):1898–904.
Sci. 2009;50(6):2686–94. 37. Barnett BP, et al. Magnetic resonance-guided, real-­
19. Bahn CF, et  al. Penetrating keratoplasty in the cat. time targeted delivery and imaging of magnetocap-
A clinically applicable model. Ophthalmology. sules immunoprotecting pancreatic islet cells. Nat
1982;89(6):687–99. Med. 2007;13(8):986–91.
20. Koizumi N, et  al. Cultivated corneal endothelial 38. Forss-Petter S, et  al. Transgenic mice express-
cell sheet transplantation in a primate model. Invest ing beta-galactosidase in mature neurons under
Ophthalmol Vis Sci. 2007;48(10):4519–26. neuron-specific enolase promoter control. Neuron.
21. Koizumi N. Cultivated corneal endothelial cell sheet 1990;5(2):187–97.
transplantation in a primate model. Nippon Ganka 39. Himes SR, Shannon MF.  Assays for transcrip-
Gakkai Zasshi. 2009;113(11):1050–9. tional activity based on the luciferase reporter gene.
22. Insler MS, Lopez JG.  Transplantation of cultured Methods Mol Biol. 2000;130:165–74.
human neonatal corneal endothelium. Curr Eye Res. 40. Barnett BP, et  al. Use of perfluorocarbon nanopar-
1986;5(12):967–72. ticles for non-invasive multimodal cell tracking
23. Insler MS, Lopez JG. Heterologous transplantation of human pancreatic islets. Contrast Media Mol
versus enhancement of human corneal endothelium. Imaging. 2011;6(4):251–9.
Cornea. 1991;10(2):136–48. 41. Walczak P, et  al. Applicability and limitations of
24. Insler MS, Lopez JG.  Extended incubation MR tracking of neural stem cells with asymmetric
times improve corneal endothelial cell trans- cell division and rapid turnover: the case of the shiv-
plantation success. Invest Ophthalmol Vis Sci. erer dysmyelinated mouse brain. Magn Reson Med.
1991;32(6):1828–36. 2007;58(2):261–9.
25. Armitage WJ, Tullo AB, Larkin DF.  The first suc- 42. Walczak P, et al. Magnetoelectroporation: improved
cessful full-thickness corneal transplant: a commen- labeling of neural stem cells and leukocytes for cellu-
452 B. P. Barnett and A. S. Jun

lar magnetic resonance imaging using a single FDA-­ 59. Sonoda Y, Streilein JW.  Impaired cell-mediated
approved agent. Nanomedicine. 2006;2(2):89–94. immunity in mice bearing healthy orthotopic corneal
43. Haldi M, et al. Human melanoma cells transplanted allografts. J Immunol. 1993;150(5):1727–34.
into zebrafish proliferate, migrate, produce melanin, 60. Akl A, Luo S, Wood KJ. Induction of transplantation
form masses and stimulate angiogenesis in zebrafish. tolerance-the potential of regulatory T cells. Transpl
Angiogenesis. 2006;9(3):139–51. Immunol. 2005;14(3–4):225–30.
44. de la Zerda A, et  al. Optical coherence contrast 61. Kang SM, Tang Q, Bluestone JA. CD4+CD25+ regu-
imaging using gold nanorods in living mice eyes. latory T cells in transplantation: progress, challenges
Clin Exp Ophthalmol. 2015;43(4):358–66. and prospects. Am J Transplant. 2007;7(6):1457–63.
45. Omoto M, et  al. Mesenchymal stem cells home to 62. Tao R, Hancock WW.  Regulating regulatory T
inflamed ocular surface and suppress allosensitiza- cells to achieve transplant tolerance. Hepatobiliary
tion in corneal transplantation. Invest Ophthalmol Pancreat Dis Int. 2007;6(4):348–57.
Vis Sci. 2014;55(10):6631–8. 63. Yang XF.  Factors regulating apoptosis and homeo-
46. Proulx S, et  al. Optimization of culture conditions stasis of CD4+ CD25(high) FOXP3+ regulatory
for porcine corneal endothelial cells. Mol Vis. T cells are new therapeutic targets. Front Biosci.
2007;13:524–33. 2008;13:1472–99.
47. Sonoda Y, Streilein JW.  Orthotopic corneal trans- 64. Niederkorn JY, Mellon J.  Anterior chamber-­
plantation in mice  – evidence that the immunoge- associated immune deviation promotes corneal
netic rules of rejection do not apply. Transplantation. allograft survival. Invest Ophthalmol Vis Sci.
1992;54(4):694–704. 1996;37(13):2700–7.
48. Plskova J, et  al. Evaluation of corneal graft 65. The collaborative corneal transplantation
rejection in a mouse model. Br J Ophthalmol. studies (CCTS). Effectiveness of histocom-
2002;86(1):108–13. patibility matching in high-risk corneal transplan-
49. Plskova J, et al. Quantitative evaluation of the cor- tation. The Collaborative Corneal Transplantation
neal endothelium in the mouse after grafting. Br J Studies Research Group. Arch Ophthalmol.
Ophthalmol. 2004;88(9):1209–16. 1992;110(10):1392–403.
50. Khodadoust AA, Silverstein AM.  Local graft ver- 66. Khodadoust AA. Penetrating keratoplasty in the rab-
sus host reactions within the anterior chamber of bit. Am J Ophthalmol. 1968;66(5):899–905.
the eye: the formation of corneal endothelial pocks. 67. Amano S, Sawa M, Ishii Y. Keratoepithelioplasty in
Investig Ophthalmol. 1975;14(9):640–7. rat: development of a model and histological study.
51. Khodadoust AA, Silverstein AM.  Transplantation Jpn J Ophthalmol. 1992;36(4):407–16.
and rejection of individual cell layers of the cornea. 68. Mimura T, et  al. Transplantation of corneas
Investig Ophthalmol. 1969;8(2):180–95. reconstructed with cultured adult human cor-
52. Khodadoust AA, Silverstein AM.  Studies on the neal endothelial cells in nude rats. Exp Eye Res.
nature of the privilege enjoyed by corneal allografts. 2004;79(2):231–7.
Investig Ophthalmol. 1972;11(3):137–48. 69. He YG, Ross J, Niederkorn JY. Promotion of murine
53. Wu M, et al. Age-related changes of corneal endothe- orthotopic corneal allograft survival by systemic
lial cell in healthy Chinese tree shrew measured by administration of anti-CD4 monoclonal antibody.
non-contact specular microscope. Int J Ophthalmol. Invest Ophthalmol Vis Sci. 1991;32(10):2723–8.
2017;10(12):1798–804. 70. Joo CK, Pepose JS, Stuart PM.  T-cell mediated
54. Ang M, et  al. Evaluation of a micro-optical coher- responses in a murine model of orthotopic cor-
ence tomography for the corneal endothelium in an neal transplantation. Invest Ophthalmol Vis Sci.
animal model. Sci Rep. 2016;6:29769. 1995;36(8):1530–40.
55. Streilein JW, Niederkorn JY.  Induction of ante- 71. Haskova Z, Filipec M, Holan V.  The role of major
rior chamber-associated immune deviation and minor histocompatibility antigens in orthotopic
requires an intact, functional spleen. J Exp Med. corneal transplantation in mice. Folia Biol (Praha).
1981;153(5):1058–67. 1996;42(3):105–11.
56. Streilein JW, et  al. Immunosuppressive properties 72. Zhang EP, Schrunder S, Hoffmann F.  Orthotopic
of tissues obtained from eyes with experimentally corneal transplantation in the mouse--a new surgical
manipulated corneas. Invest Ophthalmol Vis Sci. technique with minimal endothelial cell loss. Graefes
1996;37(2):413–24. Arch Clin Exp Ophthalmol. 1996;234(11):714–9.
57. Perez VL, et  al. The anterior chamber of the eye 73. Yamagami S, Tsuru T.  Increase in orthotopic
as a clinical transplantation site for the treatment murine corneal transplantation rejection rate with
of diabetes: a study in a baboon model of diabetes. anterior synechiae. Invest Ophthalmol Vis Sci.
Diabetologia. 2011;54(5):1121–6. 1999;40(10):2422–6.
58. Hayashi T, et al. Immunologic mechanisms of cor- 74. Hori J, Streilein JW. Dynamics of donor cell persis-
neal allografts reconstituted from cultured alloge- tence and recipient cell replacement in orthotopic
neic endothelial cells in an immune-privileged site. corneal allografts in mice. Invest Ophthalmol Vis
Invest Ophthalmol Vis Sci. 2009;50(7):3151–8. Sci. 2001;42(8):1820–8.
29  Corneal Endothelial Cell Transplantation: Animal Models 453

75. Sonoda KH, Taniguchi M, Stein-Streilein J.  Long-­ 95. Hill JC, Maske R. An animal model for corneal graft
term survival of corneal allografts is dependent rejection in high-risk keratoplasty. Transplantation.
on intact CD1d-reactive NKT cells. J Immunol. 1988;46(1):26–30.
2002;168(4):2028–34. 96. Mimura T, et al. Expression of vascular endothelial
76. Bourne WM, et  al. The effect of splenectomy growth factor C and vascular endothelial growth fac-
on corneal graft rejection. Investig Ophthalmol. tor receptor 3  in corneal lymphangiogenesis. Exp
1976;15(7):541–2. Eye Res. 2001;72(1):71–8.
77. Thomas JW. An experiment in Keratoplasty. Proc R 97. Eliason JA, McCulley JP.  A comparison between
Soc Med. 1930;23(10):1437–42. interrupted and continuous suturing techniques in
78. Castroviejo R.  Keratoplasty-microscopic study keratoplasty. Cornea. 1990;9(1):10–6.
of the corneal grafts. Trans Am Ophthalmol Soc. 98. Cohen RA, et  al. Confocal microscopy of corneal
1937;35:355–85. graft rejection. Cornea. 1995;14(5):467–72.
79. Stansbury FC, Wadsworth JA.  Surgical technique 99. Cho BJ, et  al. In vivo confocal microscopic analy-
of corneal transplantation in rabbits; a discussion of sis of corneal allograft rejection in rabbits. Cornea.
the problems encountered and suggestions for their 1998;17(4):417–22.
solution. Am J Ophthalmol. 1947;30(8):968–78. 100. Bahn CF, et  al. Postnatal development of cor-
80. Khodadoust AA. Lamellar corneal transplantation in neal endothelium. Invest Ophthalmol Vis Sci.
the rabbit. Am J Ophthalmol. 1968;66(6):1111–7. 1986;27(1):44–51.
81. Tseng SC, Tsai RJ.  Limbal transplantation for 101. Bartholomew LR, et  al. Ultrasound biomicroscopy
ocular surface reconstruction  – a review. Fortschr of globes from young adult pigs. Am J Vet Res.
Ophthalmol. 1991;88(3):236–42. 1997;58(9):942–8.
82. Swift GJ, et  al. Survival of rabbit limbal stem cell 102. Faber C, et al. Orthotopic porcine corneal xenotrans-
allografts. Transplantation. 1996;62(5):568–74. plantation using a human graft. Acta Ophthalmol.
83. Li QJ, et al. Long-term survival of allogeneic donor 2009;87(8):917–9.
cell-derived corneal epithelium in limbal deficient 103. Sanchez I, et  al. The parameters of the porcine
rabbits. Curr Eye Res. 2001;23(5):336–45. eyeball. Graefes Arch Clin Exp Ophthalmol.
84. Amano S.  Transplantation of corneal endo- 2011;249(4):475–82.
thelial cells. Nippon Ganka Gakkai Zasshi. 104. Nicholls SM, et al. A model of corneal graft rejec-
2002;106(12):805–35; discussion 836. tion in semi-inbred NIH miniature swine: sig-
85. Amano S. Transplantation of cultured human corneal nificant T-cell infiltration of clinically accepted
endothelial cells. Cornea. 2003;22(7 Suppl):S66–74. allografts. Invest Ophthalmol Vis Sci. 2012;
86. Quantock AJ, et al. Stromal architecture and immune 53(6):3183–92.
tolerance in additive corneal xenografts in rodents. 105. Schroeter J, et  al. Influence of temporary hypo-
Acta Ophthalmol Scand. 2005;83(4):462–6. thermia on corneal endothelial cell density during
87. Hori J, Niederkorn JY. Immunogenicity and immune organ culture preservation. Graefes Arch Clin Exp
privilege of corneal allografts. Chem Immunol Ophthalmol. 2008;246(3):369–72.
Allergy. 2007;92:290–9. 106. Majo F, et al. Oligopotent stem cells are distributed
88. Amano S, et al. Decellularizing corneal stroma using throughout the mammalian ocular surface. Nature.
N2 gas. Mol Vis. 2008;14:878–82. 2008;456(7219):250–4.
89. Engelmann K, Drexler D, Bohnke M. Transplantation 107. Sweatt AJ, Ford JG, Davis RM. Wound healing fol-
of adult human or porcine corneal endothelial lowing anterior keratectomy and lamellar kerato-
cells onto human recipients in  vitro. Part I: cell plasty in the pig. J Refract Surg. 1999;15(6):636–47.
culturing and transplantation procedure. Cornea. 108. Asano-Kato N, et al. Epithelial ingrowth after laser
1999;18(2):199–206. in situ keratomileusis: clinical features and possible
90. Oh JY, et al. Corneal cell viability and structure after mechanisms. Am J Ophthalmol. 2002;134(6):801–7.
transcorneal freezing-thawing in the human cornea. 109. Hwang H, Kim M. Endothelial damage of a donor
Clin Ophthalmol. 2010;4:477–80. cornea depending on the donor insertion method
91. Mimura T, et  al. Long-term outcome of iron-­ during Descemet-stripping automated endothelial
endocytosing cultured corneal endothelial cell trans- keratoplasty in porcine eyes. Jpn J Ophthalmol.
plantation with magnetic attraction. Exp Eye Res. 2009;53(5):523–30.
2005;80(2):149–57. 110. Mohay J, et  al. Transplantation of corneal endo-
92. Patel SV, et  al. Human corneal endothelial cell thelial cells using a cell carrier device. Cornea.
transplantation in a human ex  vivo model. Invest 1994;13(2):173–82.
Ophthalmol Vis Sci. 2009;50(5):2123–31. 111. Bahn CF, et al. Complications associated with bovine
93. Yokoo S, et al. Human corneal endothelial cell pre- corneal endothelial cell-lined homografts in the cat.
cursors isolated by sphere-forming assay. Invest Invest Ophthalmol Vis Sci. 1982;22(1):73–90.
Ophthalmol Vis Sci. 2005;46(5):1626–31. 112. Ling TL, Vannas A, Holden BA. Long-term changes
94. Lai JY, Li YT. Functional assessment of cross-linked in corneal endothelial morphology following
porous gelatin hydrogels for bioengineered cell sheet wounding in the cat. Invest Ophthalmol Vis Sci.
carriers. Biomacromolecules. 2010;11(5):1387–97. 1988;29(9):1407–12.
454 B. P. Barnett and A. S. Jun

113. Bourne WM, et  al. Long-term observation of mor- 123. Ohno K, et  al. Transplantation of cryopreserved
phologic and functional features of cat corneal endo- human corneas in a xenograft model. Cryobiology.
thelium after wounding. Invest Ophthalmol Vis Sci. 2002;44(2):142–9.
1994;35(3):891–9. 124. Brunette I, et  al. Comparison of the pig and feline
114. Moodie KL, et al. Postnatal development of corneal models for full thickness corneal transplantation. Vet
curvature and thickness in the cat. Vet Ophthalmol. Ophthalmol. 2011;14(6):365–77.
2001;4(4):267–72. 125. Proulx S, Brunette I. Methods being developed for
115. Kafarnik C, Fritsche J, Reese S.  In vivo confocal preparation, delivery and transplantation of a tissue-­
microscopy in the normal corneas of cats, dogs and engineered corneal endothelium. Exp Eye Res.
birds. Vet Ophthalmol. 2007;10(4):222–30. 2012;95(1):68–75.
116. Reichard M, et  al. Comparative in  vivo confo- 126. Okumura N, et  al. The new therapeutic concept of
cal microscopical study of the cornea anatomy using a rho kinase inhibitor for the treatment of cor-
of different laboratory animals. Curr Eye Res. neal endothelial dysfunction. Cornea. 2011;30(Suppl
2010;35(12):1072–80. 1):S54–9.
117. Van Horn DL, Hyndiuk RA. Endothelial wound repair 127. Zurawski CA, et al. Corneal biometrics of the rhe-
in primate cornea. Exp Eye Res. 1975;21(2):113–24. sus monkey (Macaca mulatta). J Med Primatol.
118. Petroll WM, et  al. Organization of junctional pro- 1989;18(6):461–6.
teins in proliferating cat corneal endothelium during 128. Jackson AJ, Gardiner T, Archer DB. Morphometric
wound healing. Cornea. 2001;20(1):73–80. analysis of corneal endothelial giant cells in normal
119. Melles GR, et al. A surgical technique for posterior and traumatized corneas. Ophthalmic Physiol Opt.
lamellar keratoplasty. Cornea. 1998;17(6):618–26. 1995;15(4):305–10.
120. Lopatin DE, et  al. Changes in aqueous immu- 129. Ollivier FJ, et  al. Corneal thickness and endothe-
noglobulin and albumin levels following pen- lial cell density measured by non-contact specular
etrating keratoplasty. Invest Ophthalmol Vis Sci. microscopy and pachymetry in rhesus macaques
1989;30(1):122–31. (Macaca mulatta) with laser-induced ocular hyper-
121. Tripoli NK, Cohen KL, Proia AD. Cat keratoplasty tension. Exp Eye Res. 2003;76(6):671–7.
wound healing and corneal astigmatism. Refract 130. Matsubara M, Tanishima T. Wound-healing of cor-
Corneal Surg. 1992;8(3):196–203. neal endothelium in monkey: an autoradiographic
122. Ohno K, et  al. Keratocyte activation and apoptosis study. Jpn J Ophthalmol. 1983;27(3):444–50.
in transplanted human corneas in a xenograft model. 131. Jie Y, et al. Survival of pig-to-rhesus corneal xeno-
Invest Ophthalmol Vis Sci. 2002;43(4):1025–31. grafts prolonged by prior donor bone marrow trans-
plantation. Mol Med Rep. 2013;7(3):869–74.
Cell Based Therapy for Corneal
Endothelial Regeneration
30
Noriko Koizumi and Naoki Okumura

30.1 Introduction resolved. One alternative to corneal endothelial


keratoplasty is regenerative medicine, which may
The corneal endothelium plays an essential role represent the next-generation therapy for treating
in the maintenance of corneal transparency corneal endothelial dysfunction. In that regard,
through its barrier and pump functions. However, current research efforts are aimed at developing
the proliferative ability of human corneal endo- cell-based therapies for corneal endothelial dys-
thelial cells is severely limited in vivo, and severe function. One promising approach involves cell-­
cell loss due to dystrophy, trauma, or surgical sheet transplantation and cell-injection therapy
intervention often causes irreversible corneal using cultivated corneal endothelial cells
endothelial dysfunction [1]. Traditionally, a full-­ (Fig. 30.1).
thickness corneal transplantation has been the
only therapeutic choice for the restoration of
clear vision. The recent evolution of surgical pro- 30.2 Corneal Endothelial Cell
cedures for corneal endothelial keratoplasty, such Biology
as such as Descemet’s stripping endothelial kera-
toplasty (DSEK) or Descemet’s membrane endo- In vivo human corneal endothelial cells are
thelial keratoplasty (DMEK), has enabled less thought to be arrested at the G1 phase of the cell
invasive treatment of corneal endothelial dys- cycle [2, 3], a phase that is negatively regulated
function with better visual recovery. Despite the by the CIP/KIP family (p21 and p27), the INK4
good visual recovery obtained with corneal endo- family (p15, p16, and p19), and the p53 family of
thelial keratoplasty, the severe shortage of donor proteins (p53, Tap63) [4–6]. Several other fac-
grafts is a serious worldwide problem for per- tors, such as transforming growth factor beta
forming this operation. Additional problems (TGF-β) in the aqueous humor, the contact inhi-
associated with corneal transplantation, such as bition of cells [7], and stress-induced premature
allograft rejection, primary graft failure, and senescence [8, 9] have also been reported to
gradual decrease of cell density, have yet to be inhibit the in vivo proliferation of corneal endo-
thelial cells. Although the in  vivo existence of
stem/progenitor cells has yet to be proven, these
N. Koizumi (*) · N. Okumura endothelial cells are known to retain their prolif-
Department of Biomedical Engineering, Faculty of erative ability in vitro [10, 11]. One suggestion is
Life and Medical Sciences, Doshisha University,
Kyotanabe, Kyoto, Japan that peripheral corneal endothelial cells in cell-­
e-mail: nkoizumi@mail.doshisha.ac.jp culture conditions have a higher proliferative

© Springer Nature Switzerland AG 2019 455


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_30
456 N. Koizumi and N. Okumura

Fig. 30.1  The concept Sheet transplantation


of cell based therapy as
a future therapeutic
modality for corneal
endothelial disease Expansion culture of human
corneal endothelial cells

Donor cornea
Cell-injection therapy

Cell-suspension of
cultivated corneal
endothelial cells

ability than do central corneal endothelial cells; 18]. The establishment of optimal conditions for
another is that a high density of stem/progenitor the cultivation of human corneal endothelial cells
cells of corneal endothelium exists in the periph- is critical for the clinical application of a cell
eral region [12]. These cell clusters located in the based therapy for corneal regeneration. Recently,
extreme peripheral region are considered poten- the usefulness of a TGF-β receptor inhibitor [17]
tial stem cell niches [13]. Other reported findings and p38 MAP kinase inhibitor [19] in the cell
also appear to indicate that human corneal-­ culture medium was reported in terms of prevent-
endothelial stem/progenitor cells are regulated by ing the EMT and cellular senescence. Other tech-
the leucine-rich repeat-containing G protein-­ niques, including the use of Rho-associated
coupled receptor 5 (LGR5) via the Hedgehog and protein kinase (ROCK) inhibitor [20–22] or the
Wnt signaling pathways [14]. use of a culture substrate with a specific extracel-
lular matrix similar to the Descemet’s membrane,
were also reported to promote the cell prolifera-
30.3 C
 orneal Endothelial Cell tion and to support human corneal endothelial
Culture for Clinical cell cultures that show high cell density [23].
Application

The in  vivo and in  vitro proliferative ability of 30.4 Cultivated Corneal
human corneal endothelial cells is, in general, Endothelial Cell Sheet
severely controlled, as is observed in monkey Transplantation
corneal endothelial cells [15, 16]. Thus, a com-
plete understanding of the biological characteris- A cultivated human corneal endothelial cell
tics of cultivated human corneal endothelial cells sheet, with [24–29] or without [30] carrier mate-
is essential for developing translational research rials, has now been created, and it demonstrates
for cell based therapy. However, most researchers in vivo functionality in both rabbit and monkey
have encountered difficulties in consistently models. We previously reported the temporary
expanding a clinically applicable quantity and attachment of a cultivated monkey corneal endo-
quality of cultivated human corneal endothelial thelial cell sheet (on type 1 collagen) onto the
cells with healthy biological characteristics, Descemet’s membrane. This sheet slipped off
because human corneal endothelial cells readily into the anterior chamber and, surprisingly,
enter into an epithelial mesenchymal transition resulted in the recovery of corneal transparency
(EMT) to develop a fibroblastic phenotype [17, [25, 26]. This quite unexpected finding led us to
30  Cell Based Therapy for Corneal Endothelial Regeneration 457

speculate that cultivated primate corneal endo- cultivated corneal endothelial cells into the ante-
thelial cells migrate and proliferate to some rior chamber might provide several advantages,
extent in vivo, and a subsequent series of experi- including a better optical quality of the postop-
ments provided us with a potential new concept erative cornea, less direct manipulation of the
for the treatment of corneal endothelial dysfunc- cultivated cells, a minimally invasive surgery,
tion. We postulated that, rather than transplanting and the potential for a mass application from one
a cultivated corneal endothelial sheet, we could donor cornea. Attempts to develop an effective
transplant individual endothelial cells that show a method to deliver cultivated corneal endothelial
renewed ability to proliferate in  vivo. cells to the posterior cornea have included the use
Interestingly, a similar event was reported in clin- of magnetic attachment of iron-powder [33] or
ical practice with human corneal endothelial cells superparamagnetic microspheres incorporated
following a DMEK procedure [31] or the sponta- into cultivated corneal endothelial cells [34] in a
neous clearing of the cornea after Descemet’s rabbit transplantation model and in an organ cul-
rhexis without endothelial grafting [32]. ture model of the human eye.
The recent discovery of the effects of ROCK
inhibitor on human corneal endothelial cells has
30.5 Cell-Injection Therapy also shed some light on the development of a new
concept for treatments of corneal endothelial dys-
When compared to transplantation of cultivated function, such as eye drop treatments and the cell-
corneal endothelial sheets, the direct injection of injection therapy [35–37] (Fig. 30.2). Preclinical

Fig. 30.2  Surgical procedure of cell-injection therapy. (1) chamber, (3) Patient in the face-down position, to allow the
Patients’ corneal endothelial cells were gently scraped off corneal endothelial cells to sink down to the anterior cham-
with a silicone needle, (2) Co-injection of cultured corneal ber side of the cornea, (4) Regeneration of the corneal endo-
endothelial cells and a ROCK inhibitor into the anterior thelium by the injected cultured corneal endothelial cells
458 N. Koizumi and N. Okumura

experiments involving rabbits and monkeys caused by Fuchs endothelial corneal dystrophy,
showed that a cell-injection therapy, combined pseudophakic bullous keratopathy, and the argon-
with the use of a ROCK inhibitor and followed by laser iridotomy-­ induced bullous keratopathy,
3 h in a face-down position, promoted cultivated which is a quite common cause of corneal endo-
corneal endothelial cell adhesion onto the poste- thelial dysfunction in Japan (Fig. 30.3). Notably,
rior cornea, ultimately resulting in the recovery of the replacement of corneal endothelial cells of
corneal transparency in both animals [38]. In fact, failed grafts with cultivated corneal endothelial
the monkey eyes receiving allogeneic cultivated cells was also effective in recovering the clarity
monkey corneal endothelial cell injection with of the full-­thickness graft (Fig. 30.4). Similar to
ROCK inhibitor exhibited complete corneal trans- the observation in animal experiments, a beauti-
parency post injection, and the transparency and ful monolayer of attached cells was observed on
normal corneal thickness were maintained the Descemet’s membrane when viewing the
throughout a one-year observation period. The postoperative patient’s eyes with specular micros-
safety and efficacy of cell-injection therapy was copy (Fig. 30.5). Clinically, reasonable recovery
also confirmed by the transplantation of cultivated of corneal endothelial functions and vision were
human corneal endothelial cells into a monkey observed in all cases, without any postoperative
corneal endothelial dysfunction model in preclini- adverse reactions such as elevated intraocular
cal research [39]. pressure, uveitis, or allograft rejection.
Based on the data obtained from the estab-
lished culture protocol for human corneal endo-
thelial cells and from animal experiments, a 30.6 Toward the Product
first-in-man clinical trial of cell-injection therapy Development of Cell-­
was initiated in December 2013 under official Injection Therapy
approval from the Regenerative Medicine
Committee of the Japanese government [40, 41]. Cell based therapy for corneal endothelial dys-
The initial results of this cell-injection therapy functions is a “long-cherished dream” of cornea
appear to be promising for the treatment of surgeons, as well as researchers. Recently devel-
patients with corneal endothelial dysfunction oped cell-injection therapies will deliver many

Fig. 30.3  Cell-injection therapy for bullous keratopathy tured human corneal endothelial cells, together with a
patient. Slit lamp microscopy images of the first patient, a ROCK inhibitor, were injected into the anterior chamber.
Japanese female with corneal endothelial decompensation The postoperative visual acuity recovered to 1.0, together
induced by argon laser iridotomy (left). Preoperative with the associated recovery of corneal transparency
visual acuity was 0.04 due to the edema in the corneal (right). (Reproduced from Okumura et  al. J
epithelium and stroma. After mechanical removal of an Ophthalmology, 2017. CC BY 4.0 https://creativecom-
8  mm diameter section of the corneal endothelium, cul- mons.org/licenses/by/4.0/)
30  Cell Based Therapy for Corneal Endothelial Regeneration 459

90° 90°
° 60 OD ° 60 OD
9mm 120 ° 9mm 120 °


15

15
30

30
°

°
180°

180°




33

33
21

21

24 0° 24 0°
0° 30 0° 30
T 270° N T 270° N

Fig. 30.4  Endothelial replacement with cell-injection months after cell-injection surgery, the graft edema
therapy in a failed graft eye. Pre-treatment image of a improved and corneal clarity was recovered (right).
patient with a failed graft due to allograft rejection. Severe (Reproduced from Koizumi, Ganka. Ophthalmology,
corneal edema with opacity was observed (left). Eighteen 2017)

novel findings and knowledge regarding cell based centers to hospitals around the world. Recent
therapy. Clinical evidence of the safety and effi- research is also now aimed at establishing less
cacy of cell-injection therapy should be accumu- invasive evaluation methods for cultivated human
lated and evaluated very carefully, and technical corneal endothelial cells and the technical refine-
improvements in surgery and cell culture might be ment of cell production under ­ GMP-­ compliant
necessary for the development of cell-­ injection systems [39, 42–47]. If cell products are devel-
therapy products. Product development by phar- oped by companies and made available in many
maceutical or biotechnology companies for mass countries, cell based therapy might represent a
application of this therapy will require technical paradigm shiftin corneal endothelial treatments in
breakthroughs for frozen preservation of culti- the near future.
vated human corneal endothelial cells for creating
a master cell bank and for the transport/distribu- Conflict of Interest  Noriko Koizumi and Naoki Okumura
tion of cell products from company cell processing declare that they have no conflict of interest.” (If any
460 N. Koizumi and N. Okumura

Number 212
CD /mm2 2526
AVG um2 396
SD um2 182
CV % 46
Max um2 1251
Min um2 110

CCT 548 um

Area (Polymethism) Apex (Pleomorphism)


[um2] % 50 100 % 50 100
0-100 0 3 0
100-200 13 4 12
200-300 21 5 30
300-400 22 6A 26
400-500 22 7 19
500-600 9 8 10
600-700 8 9 2
700-800 3 10- 0
800-900 1
900- 1 Customize

Fig. 30.5  Specular microscopy of the post cell-injection 2500 cells/mm2 were observed with non-contact specular
eyes. Eighteen months after cell-injection therapy, homo- microscopy (same patient as in Fig.  30.4). (Reproduced
geneous corneal endothelial cells with a density more than from Koizumi, Ganka. Ophthalmology, 2017)

authors do have a conflict of interest it must be spelled out 4. Kim TY, Kim WI, Smith RE, Kay ED.  Role of
specifically). p27(Kip1) in cAMP- and TGF-beta2-mediated anti-
proliferation in rabbit corneal endothelial cells. Invest
Ophthalmol Vis Sci. 2001;42(13):3142–9.
Informed Consent All procedures followed were in
5. Paull AC, Whikehart DR. Expression of the p53 fam-
accordance with the ethical standards of the responsible
ily of proteins in central and peripheral human corneal
committee on human experimentation (institutional and
endothelial cells. Mol Vis. 2005;11:328–34.
national) and with the Helsinki Declaration of 1975, as
6. Lee JG, Kay EP.  Involvement of two distinct ubiq-
revised in 2000. Informed consent was obtained from all
uitin E3 ligase systems for p27 degradation in cor-
patients for being included in the study.
neal endothelial cells. Invest Ophthalmol Vis Sci.
2008;49(1):189–96.
Animal Studies  No animal studies were carried out by 7. Joyce NC, Harris DL, Mello DM.  Mechanisms of
the authors for this article. mitotic inhibition in corneal endothelium: contact
inhibition and TGF-beta2. Invest Ophthalmol Vis Sci.
2002;43(7):2152–9.
8. Joyce NC, Zhu CC, Harris DL. Relationship among
References oxidative stress, DNA damage, and proliferative
capacity in human corneal endothelium. Invest
1. Kaufman HE, Katz JI. Pathology of the corneal endo- Ophthalmol Vis Sci. 2009;50(5):2116–22.
thelium. Invest Ophthalmol Vis Sci. 1977;16(4):265–8. 9. Konomi K, Joyce NC.  Age and topographical com-
2. Joyce NC. Proliferative capacity of the corneal endo- parison of telomere lengths in human corneal endo-
thelium. Prog Retin Eye Res. 2003;22(3):359–89. thelial cells. Mol Vis. 2007;13:1251–8.
3. Joyce NC, Meklir B, Joyce SJ, Zieske JD.  Cell 10. Engelmann K, Bohnke M, Friedl P.  Isolation

cycle protein expression and proliferative status in and long-­ term cultivation of human corneal
human corneal cells. Invest Ophthalmol Vis Sci. endothelial cells. Invest Ophthalmol Vis Sci.
1996;37(4):645–55. 1988;29(11):1656–62.
30  Cell Based Therapy for Corneal Endothelial Regeneration 461

11. Miyata K, Drake J, Osakabe Y, et al. Effect of donor 26. Koizumi N, Sakamoto Y, Okumura N, et al. Cultivated
age on morphologic variation of cultured human cor- corneal endothelial transplantation in a primate: pos-
neal endothelial cells. Cornea. 2001;20(1):59–63. sible future clinical application in corneal endothe-
12. Konomi K, Zhu C, Harris D, Joyce NC. Comparison lial regenerative medicine. Cornea. 2008;27(Suppl
of the proliferative capacity of human corneal endo- 1):S48–55.
thelial cells from the central and peripheral areas. 27. Mimura T, Yamagami S, Yokoo S, et  al. Cultured
Invest Ophthalmol Vis Sci. 2005;46(11):4086–91. human corneal endothelial cell transplantation with a
13. He Z, Campolmi N, Gain P, et  al. Revisited micro- collagen sheet in a rabbit model. Invest Ophthalmol
anatomy of the corneal endothelial periphery: Vis Sci. 2004;45(9):2992–7.
new evidence for continuous centripetal migra- 28. Kimoto M, Shima N, Yamaguchi M, Hiraoka Y, Amano
tion of endothelial cells in humans. Stem Cells. S, Yamagami S. Development of a bioengineered cor-
2012;30(11):2523–34. neal endothelial cell sheet to fit the corneal curvature.
14. Hirata-Tominaga K, Nakamura T, Okumura N, et al. Invest Ophthalmol Vis Sci. 2014;55(4):2337–43.
Corneal endothelial cell fate is maintained by LGR5 29. Yamaguchi M, Shima N, Kimoto M, Ebihara N,

through the regulation of hedgehog and Wnt pathway. Murakami A, Yamagami S. Optimization of cultured
Stem Cells. 2013;31(7):1396–407. human corneal endothelial cell sheet transplantation
15. Van Horn DL, Hyndiuk RA. Endothelial wound repair and post-operative sheet evaluation in a rabbit model.
in primate cornea. Exp Eye Res. 1975;21(2):113–24. Curr Eye Res. 2016;41(9):1178–84.
16. Matsubara M, Tanishima T.  Wound-healing of the 30. Sumide T, Nishida K, Yamato M, et  al. Functional
corneal endothelium in the monkey: a morphometric human corneal endothelial cell sheets harvested from
study. Jpn J Ophthalmol. 1982;26(3):264–73. temperature-responsive culture surfaces. FASEB J.
17. Okumura N, Kay EP, Nakahara M, Hamuro J,
2006;20(2):392–4.
Kinoshita S, Koizumi N. Inhibition of TGF-beta sig- 31. Jacobi C, Zhivov A, Korbmacher J, et  al. Evidence
naling enables human corneal endothelial cell expan- of endothelial cell migration after descemet mem-
sion in vitro for use in regenerative medicine. PLoS brane endothelial keratoplasty. Am J Ophthalmol.
One. 2013;8(2):e58000. 2011;152(4):537–542.e532.
18. Peh GS, Beuerman RW, Colman A, Tan DT, Mehta 32.
Shah RD, Randleman JB, Grossniklaus
JS.  Human corneal endothelial cell expansion for HE.  Spontaneous corneal clearing after Descemet's
corneal endothelium transplantation: an overview. stripping without endothelial replacement.
Transplantation. 2011;91(8):811–9. Ophthalmology. 2012;119(2):256–60.
19. Hongo A, Okumura N, Nakahara M, Kay EP, Koizumi 33. Mimura T, Shimomura N, Usui T, et  al. Magnetic
N.  The effect of a p38 mitogen-activated protein attraction of iron-endocytosed corneal endothe-
kinase inhibitor on cellular senescence of cultivated lial cells to Descemet's membrane. Exp Eye Res.
human corneal endothelial cells. Invest Ophthalmol 2003;76(6):745–51.
Vis Sci. 2017;58(9):3325–34. 34. Patel SV, Bachman LA, Hann CR, Bahler CK, Fautsch
20. Okumura N, Ueno M, Koizumi N, et al. Enhancement MP. Human corneal endothelial cell transplantation in
on primate corneal endothelial cell survival in  vitro a human ex  vivo model. Invest Ophthalmol Vis Sci.
by a ROCK inhibitor. Invest Ophthalmol Vis Sci. 2009;50(5):2123–31.
2009;50(8):3680–7. 35. Koizumi N, Okumura N, Ueno M, Nakagawa H,

21. Okumura N, Nakano S, Kay EP, et  al. Involvement Hamuro J, Kinoshita S.  Rho-associated kinase
of cyclin D and p27 in cell proliferation mediated by inhibitor eye drop treatment as a possible ­medical
ROCK inhibitors Y-27632 and Y-39983 during cor- treatment for Fuchs corneal dystrophy. Cornea.
neal endothelium wound healing. Invest Ophthalmol 2013;32(8):1167–70.
Vis Sci. 2014;55(1):318–29. 36. Okumura N, Koizumi N, Kay EP, et al. The ROCK inhib-
22. Peh GS, Adnan K, George BL, et  al. The effects of itor eye drop accelerates corneal endothelium wound
rho-associated kinase inhibitor Y-27632 on primary healing. Invest Ophthalmol Vis Sci. 2013;54:2493.
human corneal endothelial cells propagated using a 37. Okumura N, Kinoshita S, Koizumi N.  The Role

dual media approach. Sci Rep. 2015;5:9167. of Rho Kinase Inhibitors in Corneal Endothelial
23. Okumura N, Kakutani K, Numata R, et al. Laminin-511 Dysfunction. Curr Pharm Des. 2017;23(4):660–6.
and -521 enable efficient in vitro expansion of human 38. Okumura N, Koizumi N, Ueno M, et al. ROCK inhibi-
corneal endothelial cells. Invest Ophthalmol Vis Sci. tor converts corneal endothelial cells into a phenotype
2015;56(5):2933–42. capable of regenerating in vivo endothelial tissue. Am
24. Ishino Y, Sano Y, Nakamura T, et al. Amniotic mem- J Pathol. 2012;181(1):268–77.
brane as a carrier for cultivated human corneal endo- 39. Okumura N, Sakamoto Y, Fujii K, et al. Rho kinase
thelial cell transplantation. Invest Ophthalmol Vis Sci. inhibitor enables cell-based therapy for corneal endo-
2004;45(3):800–6. thelial dysfunction. Sci Rep. 2016;6:26113.
25. Koizumi N, Sakamoto Y, Okumura N, et  al.
40. Okumura N, Kinoshita S, Koizumi N.  Application
Cultivated corneal endothelial cell sheet transplanta- of Rho Kinase Inhibitors for the Treatment of
tion in a primate model. Invest Ophthalmol Vis Sci. Corneal Endothelial Diseases. J Ophthalmol.
2007;48(10):4519–26. 2017;2017:2646904.
462 N. Koizumi and N. Okumura

41. Kinoshita S, Koizumi N, Ueno M, et al. Injection of 45. Okumura N, Inoue R, Kakutani K, et  al. Corneal
cultured cells with a ROCK inhibitor for bullous kera- endothelial cells have an absolute requirement for
topathy. New Engl J Med. 2018;378(11):995–1003. cysteine for survival. Cornea. 2017;36(8):988–94.
42. Okumura N, Hirano H, Numata R, et al. Cell surface 46. Toda M, Ueno M, Hiraga A, et  al. Production of
markers of functional phenotypic corneal endothelial homogeneous cultured human corneal endothelial
cells. Invest Ophthalmol Vis Sci. 2014;55(11):7610–8. cells indispensable for innovative cell therapy. Invest
43. Okumura N, Kusakabe A, Hirano H, et al. Density-­ Ophthalmol Vis Sci. 2017;58(4):2011–20.
gradient centrifugation enables the purification of 47. Ueno M, Asada K, Toda M, et  al. MicroRNA pro-
cultured corneal endothelial cells for cell therapy by files qualify phenotypic features of cultured human
eliminating senescent cells. Sci Rep. 2015;5:15005. corneal endothelial cells. Invest Ophthalmol Vis Sci.
44. Okumura N, Ishida N, Kakutani K, et al. Development 2016;57(13):5509–17.
of cell analysis software for cultivated corneal endo-
thelial cells. Cornea. 2017;36(11):1387–94.
Corneal Endothelium
Regeneration: Future Prospects
31
Wei-Ting Ho, Hsin-Yu Liu, Fung-Rong Hu,
and I-Jong Wang

The human corneal endothelium lines the in cultured corneas [5, 6]. Contact inhibition
­innermost layer of the human cornea and consists ­preventing transition to the S-phase and low ratio
of a monolayer of hexagonal cells. This single of positive versus negative growth factors con-
layer of cells is the key to maintain corneal trans- centration in the anterior chamber both prevent
parency and health by regulating the hydration of the endothelial cells from dividing [5, 7–9].
cornea. Based on the pump-leak mechanism [1],
the endothelial cells are permeable to nutrients
and other metabolites from the aqueous humor 31.1 H
 uman Corneal Endothelial
while actively pump out ions to move water Cell Dysfunction and Current
osmotically into the anterior chamber. Managements
At birth, endothelial density (ECD) is approx-
imately 6000 cells/mm2 [2], and declines thereaf- Human corneal endothelial cell dysfunction will
ter as the eyes grow [3, 4]. In young adults, the result in increased imbibition of water and thus
ECD is about 3500 cells/mm2.The ECD gradu- corneal edema [10]. Among the relatively few pri-
ally stabilizes in adulthood and drops at a rate of mary corneal endothelial diseases, Fuchs’ endo-
0.6% annually [3, 4]. The human corneal endo- thelial corneal dystrophy (FECD) is the most
thelial cells (HCECs) are considered to have lim- common [11]. In FECD, the basement membrane
ited ability in regeneration under normal of the endothelium, Descemet’s membrane
condition, despite induced mitosis are observed becomes diffusely thickened and develops guttae
which are visible with slit-lamp and specular
microscopy. FECD can be inherited and tends to
W.-T. Ho be bilateral [12]. The endothelial cells dysfunc-
Department of Ophthalmology, Far Eastern Memorial tion often becomes evident in middle age or later.
Hospital, New Taipei City, Taiwan In secondary corneal endothelial diseases,
H.-Y. Liu there is a recognizable cause that contributes to
Department of Ophthalmology, National Taiwan the malfunction of corneal endothelium. These
University Hospital, Taipei, Taiwan
includes contact lens-induced endotheliopathy,
F.-R. Hu · I.-J. Wang (*) iatrogenic endotheliopathy and endotheliopathy
Department of Ophthalmology, National Taiwan
University Hospital, Taipei, Taiwan secondary to ocular inflammation [13, 14].
Contact lens use increases the polymegathism
College of Medicine, National Taiwan University,
Taipei, Taiwan and pleomorphism of endothelial cells. These
e-mail: ijong@ntu.edu.tw changes are considered to be hypoxia-related

© Springer Nature Switzerland AG 2019 463


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_31
464 W.-T. Ho et al.

[15]. Endothelial decompensation has been plications drive the need for investigating alterna-
reported after ocular procedures or surgeries. The tives to surgery. Strategies targeting corneal
impact on endothelium after laser iridotomy and endothelial dysfunction can be divided into three
cataract extraction has been reported and exten- parts: preventing corneal endothelial loss, in vivo
sively studied [16–18]. Mechanism proposed for stimulation of human corneal endothelial cell pro-
endothelial damage during laser iridotomy liferation and/or function, cell replacement with
include direct focal injury, thermal damage, cultivated human corneal endothelial cells.
mechanical shock waves, inflammation and iris
pigment dispersion. During cataract surgery,
endothelial injury may occur due to many fac- 31.2.1 Preventing Corneal
tors, such as corneal distortion, collision of lens Endothelial Cells Loss
fragments with corneal endothelium [19], local-
ized temperature rise [20] and increased oxida- 31.2.1.1 Sulforaphane
tive stress by free radicals [21]. After initial cell Jurkunas et al. described the involvement of oxi-
loss, the ECD will continue to decrease for at dative stress in the pathogenesis of FECD and led
least 10 years after surgery [22]. to the identification of sulforaphane. By upregu-
For mild endothelial dysfunction, observation lating major antioxidant response element-­
with lubrication and hyperosmotic topical medi- dependent antioxidants, sulforaphane can
cations may be sufficient. Once moderate or ameliorate oxidative stress-induced apoptosis in
severe endothelial decompensation occurs, surgi- ex  vivo FECD sample and in Fuchs’ corneal
cal replacement with donor cells is required to endothelial cell lines. Therefore, it may prevent
restore corneal transparency. Both penetrating further cell loss in FECD [26, 27].
keratoplasty (PK) and endothelial keratoplasty
(EK) are treatment choices for endothelial dys- 31.2.1.2 R  TA 408 Ophthalmic
function [23, 24]. EK may be superior in terms of Suspension
less invasive manipulation of recipients and RTA 408 Ophthalmic Suspension (Reata
quicker recovery. Thus, the surgical treatment for Pharmaceuticals; Irving, Texas) is currently in a
endothelial diseases dramatically evolved toward Phase II trial for prevention of endothelial cell
EK in recent years. EK procedures have evolved loss after cataract surgery through reducing oxi-
recently and now contain different techniques, dative stress-induced damage [28].
including Descemet’s stripping endothelial kera-
toplasty (DSEK) and Descemet’s membrane
endothelial keratoplasty (DMEK). Despite DSEK 31.2.2 In Vivo Stimulation of Corneal
and DMEK implement the concept of selective Endothelial Regeneration
tissue transplantation, risk of greater endothelial
loss due to donor tissue manipulation and techni- 31.2.2.1 Pharmacological Treatment
cal difficulty still pose challenges in performing
EK.  To date, there is no strong evidence from Growth Factors
randomized controlled trials of difference in Growth factors have been detected in aqueous
long-term endothelial cell loss and final visual humor and Descemet’s membrane, and even cor-
outcome between EK and PK [25]. neal endothelial cells themselves. Fibroblast
growth factor (FGF) [29–31], insulin-like growth
factor [31], insulin-like growth factor binding
31.2 Strategy for Corneal protein and hepatocyte growth factor (HGF) [31]
Endothelial Cell have all been reported as potential positive
Regeneration growth factors while TGF -β appeared to negative
and Reconstruction growth factor [32–35]. Among the many poten-
tial growth factors, FGF-2 has been proven to
Regarding keratoplasty, the difficulties such as stimulate endothelial cell proliferation. However,
limited donor availability and post-operative com- the application of FGF-2 as therapeutics is
31  Corneal Endothelium Regeneration: Future Prospects 465

l­imited by endothelial mesenchymal transforma- Another potential indication for ROCK inhibi-
tion mediated by FGF-2 [36, 37]. In 2015, Trefoil tor eye drop is acute corneal endothelial damage,
Therapeutics have announced supporting data for such as following cataract surgery. Okumura
the utility of engineered FGF-1 in corneal endo- et  al. (2015) reported 3 cases of half or more
thelial regeneration with less undesired side endothelial loss during cataract surgery whose
effects and is now moving toward an investiga- corneal edema reduced after ROCK inhibitor eye
tional new drug program [38, 39]. drop for 6  months [48]. Moloney et  al. (2017)
also described the use of a ROCK inhibitor,
ROCK Inhibitors Ripasudil, as a salvage agent in FECD cases who
Rho is a GTPase which is activated by guanine failed to regain corneal clarity after descemeto-
nucleotide exchange factors (GEFs). RhoA then rhexis without grafting [49].
activates ROCK, a serine/threonine kinase that
phosphorylates various substrates. ROCK signal- 31.2.2.2 Surgical Treatment
ing involves a wide spectrum of essential cellular
events, such as cell adhesion, motility, prolifera- Simple Descemetorhexis With or Without
tion, differentiation, and apoptosis [40]. Grafting
Therefore, it has become a potential therapeutic Unexpectedly, spontaneous recovery of corneal
target for many diseases. endothelial function after complications arising
Okumura N et al. demonstrated that the use of from EK or noncorneal surgeries have led to the
Y-27632, a selective ROCK inhibitor, can pro- development of new surgical approaches for cor-
mote the proliferation of corneal endothelial cells neal endothelial reconstruction which challenge
through stimulation of cell cycle regulators [41]. the conventional paradigms on endothelial wound
Subsequent researchers also confirmed the use of healing. Recovery of corneal clearance and
ROCK inhibitor can improve the survival of cor- pachymetry values were found in EK cases with
neal endothelial cell culture and promote corneal endothelial graft detachment from limited area to
endothelial wound healing in both rabbit and pri- subtotal detachments [50–53]. Shah et al. (2012)
mate animal models [42, 43]. reported the endothelial migration and repopula-
In 2013, results of the first clinical trial to tion after Descemet’s stripping without endothe-
determine the effect of ROCK inhibitor eye drops lial graft [45]. Giving evidence in the above
in human was published. Total 8 patients with matter, newer approaches have been tried based
corneal endothelial dysfunction scheduled for on the hypothesis that endothelial cells have
EK (4 FECD cases and 4 bullous keratopathy capacity to migrate and repopulate. In 2012,
cases) were enrolled [44]. Patients underwent Descemet’s membrane endothelial transfer
transcorneal freezing, creating a central endothe- (DMET) was reported, which was similar to
lial wound of 2-mm diameter, followed by DMEK but without the unrolling part. Dirisamer
10 mM ROCK inhibitor eye drops 6 times daily et al. enrolled 7 FECD patients and 5 pseudopha-
for 7 days. A clear trend of reduction in central kic bullous keratopathy (PBK) patients in a pro-
corneal thickness was observed in the group with spective DMET study [54]. At 6  months
central corneal edema group, but not in the dif- post-operatively, all eyes with FECD showed re-­
fuse edema group. Despite the small case number endothelialization and clinical improvement
and the absence of control group in this pilot while none of the eyes with PBK had clinical
research, the positive results prompted several improvement. This observation may suggest a
researchers to test the efficacy of ROCK inhibitor depletion of recipient endothelial cells such as
eye drop as a treatment for early-stage FECD fol- PBK requires anatomical positioning of endothe-
lowing Descemet’s membrane removal [45–47]. lial transplant while dysfunctional recipient
Further studies to determine whether corneal endothelial cells such as FECD may need only to
endothelial cells can repopulate on the Descemet’s remove the Descemet’s membrane and its guttae
membrane with guttae and validate the effective- without transplant. Hemi- and quarter-DMEK
ness of ROCK inhibitors eye drops in larger have also been proposed [55, 56]. The reported
cohorts as treatment for FECD are necessary. outcome showed gradual clearance of cornea
466 W.-T. Ho et al.

from center to periphery in first 12 months post-­ at ex vivo expansion of HCECs on the carriers or
operatively with a low but steady ECD years after scaffolds to form tissue-engineered cell sheet,
surgery. which can be used for further transplantation.
Some studies have tried to explore the possi- Since its development, different materials have
bility of the regenerative capacity of endothelial been used as carriers or scaffolds for the cells
cells further by descemetorhexis without donor [62]. The first category is naturally grown mem-
material. The rationale behind this surgery was to branes, which can be dated back to late 1970s
remove the dysfunction endothelial cells in the when Gospodarowicz and Greenburg cultured
center and to allow the healthy endothelial cell in bovine CECs onto bovine and rabbit denuded
the periphery to be free from contact inhibition, corneal stroma [63]. Subsequently, other natu-
thus to migrate and repopulate. FECD with cen- rally grown membranes were also used as the
tral guttae was the specific treatment target with HCEC carrier and has been tested in the animal
various outcomes reported [47, 57–59], mainly in model to restore the corneal transparency, includ-
case reports or case series. All patients undergone ing denuded Descemet’s membrane and amniotic
descemetorhexis were FECD and one case was a membrane [64, 65].
dual combination of FECD and posterior poly- The second category is biological polymers.
morphous corneal dystrophy proven by Compared with naturally grown membranes, bio-
Descemet’s membrane specimens [45]. In those logical polymers are composed of definite ele-
reporting spontaneous corneal clearance after ments, have more standardized properties, and
descemetorhexis without grafting, most were can avoid contamination by infectious agents or
observed between post-operative month 1 to undesirable biological substances. Biological
month 6 [49, 58]. Moloney G. et al. (2017) also polymers used as HCECs carriers include gelatin
reported using topical Ripasudil as salvage treat- and cross-linked type I collagen [66, 67]. In addi-
ment in those who failed to clear. Among the 3 tion, silk fibroin has also been used as the sub-
cases using topical Ripasudil, 2 eyes resulted stratum of HCECs [68]. Because silk fibroin has
complete corneal clearance after 2 weeks of sal- the advantage of optical transparency, nonimmu-
vage therapy while 1 eye still failed to clear and nogenicity, controllable degradation rate, and
received EK at the end [57]. In those with unsat- robust mechanical properties, it has been devel-
isfactory outcomes, residual stromal edema, slow oped as an artificial endothelial graft for DMEK
visual recovery and even posterior stromal scar- surgery [69].
ring with irregular topography were observed In addition to naturally grown membranes and
[59–61]. Despite that simple descemetorhexis biological polymers, synthetic materials comprise
proves to stimulate corneal endothelial wound the third category of carriers for HCECs.
healing in vivo, further clinical study is required Examples include polyvinylidene fluoride and
to determine the adequate patient population, the biodegradable polymers, such as poly-l-lactic and
safe extent of tissue removal and the potential poly-dl-lactic-co-glycolic acid [70, 71]. Moreover,
role of topical Ripasudil of the surgery. synthetic materials can be blended with biological
polymers to generate carriers. For instance, chito-
san and polycaprolactone (PCL), both of which
31.2.3 Ex Vivo Expansion of Corneal are biodegradable biomaterials approved by Food
Endothelial Cell and Drug Administration (FDA), can be mixed
and Transplantation into 75:25 (chitosan:PCL) blends that have better
transparency than 100% PCL, and better cell
31.2.3.1 Corneal Endothelial growth after seeding bovine CECs than 100% chi-
Regeneration with Tissue-­ tosan [72]. Another group of synthetic materials is
Engineered Cell Sheet thermos-responsive polymers. Made of homopol-
Regenerated HCECs can be transplanted into the ymer of poly(N-­isopropyl acrylamide) or copoly-
eye by using several strategies. One strategy aims mer of N-isopropyl acrylamide with
31  Corneal Endothelium Regeneration: Future Prospects 467

2-carboxyisopropylacrylamide, the carrier can trypsin, collagenase A, and dispase. Compared


undergo phase transition from hydrophobic to with others, collagenase A treatment can preserve
hydrophilic state when the temperature is below cell junctions and form cell aggregates with
the critical temperature, which can allow the cul- higher viability [78]. After enzymatic digestion,
tured HCECs detached from the carrier and form the cell aggregates are seeded onto the culture
the cell sheet ready for transplantation. The fabri- plate. Evidences suggest that culture plates
cated cell sheet has been tested for the ability to coated with extracellular matrices, such as type
reconstruct corneal endothelium in vivo [73, 74]. IV collagen, fibronectin, FNC mix, laminin-511
and -521 improve the proliferation of HCECs
[75, 79, 80]. Furthermore, supplemented media
31.2.4 Cell-Based Therapy containing EGF, insulin, transferrin, basic fibro-
blast growth factor, nerve growth factor, and pitu-
Although various materials can be used as carri- itary extracts have been used during the culture
ers or scaffolds to generate HCECs cell sheet, [79]. In addition, other strategies have also been
which can be further used for corneal endothe- developed to further overcome the limited prolif-
lium reconstruction, this strategy has some chal- erative capacity. For example, Zhu et al. demon-
lenges. For example, the sheet material may strated that the knockdown of p120 catenin can
reduce corneal transparency after transplantation, promote HCEC proliferation by disrupting the
change the cell behavior, or may not be biocom- contact inhibition [81]. ROCK inhibitors have
patible in the eye and cause excessive intraocular also been used to enhance HCEC proliferation
inflammation. Moreover, to transplant a thin cell both in vitro and in vivo [41, 82]. To further sus-
sheet onto the posterior corneal surface, sophisti- tain cell growth and prevent cell senescence,
cated surgical skill is indispensable. Direct trans- mesenchymal stem cell-conditioned medium was
plantation of cultivated corneal endothelial cells used in HCEC culture [83]. The development of
through “cell injection into the anterior chamber” these culture techniques has markedly improved
ideally bypass the difficulties in managing the the efficiency of the ex vivo expansion of HCECs.
fragile graft in EK.  To this attempt, Okumura Although HCEC proliferation can be enhanced
et al. demonstrated that direct injection of CECs through incubation with growth-stimulating
combined with a ROCK inhibitor into the ante- agents, a major concern is that the cells may
rior chamber can regenerate a corneal endothelial undergo phenotypic changes and exhibit various
monolayer and restore corneal clarity without side populations during culture, particularly in the
causing elevated intraocular pressure or exces- presence of these agents [84]. The so-called endo-
sive inflammation in rabbit and primate models thelial–mesenchymal transition (EnMT), which
[75, 76]. Another study also reported the efficacy resembles epithelial–mesenchymal ­ transition
of cell injection in restoring corneal clarity in a (EMT), is characterized by cell junction destabili-
rabbit model [77]. zation, loss of apical–basal polarity, cytoskeletal
After proving the efficacy and the safety of rearrangement, expression of alpha-­smooth mus-
cell injection therapy, there are still challenges to cle actin, and secretion of type I collagen [85].
be overcome. As aforementioned, HCECs are These characteristics may compromise the func-
considered to have limited proliferative potential. tion of HCECs, hampering the use of ex  vivo
To promote the efficiency of ex vivo expansion of expanded cells in tissue engineering. To further
HCECs, various techniques have been developed, optimize the efficacy of cell injection therapy,
including isolation of HCECs, extracellular there are several goals to be accomplished:
matrices coating, and culture media. After being
peeled off from the research grade cornea or 31.2.4.1 I dentification of the EnMT or
residual corneal-scleral rim after keratoplasty, the Side Population Markers
Descement’s membranes are subjected to enzy- The hallmark of EMT includes loss of epithelial
matic digestion to dislodge the HCECs, such as markers, such as E-cadherin, and gain of mesen-
468 W.-T. Ho et al.

chymal markers including N-cadherin and ical data is required to completely assess the
vimentin [86]. Due to its neural crest origin, long-term safety and effectiveness of cell-­
CECs constantly express N-cadherin and vimen- injection therapy in the future.
tin [87, 88], rendering these typical EMT markers
unsuitable for defining the EnMT process.
Studies have identified that the morphological 31.2.5 Stem Cell Approach
and functional change of CECs during EnMT are to Regenerate Corneal
accompanied with the activation of some EMT-­ Endothelium
triggering pathways, including FGF-2/PI3K,
Wnt/β-catenin, TGF-β1/Smad2, and Notch sig- Limited donor resource has always been an issue
naling pathway during the EnMT process [81, 84, in corneal transplantation. Although ex  vivo
89–91]. To more comprehensively identify side expansion of HCECs combined with transplanta-
populations, Hamuro et  al. used a panel of cell tion by tissue-engineered sheet or direct cell
surface markers to discriminate different HCECs, injection can ameliorate this problem, the cul-
and identified that effector cells, ie. functional tured HCECs rarely maintain normal phenotype
HCECs, express CD166  +  CD105-­ CD44-­ beyond higher passage and will undergo EnMT
CD26-CD24- among heterogeneous HCECs eventually [98]. To address this unmet need, stem
[92]. The same group further analyzed microRNA cells-based therapies have been proposed because
or exosome proteins to identify functional of its capabilities of unlimited proliferation, self-­
HCECs [93–95]. renewal and differentiation into functional cells.
The sources of stem cells for corneal endothelial
31.2.4.2 Increase of the Percentage regeneration can be classified according to the
of Functional HCECs cell potency.
Currently, several strategies have been developed
to increase the functional HCECs percentage. 31.2.5.1 Pluripotent Stem Cells
Peh et al. used dual media approach to culture the Embryonic stem cells (ESCs) harbor the pluripo-
cells, one is Ham’s F12/M199-based medium tency that can differentiate into three germ layers.
supplemented with 5% FBS and 10 ng/ml bFGF To induce the differentiation, ESCs are first
(M4 medium), and the other is human endothelial-­ induced to differentiate into neural crest cells and
SFM-­based medium supplemented with 5% FBS then into CECs, mimicking the embryonic devel-
(M5 medium). Compared with M4 medium opment process [99, 100]. The differentiated
alone, cells grown in M4 followed by mainte- cells are further validated by cell morphology,
nance in M5 exhibited better phenotype and CEC markers, and transcriptomics analysis,
higher expression of CEC specific markers [96]. which show high similarity between ESC-­derived
Alternatively, small chemicals targeting different CECs and primary HCECs [99, 100].
EnMT signaling pathway have been shown to The second source of pluripotent stem cells
suppress EnMT and preserve the phenotype and comes from induced-pluripotent stem cells
function of CECs, including TGFβ inhibitor, (iPSCs). Developed by Yamanaka et  al., iPSCs
Notch inhibitor, and MMP inhibitor [89, 91, 97]. have the advantage over ESCs that it can be cus-
These approaches, combined with selection tomized, ie. derived from the patient’s own cells,
and enrichment of functional HCECs by surface obviating the risk of immune rejection, and it
markers, microRNA, exosome proteins, may avoids the ethical issue that using cells from the
extensively improve the treatment effect of cell embryo [101]. Using iPSCs as the tool, CECs or
injection therapy. In 2013, the Kinoshita group corneal organoids can be generated by sequential
reported the first cohort of patients treated with round of differentiation program [102, 103].
cell injection therapy (UMIN000012534) and a Another difference from ESCs is that iPSCs har-
positive outcome was observed. Despite current bor epigenetic memories, as iPSCs generated
evidence has suggest a great potential, more clin- from a somatic cell type exhibit tendency for
31  Corneal Endothelium Regeneration: Future Prospects 469

spontaneous re-differentiation into the same cell further differentiation, these progenitor cells can
[104]. This property may be utilized to generate be served as a promising source for regenerating
iPSCs derived from CECs that can be further re-­ corneal endothelium.
differentiated into functional cells.

31.2.5.2 Multipotent Stem Cells 31.3 Conclusion


Multipotent stem cells, with more restricted dif-
ferentiation capacity compared with pluripotent The field of CEC regeneration evolves rapidly.
cells, are implicated to regenerate corneal endo- Newer surgical techniques greatly improve the
thelium. Bone marrow-derived endothelial pro- safety and treatment outcome. Emerging medica-
genitor cells (BEPCs), co-cultured with CECs tions can prevent CEC loss or may enhance cell
using transwell system, can develop polygonal proliferation in  vivo. Advances in cell culture
cell shape and markers of CECs [105]. Likewise, technology also help ex vivo expansion of CECs
human umbilical cord blood mesenchymal stem that once considered poorly proliferated.
cells (hUCB MSCs) can be induced to differenti- Different modalities, either tissue-engineered
ate into CEC [106]. Direct injection of hUCB cell sheet or direct cell injection, facilitate the
MSCs into anterior chamber has also been shown implantation of CECs onto the posterior cornea
to repair corneal endothelial defects in rabbit and regeneration of CEC monolayer. Finally,
model [107]. stem cell-based treatment may fundamentally
solve the issue of organ shortage or even contrib-
31.2.5.3 CEC Progenitor Cell ute to customized regenerative medicine. These
HCECs are largely believed to have limited pro- developments pave the way for better treatment
liferative capacity in vivo. However, there are sev- and improvement in live quality for those afflicted
eral evidences indicating that corneal endothelium by corneal endothelial diseases.
stem cells may locate adjacent to the endothelial
periphery in the posterior limbus. First, HECEs Conflict of Interest Wei-Ting Ho, Hsin-Yu Liu, Fung-
derived from the peripheral part of the cornea Rong Hu and I-Jong Wang declare that they have no con-
flict of interest.
showed more vibrant mitotic activity compared
with cells from central part [108]. Besides, several
Informed Consent  No human studies were carried out
stem cell markers were detected in the trabecular by the authors for this article.
meshwork and the transition zone of the human
posterior limbus, including Oct-3/4, Wnt-1, Animal Studies  All institutional and national guidelines
telomerase, alkaline phosphotase, nestin, Pax-6, for the care and use of laboratory animals were followed.
and Sox-2 [109, 110]. Moreover, cultured HCECs
can form spherical colonies in sphere forming
assay, suggesting the existence of precursor cells Reference
[111]. These cells may contribute to replacement
of HCECs, but its slow proliferation in physiolog- 1. Barry PA, Petroll WM, Andrews PM, Cavanagh HD,
ical environment may be insufficient to compen- Jester JV. The spatial organization of corneal endo-
thelial cytoskeletal proteins and their relationship to
sate for the cell loss that leads to corneal edema. the apical junctional complex. Invest Ophthalmol
Nevertheless, isolating these precursors and Vis Sci. 1995;36(6):1115–24.
expanding in vitro are attracting choices for cor- 2. Bahn CF, Glassman RM, MacCallum DK, Lillie JH,
neal endothelial regeneration. Katikireddy et  al. Meyer RF, Robinson BJ, et  al. Postnatal develop-
ment of corneal endothelium. Invest Ophthalmol Vis
have isolated a subpopulation of cells from human Sci. 1986;27(1):44–51.
corneal endothelium exhibiting rapid prolifera- 3. Bourne WM, Nelson LR, Hodge DO.  Central cor-
tion and neural crest markers, and can be cultured neal endothelial cell changes over a ten-year period.
in  vitro up to 80 passages [112]. Coupled with Invest Ophthalmol Vis Sci. 1997;38(3):779–82.
470 W.-T. Ho et al.

4. Wilson RS, Roper-Hall MJ.  Effect of age on the tion. Ophthalmology. 1994;101(6):1014–22. discus-
endothelial cell count in the normal eye. Br J sion 22-3
Ophthalmol. 1982;66(8):513–5. 23. Armitage WJ, Tullo AB, Larkin DF.  The first suc-
5. Senoo T, Joyce NC.  Cell cycle kinetics in corneal cessful full-thickness corneal transplant: a commen-
endothelium from old and young donors. Invest tary on Eduard Zirm's landmark paper of 1906. Br J
Ophthalmol Vis Sci. 2000;41(3):660–7. Ophthalmol. 2006;90(10):1222–3.
6. Senoo T, Obara Y, Joyce NC. EDTA: a promoter of 24. Crawford AZ, Patel DV, McGhee C. A brief history
proliferation in human corneal endothelium. Invest of corneal transplantation: from ancient to modern.
Ophthalmol Vis Sci. 2000;41(10):2930–5. Oman J Ophthalmol. 2013;6(Suppl 1):S12–7.
7. Joyce NC. Proliferative capacity of corneal endothe- 25. Nanavaty MA, Wang X, Shortt AJ. Endothelial kera-
lial cells. Exp Eye Res. 2012;95(1):16–23. toplasty versus penetrating keratoplasty for Fuchs
8. Yoshida K, Kase S, Nakayama K, Nagahama H, endothelial dystrophy. Cochrane Database Syst Rev.
Harada T, Ikeda H, et al. Involvement of p27KIP1 in 2014;2:CD008420.
the proliferation of the developing corneal endothe- 26. Jurkunas UV, Bitar MS, Funaki T, Azizi B. Evidence
lium. Invest Ophthalmol Vis Sci. 2004;45(7):2163–7. of oxidative stress in the pathogenesis of fuchs
9. Joyce NC, Harris DL, Mello DM.  Mechanisms of endothelial corneal dystrophy. Am J Pathol.
mitotic inhibition in corneal endothelium: contact 2010;177(5):2278–89.
inhibition and TGF-beta2. Invest Ophthalmol Vis 27. Ziaei A, Schmedt T, Chen Y, Jurkunas
Sci. 2002;43(7):2152–9. UV.  Sulforaphane decreases endothelial cell
10. Teruo Nishida SSaNM. In: Mannis MJ, Holland EJ, apoptosis in fuchs endothelial corneal dystro-
editors. Cornea. 4th ed. St. Louis, MO.: Elsevier; phy: a novel treatment. Invest Ophthalmol Vis Sci.
2017. p. 1–22. 2013;54(10):6724–34.
11. Bourne WM.  Primary corneal endotheliopathies. 28. Liu X, Ward K, Xavier C, Jann J, Clark AF, Pang
Am J Ophthalmol. 1983;95(6):852–4. IH, et  al. The novel triterpenoid RTA 408 protects
12. Magovern M, Beauchamp GR, McTigue JW, Fine human retinal pigment epithelial cells against H2O2-­
BS, Baumiller RC. Inheritance of Fuchs' combined induced cell injury via NF-E2-related factor 2 (Nrf2)
dystrophy. Ophthalmology. 1979;86(10):1897–923. activation. Redox Biol. 2016;8:98–109.
13. Bourne WM. Biology of the corneal endothelium in 29. Schulz MW, Chamberlain CG, de Iongh RU,
health and disease. Eye (Lond). 2003;17(8):912–8. McAvoy JW.  Acidic and basic FGF in ocular
14. Inoue Y.  Review of clinical and basic approaches media and lens: implications for lens polar-
to corneal endotheliitis. Cornea. 2014;33(Suppl ity and growth patterns. Development. 1993;
11):S3–8. 118(1):117–26.
15. Carlson KH, Ilstrup DM, Bourne WM, Dyer 30. Tripathi RC, Borisuth NS, Tripathi BJ.  Detection,
JA.  Effect of silicone elastomer contact lens wear quantification, and significance of basic fibroblast
on endothelial cell morphology in aphakic eyes. growth factor in the aqueous humor of man, cat, dog
Cornea. 1990;9(1):45–7. and pig. Exp Eye Res. 1992;54(3):447–54.
16. Ang LP, Higashihara H, Sotozono C, 31. Rieck P, Oliver L, Engelmann K, Fuhrmann G,
Shanmuganathan VA, Dua H, Tan DT, et  al. Hartmann C, Courtois Y.  The role of exogenous/
Argon laser iridotomy-induced bullous keratopa- endogenous basic fibroblast growth factor (FGF2)
thy a growing problem in Japan. Br J Ophthalmol. and transforming growth factor beta (TGF beta-1)
2007;91(12):1613–5. on human corneal endothelial cells proliferation
17. Wang PX, Koh VT, Loon SC. Laser iridotomy and in vitro. Exp Cell Res. 1995;220(1):36–46.
the corneal endothelium: a systemic review. Acta 32. van Setten GB, Fagerholm P, Philipson B, Schultz
Ophthalmol. 2014;92(7):604–16. G. Growth factors and their receptors in the anterior
18. Takahashi H. Corneal endothelium and phacoemul- chamber. Absence of epidermal growth factor and
sification. Cornea. 2016;35(Suppl 1):S3–7. transforming growth factor alpha in human aqueous
19. Miyata K, Nagamoto T, Maruoka S, Tanabe T, humor. Ophthalmic Res. 1996;28(6):361–4.
Nakahara M, Amano S.  Efficacy and safety of the 33. Wilson SE, Schultz GS, Chegini N, Weng J, He
soft-shell technique in cases with a hard lens nucleus. YG. Epidermal growth factor, transforming growth
J Cataract Refract Surg. 2002;28(9):1546–50. factor alpha, transforming growth factor beta, acidic
20. Suzuki H, Oki K, Igarashi T, Shiwa T, Takahashi fibroblast growth factor, basic fibroblast growth fac-
H.  Temperature in the anterior chamber dur- tor, and interleukin-1 proteins in the cornea. Exp Eye
ing phacoemulsification. J Cataract Refract Surg. Res. 1994;59(1):63–71.
2014;40(5):805–10. 34. Jampel HD, Roche N, Stark WJ, Roberts
21. Murano N, Ishizaki M, Sato S, Fukuda Y, Takahashi AB.  Transforming growth factor-beta in human
H.  Corneal endothelial cell damage by free radi- aqueous humor. Curr Eye Res. 1990;9(10):963–9.
cals associated with ultrasound oscillation. Arch 35. Cousins SW, McCabe MM, Danielpour D, Streilein
Ophthalmol. 2008;126(6):816–21. JW. Identification of transforming growth factor-beta
22. Bourne WM, Nelson LR, Hodge DO.  Continued as an immunosuppressive factor in aqueous humor.
endothelial cell loss ten years after lens implanta- Invest Ophthalmol Vis Sci. 1991;32(8):2201–11.
31  Corneal Endothelium Regeneration: Future Prospects 471

36. Gu X, Seong GJ, Lee YG, Kay EP.  Fibroblast 50. Balachandran C, Ham L, Verschoor CA, Ong TS, van
growth factor 2 uses distinct signaling pathways der Wees J, Melles GR. Spontaneous corneal clear-
for cell proliferation and cell shape changes in cor- ance despite graft detachment in descemet mem-
neal endothelial cells. Invest Ophthalmol Vis Sci. brane endothelial keratoplasty. Am J Ophthalmol.
1996;37(11):2326–34. 2009;148(2):227–34. e1
37. Lee JG, Ko MK, Kay EP.  Endothelial mesenchy- 51. Dirisamer M, Dapena I, Ham L, van Dijk K, Oganes
mal transformation mediated by IL-1beta-induced O, Frank LE, et al. Patterns of corneal endothelial-
FGF-2  in corneal endothelial cells. Exp Eye Res. ization and corneal clearance after descemet mem-
2012;95(1):35–9. brane endothelial keratoplasty for fuchs endothelial
38. Xia X, Babcock JP, Blaber SI, Harper KM, Blaber dystrophy. Am J Ophthalmol. 2011;152(4):543–55.
M.  Pharmacokinetic properties of 2nd-generation e1
fibroblast growth factor-1 mutants for therapeutic 52. Dirisamer M, van Dijk K, Dapena I, Ham L,
application. PLoS One. 2012;7(11):e48210. Oganes O, Frank LE, et  al. Prevention and man-
39. Degirolamo C, Sabba C, Moschetta A. Therapeutic agement of graft detachment in descemet mem-
potential of the endocrine fibroblast growth factors brane endothelial keratoplasty. Arch Ophthalmol.
FGF19, FGF21 and FGF23. Nat Rev Drug Discov. 2012;130(3):280–91.
2016;15(1):51–69. 53. Zafirakis P, Kymionis GD, Grentzelos MA, Livir-­
40. Riento K, Ridley AJ.  Rocks: multifunctional Rallatos G.  Corneal graft detachment without cor-
kinases in cell behaviour. Nat Rev Mol Cell Biol. neal edema after descemet stripping automated
2003;4(6):446–56. endothelial keratoplasty. Cornea. 2010;29(4):456–8.
41. Okumura N, Ueno M, Koizumi N, Sakamoto Y, 54. Dirisamer M, Yeh RY, van Dijk K, Ham L, Dapena
Hirata K, Hamuro J, et  al. Enhancement on pri- I, Melles GR. Recipient endothelium may relate to
mate corneal endothelial cell survival in  vitro by corneal clearance in descemet membrane endothelial
a ROCK inhibitor. Invest Ophthalmol Vis Sci. transfer. Am J Ophthalmol. 2012;154(2):290–6 e1.
2009;50(8):3680–7. 55. Lam FC, Baydoun L, Dirisamer M, Lie J, Dapena I,
42. Okumura N, Nakano S, Kay EP, Numata R, Ota A, Melles GR. Hemi-Descemet membrane endothelial
Sowa Y, et  al. Involvement of cyclin D and p27  in keratoplasty transplantation: a potential method for
cell proliferation mediated by ROCK inhibitors increasing the pool of endothelial graft tissue. JAMA
Y-27632 and Y-39983 during corneal endothe- Ophthalmol. 2014;132(12):1469–73.
lium wound healing. Invest Ophthalmol Vis Sci. 56. Lie JT, Lam FC, Groeneveld-van Beek EA, van der
2014;55(1):318–29. Wees J, Melles GR.  Graft preparation for hemi-­
43. Peh GS, Adnan K, George BL, Ang HP, Seah XY, Descemet membrane endothelial keratoplasty (hemi-­
Tan DT, et al. The effects of Rho-associated kinase DMEK). Br J Ophthalmol. 2016;100(3):420–4.
inhibitor Y-27632 on primary human corneal endo- 57. Moloney G, Petsoglou C, Ball M, Kerdraon Y,
thelial cells propagated using a dual media approach. Hollhumer R, Spiteri N, et  al. Descemetorhexis
Sci Rep. 2015;5:9167. without grafting for Fuchs endothelial dystrophy-­
44. Okumura N, Koizumi N, Kay EP, Ueno M, Sakamoto supplementation with topical ripasudil. Cornea.
Y, Nakamura S, et al. The ROCK inhibitor eye drop 2017;36(6):642–8.
accelerates corneal endothelium wound healing. 58. Borkar DS, Veldman P, Colby KA.  Treatment of
Invest Ophthalmol Vis Sci. 2013;54(4):2493–502. Fuchs endothelial dystrophy byddescemet strip-
45. Shah RD, Randleman JB, Grossniklaus ping without endothelial keratoplasty. Cornea.
HE. Spontaneous corneal clearing after Descemet's 2016;35(10):1267–73.
stripping without endothelial replacement. 59. Iovieno A, Neri A, Soldani AM, Adani C, Fontana
Ophthalmology. 2012;119(2):256–60. L.  Descemetorhexis without graft placement for
46. Galvis V, Tello A, Miotto G.  Human corneal the treatment of Fuchs endothelial dystrophy: pre-
endothelium regeneration. Ophthalmology. liminary results and review of the literature. Cornea.
2012;119(8):1714–5. 2017;36(6):637–41.
47. Bleyen I, Saelens IE, van Dooren BT, van Rij 60. Arbelaez JG, Price MO, Price FW Jr. Long-term
G.  Spontaneous corneal clearing after Descemet's follow-up and complications of stripping des-
stripping. Ophthalmology. 2013;120(1):215. cemet membrane without placement of graft in
48. Okumura N, Inoue R, Okazaki Y, Nakano S, eyes with Fuchs endothelial dystrophy. Cornea.
Nakagawa H, Kinoshita S, et  al. Effect of the rho 2014;33(12):1295–9.
kinase inhibitor Y-27632 on corneal endothe- 61. Koenig SB. Planned descemetorhexis without endo-
lial wound healing. Invest Ophthalmol Vis Sci. thelial keratoplasty in eyes with Fuchs corneal endo-
2015;56(10):6067–74. thelial dystrophy. Cornea. 2015;34(9):1149–51.
49. Moloney G, Chan UT, Hamilton A, Zahidin AM, 62. Teichmann J, Valtink M, Nitschke M, Gramm S,
Grigg JR, Devasahayam RN.  Descemetorhexis Funk RH, Engelmann K, et al. Tissue engineering of
for Fuchs' dystrophy. Can J Ophthalmol. the corneal endothelium: a review of carrier materi-
2015;50(1):68–72. als. J Funct Biomater. 2013;4(4):178–208.
472 W.-T. Ho et al.

63. Gospodarowicz D, Greenburg G, Alvarado capable of regenerating in  vivo endothelial tissue.
J. Transplantation of cultured bovine corneal endo- Am J Pathol. 2012;181(1):268–77.
thelial cells to rabbit cornea: clinical implications 76. Okumura N, Sakamoto Y, Fujii K, Kitano J, Nakano
for human studies. Proc Natl Acad Sci U S A. S, Tsujimoto Y, et  al. Rho kinase inhibitor enables
1979;76(1):464–8. cell-based therapy for corneal endothelial dysfunc-
64. Proulx S, Bensaoula T, Nada O, Audet C, d'Arc tion. Sci Rep. 2016;6:26113.
Uwamaliya J, Devaux A, et  al. Transplantation of 77. Sabater AL, Andreu EJ, Garcia-Guzman M, Lopez
a tissue-engineered corneal endothelium recon- T, Abizanda G, Perez VL, et  al. Combined PI3K/
structed on a devitalized carrier in the feline model. Akt and Smad2 activation promotes corneal endo-
Invest Ophthalmol Vis Sci. 2009;50(6):2686–94. thelial cell proliferation. Invest Ophthalmol Vis Sci.
65. Fan T, Zhao J, Ma X, Xu X, Zhao W, Xu 2017;58(2):745–54.
B.  Establishment of a continuous untransfected 78. Li W, Sabater AL, Chen YT, Hayashida Y, Chen SY,
human corneal endothelial cell line and its biocom- He H, et al. A novel method of isolation, preserva-
patibility to denuded amniotic membrane. Mol Vis. tion, and expansion of human corneal endothelial
2011;17:469–80. cells. Invest Ophthalmol Vis Sci. 2007;48(2):614–20.
66. Watanabe R, Hayashi R, Kimura Y, Tanaka Y, 79. Peh GS, Beuerman RW, Colman A, Tan DT, Mehta
Kageyama T, Hara S, et al. A novel gelatin hydrogel JS.  Human corneal endothelial cell expansion for
carrier sheet for corneal endothelial transplantation. corneal endothelium transplantation: an overview.
Tissue Eng Part A. 2011;17(17–18):2213–9. Transplantation. 2011;91(8):811–9.
67. Mimura T, Yamagami S, Yokoo S, Usui T, Tanaka 80. Okumura N, Kakutani K, Numata R, Nakahara M,
K, Hattori S, et  al. Cultured human corneal endo- Schlotzer-Schrehardt U, Kruse F, et al. Laminin-511
thelial cell transplantation with a collagen sheet and -521 enable efficient in  vitro expansion of
in a rabbit model. Invest Ophthalmol Vis Sci. human corneal endothelial cells. Invest Ophthalmol
2004;45(9):2992–7. Vis Sci. 2015;56(5):2933–42.
68. Madden PW, Lai JN, George KA, Giovenco T, 81. Zhu YT, Chen HC, Chen SY, Tseng SCG.  Nuclear
Harkin DG, Chirila TV. Human corneal endothelial p120 catenin unlocks mitotic block of contact-­
cell growth on a silk fibroin membrane. Biomaterials. inhibited human corneal endothelial monolayers
2011;32(17):4076–84. without disrupting adherent junctions. J Cell Sci.
69. Vazquez N, Rodriguez-Barrientos CA, Aznar-­ 2012;125(15):3636–48.
Cervantes SD, Chacon M, Cenis JL, Riestra AC, 82. Okumura N, Koizumi N, Ueno M, Sakamoto Y,
et al. Silk fibroin films for corneal endothelial regen- Takahashi H, Hamuro J, et al. The new therapeutic
eration: transplant in a rabbit descemet membrane concept of using a rho kinase inhibitor for the treat-
endothelial keratoplasty. Invest Ophthalmol Vis Sci. ment of corneal endothelial dysfunction. Cornea.
2017;58(9):3357–65. 2011;30(Suppl 1):S54–9.
70. Wang TJ, Wang IJ, Chen YH, Lu JN, Young 83. Nakahara M, Okumura N, Kay EP, Hagiya M,
TH.  Polyvinylidene fluoride for prolifera- Imagawa K, Hosoda Y, et  al. Corneal endothelial
tion and preservation of bovine corneal endo- expansion promoted by human bone marrow mes-
thelial cells by enhancing type IV collagen enchymal stem cell-derived conditioned medium.
production and deposition. J Biomed Mater Res PLoS One. 2013;8(7):e69009.
A. 2012;100(1):252–60. 84. Lee JG, Kay EP.  FGF-2-mediated signal transduc-
71. Hadlock T, Singh S, Vacanti JP, McLaughlin tion during endothelial mesenchymal ­transformation
BJ.  Ocular cell monolayers cultured on biodegrad- in corneal endothelial cells. Exp Eye Res.
able substrates. Tissue Eng. 1999;5(3):187–96. 2006;83(6):1309–16.
72. Wang TJ, Wang IJ, Lu JN, Young TH.  Novel 85. Roy O, Leclerc VB, Bourget JM, Theriault M, Proulx
chitosan-­polycaprolactone blends as potential scaf- S. Understanding the process of corneal endothelial
fold and carrier for corneal endothelial transplanta- morphological change in  vitro. Invest Ophthalmol
tion. Mol Vis. 2012;18:255–64. Vis Sci. 2015;56(2):1228–37.
73. Sumide T, Nishida K, Yamato M, Ide T, Hayashida 86. Lamouille S, Xu J, Derynck R.  Molecular mecha-
Y, Watanabe K, et  al. Functional human cor- nisms of epithelial-mesenchymal transition. Nat Rev
neal endothelial cell sheets harvested from Mol Cell Biol. 2014;15(3):178–96.
temperature-­responsive culture surfaces. FASEB J. 87. Jakobiec FA, Bhat P.  Retrocorneal membranes: a
2006;20(2):392–4. comparative immunohistochemical analysis of kera-
74. Lai JY, Chen KH, Hsiue GH.  Tissue-engineered tocytic, endothelial, and epithelial origins. Am J
human corneal endothelial cell sheet transplanta- Ophthalmol. 2010;150(2):230–42. e2
tion in a rabbit model using functional biomaterials. 88. Zhu YT, Hayashida Y, Kheirkhah A, He H, Chen
Transplantation. 2007;84(10):1222–32. SY, Tseng SC. Characterization and comparison of
75. Okumura N, Koizumi N, Ueno M, Sakamoto Y, intercellular adherent junctions expressed by human
Takahashi H, Tsuchiya H, et  al. ROCK inhibitor corneal endothelial cells in vivo and in vitro. Invest
converts corneal endothelial cells into a phenotype Ophthalmol Vis Sci. 2008;49(9):3879–86.
31  Corneal Endothelium Regeneration: Future Prospects 473

89. Ho WT, Chang JS, Su CC, Chang SW, Hu FR, Jou cells to corneal endothelial cell-like cells: a tran-
TS, et  al. Inhibition of matrix metalloproteinase scriptomic analysis. Exp Eye Res. 2016;151:107–14.
activity reverses corneal endothelial-mesenchymal 101. Takahashi K, Yamanaka S.  Induction of plu-
transition. Am J Pathol. 2015;185(8):2158–67. ripotent stem cells from mouse embryonic and
90. Okumura N, Kay EP, Nakahara M, Hamuro J, adult fibroblast cultures by defined factors. Cell.
Kinoshita S, Koizumi N.  Inhibition of TGF-beta 2006;126(4):663–76.
signaling enables human corneal endothelial cell 102. Foster JW, Wahlin K, Adams SM, Birk DE,
expansion in vitro for use in regenerative medicine. Zack DJ, Chakravarti S.  Cornea organoids from
PLoS One. 2013;8(2):e58000. human induced pluripotent stem cells. Sci Rep.
91. Li C, Dong F, Jia Y, Du H, Dong N, Xu Y, et al. Notch 2017;7:41286.
signal regulates corneal endothelial-to-mesenchymal 103. Zhao JJ, Afshari NA. Generation of human corneal
transition. Am J Pathol. 2013;183(3):786–95. endothelial cells via in  vitro ocular lineage restric-
92. Hamuro J, Toda M, Asada K, Hiraga A, Schlotzer-­ tion of pluripotent stem cells. Invest Ophthalmol Vis
Schrehardt U, Montoya M, et al. Cell homogeneity Sci. 2016;57(15):6878–84.
indispensable for regenerative medicine by cultured 104. Hu Q, Friedrich AM, Johnson LV, Clegg
human corneal endothelial cells. Invest Ophthalmol DO.  Memory in induced pluripotent stem cells:
Vis Sci. 2016;57(11):4749–61. reprogrammed human retinal-pigmented epithelial
93. Ueno M, Asada K, Toda M, Hiraga A, Montoya cells show tendency for spontaneous redifferentia-
M, Sotozono C, et  al. MicroRNA profiles qual- tion. Stem Cells. 2010;28(11):1981–91.
ify phenotypic features of cultured human cor- 105. Shao C, Fu Y, Lu W, Fan X. Bone marrow-derived
neal endothelial cells. Invest Ophthalmol Vis Sci. endothelial progenitor cells: a promising therapeutic
2016;57(13):5509–17. alternative for corneal endothelial dysfunction. Cells
94. Ueno M, Asada K, Toda M, Nagata K, Sotozono C, Tissues Organs. 2011;193(4):253–63.
Kosaka N, et al. Concomitant evaluation of a panel 106. Joyce NC, Harris DL, Markov V, Zhang Z, Saitta
of exosome proteins and MiRs for qualification of B. Potential of human umbilical cord blood mesen-
cultured human corneal endothelial cells. Invest chymal stem cells to heal damaged corneal endothe-
Ophthalmol Vis Sci. 2016;57(10):4393–402. lium. Mol Vis. 2012;18:547–64.
95. Ueno M, Asada K, Toda M, Schlotzer-Schrehardt 107. Shao C, Chen J, Chen P, Zhu M, Yao Q, Gu P, et al.
U, Nagata K, Montoya M, et  al. Gene signature- Targeted transplantation of human umbilical cord
based development of ELISA assays for repro- blood endothelial progenitor cells with immunomag-
ducible qualification of cultured human corneal netic nanoparticles to repair corneal endothelium
endothelial cells. Invest Ophthalmol Vis Sci. defect. Stem Cells Dev. 2015;24(6):756–67.
2016;57(10):4295–305. 108. Bednarz J, Rodokanaki-von Schrenck A, Engelmann
96. Peh GS, Chng Z, Ang HP, Cheng TY, Adnan K, Seah K. Different characteristics of endothelial cells from
XY, et al. Propagation of human corneal endothelial central and peripheral human cornea in primary
cells? A novel dual media approach. Cell Transplant. culture and after subculture. In Vitro Cell Dev Biol
2015;24(2):287–304. Anim. 1998;34(2):149–53.
97. Okumura N, Kay EP, Nakahara M, Hamuro J, 109. Whikehart DR, Parikh CH, Vaughn AV, Mishler K,
Kinoshita S, Koizumi N. Inhibition of TGF-β signal- Edelhauser HF. Evidence suggesting the existence of
ing enables human corneal endothelial cell expan- stem cells for the human corneal endothelium. Mol
sion in vitro for use in regenerative medicine. PLoS Vis. 2005;11:816–24.
One. 2013;8(2):e58000. 110. McGowan SL, Edelhauser HF, Pfister RR, Whikehart
98. Bartakova A, Alvarez-Delfin K, Weisman AD, DR. Stem cell markers in the human posterior lim-
Salero E, Raffa GA, Merkhofer RM Jr, et al. Novel bus and corneal endothelium of unwounded and
identity and functional markers for human cor- wounded corneas. Mol Vis. 2007;13:1984–2000.
neal endothelial cells. Invest Ophthalmol Vis Sci. 111. Yokoo S, Yamagami S, Yanagi Y, Uchida S, Mimura
2016;57(6):2749–62. T, Usui T, et  al. Human corneal endothelial cell
99. McCabe KL, Kunzevitzky NJ, Chiswell BP, Xia precursors isolated by sphere-forming assay. Invest
X, Goldberg JL, Lanza R.  Efficient generation of Ophthalmol Vis Sci. 2005;46(5):1626–31.
human embryonic stem cell-derived corneal endo- 112. Katikireddy KR, Schmedt T, Price MO, Price FW,
thelial cells by directed differentiation. PLoS One. Jurkunas UV.  Existence of neural crest-derived
2015;10(12):e0145266. progenitor cells in normal and Fuchs endothe-
100. Song Q, Yuan S, An Q, Chen Y, Mao FF, Liu Y, et al. lial dystrophy corneal endothelium. Am J Pathol.
Directed differentiation of human embryonic stem 2016;186(10):2736–50.
Part VI
Bioengineering Cornea Surgery
Umbilical Cord Stem Cells
in the Treatment of Corneal
32
Diseases

Mohammed Ziaei, Jie Zhang, Dipika V. Patel,


and Charles N. J. McGhee

32.1 Introduction Stem cells can be classified according to their


origin or potency. On the basis of trans-­
The inbuilt capacity of self-renewal and differen- differentiation potency stem cells can be either
tiation of stem cells has rapidly established stem totipotent, pluripotent, multipotent, oligopotent
cell-based therapies as a major new direction in and unipotent.
the investigation of novel therapies for human Totipotent stem cells can differentiate into
disease. The field of ophthalmology, and in par- embryonic and extra-embryonic cell types and
ticular corneal surgery, has witnessed great prog- thus can construct a complete, viable organism.
ress in the development of stem cell therapy for Pluripotent stem cells are descendants of totipo-
tissue engineering applications in recent years. tent cells and have the potential to differentiate
This is of key importance as specialists deal with into any of the three germ layers. Multipotent
the ever-increasing burden of corneal disease and stem cells can differentiate into a limited number
a chronic global shortage of suitable donor mate- of cells i.e. those of a closely related family of
rial in their efforts to treat >1.5 million new cases cells.
of monocular corneal blindness annually [1]. Oligopotent stem cells can differentiate into
only a few cell lineages whereas unipotent cells
possess self-renewal properties but can produce
32.2 Stem Cells only one cell type.
Human stem cells can be isolated from the fol-
Stem cells are cells with three basic properties: lowing four sources:
they are capable of self-renewal, they have the
capacity to undergo differentiation to become spe- 1. Embryonic tissue: these stem cells are pluripo-
cialized progeny cells, and they have the potential tent, capable of differentiating into embryonic
to renew the tissue that they populate [2]. tissue such as ectoderm, mesoderm and endo-
derm and are harvested from the blastocyst of
5-day-old pre-implantation embryos [3].
2. Foetal tissue: these stem cells are multipotent
M. Ziaei · J. Zhang (*) · D. V. Patel · C. N. J. McGhee cells capable of differentiating into a limited
Department of Ophthalmology, New Zealand number of cell types, as dictated by the degree
National Eye Centre, Faculty of Medical and Health of prior differentiation and can be isolated
Sciences, The University of Auckland,
Auckland, New Zealand from two distinct sources; the foetus proper
e-mail: jie.zhang@auckland.ac.nz and supportive extra-embryonic tissues, such

© Springer Nature Switzerland AG 2019 477


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_32
478 M. Ziaei et al.

as amniotic fluid, umbilical cord, placenta and 32.3 H


 uman Umbilical Cord Stem
amnion. Cells
3. Adult tissue: these stem cells are multipo-
tent cells capable of differentiating into a The umbilical cord was the first foetal tissue to be
more limited number of cell types than foe- explored for the presence of stem cells. The cord
tal stem cells and have been isolated from contains two arteries and a vein, wrapped in a
the corneal limbus, stroma, and transition sheath of proteoglycan rich, mucoid, porous con-
zone (periphery of Schwalbe’s line to the nective tissue (Wharton’s jelly), and covered by a
anterior portion of the trabecular mesh- simple epithelial layer believed to be derived
work) [4]. from amniotic membrane epithelium (Fig. 32.1).
4. Induced pluripotent stem cells (iPSCs): adult A reservoir of blood is present in the umbilical
somatic cells can be genetically repro- cord vein post-partum.
grammed to an embryonic stem cell-like state Umbilical cord-derived stem cells represent a
by forcing expression of key transcription fac- class of stem cells that have a number of advan-
tors such as OCT4, SOX2, KLF4, and c-Myc; tages over other traditional sources of stem cells.
important for maintaining the defining proper- The advantages and disadvantages of umbilical
ties of embryonic stem cells. The resulting cord-derived stem cells are summarized in
cells can give rise to all tissue types [5]. Table 32.1 [7].

Fig. 32.1  Cross-section of the human umbilical cord. A Umbilical vein: Umbilical cord blood derived mesenchy-
artery, V vein, WJ Wharton’s jelly, UCL umbilical cord mal stem cells (UCBMSCs) & Human umbilical cord
lining, SA, IV, and PV subamnion, intervascular, and peri- blood endothelial progenitor cells (UCB EPCs). Wharton’s
vascular zones of Wharton’s jelly, VW blood vessel wall. jelly: Umbilical Cord Mesenchymal Stem Cells
Hematoxylin and eosin staining, scale bar 200 μm. Stem (UCMSCs). (Reproduced from [6] with permission.
cells found in the umbilical cord: Cord lining: Mucin-­ CC-BY 4.0 https://creativecommons.org/licenses/by/4.0/)
expressing cord lining epithelial cell (CLEC-muc).
32  Umbilical Cord Stem Cells in the Treatment of Corneal Diseases 479

Table 32.1  A summary of the advantages and disadvantages of therapy utilizing stem cells derived from the umbilical
cord [7]
Advantages
Easy, non-invasive and ethically acceptable collection of cells.
Widely available stem cell banks in the public and private setting.
Cryopreservation of cells possible.
Multiple cell lineages available in the umbilical cord.
Low risk of teratoma formation.
UCMSCs exhibit a gene expression profile more similar to that of embryonic stem cells and possess faster
self-renewal compared to other mesenchymal stem cells.
A substantial number of UCMSCs can be derived after ex vivo expansion.
UCMSCs can be considered for autologous/allogeneic use.
UCMSCs retain their multipotency for longer periods than other mesenchymal stem cells.
UCMSCs are considered to have low immunogenicity with good immunosuppressive properties in vitro and in vivo
and therefore pose a lower risk for rejection and graft-vs-host disease.
UCMSCs can act as a feeder layer for other pluripotent stem cells, and exert a non-tumorigenic effect on other cell
types.
Disadvantages
The differentiation of UCMSCs can be partial for some cell lineages.
Transplantation can lead to transfer of rare infectious or genetic diseases.
Banking of umbilical cord derived stem cells is expensive, lacks stringent regulation and long-term health of stored
cells remains unknown.
Current methods for isolation and culture of stem cells from the umbilical cord are cumbersome and result in a low
yield of primary cells.

Currently, there are numerous methods for 32.4 Corneal Regeneration


isolation and culture of stem cells from the
umbilical cord. These approaches are mainly 32.4.1 Epithelial Regeneration
based on enzymatic treatment or explantation
techniques in order to isolate different stem cell Corneal epithelium is an ectoderm-derived, non-­
populations from the cord, with subsequent cul- keratinized, stratified layer with a unique cyto-
ture in specific media to encourage differentia- keratin expression pattern and a reported
tion of cell types. Stem cells can be derived from: thickness of between 48–53 μm [10]. In health,
the corneal epithelium is capable of rapid regen-
• Umbilical cord blood: a source of primarily eration and is maintained by the integrity and
hematopoietic lineage cells with a small mul- functionality of a specialized stem cell popula-
tipotent cell population (0.00003% of all tion, known as limbal stem cells (LSCs). The
extracted nucleated cells). LSCs are located in the basal region of the lim-
• Umbilical cord lining membrane: a source for bus, a narrow transition zone between the periph-
umbilical cord blood derived mesenchymal eral cornea and conjunctiva [11].
stem cells (UCBMSCs) which have the poten- Regeneration of the corneal epithelial sur-
tial to differentiate into multiple cell types, in face has traditionally been thought to involve
addition to generating ectodermal and endo- division, migration, and maturation of LSCs.
dermal lineages by crossing the germline bar- When this process breaks down due to destruc-
rier [8]. tion of the delicate micro-environmental niche
• Wharton’s jelly: a rich source of umbilical of the LSC, limbal stem cell deficiency (LSCD)
cord mesenchymal stem cells (UCMSCs), results. Advanced LSCD has a reported inci-
approximately 400,000 cells per umbilical dence of 3.81  ×  10−6 cases per million,
cord compared to 1 mesenchymal stem cell in per  annum and is associated with devastating
10,000 nucleated bone marrow cells [9]. visual loss [12, 13].
480 M. Ziaei et al.

A variety of corneal limbal stem cell grafting The success of the above in  vivo studies has
procedures have been utilized to treat LSCD. They led to trials for the treatment of persistent corneal
involve either direct transplantation of limbal tis- epithelial defects in humans. Unpublished reports
sue or transplantation of in vitro expanded cells suggest that allogeneic CLEC-muc sheets cul-
on a variety of biological or synthetic carrier tured on contact lens-shaped plastic can lead to
materials. These techniques include conjunctival corneal epithelialization in 95% of patients with
limbal autograft (CLAU), living-related conjunc- chronic corneal ulcers but further details remain
tival allograft (Lr-CAL), keratolimbal allograft elusive at this stage and are keenly awaited.
(KLAL), autologous ex  vivo cultivated limbal The use of umbilical stem cell derived epithe-
epithelial transplantation (CLET), simple limbal lial cells would circumvent the problem of the
epithelial transplantation (SLET), cultivated oral limited amount of limbal tissue available for
mucosal epithelial transplantation (COMET) and transplantation, especially in cases of bilateral
transplantation of peripheral corneal cells. LSCD, but further study is still required.
Currently these techniques are challenging to One significant barrier that needs to be over-
perform, expensive, and have variable long term come to allow the clinical application of umbili-
success rates. cal stem cell reconstruction of the ocular surface
Cells derived from umbilical cord stem cells is the lack of cell surface marker exploitation
are capable of forming a stratified epithelial layer techniques. These can be used for enriching spe-
when seeded on artificial matrices such as colla- cific cell populations when compared to other
gen gels populated with fibroblasts as feeder stem cell populations (e.g. hematopoeitic stem
cells. Researchers have therefore investigated the cells). There is also the lack of a standardized
feasibility of umbilical cord-derived stem cells cultivation process that consists of only clinical-­
for ocular surface reconstruction. grade, xeno-free, products in good manufactur-
Researchers have transdifferentiated UCMSCs ing practice (GMP) grade conditions [15].
into corneal epithelial cells on a 3D human artifi- Further work is also required to determine the
cial anterior cornea model and confirmed the best composition of the substrate matrix used to
capacity of UCMSCs to form a stratified epithe- differentiate mesenchymal stem cells into epithe-
lial layer with favourable ultrastructural charac- lial cells. Current techniques which attempt to
teristics such as expression of epithelial markers recreate the microenvironmental niche of the host
(cytokeratin 3/12, plakoglobin), tight junction tissue by seeding the stem cells on a fibrin–aga-
protein zonula occludens 1 (ZO-1), and gap junc- rose stromal substitute containing human kerato-
tion protein Connexin 43. cytes, fail to allow full in vivo differentiation of
A novel cell type of the human umbilical an epithelial layer which suggests that a number
cord lining membrane that expresses mucin1, of different cell signals are required for complete
hence termed mucin-expressing cord lining epi- cell differentiation and tissue morphogenesis.
thelial cell (CLEC-muc) on human amniotic
membrane has also been used successfully in an
in vivo rabbit model of LSCD. Complete regen- 32.4.2 Stromal Regeneration
eration of the ocular surface was observed in 4
out of the 6 animals treated with CLEC-mucs The corneal stroma is a precisely organized avas-
with all controls demonstrating signs of total cular tissue which comprises approximately 90%
limbal stem cell failure at 10-weeks postopera- of corneal thickness and provides the mechanical
tively. In the treatment group, one eye devel- integrity and transparency of the cornea. It con-
oped mild superficial inflammation whilst tains 200–250 distinct lamellae, each layer
another developed severe n­eovascularization. arranged at approximately right angles relative to
Molecular analysis of transplanted eyes revealed fibers in adjacent lamellae.
a similar gene expression profile to intact nor- Keratocytes are the major cell type in the
mal cornea controls [14]. stroma and are embryologically derived from
32  Umbilical Cord Stem Cells in the Treatment of Corneal Diseases 481

mesenchymal cells lying between the corneal bition of inflammatory cell adhesion/invasion
epithelium and endothelium. Keratocytes are [19]. Another potential difficulty with intrastro-
capable of synthesizing collagen molecules, gly- mal injection of stem cells or transplantation of
cosaminoglycans, and matrix metalloproteases stroma-like tissue sheets in hazy or scarred cor-
(MMPs), and are crucial in maintaining local neas is the possibility of an inflammatory or scar-
homeostasis [16]. Keratocytes, unlike corneal ring response in an already compromised cornea.
epithelial cells, are difficult to propagate ex vivo Ultimately a great deal of additional research
as they have a low proliferative capacity and tend is required to fully elucidate the therapeutic
to irreversibly transform into fibroblasts. potential for curative therapy in animal models
Recent work has suggested that UCMSC before consideration of treatment of corneal stro-
transplantation may be a feasible alternative to mal disease in humans.
keratoplasty in treating congenital disorders of
the cornea secondary to keratocyte dysfunction
such as mucopolysaccharidoses. Intrastromal 32.4.3 Endothelial Regeneration
injection of UCMSC has been shown to be effec-
tive in treating Mucopolysaccharidosis VII (Sly Human corneal endothelial cells (HCECs) con-
syndrome) and Lumican deficiency in genetically sist of a 4 μm thick monolayer of hexagonal cells
modified mouse models [17]. Stem cell treatment lining the posterior aspect of the cornea. To per-
led to a reduction in corneal haze as assessed by form their essential pump function, a minimal
in  vivo confocal microscopy with an overall HCEC functional capacity is required, which
decrease in the corneal content of glycosamino- depends on both HCEC density and quality.
glycans. Transplanted stem cells survived the Endothelial cells do not undergo mitosis in vivo
host immune response and ultimately differenti- as they are arrested in the G1-phase of the cell
ated into resident stromal cells without the need cycle [20]. As a result, damaged endothelial cells
for immunosuppressive therapy. Of particular are rapidly replaced by enlargement of the sur-
note early treatment of affected mice led to sig- rounding cells and centripetal migration into the
nificantly improved corneal integrity, suggesting injured region. As the endothelial cell density
that prophylactic treatment may prevent the cor- falls below a threshold level of 400–800 cells/
neal haze seen in congenital corneal dystrophies. mm2, the cornea fails to maintain its transparency
UCMSCs have also been shown to be useful through loss of deturgescence and subsequent
in treating acquired disorders of the cornea such corneal edema [21].
as corneal opacification secondary to iatrogenic In recent years, investigators have evaluated
trauma. In a mouse model created by a central the feasibility of transforming UCMSCs into
keratectomy wound, UCMSC treatment resulted HCECs. Corneal endothelial cells are embryo-
in partial recovery of corneal transparency and logically differentiated from mesenchymal cells
thickness [18]. that migrate between the surface epithelium and
The use of umbilical stem cell derived stromal lens placode [22]. Researchers have used lens
substitutes is an intriguing therapeutic possibility, epithelial cell-conditioned medium (LECCM) to
especially given the longevity of mesenchymal mimic the effect of the lens during corneal endo-
stem cell derivatives in corneal tissue. However, thelial differentiation.
further work is required to refine techniques to Ex vivo models have shown that UCBMSCs
create a reproducible and homogenous cell popu- could be transformed into HCECs through
lation that can be used for clinical purposes. LECCM treatment and exhibit a close phenotypi-
Currently it remains uncertain how long UCMSC cal resemblance to HCECs (monolayer of
derived cells survive in the cornea and whether closely-packed, elongated cells), possess specific
they have potential to cause long term adverse cell markers such as N-cadherin and ZO-1 and
effects e.g. through modulation of the local “home in” and attach to damaged HCEC with
immune response that can occur through the inhi- great efficiency [23].
482 M. Ziaei et al.

An in vivo rabbit model has also demonstrated structure and to confirm the presence of a
the ability of expanded human umbilical cord confluent and adherent layer of cells and pro-
blood endothelial progenitor cells (UCBEPCS) gressive closure of the tight junctions between
to restore function and transparency to decom- adjacent cells [27]. The use of corneal endothe-
pensated corneas. After a central 9 mm surgically lial cell biomarkers such as ZO-1 and N-cadherin
created Descemetorhexis was performed, to identify corneal endothelial cells in  vitro is
UCBEPCS bound with immuno-magnetic CD34 problematic as they are also expressed in other
nanoparticles were injected into the anterior lineages, and a step toward a better characteriza-
chamber and a magnet was placed on the rabbits tion of endothelial cells might come from the
eyelid for 12 hours to attract the injected cells to identification of novel cell markers such as
the host stroma [24]. The injected cells were Wnt5a, S100A4, S100A6 and IER3 [28].
noted to migrate towards the magnet and the
treated corneas expressed aquaporin-1 (AQP1),
indicating intact pump function. At 4  months 32.5 Conclusion
postoperatively, the experimental group had little
to no corneal edema, and corneal thickness mea- Umbilical stem cells have a number of qualities
surements comparable to normal eyes, suggest- which make them attractive in the field of regen-
ing restoration of endothelial pump function. erative medicine. They are easy to access, com-
The aforementioned studies are fascinating as prise large numbers of newborn stem cells with
they open up the possibility of a cell-based treat- excellent proliferative capacity and cell extrac-
ment option for endothelial dysfunction, which is tion does not lead to any morbidity or mortality
the most common reason for corneal transplanta- to either mother or infant (nor do they carry the
tion in industrialized countries [25]. However, contentious ethical issues of embryo derived
there are many obstacles that need to be over- cells). This makes them an economically viable
come before such therapy can become a clinical cell source for translational clinical applications.
reality. The potential for developing cornea-like bioma-
Currently our knowledge regarding the molec- terials has been greatly advanced with the use
ular pathways that govern development and dif- and availability of stem cells, with preclinical
ferentiation of corneal endothelial cells is limited. studies suggesting umbilical stem cells to be a
In particular, identification is required of specific practical and realistic alternative to traditional
micro-environmental conditions that enable stem cell sources in the treatment of corneal dis-
inductivity of HCEC from umbilical stem cells to ease. The field is still in its infancy, and further
replace damaged or diseased cells. Although studies are required to determine the long-term
advances have been made with the discovery of outcome of these transplants.
the role of TGF-β2 in HCEC development, unsur- Specifically, additional research is required
prisingly the process appears to be complex with regarding the phenotypical and transformational
differentiation requiring the interaction of a num- quality control of expanded cells before applica-
ber of signals from other cell types from the ante- tion to the eye and the microbiological safety of
rior segment [26]. such an intervention. There are also issues regard-
It is also worth noting that whilst the ability ing the assessment of the functional capacity of
of injected HCEC to deturgesce the cornea has differentiated cells in vitro.
been borne out in the short term, the longevity Whilst there are many avenues that are yet to
of effect of such treatment remains undeter- be explored, the evidence suggests that human
mined. Additionally, studies to date have not umbilical stem cells are excellent candidates for
rigorously examined the characteristics and regenerative medicine and may well play an inte-
functions of transplanted endothelial cells. gral role in stem cell translational therapy.
Ideally, transmission electron microscopy Preliminary data support the potential use of
(TEM) studies are required to allow fine charac- human umbilical stem cell derived tissue in the
terization of the reconstructed endothelial ultra- treatment of human limbal stem cell deficiency,
32  Umbilical Cord Stem Cells in the Treatment of Corneal Diseases 483

keratocyte dysfunction and corneal endothelial dence of severe limbal stem cell deficiency in
Australia and New Zealand. Clin Exp Ophthalmol.
disease. Developments in this field offer hope of 2017;45(2):174–81.
future cellular therapy for the millions whose 14. Reza HM, Ng BY, Phan TT, Tan DT, Beuerman RW,
lives are affected by visually-impairing corneal Ang LP.  Characterization of a novel umbilical cord
disease, currently untreatable or only amenable lining cell with CD227 positivity and unique pat-
tern of P63 expression and function. Stem Cell Rev.
to corneal transplantation. 2011;7(3):624–38.
15. Wang J, Hao J, Bai D, Gu Q, Han W, Wang L, et al.
Compliance with Ethical Requirements Mohammed Generation of clinical-grade human induced pluripo-
Ziaei, Jie Zhang, Dipika V. Patel, and Charles N.  J. tent stem cells in Xeno-free conditions. Stem Cell Res
McGhee declare that they have no conflicts of interest. No Ther. 2015;6:223.
human or animal studies were carried out by the authors 16. DelMonte DW, Kim T. Anatomy and physiology of the
for this article. cornea. J Cataract Refract Surg. 2011;37(3):588–98.

17. Coulson-Thomas VJ, Caterson B, Kao
WW. Transplantation of human umbilical mesenchymal
stem cells cures the corneal defects of mucopolysac-
References charidosis VII mice. Stem Cells. 2013;31(10):2116–26.
18. Kao WW, Call M, Chang S-H, Birk DE.  Human

1. Oliva MS, Schottman T, Gulati M.  Turning the umbilical mesenchymal stem cells treat acquired and
tide of corneal blindness. Indian J Ophthalmol. congenital corneal opacity. Invest Ophthalmol Vis
2012;60(5):423–7. Sci. 2015;56:1304.
2. Weiss ML, Troyer DL.  Stem cells in the umbilical 19. Coulson-Thomas VJ, Gesteira TF, Hascall V, Kao

cord. Stem Cell Rev. 2006;2(2):155–62. W.  Umbilical cord mesenchymal stem cells sup-
3. Thomson JA, Itskovitz-Eldor J, Shapiro SS, Waknitz press host rejection: the role of the glycocalyx. J Biol
MA, Swiergiel JJ, Marshall VS, et al. Embryonic stem Chem. 2014;289(34):23465–81.
cell lines derived from human blastocysts. Science. 20. Joyce NC. Proliferative capacity of the corneal endo-
1998;282(5391):1145–7. thelium. Prog Retin Eye Res. 2003;22(3):359–89.
4. Takacs L, Toth E, Berta A, Vereb G. Stem cells of the 21. Joyce NC. Proliferative capacity of corneal endothe-
adult cornea: from cytometric markers to therapeutic lial cells. Exp Eye Res. 2012;95(1):16–23.
applications. Cytometry A. 2009;75(1):54–66. 22. Cvekl A, Tamm ER.  Anterior eye development and
5. Takahashi K, Tanabe K, Ohnuki M, Narita M, Ichisaka ocular mesenchyme: new insights from mouse models
T, Tomoda K, et al. Induction of pluripotent stem cells and human diseases. BioEssays. 2004;26(4):374–86.
from adult human fibroblasts by defined factors. Cell. 23. Joyce NC, Harris DL, Markov V, Zhang Z, Saitta
2007;131(5):861–72. B.  Potential of human umbilical cord blood mesen-
6. Arutyunyan I, Elchaninov A, Makarov A, Fatkhudinov chymal stem cells to heal damaged corneal endothe-
T.  Umbilical cord as prospective source for mes- lium. Mol Vis. 2012;18:547–64.
enchymal stem cell-based therapy. Stem Cells Int. 24. Shao C, Chen J, Chen P, Zhu M, Yao Q, Gu P, et al.
2016;2016:6901286. Targeted transplantation of human umbilical cord
7. Ziaei M, Zhang J, Patel DV, McGhee CNJ. Umbilical blood endothelial progenitor cells with immunomag-
cord stem cells in the treatment of corneal disease. netic nanoparticles to repair corneal endothelium
Surv Ophthalmol. 2017;62(6):803–15. defect. Stem Cells Dev. 2015;24(6):756–67.
8. Chang YJ, Tseng CP, Hsu LF, Hsieh TB, Hwang 25. Park CY, Lee JK, Gore PK, Lim CY, Chuck

SM.  Characterization of two populations of mesen- RS.  Keratoplasty in the United States: a 10-year
chymal progenitor cells in umbilical cord blood. Cell review from 2005 through 2014. Ophthalmology.
Biol Int. 2006;30(6):495–9. 2015;122(12):2432–42.
9. Chamberlain G, Fox J, Ashton B, Middleton J. Concise 26. Ittner LM, Wurdak H, Schwerdtfeger K, Kunz T, Ille
review: mesenchymal stem cells: their phenotype, dif- F, Leveen P, et al. Compound developmental eye dis-
ferentiation capacity, immunological features, and orders following inactivation of TGFbeta signaling in
potential for homing. Stem Cells. 2007;25(11):2739–49. neural-crest stem cells. J Biol. 2005;4(3):11.
10. Oie Y, Nishida K. Regenerative medicine for the cor- 27. Proulx S, Brunette I.  Methods being developed for
nea. Biomed Res Int. 2013;2013:428247. preparation, delivery and transplantation of a tissue-­
11. Dua HS, Forrester JV.  The corneoscleral limbus
engineered corneal endothelium. Exp Eye Res.
in human corneal epithelial wound healing. Am J 2012;95(1):68–75.
Ophthalmol. 1990;110(6):646–56. 28. Chen Y, Huang K, Nakatsu MN, Xue Z, Deng SX,
12. Shortt AJ, Tuft SJ, Daniels JT. Corneal stem cells in Fan G.  Identification of novel molecular markers
the eye clinic. Br Med Bull. 2011;100:209–25. through transcriptomic analysis in human fetal and
13. Bobba S, Di Girolamo N, Mills R, Daniell M,
adult corneal endothelial cells. Hum Mol Genet.
Chan E, Harkin DG, et  al. Nature and inci- 2013;22(7):1271–9.
Dysfunctional Corneal
Endothelium: Delivery of Cell
33
Therapy

Stephen Wahlig, Gary Swee-Lim Peh,
Matthew Lovatt, and Jodhbir S. Mehta

33.1 Introduction an estimated 12.7 million individuals globally are


awaiting corneal transplantation, less than 2%
Corneal endothelial transplantation has seen a receive their needed treatment [8].
significant series of advances over the past sev- In the face of this shortage, cell based therapies
eral decades. Instead of full thickness penetrating provide an alternative source of tissue for treat-
keratoplasty (PK), partial thickness techniques ment of endothelial dysfunction. While originally
like Descemet’s stripping automated endothelial thought to be non-proliferative, human corneal
keratoplasty (DSAEK) and Descemet’s mem- endothelial cells (HCEnCs) have been success-
brane endothelial keratoplasty (DMEK) are now fully propagated in vitro [9, 10]. These develop-
preferred for treating endothelial pathology. ments are expanded upon in greater detail within
These newer methods have improved the recov- [CITE CELL CULTURE CHAPTER XXX]. In
ery speed and visual outcomes for patients with addition to propagation of these primary corneal
endothelial dysfunction [1]. However, graft rejec- endothelial cells (CEnCs), a robust method of
tion is still a concern: graft viability is only 83.3% delivering HCEnCs to the dysfunctional endothe-
at 3  years [2] and 76% at 5  years [3] for non-­ lium is necessary for clinical applications. While
Fuch’s dystrophy indications. While DMEK is some have attempted to mimic a donor corneal
thought to have reduced graft rejection compared graft using scaffolds seeded with HCEnCs, novel
to PK, some centers have reported rejection rates approaches like cellular injection also show prom-
up to 5.1% after 1  year [4], with early steroid ise as regenerative therapies (Fig. 33.1) [11, 12].
tapering thought to contribute to these elevated This chapter will initially review the pre-clinical
rates [5]. In addition, demand for graft tissue is animal models used to evaluate endothelial thera-
rapidly outstripping our limited supply. The num- pies, then proceed to discuss the principles and
ber of transplants in the United States has efficacy of several vehicles for delivery of propa-
increased over 60% since 2005 [6, 7]. Although gated endothelial cells to a dysfunctional cornea.

S. Wahlig · G. S.-L. Peh · M. Lovatt 33.2 Animal Models


Tissue Engineering and Stem Cell Group, The of Endothelial Dysfunction
Academia, Singapore Eye Research Institute,
Singapore, Singapore
Prior to initiation of human clinical trials,
J. S. Mehta (*)
Singapore National Eye Centre, Department of promising therapies must demonstrate safety
Cornea and External Disease, Singapore, Singapore and efficacy in  vivo. To create animal models

© Springer Nature Switzerland AG 2019 485


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2_33
486 S. Wahlig et al.

a
Endothelial dysfunction

Healthy cornea Edematous cornea

b
HCEnC isolation HCEnC expansion
Donor tissue

c Removal of
DM and endothelium
Detached monolayer
1
Critical Endothelial
temperature keratoplasty

Removal of
DM and endothelium
2

HCEnC Endothelial
seeding keratoplasty

3
Legend Removal of endothelium
Cornea
Descemet’s membrane
HCEnC Cell
Endothelial cell (CEnC)
dissociation injection
Dysfunctional CEnC
Cell scaffold

Fig. 33.1  Overview schematic of endothelial cell ther- in vitro. (c) Expanded HCEnCs can potentially be deliv-
apy. (a) The dynamics of corneal hydration are altered by ered via (1) direct implantation of the cultured monolayer,
endothelial dysfunction or cell loss, resulting in swelling (2) seeding cells onto an engineered scaffold to be
and edema. (b) To generate cells for therapeutic applica- implanted via a DSAEK-like technique, or (3) injected
tions, HCEnCs are harvested from donor corneas via DM directly into the anterior chamber
stripping and digestion and subsequently expanded

for this in  vivo evaluation, dysfunction of the and producing an edematous cornea [18]. The
native endothelium must first be induced. This most commonly used animal model for corneal
can be accomplished by mechanical scraping of endothelial cell therapy is the rabbit [20–25].
the endothelium [13–15], cryoinjury [16], Murine [16] and rat [26] models have also been
phacoemulsification [17], Nd:YAG laser abla- described, but the small size of these animals’
tion [18], or induction of bullous keratopathy eyes causes difficulty with surgical manipula-
with injection of benzalkonium chloride (BAK) tion and therapeutic delivery. Rabbit eyes are
[19]. While these procedures carry some risk of similar in size to humans, which makes them
inducing side effects like anterior chamber ideal models for ocular therapy [27]. However,
inflammation or glaucoma, they are all effective the rabbit cornea possesses CEnCs)that retain
at disrupting normal endothelial pump function extensive proliferative capacity in  vivo, such
33  Dysfunctional Corneal Endothelium: Delivery of Cell Therapy 487

that even damage to 50% of the central cornea method does not require additional carrier mate-
can be repaired within 10  days [28]. As such, rial to be transplanted, thus reducing the risk of
several in vivo trials are now using primate [12, immunogenic complications. However, the
24, 29] and feline [30] models, which like ultra-­thin nature of the monolayer causes
human eyes lack a proliferative endothelium extreme difficulty with manipulation and han-
[28, 31]. While rabbits are more cost effective dling of the implant, potentially damaging the
to purchase and maintain than primates or delicate endothelial cells and limiting the practi-
felines, trials in rabbit models must include cality of its use. As compensation for the mono-
careful analysis of negative controls to ensure layer fragility, gelatin carriers have been used to
any observed endothelial regeneration is not the provide stability during surgical manipulation
product of native CEnC proliferation. [37]. This carrier is implanted with the mono-
layer and provides mechanical support from
within the anterior chamber until it eventually
33.3 Direct Monolayer degrades. The pores in the carrier facilitate nutri-
Implantation ent transport while the gelatin is in place [37–
39]. However, the need for a gelatin carrier
HCEnCs can be propagated in vitro as an adher- circumvents the benefits of direct monolayer
ent culture, producing a polygonal monolayer transplantation compared to seeding HCEnCs
morphologically similar to the in  vivo corneal directly onto an engineered carrier scaffold.
endothelium. One delivery strategy involves har- However, we feel that an endothelial monolayer
vesting this cultured cell monolayer and associ- is too fragile and its transplantation too techni-
ated extracellular matrix (ECM) for direct cally complex for widespread clinical use.
transplantation without need for a scaffold or fur- Instead, HCEnCs seeded on a tuneable scaffold
ther modification. The first implementation of is the most feasible approach.
this approach utilized culture surfaces coated
with the thermoresponsive polymer poly(N-­
isopropylacrylamide) (PNIPAAm), that detach 33.4 Biological Cell Carriers
the HCEnC monolayer when cooled below a
critical temperature [32]. This method was subse- Cultured CEnCs can be seeded onto a biocom-
quently used to transplant functional HCEnC patible scaffold, generating a tissue engineered
monolayers into a rabbit model, reducing corneal construct prior to implantation into the anterior
edema [36]). Use of thermoresponsive polymer chamber. A variety of scaffolds tested in animal
culture surfaces requires a careful balance models are detailed in Table 33.1. Implantation
between HCEnC attachment for monolayer for- of the seeded scaffold is nearly identical to the
mation and effective stimulated detachment. existing Descemet Stripping Automated
Improvements to this approach have come Endothelial Keratoplasty (DSAEK), greatly sim-
through development of superior polymers, plifying clinical integration of these constructs.
including poly(vinyl methyl ether) (PVME) and The HCEnC density on the scaffold can also be
N-isopropylacrylamide-co-glycidylmethacrylate precisely controlled to optimize graft function
(NGMA) [34, 35]. Both PVME and NGMA have while minimizing the quantity of required cells.
tuneable physical properties (e.g. film thickness, Of note, such a scaffold needs to satisfy certain
stiffness) that in turn modulate HCEnC attach- criteria such as biocompatibility, optically trans-
ment and ease of detachment [34, 35]. parency, easy handling for surgical manipula-
One of the advantages of this approach is the tion, and demonstration of similar mechanical
ultra-thin nature of the monolayer, since the abil- properties to the native cornea. Biological scaf-
ity of corneal endothelium to maintain deturges- folds can mimic the native endothelial environ-
cence (stromal dehydration) is reduced with ment, which should promote superior graft
increasing membrane thickness [36]. This integration.
488 S. Wahlig et al.

Table 33.1  Animal studies of carrier implantation


Animal Surgical Follow-up
Carrier material CEnC source model technique duration Reference
Biological
100-μm-thick Human Stroma/ Human Rabbit DSEK 4 weeks Peh et al. [55]
DM Lenticules
Human Denuded Amninotic Human cell line Feline PK 3 months Fan et al. [41]
Membrane (C3B)
Human Stromal Lenticules Human Rabbit DSEK 4 weeks Honda et al. [11]
Human Denuded Amniotic Human Rabbit PK 1 week Ishino et al. [42]
Membrane
Bovine DM Rabbit Rabbit PK 4 months Lange et al. [90]
Synthetic
10-μm-thick Silk Fibroin Rabbit Rabbit DSEK 6 weeks Vazquez et al. [91]
Membrane
6–8-μm-thick Aloe Vera-Silk Rabbit Rabbit DSEK 4 weeks Kim do et al. [20]
Fibroin Membrane
200-μm-thick Gelatin None Rabbit DSEK 8 weeks Niu et al. [23]
Membrane
20-μm-thick Spherically Curved Monkey Monkey DSEK 4 weeks Kimoto et al. [29]
Gelatin Membrane
50-μm-thick PEG-based None Sheep N/Aa 4 weeks Ozcelik et al. [69]
Hydrogel
700-μm-thick Gelatin Disc None Rabbit N/Aa 3 days Lai et al. [39]
Collagen I Membrane Monkey Monkey DSEK 2 years Koizumi et al. [14]
700–800-μm-thick Gelatin Disc Human Rabbit DSEK 3 months Hsiue et al. [32]
40–50-μm-thick Collagen I Human Rabbit DSEK 1 month Mimura et al. [92]
Membrane
100-μm-thick Hydrogel Lens Rabbit Rabbit PK 2 months Mohay et al. [57]
100-μm-thick Hydrogel Lens Feline Feline PK 2 months Mohay et al. [57]
1–5-μm-thick Gelatin Rabbit Rabbit PK 7 months McCulley et al. [77]
Membrane
DSEK technique implies use of a partial-thickness graft, while PK technique implies use of a full-thickness graft. CEnC
corneal endothelial cell, DSEK Descemet’s Stripping Endothelial Keratoplasty, DM Descemet’s membrane, PEG
poly(ethylene glycol), PK Penetrating Keratoplasty
a
Scaffolds were implanted into the anterior chamber without alteration of host endothelium

One of the earliest biological scaffolds utilized buttons and successful transplanted via PK into
was Descemet’s membrane. As the natural sub- both rabbit and feline model [41, 42]. However,
strate for endothelial cells, DM is an intuitive cell the quality of amniotic membrane may not be
carrier choice. Rabbit CEnCs were seeded onto consistent, and it is not fully transparent. As such,
bovine DM, which was sutured onto a rabbit cor- it is not likely to be a carrier of choice for a stan-
neal button and implanted using a PK approach dardized tissue engineering protocol.
into a rabbit endothelial deficiency model with Another intuitive option is the use of pre-­
resolution of corneal edema [40]. Since animal existing corneal tissue. Corneal tissue is both
DM and xeno-transplantation introduces potential transparent and highly compatible with endothe-
immunogenic complications, and human DM is lial cells, although donor cells should be ­eliminated
already in limited supply, this is not a realistic car- to reduce immunogenic complications [25, 43].
rier for clinical practice. Another option, amniotic Decellularized bovine and porcine stroma has
membrane, is already used for a variety of sys- served as scaffolds for CEnCs with formation of
temic therapies [40]. Human amniotic membrane regular polygonal monolayers that demonstrate
has functioned as a scaffold for human endothe- pump functionality in vitro [44–48]. Reassuringly
lial cells, with the constructs sutured to corneal for xenotransplantation, porcine constructs did not
33  Dysfunctional Corneal Endothelium: Delivery of Cell Therapy 489

induce an immune response in a rabbit model [25]. generally easier to mass produce and can be mod-
One challenge with these animal corneas is unde- ified to acquire desired properties for implanta-
sirable alteration of tissue properties after decel- tion. The first synthetic scaffolds were simple
lularization. A variety of protocols, including biomaterials like gelatin and soft hydrogel con-
freeze/thaw cycles, high pressure exposure, and tact lenses [56, 57]. While capable of adhering to
chemical detergents, have been used for removal endothelial cells, these early attempts were com-
of host cells [45, 49, 50]. Some of these techniques plicated by graft rejection in animal models [57].
incompletely remove host cellular material, while To reduce the incidence of these immunologic
others alter the stromal architecture. The com- complications, synthetic carriers can be formed
monly used ionic detergent SDS increases corneal from bioactive substances. Hyaluronic acid, an
swelling and diminishes transparency, although ECM protein with desirable biocompatibility, has
this can be reversed through immersion in 100% been processed to form a transparent hydrogel
glycerol [51]. In addition to the need for a stan- discs for HCEnC transplantation [58], although
dardized decellularization protocol, transplanta- some reports indicate that endothelial cells do not
tion of animal tissue may raise regulatory concern. adhere well to this carrier [59]. Collagen is a par-
Despite the use of implanted decellularized por- ticularly attractive substrate, as it is a major com-
cine tissue for approved cardiac valvular indica- ponent of Descemet’s membrane [60]. Monkey
tions [52, 53], this will be an obstacle to CECs cultured on collagen I sheets have been
implementation in humans. successfully transplanted into a monkey endothe-
Use of human corneal tissue circumvents lial deficiency model, demonstrating preserved
many of these concerns around xenotransplanta- corneal clarity at 2  years [13, 14]. For further
tion. A construct generated from non-­improvement in graft transparency, collagen III
decellularized human stromal discs and HCEnCs could be substituted for collagen I [61]. These
has been implanted into a rabbit model with min- scaffolds appear highly biocompatible in humans,
imal post-operative inflammation [11]. This tech- as demonstrated by clinical trial results of a col-
nique only harvested 2–3 stromal discs per donor lagen III anterior corneal implant which reported
cornea, indicating that tissue shortages will still no rejection episodes over 4 years [62]. Collagen
be an obstacle with the use of human stroma [11]. sheets can be easily and reproducibly manufac-
More recently, He et  al. published a method of tured using techniques like plastic compression,
producing thin decellularized human stromal acquiring the mechanical properties necessary
lamellae with a femtosecond laser [54]. Using the for surgical handling [63]. However, while highly
immortalized cell line HCEC-B4G12, these stro- biocompatible and easy to produce, pure collagen
mal lamellae were able to support an endothelial scaffolds have considerably reduced mechanical
cell density > 6000 cells/mm2 with minimal cell strength compared to native cornea [64].
loss upon surgical handling. These thin lamellae Another method to promote graft integration
are also compatible with primary human cells, as is the use of bioactive surface coatings. Silk
transplantation of HCEnC-seeded grafts have fibroin, a transparent biomaterial with high ten-
successfully reversed corneal edema in a rabbit sile strength, demonstrates significantly improved
model (Fig. 33.2), although the use of native DM endothelial attachment and morphology after
in these constructs restricts the number of obtain- coating with collagen or FNC [65, 66]. This idea
able grafts per donor cornea [55]. of combining properties from multiple biomateri-
als has yielded an enormous variety of blended
polymer substrates. These blends combine bioac-
33.5 Synthetic Cell Carriers tive materials (e.g. collagen, chitosan) with
mechanically strong substances (e.g. phosphoryl-
A variety of synthetic carriers have been pro- choline, polycaprolactone) [20, 67–71]. By
posed as endothelial scaffolds. Although syn- adjusting the relative levels of each material in a
thetic materials do not possess the innate blend, scaffold properties can be tuned for opti-
biocompatibility of biological scaffolds, they are mal function [20, 72].
490 S. Wahlig et al.

a b

c d

e f

Fig. 33.2  Descemet’s stripping endothelial keratoplasty into a Tan EndoGlide (b) Scleral tunnel is created (d) Host
(DSEK) in a primate model using a biologic graft (denuded DM and endothelium is peeled and removed (e) Graft is
Descemet’s membrane) seeded with HCEnCs. (a) inserted into the anterior chamber (f) Unfolded graft is in
Transparent graft prior to implantation (b) Graft is loaded place on the posterior cornea, supported by an air bubble

In addition to the materials used to produce fibrils that form corneal stroma, an electrospin-
the scaffold, reproduction of the corneal microar- ning technique creating a poly (glycerol
chitecture has helped synthetic scaffolds mimic sebacate)-poly (ε-caprolactone) (PGS-PCL)
the transparency and biomechanical properties of nanofibrous scaffold with similar physical and
native cornea. Instead of the aligned collagen chemical properties to native cornea [73, 74].
33  Dysfunctional Corneal Endothelium: Delivery of Cell Therapy 491

This PGS-PCL nanofibrous scaffold also sup- chemical and physical cross-linking to signifi-
ports HCEnC monolayer formation and does not cantly improve its mechanical properties [76].
alter human granulocyte activation, encouraging The development of synthetic scaffolds has given
evidence of its biocompatibility [75]. rise to tremendously customizable materials,
An array of production modifications can also which show promising compatibility with
enhance the viability of synthetic scaffolds. Use HCEnCs and the mechanical requirements of
of curved molds produces a 3D scaffold shape transplantation. Of note, while many of these
similar to the cornea, reducing complications of polymers have demonstrated encouraging repro-
graft folding and non-adherence that can arise duction of corneal properties, head to head com-
with flat scaffolds [29]. Scaffolds can be func- parisons between synthetic scaffolds is still
tionalized by integrating heparin, which binds needed to clarify the optimal carrier material.
growth factors that are slowly released to pro-
mote HCEnC growth [23]. Another useful modi-
fication is the use of topographical patterning. 33.6 Cellular Injection
This approach mimics the biophysical interaction
between endothelial cells and Descemet’s Direct HCEnC injection of single cells into the
Membrane. A gelatin methacrylate (GelMA+) anterior chamber is one of the original approaches
based hydrogel was patterned with 1  μm and tested for endothelial cell therapy; a collection of
250 nm cylindrical wells and pillars, which dem- animal trials are listed in Table 33.2. The simplic-
onstrated improved endothelial homogeneity and ity of HCEnC injection is its strongest feature:
ZO-1 expression (Fig. 33.3) [76]. In addition, the corneal regeneration therapy could be imple-
GelMA+ polymer was manipulated through mented globally, even in areas without access to
highly skilled corneal surgeons (Fig.  33.4). A
1980 publication by McCulley et  al. describes
injection of cultured rabbit CECs into a rabbit
a model, intending to produce a monolayer on the
posterior cornea for replacement of the removed
endothelium [77]. However, the endothelium did
not regenerate, and ectopic CEC deposits were
visible on the iris and lens in addition to the cor-
nea. This initial failure highlights a major chal-
lenge with the direct injection approach:
localizing cells to the posterior cornea, rather
than other anterior chamber structures. Ectopic
cell localization diminishes efficient endothelial
b monolayer formation, necessitating a higher
number of injected cells. It is plausible that unat-
tached cells would follow aqueous flow into the
trabecular meshwork (TM). Ectopic cellular
deposits in the TM could produce an obstruction
and increase intraocular pressure (IOP), poten-
tially causing secondary glaucoma [21, 78].
One simple approach to cell localization is
having the recipient assume a face down position,
so that gravity brings the injected cells down onto
the DM [79]. Studies in a rabbit model of bullous
keratopathy have been successful in using cellu-
Fig. 33.3  Scanning electron microscopy (SEM) images
of GelMa scaffold with 1  μm topographical patterning, lar injection, followed by 6  h in the face down
depicted in oblique (a) and overhead (b) views position, to generate a new endothelial layer and
492 S. Wahlig et al.

Table 33.2  Animal studies of cell injection


CEnC Animal Cell ROCK Localization Follow-up
source model number inhibitor technique duration Reference
Monkey Monkey 5 × 105 Yes Face-down 1 year Okumura et al.
Human Monkey 5 × 105 Yes Face-down 3 months [12]
Rabbit Rabbit 5 × 105 Yes Face-down 2 weeks Okumura et al.
Human Rabbit 1 × 106 Yes Face-down 48 h [87]
Feline Feline 2 × 105 Yes Face-down 1 month Bostan et al. [30]
1 × 106
Rabbit Rabbit 2 × 105 Yes Face-down 2 weeks Okumura et al.
Monkey Monkey 2 × 105 Yes Face-down 3 months [24]
Human Rabbit 150 No Face-down 1 month Mimura et al.
spheres [80]
Human Rabbit 1 × 107 No Face-down 1 month Mimura et al.
150 [79]
spheres
Rabbit Rabbit 5 × 105 No Magnet 1 year Mimura et al.
[22]
Rabbit Rabbit 5 × 105 No Magnet 2 months Mimura et al.
[21]
CEnC corneal endothelial cell

reduce corneal edema [80]. However, even slight net [21], whether this is appropriate for humans
movements can interfere with HCEnC localiza- is unknown. Additionally, the duration of mag-
tion, so this is not a perfect method of preventing netic field application needs clarification, as the
ectopic endothelial deposition. Another potential 2–7  days used in an ex  vivo human eye model
technique for localization integrates magnetic [78] could cause difficulties with patient compli-
particles into HCEnCs, and application of an ance if the magnet is bulky or uncomfortable. A
external magnetic field attracts the cells towards variety of alternative techniques have been pro-
the DM.  Initial studies successfully used ferro- posed for future consideration, including suspen-
magnetic iron particles loaded in HCEnCs in a sion of HCEnCs in a viscous material prior to
rabbit model, but this approach carries some con- injection to limit uncontrolled dispersal of free-­
cern of ocular toxicity [21, 22]. Ferromagnetic floating cells. Yet to date the only two viable
material can retain magnetic properties even after methods are the gravity and magnetic nanoparti-
the external field is removed, potentially causing cle techniques, both of which have demonstrated
self-aggregation and poor endothelial formation success with in  vitro and in  vivo experiments.
[81, 82]. In response, superparamagnetic micro- However, the gravity approach is significantly
spheres and nanoparticles have been developed simpler and more practical, with no need to inte-
which do not demonstrate this ‘memory’ of the grate magnetic particles into delicate HCEnCs or
external magnetic field [78, 83]. These nanopar- use an external magnet. The long term fate and
ticles also have minimal impact on HCEnC func- safety implications of these nanoparticles after
tion or viability after uptake [84]. However, more injection is also not well understood. Still, local-
work is needed to determine the optimal device ization of HCEnCs must be further be optimized
for magnetic field generation, as both commer- to minimize cell seeding numbers and potential
cially available permanent magnets [21, 22] and side effects of the injection approach.
custom electromagnets [83] have been used in In addition to localizing cells to the posterior
existing animal trials. The size and shape of the cornea, the HCEnCs must attach to the corneal
magnet must also be defined; although rabbit surface to form a functional monolayer. Inhibition
experiments have been successful using a small of Rho-associated kinase (ROCK) has been
(10 mm diameter × 5 mm length cylinder) mag- hypothesized to modulate HCEnC cytoskeleton
33  Dysfunctional Corneal Endothelium: Delivery of Cell Therapy 493

a b

c d

Fig. 33.4  Cell injection using a rabbit endothelial dys- ough removal of host endothelium (d) HCEnCs are
function model. (a) Clump of endothelial cells (stained injected into the anterior chamber. The rabbits are then
blue) kept in a tube prior to injection (b) Rabbit endothe- kept in a prone position for 3–4 h to facilitate cell adher-
lium is scraped off (c) Trypan blue is used to ensure thor- ence to host DM

function, and is noted to improve endothelial cell ROCK inhibitor [30]. In this model, increasing
adhesion to substrates, as well as proliferation, the number of injected CEnCs led to formation of
migration, and wound healing [85]. Co-injection abnormal multilayered endothelium that
of monkey CEnCs with the prototypic ROCK expressed non-endothelial markers collagen I
inhibitor Y-27632  in a primate model reduced and alpha smooth muscle actin (α-SMA), sug-
corneal thickness significantly more than injec- gesting that injection of excess cells may reduce
tion of CEnCs alone [12, 24]. However, conflict- treatment efficacy [30]. Further studies with
ing data has emerged from trials in a feline model, HCEnC and ROCK inhibitor injection are neces-
which suggested that Y-27632 alone may be more sary to clarify the relative contribution of each
beneficial than co-administration of CEnCs and therapy to endothelial regeneration. In 2013
494 S. Wahlig et al.

Okumura et al. were approved to begin a human given the monolayer fragility. In response, a vari-
clinical trial of HCEnC injection therapy with ety of biological and synthetic substrates have
ROCK inhibition. been created to maximize ease of handling and
As clinical trials of HCEnC injection begin, post-operative integration within the host cornea.
there are still a number of safety and efficacy One of the most promising delivery approaches is
questions that must be answered. Even with direct injection of expanded HCEnCs into the
ROCK inhibition, only 40–50% of injected cells anterior chamber. This injection technique can be
were adherent to the cornea [12]. Not only does easily utilized by ophthalmologists globally, even
this hinder endothelial regeneration by reducing those without specialized training in corneal sur-
the effective cell seeding density, free floating gery. While we are still awaiting results from the
cells can potentially pass into venous circulation first human trial for endothelial cell therapy, this
via aqueous flow and cause systemic side effects emerging technology holds significant promise
(eg immune rejection, pulmonary embolism). for future clinical applications.
Use of ROCK inhibitor-based therapy in the US
will also require an FDA-approved pharmaceuti- Compliance Statements J.  S. Mehta, M.  Lovatt,
cal; netarsudil is currently awaiting FDA approval G. Peh, and S. Wahlig declare that they have no conflict
of interest. All procedures performed by the authors
while ripasudil has been approved in Japan since followed and were in accordance with the ethical stan-
2014 [86]. Additionally, though multiple in vivo dards of the responsible committee on human experi-
studies have shown no increase in IOP [21, 22, mentation (institutional and national) and with the
24] despite ectopic localization of non-adherent Helsinki Declaration of 1975, as revised in 2000.
Informed consent was obtained from all patients for
HCEnCs [30], long term monitoring for intraocu- being included in the study. All institutional and
lar hypertension will also be necessary to confirm national guidelines for the care and use of laboratory
safety in humans. Also, effectiveness of the animals were followed.
injected HCEnCs relies on an intact host DM for Conflict of Interest: No conflicting relationship exists
for any author.
endothelial cell adherence [24, 87]. Fuch’s endo- Financial disclosure: No financial disclosures.
thelial dystrophy, the most common indication
for endothelial transplantation in the United
States [8], produces a dysfunctional DM marked
by excressences known as guttata [88]. Although References
removal of the pathological DM is necessary for
1. Tan DT, Dart JK, Holland EJ, et al. Corneal transplan-
visual recovery in Fuchs dystrophy, it may hinder tation. Lancet. 2012;379(9827):1749–61.
the efficacy of an injection based cell therapy 2. Ang M, Mehta JS, Lim F, et al. Endothelial cell loss
[89]. Given these limitations, there may be a role and graft survival after Descemet’s stripping auto-
mated endothelial keratoplasty and penetrating kera-
for both scaffold and injection based delivery
toplasty. Ophthalmology. 2012;119(11):2239–44.
methods for regenerative endothelial therapies. 3. Price MO, Fairchild KM, Price DA, et al. Descemet’s
stripping endothelial keratoplasty five-year graft
survival and endothelial cell loss. Ophthalmology.
2011;118(4):725–9.
33.7 Conclusion 4. Guerra FP, Anshu A, Price MO, et  al. Descemet’s
membrane endothelial keratoplasty: prospective study
Current supply of graft tissue for corneal trans- of 1-year visual outcomes, graft survival, and endothe-
plantation is insufficient to address endothelial lial cell loss. Ophthalmology. 2011;118(12):2368–73.
5. Ang M, Wilkins MR, Mehta JS, et  al. Descemet
dysfunction globally. Major advances in HCEnC
membrane endothelial keratoplasty. Br J Ophthalmol.
expansion in vitro have opened the door for cel- 2016b;100(1):15–21.
lular therapies to meet this need [CITE CELL 6. Eye Bank Association of America. 2010; 2009 Eye
CULTURE CHAPTER 8]. In order to apply this Banking Statistical Report.
7. Eye Bank Association of America. 2016; 2015 Eye
technology clinically, a robust method for delivery
Banking Statistical Report.
of therapies to the posterior corneal surface must 8. Gain P, Jullienne R, He Z, et  al. Global survey of
be developed. Application of endothelial mono- corneal transplantation and eye banking. JAMA
layers is possible, although technically difficult Ophthalmol. 2016;134(2):167–73.
33  Dysfunctional Corneal Endothelium: Delivery of Cell Therapy 495

9. Baum JL, Niedra R, Davis C, et al. Mass culture of 25. Zhou Y, Wu Z, Ge J, et al. Development and charac-
human corneal endothelial cells. Arch Ophthalmol. terization of acellular porcine corneal matrix using
1979;97(6):1136–40. sodium dodecylsulfate. Cornea. 2011;30(1):73–82.
10. Peh GS, Beuerman RW, Colman A, et al. Human cor- 26. Tuft SJ, Williams KA, Coster DJ.  Endothelial

neal endothelial cell expansion for corneal endothe- repair in the rat cornea. Invest Ophthalmol Vis Sci.
lium transplantation: an overview. Transplantation. 1986;27(8):1199–204.
2011;91(8):811–9. 27. Werner L, Chew J, Mamalis N. Experimental evalua-
11. Honda N, Mimura T, Usui T, et  al. Descemet strip- tion of ophthalmic devices and solutions using rabbit
ping automated endothelial keratoplasty using cul- models. Vet Ophthalmol. 2006;9(5):281–91.
tured corneal endothelial cells in a rabbit model. Arch 28. Van Horn DL, Sendele DD, Seideman S, et  al.

Ophthalmol. 2009;127(10):1321–6. Regenerative capacity of the corneal endothe-
12. Okumura N, Sakamoto Y, Fujii K, et al. Rho kinase lium in rabbit and cat. Invest Ophthalmol Vis Sci.
inhibitor enables cell-based therapy for corneal 1977;16(7):597–613.
endothelial dysfunction. Sci Rep. 2016b;6:26113. 29. Kimoto M, Shima N, Yamaguchi M, et al. Development
13. Koizumi N, Sakamoto Y, Okumura N, et  al.
of a bioengineered corneal endothelial cell sheet to
Cultivated corneal endothelial cell sheet transplanta- fit the corneal curvature. Invest Ophthalmol Vis Sci.
tion in a primate model. Invest Ophthalmol Vis Sci. 2014;55(4):2337–43.
2007;48(10):4519–26. 30. Bostan C, Theriault M, Forget KJ, et  al. In vivo

14. Koizumi N, Sakamoto Y, Okumura N, et al. Cultivated functionality of a corneal endothelium transplanted
corneal endothelial transplantation in a primate: pos- by cell-injection therapy in a feline model. Invest
sible future clinical application in corneal endothe- Ophthalmol Vis Sci. 2016;57(4):1620–34.
lial regenerative medicine. Cornea. 2008;27(Suppl 31. Van Horn DL, Hyndiuk RA. Endothelial wound repair
1):S48–55. in primate cornea. Exp Eye Res. 1975;21(2):113–24.
15. Maurice DM, Giardini AA.  Swelling of the cornea 32. Hsiue GH, Lai JY, Chen KH, et al. A novel strategy
in vivo after the destruction of its limiting layers. Br J for corneal endothelial reconstruction with a bioengi-
Ophthalmol. 1951;35(12):791–7. neered cell sheet. Transplantation. 2006;81(3):473–6.
16. Han SB, Ang H, Balehosur D, et al. A mouse model of 33. Sumide T, Nishida K, Yamato M, et  al. Functional
corneal endothelial decompensation using cryoinjury. human corneal endothelial cell sheets harvested from
Mol Vis. 2013;19:1222–30. temperature-responsive culture surfaces. FASEB J.
17. Wu M, Hu ZL, Sun XM, et al. In vivo confocal micro- 2006;20(2):392–4.
scopic evaluation of corneal endothelial dysfunction 34.
Madathil BK, Kumar PR, Kumary TV.
induced by phacoemulcification in rhesus monkey N-isopropylacrylamide-co-glycidylmethacrylate
models. Zhongguo Yi Xue Ke Xue Yuan Xue Bao. as a thermoresponsive substrate for corneal
2016;38(1):42–8. endothelial cell sheet engineering. Biomed Res In.
18. Zhang W, Hu Y, Lu L, et al. Rabbit model of corneal 2014;2014:450672.
endothelial injury established using the Nd: YAG 35. Teichmann J, Valtink M, Gramm S, et al. Human cor-
laser. Cornea. 2017;36(10):1274–81. neal endothelial cell sheets for transplantation: thermo-
19. Ang M, Konstantopoulos A, Goh G, et al. Evaluation responsive cell culture carriers to meet cell-specific
of a micro-optical coherence tomography for the requirements. Acta Biomater. 2013;9(2):5031–9.
corneal endothelium in an animal model. Sci Rep. 36. Tsaousis KT, Kopsachilis N, Tsinopoulos IT, et  al.
2016a;6:29769. In vitro study of the deturgescence ability of cul-
20. Kim do K, Sim BR, Khang G.  Nature-derived Aloe tivated human corneal endothelial cells. Cornea.
Vera gel blended silk fibroin film scaffolds for cornea 2016;35(5):669–72.
endothelial cell regeneration and transplantation. ACS 37. Lai JY, Lu PL, Chen KH, et  al. Effect of charge
Appl Mater Interfaces. 2016;8(24):15160–8. and molecular weight on the functionality of gela-
21. Mimura T, Shimomura N, Usui T, et  al. Magnetic tin carriers for corneal endothelial cell therapy.
attraction of iron-endocytosed corneal endothe- Biomacromolecules. 2006;7(6):1836–44.
lial cells to Descemet’s membrane. Exp Eye Res. 38. Lai JY, Li YT. Functional assessment of cross-linked
2003;76(6):745–51. porous gelatin hydrogels for bioengineered cell sheet
22. Mimura T, Yamagami S, Usui T, et al. Long-term out- carriers. Biomacromolecules. 2010;11(5):1387–97.
come of iron-endocytosing cultured corneal endothe- 39. Lai JY, Ma DH, Lai MH, et  al. Characterization of
lial cell transplantation with magnetic attraction. Exp cross-linked porous gelatin carriers and their interac-
Eye Res. 2005a;80(2):149–57. tion with corneal endothelium: biopolymer concentra-
23.
Niu G, Choi JS, Wang Z, et  al. Heparin- tion effect. PLoS One. 2013;8(1):e54058.
modified gelatin scaffolds for human corneal 40. Dua HS, Azuara-Blanco A.  Amniotic membrane

endothelial cell transplantation. Biomaterials. transplantation. Br J Ophthalmol. 1999;83(6):748–52.
2014;35(13):4005–14. 41. Fan T, Ma X, Zhao J, et al. Transplantation of tissue-
24. Okumura N, Koizumi N, Ueno M, et al. ROCK inhibi- engineered human corneal endothelium in cat models.
tor converts corneal endothelial cells into a phenotype Mol Vis. 2013;19:400–7.
capable of regenerating in vivo endothelial tissue. Am 42. Ishino Y, Sano Y, Nakamura T, et  al. Amniotic

J Pathol. 2012;181(1):268–77. membrane as a carrier for cultivated human corneal
496 S. Wahlig et al.

endothelial cell transplantation. Invest Ophthalmol corneal endothelial cells. J Biomater Sci Polym Ed.
Vis Sci. 2004;45(3):800–6. 2008;19(1):1–18.
43. Zhang C, Nie X, Hu D, et al. Survival and integration 59. Watanabe R, Hayashi R, Kimura Y, et  al. A

of tissue-engineered corneal stroma in a model of cor- novel gelatin hydrogel carrier sheet for corneal
neal ulcer. Cell Tissue Res. 2007;329(2):249–57. endothelial transplantation. Tissue Eng Part A.
44. Bayyoud T, Thaler S, Hofmann J, et al. Decellularized 2011;17(17–18):2213–9.
bovine corneal posterior lamellae as carrier matrix for 60. Kabosova A, Azar DT, Bannikov GA, et  al.

cultivated human corneal endothelial cells. Curr Eye Compositional differences between infant and
Res. 2012;37(3):179–86. adult human corneal basement membranes. Invest
45. Du L, Wu X, Pang K, et al. Histological evaluation and Ophthalmol Vis Sci. 2007;48(11):4989–99.
biomechanical characterisation of an acellular porcine 61. Merrett K, Fagerholm P, McLaughlin CR, et al. Tissue-
cornea scaffold. Br J Ophthalmol. 2011;95(3):410–4. engineered recombinant human collagen-based cor-
46. Fu Y, Fan X, Chen P, et  al. Reconstruction of a tis- neal substitutes for implantation: performance of type
sue-engineered cornea with porcine corneal acel- I versus type III collagen. Invest Ophthalmol Vis Sci.
lular matrix as the scaffold. Cells Tissues Organs. 2008;49(9):3887–94.
2010;191(3):193–202. 62. Fagerholm P, Lagali NS, Ong JA, et  al. Stable cor-
47. Ju C, Gao L, Wu X, et al. A human corneal endothe- neal regeneration four years after implantation of
lium equivalent constructed with acellular porcine cor- a cell-free recombinant human collagen scaffold.
neal matrix. Indian J Med Res. 2012;135(6):887–94. Biomaterials. 2014;35(8):2420–7.
48.
Yoeruek E, Bayyoud T, Maurus C, et  al. 63. Levis HJ, Peh GS, Toh KP, et al. Plastic compressed
Decellularization of porcine corneas and repopulation collagen as a novel carrier for expanded human cor-
with human corneal cells for tissue-engineered xeno- neal endothelial cells for transplantation. PLoS One.
grafts. Acta Ophthalmol. 2012;90(2):e125–31. 2012;7(11):e50993.
49.
Amano S, Shimomura N, Yokoo S, et  al. 64. Hayes S, Lewis P, Islam MM, et  al. The structural
Decellularizing corneal stroma using N2 gas. Mol Vis. and optical properties of type III human collagen
2008;14:878–82. biosynthetic corneal substitutes. Acta Biomater.
50. Sasaki S, Funamoto S, Hashimoto Y, et  al. In vivo 2015;25:121–30.
evaluation of a novel scaffold for artificial corneas pre- 65. Kim EY, Tripathy N, Cho SA, et  al. Bioengineered
pared by using ultrahigh hydrostatic pressure to decel- neo-corneal endothelium using collagen type-I coated
lularize porcine corneas. Mol Vis. 2009;15:2022–8. silk fibroin film. Colloids Surf B Biointerfaces.
51. Pang K.  Differentiation of human embryonic stem 2015;136:394–401.
cells to corneal epithelium and endothelium like cells 66. Madden PW, Lai JN, George KA, et al. Human cor-
for cornea replacement construction. Invest Ophth Vis neal endothelial cell growth on a silk fibroin mem-
Sci. 2015;56:5831. brane. Biomaterials. 2011;32(17):4076–84.
52. Bloomfield P, Wheatley DJ, Prescott RJ, et al. Twelve- 67. Chen J, Yan C, Zhu M, et al. Electrospun nanofibrous
year comparison of a Bjork-Shiley mechanical heart SF/P(LLA-CL) membrane: a potential substratum
valve with porcine bioprostheses. N Engl J Med. for endothelial keratoplasty. Int J Nanomedicine.
1991;324(9):573–9. 2015;10:3337–50.
53. Booth C, Korossis SA, Wilcox HE, et  al. Tissue 68. McLaughlin CR, Acosta MC, Luna C, et  al.

engineering of cardiac valve prostheses I: devel- Regeneration of functional nerves within full thick-
opment and histological characterization of an ness collagen-phosphorylcholine corneal sub-
acellular porcine scaffold. J Heart Valve Dis. stitute implants in guinea pigs. Biomaterials.
2002;11(4):457–62. 2010;31(10):2770–8.
54. He Z, Forest F, Bernard A, et al. Cutting and decel- 69. Ozcelik B, Brown KD, Blencowe A, et al. Ultrathin
lularization of multiple corneal stromal lamellae chitosan-poly(ethylene glycol) hydrogel films
for the bioengineering of endothelial grafts. Invest for corneal tissue engineering. Acta Biomater.
Ophthalmol Vis Sci. 2016;57(15):6639–51. 2013;9(5):6594–605.
55. Peh GSL, Ang HP, Lwin CN, et al. Regulatory com- 70. Wang TJ, Wang IJ, Chen S, et  al. The phenotypic
pliant tissue-engineered human corneal endothelial response of bovine corneal endothelial cells on chi-
grafts restore corneal function of rabbits with bullous tosan/polycaprolactone blends. Colloids Surf B
keratopathy. Sci Rep. 2017;7(1):14149. Biointerfaces. 2012a;90:236–43.
56. Jumblatt MM, Maurice DM, Schwartz BD. A gelatin 71. Wang TJ, Wang IJ, Lu JN, et  al. Novel chitosan-
membrane substrate for the transplantation of tissue polycaprolactone blends as potential scaffold and car-
cultured cells. Transplantation. 1980;29(6):498–9. rier for corneal endothelial transplantation. Mol Vis.
57. Mohay J, Lange TM, Soltau JB, et al. Transplantation 2012b;18:255–64.
of corneal endothelial cells using a cell carrier device. 72. Wang TJ, Wang IJ, Hu FR, et al. Applications of bio-
Cornea. 1994;13(2):173–82. materials in corneal endothelial tissue engineering.
58. Lu PL, Lai JY, Ma DH, et  al. Carbodiimide cross- Cornea. 2016;35(Suppl 1):S25–30.
linked hyaluronic acid hydrogels as cell sheet deliv- 73. Salehi S, Bahners T, Gutmann J, et al. Characterization
ery vehicles: characterization and interaction with of structural, mechanical and nano-mechanical prop-
33  Dysfunctional Corneal Endothelium: Delivery of Cell Therapy 497

erties of electrospun PGS/PCL fibers. RSC Adv. 83. Moysidis SN, Alvarez-Delfin K, Peschansky VJ,

2014a;4(33):16951–7. et  al. Magnetic field-guided cell delivery with nan-
74. Salehi S, Fathi M, Javanmard SH, et  al. Generation oparticle-loaded human corneal endothelial cells.
of PGS/PCL blend nanofibrous scaffolds mimick- Nanomedicine. 2015;11(3):499–509.
ing corneal stroma structure. Macromol Mater Eng. 84. Bi YL, Wu MF, Lu LX, et  al. Functions of corneal
2014b;299(4):455–69. endothelial cells do not change after uptake of super-
75. Salehi S, Czugala M, Stafiej P, et  al. Poly (glycerol paramagnetic iron oxide nanoparticles. Mol Med Rep.
sebacate)-poly (epsilon-caprolactone) blend nanofi- 2013;7(6):1767–72.
brous scaffold as intrinsic bio- and immunocom- 85. Okumura N, Ueno M, Koizumi N, et al. Enhancement
patible system for corneal repair. Acta Biomater. on primate corneal endothelial cell survival in  vitro
2017;50:370–80. by a ROCK inhibitor. Invest Ophthalmol Vis Sci.
76. Rizwan M, Peh GS, Ang HP, et  al. Sequentially-
2009;50(8):3680–7.
crosslinked bioactive hydrogels as nano-patterned 86. Garnock-Jones KP.  Ripasudil: first global approval.
substrates with customizable stiffness and degra- Drugs. 2014;74(18):2211–5.
dation for corneal tissue engineering applications. 87. Okumura N, Kakutani K, Inoue R, et al. Generation
Biomaterials. 2017;120:139–54. and feasibility assessment of a new vehicle for cell-
77. McCulley JP, Maurice DM, Schwartz BD.  Corneal based therapy for treating corneal endothelial dys-
endothelial transplantation. Ophthalmology. function. PLoS One. 2016a;11(6):e0158427.
1980;87(3):194–201. 88. Adamis AP, Filatov V, Tripathi BJ, et  al. Fuchs’

78. Patel SV, Bachman LA, Hann CR, et al. Human cor- endothelial dystrophy of the cornea. Surv Ophthalmol.
neal endothelial cell transplantation in a human ex vivo 1993;38(2):149–68.
model. Invest Ophthalmol Vis Sci. 2009;50(5):2123–31. 89. Rizwan M, Peh GS, Adnan K, et  al. In vitro topo-
79. Mimura T, Yokoo S, Araie M, et  al. Treatment of graphical model of Fuchs dystrophy for evaluation
rabbit bullous keratopathy with precursors derived of corneal endothelial cell monolayer formation. Adv
from cultured human corneal endothelium. Invest Healthc Mater. 2016;5(22):2896–910.
Ophthalmol Vis Sci. 2005b;46(10):3637–44. 90. Lange TM, Wood TO, McLaughlin BJ.  Corneal

80. Mimura T, Yamagami S, Usui T, et  al. Necessary endothelial cell transplantation using Descemet’s
prone position time for human corneal endothelial membrane as a carrier. J Cataract Refract Surg.
precursor transplantation in a rabbit endothelial defi- 1993;19(2):232–5.
ciency model. Curr Eye Res. 2007;32(7–8):617–23. 91.
Vazquez N, Rodriguez-Barrientos CA, Aznar-
81. Clement O, Siauve N, Cuenod CA, et al. Liver imag- Cervantes SD, et  al. Silk fibroin films for corneal
ing with ferumoxides (Feridex): fundamentals, con- endothelial regeneration: transplant in a rabbit
troversies, and practical aspects. Top Magn Reson Descemet membrane endothelial Keratoplasty. Invest
Imaging. 1998;9(3):167–82. Ophthalmol Vis Sci. 2017;58(9):3357–65.
82. Wang YX, Hussain SM, Krestin GP. Superparamagnetic 92. Mimura T, Yamagami S, Yokoo S, et  al. Cultured
iron oxide contrast agents: physicochemical charac- human corneal endothelial cell transplantation with a
teristics and applications in MR imaging. Eur Radiol. collagen sheet in a rabbit model. Invest Ophthalmol
2001;11(11):2319–31. Vis Sci. 2004;45(9):2992–7.
Index

A B
Adipose derived adult mesenchymal stem cell Barrier function, corneal endothelium, 421
(ADASC), 373, 405 Basic fibroblast growth factor (bFGF), 112
Adipose tissue-derived MSCs (AT-MSCs), 236 Biodegradable polymers, 466
Adipose tissue mesothelial cell (ATMC), 138 Biological cell carriers, 487
Adult stem cells, 42, 478 Blindness, 390
Advanced therapy medicinal products (ATMPs), 48, 91 Bone marrow-derived endothelial progenitor cells
Allogeneic limbal transplant, 287 (BEPC), 137, 489
Alpha smooth muscle actin (α-SMA), 304, 307, Bone marrow-derived MSCs (BM-MSCs), 234
308, 493 Bone marrow mesenchymal stem cell
Amniotic membrane (AM), 63, 128, 429 (BM-MSC), 404
Amniotic membranes, 293 Bone morphogenic protein 4 (BMP4), 270
Amniotic membrane transplantation, 301 Bovine pituitary extract (BPE), 112
ANGEL®, 327 Bowman’s layer, 13, 388
Animal model, 152 Bullous keratopathy (BK), 166
cat, 448
choice of, 443
monkey, 449 C
pig, 448 Calcineurin inhibitors, 294
rabbit, 444 Cancer stem cells (CSCs), 51–52
rodent, 444 Canonical Wnt signaling, 430
sheep, 449 Carbon dioxide, 422
Anterior chamber associated immune deviation Cataract extraction, 464
(ACAID), 441, 443 CD44, 117
Anterior chamber injection, 411 CD56, 423
Anterior lens capsule, 429 CD109, 423
Anti-angiogenic function, 239 CD 166, 422
Anti-inflammatory action, 238 CD200, 117, 422
Aquaporin 1 (AQP1), 482 CD248, 423
Aqueous deficient DED (ADDED), 328 Cell-based medicinal products (CBMP), 399
Artificial cell-free device, 155 Cell-based therapy, 467–468
Autologous eye-platelet rich plasma (E-PRP), 19 Cell carriers
Autologous limbal stem cell product, quality biological, 487
control for, 91 synthetic, 489
Autologous limbal stem cell transplants (ALSCT), 82 Cell-injection therapy, 457
Autologous serum eye drops (ASE), 81 for bullous keratopathy patient, 458
Autologous transplantation in failed graft eye, 459
European marketing authorisation, 93 ROCK inhibitor, 457
good manufacturing practice, 91 safety and efficacy of, 459
limbal stem cell, 92–93 specular microscopy, 460
manufactured product, 92 surgical procedure, 457
quality control, 91–92 Cell migration, 306

© Springer Nature Switzerland AG 2019 499


J. L. Alió et al. (eds.), Corneal Regeneration, Essentials in Ophthalmology,
https://doi.org/10.1007/978-3-030-01304-2
500 Index

Cell therapy Connexon, 178


with adult tissue progenitors, 264 Cornea, 419
clinical application, 271 anatomical dimensions, 3
using cultivated oral mucosal epithelial transplant biometrics and structure, 350
(COMET), 225–228 Bowman’s layer, 388
using ex vivo cultured limbal cells, 221–224 Descemet’s membrane, 388
using induced pluripotent stem cell, 263–273 diseases, 23
Cell types, 109 embryology, 3
Cellular injection endothelium, 388
animal studies of, 492 epithelium, 388
endothelial dysfunction, 491 integrity and transparency, 57
CEnPC, see Corneal endothelial progenitor cell (CEnPC) stroma, 388
Chitosan, 466 transparency, 388
CLAU, see Conjunctival limbal autograft (CLAU) Corneal endothelial cell (CEnC), 58, 109, 481, 485, 486
CLET, see Cultivated limbal epithelial transplantation biology, 455–456
(CLET) cell-injection therapy, 457–458
CLRN1, 424 for bullous keratopathy patient, 458–459
Collagen, 27, 489 in failed graft eye, 459
Collagen II production, 152 product development, 458
dystrophic cornea biomechanical property, 151 specular microscopy, 460
in human tissue, 149 surgical procedure, 457
in vivo model, 151 culture for clinical application, 456
in keratoconic tissue, 150 sheet transplantation, 456–457
Collagen type-1, 351 Corneal endothelial cell loss, prevention
COMET, see Cultivated oral mucosal epithelial RTA 408 ophthalmic suspension, 464
transplantation (COMET) sulforaphane, 464
Committee for Advanced Therapies (CAT), 49 Corneal endothelial progenitor cell (CEnPC), 58
Confocal microscopy function, 66
cell counting methods obtaining methods
ADAS cell, 373 adherent culture of single cells, 67
automated cell counting, 371 application, 69
decellularized lamina, 374 culture onto biological and biosynthetic substrate,
difficulty, 375 68
manual cell counting, 371–373 explant culture, 68
recellularized lamina, 375 suspension culture, 68
corneal density calculation, 375 study, 66
corneal stroma Corneal endothelium, 14, 26, 425
automatic cell counting, 384 changes in cell density and size, 421
patients with decellularized lamina, 377–381 3D mapping proteins, 423
patients with recellularized lamina, 380–383 embryology, 419–420
patients with transplanted ADAS, 375–376 functional activity, 8
sample, 367–371 human, 419, 420
handling difficulty, 366 immunohistochemistry of, 422–424
preparation, 364–366 mitotic activity, 9
Confocal ophthalmoscopy, 441 morphology, 420–421
Conjunctival epithelium physiology
goblet cells, 184 barrier function, 421
innervation, 186 pump function, 422
mucin, 184 secretory function, 422
tear film, 185–186 ZO-1 staining for rabbit cells, 423
Conjunctival invasion, 300 Corneal epithelial cells (CEC), 13, 77, 422, 479
Conjunctival limbal autograft (CLAU), 156 apical cell layer, 4
indication, 156 basal cell layer, 4
LSCD, 156 hypothesis, 6
ocular surface neoplasia, 157 innervation, 5
outcome, 157 limbus, 6
pterygium, 157 mass, 183
Conjunctival transplantation, 191, 195–198 regenerate, 78
Conjunctivalisation, 301 stem cell based therapy, 155
Connexins, 178 Corneal epithelial stem cells, 179
Index 501

Corneal graft rejection bone marrow, 393


clinical criteria, 440 dental pulp, 394
from animal models, 441 embryonic, 394
Corneal healing process induced pluripotent stem cell, 394
attachment, 17 neural stem cell, 394
cytokines, 14 umbilical cord, 394
endothelial regeneration, physiological mechanism stromal remodelling, 392
of, 19–21 Corneoscleral rims, 158, 302, 303, 427
epithelial cell migration, 16 Coxsackieadenovirus receptor (CAR), 423
growth factor Cryoablation, 445
autologous eye-platelet rich plasma, 19 Cultivated limbal epithelial transplantation (CLET),
epidermal, 14, 18 160–161, 222, 223
fibroblast, 15 biopsy harvesting, 285
hepatocyte, 15 culture method
insulin-like growth factor, 16, 18 scaffold, 284
keratinocyte growth factor 1, 15 technique, 285
nerve growth factor, 15, 18 graft transplantation, 285
platelet-derived growth factor, 15 post-operative management
transforming growth factor-β, 15 additional, 287
latent phases, 16 early, 286
proliferation and differentiation, 16 pre-operative assessment
Corneal layers, 176 indications, 278–281
Corneal neovascularization, 23 supplementary examinations, 281–283
Corneal opacification, 440, 441, 444 Cultivated oral mucosal epithelial transplantation
Corneal scraping, 301 (COMET), 161–163
Corneal stroma. See also Corneal stromal stem cells clinical results, 227
(CSSC), 7–8 development, 226
Corneal stromal stem cell (CSSC), 57, 404 keratoplasty, 227–228
with basic fibroblast growth factor, 64 surgery, 226–227
culturing, 104–105 Cultured limbal epithelial transplantation (CLET), 25
differentiation Current stem cell based therapy, 155
into keratocytes, 101–102 conjunctival limbal autograft, 156
into non-ocular cells, 101 indication, 156
embryonic origin, 100 LSCD, 156
expanding methods ocular surface neoplasia, 157
explant culture, 65 outcome, 157
plastic adherent culture, 65 pterygium, 157
suspension culture, 66 cultivated limbal epithelial transplantation, 160–161
function, 103 cultivated oral mucosal epithelial transplantation,
gene expression, 101 161–163
immune reactions of, 103 keratolimbal allograft, 158
interactions, 103 living-related conjunctival limbal allograft, 158–160
isolation from limbal biopsy, 104 postoperative complication, 163–164
limbal niche, confocal microscopy of, 105–106 simple limbal epithelial transplantation, 161
localization of, 102 Cyclophosphamide, 449
mesenchymal origin, 101 Cysclosporine A (CsA), 295
obtaining methods, 64 Cytokeratin(s), 178, 181
primary isolation, in humans, 100 Cytokeratin intermediate filaments, 5
progenitor potential of, 100 Cytokine, 392
properties, 101 Cytoskeleton, 178
Corneal wound healing, 237
clinical trial, 398–400
epithelial injury, 392 D
hLMSC, 395–398 Decellularization, 88, 489
limbus and the stem cell, 390–391 Decellularized lamina
MSC limitations, 394 anterior face, 377–379
myofibroblast migration and differentiation, 392 collagen matrix morphology, 378
preclinical trial, 394–395 posterior face, 377–379
stem cell-based therapy, 393 reflective peripheral structures, 380
adipose tissue, 393 transition zone, 380, 385
502 Index

Degranulation, 318 Endothelial transplantation


Demoglein, 177 animal models, 438
Dendritic cells, 8 cat, 448
Dental pulp-derived MSCs (DP-MSCs), 236 choice of, 443–444
Dental pulp stem cell (DPSC), 394 monkey, 449
Descemet membrane (DM), 8, 13, 112, 388, 420, 428, pig, 448
463, 488 rabbit, 444–448
Descemet membrane endothelial transfer rodent, 444
(DMET), 20, 465 sheep, 449
Descemet’s membrane endothelial keratoplasty central corneal thickness, changes in, 442
(DMEK), 464 corneal graft rejection
Descemet stripping automated endothelial keratoplasty clinical criteria, 440–441
(DSAEK), 445, 487 from animal models, 441–443
Descemet’s stripping endothelial keratoplasty (DSEK), in vivo and post-mortem assessment, 438–440
464, 490 Enzymatic digestion, 428
Descemetorhexis, 465 Epidermal growth factor (EGF), 14, 18, 28, 320, 335
Desmocollins, 177 Epidermal stem cell, 128
Desmoglein 3, 182 Epithelial basement membrane (EBM), 18
Desmosomes, 177 Epithelial cell culture
Devitalization, 445 assessment
Direct cell labelling methods, 440 cell counting and viability, 90
Direct cellular reprogramming, 267 colony forming efficiency assays, 90–91
Direct compression injury, 308 histology and immunostaining, 89–90
Direct monolayer implantation, 487 measure explant outgrowth expansion, 90
Donut excimer laser ablation model, 302 human amniotic membrane, 87–89
Dry eye disease, 225 limbal explant culture, 87
DSAEK, see Descemet stripping automated endothelial biopsy, 87
keratoplasty (DSAEK) culture medium preparation, 84
Dua-Fine layer, 355 HAM construct preparation, 85
Dua’s layer, 13 single cell suspension method, 83
Dulbecco's phosphate-buffered saline (DPBS), 84 3T3 fibroblast co-culture
from frozen stock initiation, 84
frozen layers preparation, 84
E frozen stocks preparation, 84
E-cadherin, 467 maintenance and subculture, 84
Ectatic disorder, 403 Epithelial keratinisation, 79
Embryonic stem cell (ESC), 41, 123, 394, 404, 405, 468 Epithelial mesenchymal transition (EMT), 456, 467
corneal endothelium regeneration, 135 Epithelial regeneration, 479–480
corneal epithelium regeneration, 124–127 Epithelial stem cell, corneal epithelium regeneration
corneal stroma regeneration, 131–133 amniotic membrane, 128
Embryonic tissue, 477 epidermis, 128
Endosialin, 423 hair follicle, 128
Endothelial cell, see Human corneal endothelial cell oral mucus, 127–128
(HCEnC) umbilical cord lining, 129
Endothelial cell density (ECD), 425, 463 Epithelial wound healing, 16–17
Endothelial cell therapy, 486 Ethylenediamine tetra-acetic acid (EDTA), 428
Endothelial decompensation, 464 European Medicines Agency (EMA), 30, 93
Endothelial dysfunction Ex vivo expansion, 83
animal models, 485–487 corneal endothelial regeneration, 466–467
carrier implantation Exosomes, 413
animal studies of, 491–494 Extracellular matrix (ECM), 24, 351, 429
biological cell carriers, 487–489 Extracorneal stem cell
synthetic cell carriers, 489–491 corneal endothelium regeneration, 135
cellular injection, 491, 493 embryonic and induced pluripotent stem cells,
animal studies of, 492 135–136
direct monolayer implantation, 487 mesenchymal stem cells, 138
management, 463–464 mesothelial cell, 138–139
Endothelial keratoplasty (EK), 426, 455, 464 neural crest stem cells, 137
Endothelial-mesenchymal transition skin-derived precursor, 137
(EMT), 110, 430, 467 vascular endothelial progenitor, 137–138
Endothelial regeneration, 481–482 corneal epithelium regeneration, 124
Index 503

epithelial stem cell, 127–129 G


immature dental pulp stem cell, 130–131 Gap junctions (GP), 178, 421–422
mesenchymal stem cell, 129–130 Gelatin, 489
pluripotent stem cell, 124–127 Gelatin disc, 447
corneal stroma regeneration, 131 Gene therapy medicinal product (GTMP), 49
embryonic stem cells and induced pluripotent Glucocorticoids, 294
stem cells, 131–133 Glycocalyx, 4
mesenchymal stem cells, 133–134 Glypican 4, 422
oral stem cell, 134–135 Gold nano-rods, 440
Extraocular mesenchymal stem cell Good manufacturing practice, 91
for corneal epithelial damage, 253–254 GPC4, 117
experimental model GRIP1, 424
amniotic membrane, 249–250 Growth factors (GF), 14, 464
with central corneal damage, 246 epidermal growth factor, 320
combined administration, 252–253 function, 319
contact lens carrier, 250 platelet-derived growth factor, 319–320
fibrin matrix, 250 synthesize, 319
intraperitoneal administration, 248 transforming growth factor beta, 320
intravenous administration, 246–248 vascular endothelial growth factor, 321
mobilization of endogenous, 246 Guanine nucleotide exchange factors (GEFs), 465
ocular topical administration, 248–249
polyamide 6/12 carrier, 251
poly-L-lactic acid carrier, 250 H
subconjunctival injection, 252 Hair follicle stem cell (HFSC), 128
therapeutic potential HCEnC, see Human corneal endothelial cell (HCEnC)
immunomodulatory ability, 240–244 Hemangioblast, 419
in vitro and in vivo studies, 233 Hematopoietic markers, 8
tissue homing capacity, 244–246 Hemidesmosomes, 178
transdifferentiation capacity, 234–237 Hemoderivatives, 292
trophic activity, 237–240 Heparin, 447
Extraocular stem cell Hepatocyte growth factor (HGF), 15, 28, 464
adipose derived adult mesenchymal stem cell, 405 Herpes simplex virus type-1 (HSV-1), 23
bone marrow mesenchymal stem cell, 404 Holoclar®, 51
embryonic stem cell, 405 Homeostasis, in stroma, 145
induced pluripotent stem cell, 406 HTR1D, 424
umbilical cord mesenchymal stem cell, 405 Human adult adipose tissue, 404
Eyelid abnormalities, 295 Human adult corneal limbal epithelial cell (HLEC), 127
Eye-platelet rich plasma (E-PRP), 325–326, 332 Human adult dental pulp cells (hDPCs), 134
Human adult dermal fibroblast (HDF), 127
Human amniotic membrane (HAM), 85–89, 198, 284
F Human bone marrow-MSC (hBM-MSC), 138
Ferromagnetic material, 492 Human corneal endothelial cell (HCEnC), 485, 487, 489,
Fetal calf serum, 432 492
Fibrin gel, 250 cellular and regenerative therapies, 110
Fibroblast growth factor (FGF), 15, 464 culture media, 430–432
Fibronectin, 321 culture conditions
Fluorescence ubiquitination cell cycle indicator (Fucci), functional cell marker, 117–118
304 growth media, 112–116
Fluorochromes, 440 surface, 116–117
Foetal tissue, 477 endothelial progenitor cell, 110–112
Food and Drug Administration (FDA), 30 extracellular matrix, 429, 430
Fuchs endothelial corneal dystrophy (FECD), 20, 167, immortalized cell lines, 428
463, 465, 466 injection, 491, 494
Future stem cell based therapy, 164 isolation, 112, 428
corneal endothelium, 166–167 morphology of, 439
corneal epithelial cell native, 110
differentiated cell source, 164–166 primary, 427
from pluripotent stem cell, 164 sources, 427–428
corneal stroma, 166 stem cells, 427–428
purified LSC, 166 substrates, 429–430
surface ectoderm-derived cell, 166 tissue engineering, 426–432
504 Index

Human eye, 299 automated cell counting, 371


Human limbal-derived stromal/mesenchymal stem cell collagen II production, 150
(hLMSC), 395–398 differentiation, 146
Human mesenchymal stem cells (hMSCs), 46 manual cell counting, 371
Human umbilical cord morphology of, 369
cross-section of, 478 treatment, 149–153
stem cells, 478–479 untapped potential of, 147, 149
Human umbilical cord blood mesenchymal stem cells Keratocyte differentiation medium (KDM), 102
(hUCB MSCs), 469 Keratocyte progenitor cells (KPC), 300
Hyaluronic acid, 489 Kerato-epithelioplasty, 191
Hydrogel, 489 Keratolimbal allograft (KLAL), 158
Keratoplasty, 99, 227–228
KLAL, see Keratolimbal allograft (KLAL)
I
Immature dental pulp stem cells, 130
Immune cells, 8 L
Immunohistochemistry, 422 Lagophthamos, 295
Immunosuppression, 241 Lens epithelial cell-conditioned medium
Impression cytology, 282 (LECCM), 111, 481
In vivo confocal microscopy (IVCM), 105, 282 LESC, see Limbal epithelial stem cell (LESC)
In vivo laser scanning confocal microscopy (IVCM), 301 Leucine-rich repeat-containing G protein-coupled
In vivo stimulation receptor 5 (LGR5), 456
pharmacological treatment Lid abnormality repair, 295–296
growth factors, 464 Lifitegrast, 294
ROCK inhibitors, 465 Limbal cells, 299
surgical treatment, 465–466 Limbal epithelial crypts (LEC), 180–183
Indirect cell labeling methods, 440 Limbal epithelial stem cell (LESC), 6, 57, 77
Induced pluripotent stem cell (iPSC), 111, 394, 406, identification, 58
468, 478 isolation methods
corneal endothelium regeneration, 135 amniotic membrane, 63
corneal epithelium regeneration, 127 culture media, 60–62
corneal stroma regeneration, 131–133 culture onto ECM coated surface, 63
mechanisms for reprogramming, 266–269 explant culture, 62
Inflammation feeder cells, 60–62
local, 294 silk fibroin, 63
systemic, 295 single-cell suspension culture, 62
Insulin-like growth factor (IGF), 16, 18, 28 sphere-forming assay, 62
Integrins, 14 tissue preparation, 60
Interleukin 1α (IL−1α), 14 properties, 58
International Society for Cellular Therapy, 237 Limbal explant biopsy, 87
Intracellular carbonic anhydrase pathway, 422 Limbal explant culture, 87
Intrastromal implantation, 407–409 Limbal niche, 77
Intravenous injection, 412 Limbal stem cell (LSC), 155, 191, 263–266,
299–300, 479
Limbal stem cell deficiency (LSCD), 25, 78, 232, 277,
J 300, 479
Junctional adhesion molecules (JAMs), 423 aetiology, 78, 277
cause and manifestation, 213–214
characteristics, 80
K conjunctivalisation, 301
Keratin, 304 destruction, 300
Keratinocytes growth factor, 28 epidemiology, 79, 221
Keratinocyte growth factor 1 (KGF-1), 15 grafting procedures, 480
Keratoconus impression cytology, 300
characteristic, 363 in vivo laser scanning confocal microscopy, 301
methods (see Confocal microscopy) management, 80–83
patients, 364 regeneration of limbus, 302
treatment, 363 semiology, 79–80
Keratocyte(s), 8, 26, 131, 353, 480 slit lamp examination, 300
activation, 145 treatment, 214, 221–223, 301–302
Index 505

Limbal stem cell failure (LSCF), management, 225–226 Negative HCEnC markers, 117
Limbal transplantation, 195 Nerve growth factor (NGF), 15, 18, 28
allograft, 202–208 Neural cell adhesión molecule (NCAM), 423
autograft, 198–202 Neural crest cell (NCC), 271, 419, 420
classification, 194–195 Neural crest-derived progenitor (NCDP), 167
conjunctival, 195–198 Neural crest progenitors (NPCs), 135
history, 191–194 N-isopropylacrylamide-co-glycidylmethacrylate
Limbus, 6 (NGMA), 487
Living-related conjunctival limbal allograft (Lr-CLAL), Non-enzymatic digestion, 428
158–160
Local inflammation
cyclosporine/tacrolimus, 294 O
lifitegrast, 294 Occludin, 177
steroids, 294 Ocular surface (OS), 176
Lr-CLAL, see Living-related conjunctival limbal ecosystem restoration, 223
allograft (Lr-CLAL) epithelium
LSCD, see Limbal stem cell deficiency (LSCD) basement membrane, 178–179
cell attachment, 177–178
conjunctiva, 183–186
M function, 176
Macula adherens, 177 limbus and limbal, 179–183
Massive injury, 57 superficial flat cell, 176
Matrix metalloproteinases (MPPs), 14 implantation, 406–407
Medical therapy, 221 optical polish, 176
Membrane-bound Na+, K+ -ATPase sites, 422 reconstruction of, 226
Mesenchymal stem cell (MSC), 25, 42, 101, 123, 233 translational application on, 269–271
autologous vs. heterologous, 412 transparent cornea, 176
corneal endothelium regeneration, 138 Oligopotent stem cells, 477
corneal epithelium regeneration Opioid growth factor (OGF), 29
adipose tissue, 130 Optical coherence tomography (OCT), 282, 301, 438, 445
bone marrow, 129 OptiMEM-I based media, 112
fetal tissue, 130 Oral mucus, 127
mesenchymal-to-epithelial transition, 129 Oral mucosa transplantation, 208
corneal stroma regeneration Oral mucosa uses, 226
adipose tissue, 133 Oral stem cell, 134
bone marrow, 133 Osmotic gradient, 422
umbilical cord, 134 Oxidative stress protection, 239
exosome, (see also Extraocular mesenchymal stem
cell), 413
immune modulatory characteristics, 241 P
immunomodulatory properties, 242–244 Palisades of Vogt, 6–7, 179, 180
types, 403 Palpebral conjunctiva, 184
Misdirected lashes, 296 Penetrating keratoplasty (PK), 426, 437, 464
Mitomycin C inhibits, 18 Peptide amphiphiles (PA), 27
Modified embryonic stem cell medium (MESCM), 432 Pericellular matrix from human decidua-derived
MRGPRX3, 424 mesenchymal cells (PCM-DM), 116
Mucin-expressing cord lining epithelial cell (CLEC-­ Periocular mesenchymal progenitors (POMPs), 135
muc), 480 Photorefractive keratectomy (PRK), 18
Mucopolysaccharides, 352 Placental stem cell, 42
Mucosa transplantation, 208–209 Platelet(s)
Multipotent, 42 alpha granule, 318
Multipotent stem cell, 42, 111, 469, 477 blood derivative
Mycophenolate mofetil, 295 inflammation, 323–324
Myofibroblast, 392 properties, 321–323
discovery, 317
electron microscopy, 318
N function, 318
National Programme for Control of Blindness (NPCB), 390 physiological condition, 318
Native human corneal endothelial cells, 110 Platelet-derived growth factor (PDGF), 15, 28,
N-cadherin, 468, 481, 482 319–320, 335
506 Index

Platelet-rich plasma (PRP), 321 Sclerocorneal limbus, 263


antimicrobial activity, 324 Secretory function, corneal endothelium, 422
autologous serum, 327–328 Serum free media (SFM), 46, 432
blood derivative, 324 cell morphology, 47
clinical applications, 328–331 cryopreservation, 47
closed system, 326 expansion, 47
ANGEL®, 327 harvest protocol, 47
PRGF Endoret®, 327 immunophenotype analysis, 47
RegenKit® BCT, 326 isolation, 47
E-PRP, 325–326 multilineage differentiation potential, 47
Pluripotency of iPSCs, 52 quality analysis, 47
Pluripotent stem cell, 42, 111, 477 viability and cell proliferation, 47
corneal epithelium regeneration Severe stem cell deficiency, 227
embryonic stem cell, 124–127 Silk fibroin (SF), 63, 489
induced, 127 Silk fibroin graft, 445
p75 neurotropin receptor (p75NTR), 111 Simple limbal epithelial transplantation (SLET), 161,
Poly (glycerol sebacate)-poly (ε-caprolactone) 214, 222
(PGS-­PCL), 490 advantages, 215
Poly(N-isopropyl acrylamide), 430, 466, 487 complication, 217–218
Polyamide nanofiber scaffold, 251 indication, 215–217
Polycaprolactone (PCL), 466 Skin-derived precursors (SKPs), 137
Poly-L-lactic acid carrier, 250 SLET, see Simple limbal epithelial transplantation (SLET)
Polymegathism, 421 Slit lamp examination, 300
Polymorphism, 421 Slit lamp microscopy, 438, 441
Polysaccharide-based materials, 27 of rabbit corneas, 446
Poly (vinylidene fluoride) membrane, 447, 466 of rat corneas, 446
Poly(vinyl methyl ether) (PVME), 487 Soft contact lenses, 293
Post-mortem techniques, 438 Somatic cell therapy medicinal product (SCTMP), 49
Pre-Descemets layer (PDL), 355–359 Somatic stem cell, 123
PRGF Endoret®, 327 Specular microscopy, 438, 441
Primordium, 271 Sphere(s)
Proliferating cell nuclear antigen (PCNA), 310 cell characteristics, 305
Pterygium, 157 direct compression injury of, 307, 308
Pump function, 422 en masse migration, 306, 307
Pump-leak mechanism, 422 from keratoconic cell, 312
implantation
into diseased central corneal tissue, 311
Q into donor corneoscleral rims, 309
Quality assurance, 399 migratory cells, 304
Quality system approach, 400 on collagen substrate, 304–306
Quantum dots, 440 Sphere formation, 303
Sphere forming assay, 68, 302, 303
Stage-specific embryonic antigens (SSEAs), 45
R Stem cells, 179
Recombinant human collagen (RHC), 29 anterior chamber injection of, 411
Regenerative therapy, 123 advanced therapies with, 48–51
RegenKit® BCT, 326 and cancer, 51–52
Rho-associated kinase (ROCK), 114, 426, 456–458, 465, CEC progenitor cell, 469
467, 492, 494 cell membrane markers, 45
Ripasudil, 465, 466 characterization, 45
Rostock Cornea Module (RCM) object, 364 classification
RTA 408 ophthalmic suspension, 464 potentiality, 42
source, 41
concepts, 41
S culture, 46–48
Scaffold, 26–28 intracellular markers, 45
bioengineered, 27 intrastromal implantation of, 407–409
in ophthalmology, 26 intravenous injection of, 412
peptide amphiphiles, 27 marker, 310
synthetic and natural based, 27 multipotent, 469, 477
Schirmer filter paper test, 282 non-coding RNAs, 45
Schwalbe’s line cells, 110, 427 ocular surface implantation of, 406–407
Index 507

oligopotent, 477 Tenascin C, 182


pluripotent, 468, 477 Terminally differentiated cells (TDC), 78
properties, 477 Tight junctions (TJ), 422
differentiation, 44 Tissue engineered product (TEP), 49
immunomodulation and immunogenicity, 44 Tissue engineering (TE)
self-renewal program, 42 clinical experiences, 29–30
spheres corneal substitute
isolation of sphere with mixed potential, 302–313 bioactive molecules, 29
keratocyte progenitor cells, 300 cells, 25
limbal stem cell, 299–300 scaffolds, 26–28
limbal stem cell deficiency, 300–301 definition, 24
regeneration of limbus, 302 international regulations, 29–30
stromal stem cells, 300 Tonopen, 152
treatment, 301–302 Topical glucocorticoids, 294
sources Total limbal stem cell deficiency (TLSCD), 80
adult tissue, 478 Totipotent stem cells, 42, 477
embryonic tissue, 477 Trabecular meshwork (TM), 491
foetal tissue, 477 Traditional corneal grafting techniques, 278
induced pluripotent stem cells, 478 Transcription factor p63 (ΔNp63α), 266
totipotent, 477 Transforming growth factor beta (TGF-β), 15, 28, 320
transplantation, 215 Transient amplifying cell (TAC), 78, 265, 391
types, 123 Trypsin, 428
Steroids
local inflammation, 294
systemic inflammation, 295 U
Steroid-sparing immunosuppression, 295 Umbilical cord, 478
Stevens-Johnsons syndrome (SJS), 79 blood, 138, 479
Stroma, 13 lining membrane, 479
Bowman’s layer, 350 stem cells in, 478
corneal transparency, 359–360 Umbilical cord blood endothelial progenitor cells
extracellular matrix, 351–352 (UCBEPCS), 482
glycosaminoglycan, 352 Umbilical cord lining epithelium, 129
keratocyte, 353–355 Umbilical cord mesenchymal stem cell
nerves, 355 (UCMSC), 405, 480
pre-descemets layer, 355–359 corneal opacification, 481
Stromal keratocytes, 428 in Lumican deficiency, 481
Stromal regeneration, 480–481 in mucopolysaccharidosis VII, 481
Stromal stem cells, 300 Umbilical cord stem cell, 42
Stromal wound healing, 17–18 Unipotential stem cell, 42
Structure-function relationships, 24 Urokinase-Type Plasminogen Activator (uPA), 14
Subconjunctival injection, 252
Sulforaphane, 464
Superficial corneal epithelial cells, 178 V
Superparamagnetic microspheres, 445 Vascular endothelial growth factor (VEGF), 321, 336
Synthetic cell carriers, 489 Vimentin, 304, 310, 312, 468
Systemic immunosuppressive therapy, 295 Vogt’s striae, 369
Systemic inflammation, 295
Systemic toxicity test, 152
W
Wharton’s jelly, 479
T Wharton’s jelly of the umbilical cord (WJ-MSCs), 236
Tacroliums, 294 Wing cells, 3
TAG-2A12, 117
Tear film health
amniotic membrane, 293 X
hemoderivatives, 292 Xenogeneic grafts, 441, 444
MGD and blepharitis treatment, 292
minimizing drop toxicity, 293
punctal occlusion, 293 Z
soft and scleral contact lenses, 293 ZO-1, 481, 482
tear supplementation, 291 ZP4, 424

You might also like