Gastrointestinal - Mucosal.repair - And.experimental - Therapeutics Ublog - TK

Gastrointestinal Mucosal Repair and Experimental Therapeutics
Frontiers of
Gastrointestinal Research
Vol. 25
Series Editors
I.M. Modlin New Haven, Conn.

P. Rozen Tel Aviv
C. Scarpignato Parma/Nantes
Gastrointestinal Mucosal
Repair and Experimental
Therapeutics
Volume Editors
C.-H. Cho Hong Kong

J.-Y. Wang Baltimore, Md.
57 figures, and 9 tables, 2002
Basel · Freiburg · Paris · London · New York ·

New Delhi · Bangkok · Singapore · Tokyo · Sydney
Frontiers of Gastrointestinal Research
Library of Congress Cataloging-in-Publication Data
Gastrointestinal mucosal repair and experimental therapeutics/volume editors, C.-H. Cho, J.-Y. Wang.
p.; cm. – (Frontiers of gastrointestinal research; vol. 25)
Includes bibliographical references and index.
ISBN 3805573820 (hard cover)
1. Peptic ulcer. 2. Gastrointestinal mucosa–Pathophysiology. 3. Wound healing.
I. Cho, C.-H. (Chi-Hin) II. Wang, J.-Y. (Jian-Ying) III. Series.
[DNLM: 1. Gastric Mucosa–injuries. 2. Gastric Mucosa–drug effects. 3. Gastric
Mucosa–pathology. 4. Wound Healing. WI 302 G2575 2002]
RC821.G375 2002
616.3⬘43–dc21
2002016244
Drug Dosage. The authors and the publisher have exerted every effort to ensure that drug selection and
dosage set forth in this text are in accord with current recommendations and practice at the time of publication.
However, in view of ongoing research, changes in government regulations, and the constant flow of information
relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for
any change in indications and dosage and for added warnings and precautions. This is particularly important
when the recommended agent is a new and/or infrequently employed drug.
All rights reserved. No part of this publication may be translated into other languages, reproduced or
utilized in any form or by any means electronic or mechanical, including photocopying, recording, microcopying,
or by any information storage and retrieval system, without permission in writing from the publisher.
© Copyright 2002 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland)

www.karger.com
Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel
ISSN 0302–0665
ISBN 3–8055–7382–0
Contents
VII Preface
Part 1: Epithelial Restitution
1 Roles of Extracellular Matrix and Cytoskeleton in Intestinal Epithelial

Restitution
Turner, J.R. (Chicago, Ill.); Basson, M.D. (Detroit, Mich.)
14 Cytokines in Restitution
Yoo, J.; Lotz, M.M.; Matthews, J.B. (Cincinnati, Ohio/Boston, Mass.)
29 Ca2ⴙ Signaling in Epithelial Restitution
Rao, J.N.; Wang, J.-Y. (Baltimore, Md.)
43 Polyamines in Intestinal Epithelial Restitution
McCormack, S.A.; Ray, R.M.; Johnson, L.R. (Memphis, Tenn.)
57 Epithelial Restitution and Physical Stress
Osada, T. (Tokyo); Watanabe, S. (Akita); Sato, N. (Tokyo)
69 The Diacylglycerol/Protein Kinase C Pathway in Gastrointestinal Mucosal
Injury and Defense
Miller, T.A.; Redlak, M.J.; Coy, L.M.; Taher, M.M. (Richmond, Va.)
Part 2: Mucosal Repair and Ulcer Healing
82 Expression of Early Primary Response Genes in Healing of

Gastrointestinal Mucosal Injury
Wang, J.-Y. (Baltimore, Md.)
101 Role of Angiogenesis and Angiogenic Growth Factors in Mucosal Repair
and Ulcer Healing
Tarnawski, A.S.; Jones, M.K. (Long Beach, Calif./Irvine, Calif.); Baatar, D.; Pai, R.
(Irvine, Calif.)
V
117 Role of Platelets in Gastric Ulcer Healing: A Delivery System for Growth
Factors
Ma, L.; Wallace, J.L. (Calgary)
129 Intestinal Mucosal Function following Ischemia/Reperfusion
Tso, P. (Cincinnati, Ohio); Wollin, A. (Saskatoon)
143 Helicobacter pylori Infection and Gastroduodenal Mucosal Damage and
Healing
Xia, H.H.X.; Wong, B.C.Y.; Lam, S.K. (Hong Kong)
Part 3: Experimental Therapeutics
158 Nitric Oxide-Releasing Agents – A New Generation of Drugs for

Gastrointestinal Diseases
Brzozowski, T. (Cracow); Konturek, P.C. (Erlangen); Konturek, S.J. (Cracow)
166 Cyclooxygenase Inhibitor, a Foe or a Friend in the Mucosal Protection
and Repair
Peskar, B.M. (Bochum)
180 Polysaccharides: A New Role in Gastrointestinal Protection
Cho, C.C.M.; Liu, E.S.L.; Shin, V.Y.; Cho, C.H. (Hong Kong)
190 Modulators of Inducible Nitric Oxide Synthase: Potential Drugs for the
Therapy of Gut Inflammation?
Whittle, B.J.R.; Cavicchi, M.; Lamarque, D. (London)
209 Peptide and Gene Therapy with Angiogenic Growth Factors bFGF,
PDGF or VEGF in Gastrointestinal Ulcers in Rats
Khomenko, T.; Deng, X.; Ishikawa, H.; Sandor, Z.; Szabo, S. (Long Beach,
Calif./Irvine, Calif.)
227 Gastrointestinal Protective Action of Prostaglandin E2 and EP Receptor
Subtypes
Takeuchi, K.; Kato, S.; Tanaka, A. (Kyoto)
243 Author Index

244 Subject Index
Contents VI
Preface
Over the last decade, considerable progress has been made in understanding
cellular and molecular mechanisms involved in mucosal injury and repair in the
gastrointestinal tract. These significant findings provide a fundamental basis to
identify the etiology and pathogenesis of various gut mucosal injury-related dis-
eases and to develop new therapeutic approaches. This book is to provide a
timely and long-lasting guide for investigators in the field of gut mucosal injury
and repair, and has been divided into three main sections: Epithelial restitution,
Mucosal repair and ulcer healing, and Experimental therapeutics. The first sec-
tion highlights the early rapid mucosal restitution and focuses on the roles of
extracellular matrix, cytoskeleton, cytokines, Ca2⫹ signaling, polyamines, and
the protein kinase C/DAG pathways. The second section is devoted to aspects of
chronic mucosal healing and concentrates on the roles of primary response gene
expression, angiogenesis and angiogenic growth factors, platelets, and the mech-
anisms of cell renewal after injury in special circumstances such as ischemia/
reperfusion and Helicobacter pylori infection. The third section is designed to
explore new therapeutic approaches that are based on the scientific development
and achievements during the last decade. We concentrate on potential clinical
applications of nitric oxide-releasing agents, polysaccharides, nitric oxide syn-
thase modulators, growth factors, prostaglandins, and cyclooxygenase inhibitors.
Therefore, this book not only covers the current state-of-the-art findings relevant
to gut mucosal injury and repair, but also provides the underlying conceptual
basis and knowledge regarding experimental therapeutics for gastrointestinal
mucosal injury-related diseases in the future.
We would like to take this opportunity to thank Karger Publishers, espe-
cially Dr. Thomas Karger and Mr. Peter Roth, who have made a great effort to
VII
make this book possible. We are indebted to all the contributors who have
shared and contributed their invaluable research experiences and knowledge
with us and to the medical community at large. And last but not least, we
express our sincere thanks to our families for their generous support through-
out the years.
Chi-Hin Cho, PhD

Jian-Ying Wang, MD, PhD
Preface VIII
Part 1: Epithelial Restitution
Cho C-H, Wang J-Y (eds): Gastrointestinal Mucosal Repair and Experimental Therapeutics.
Front Gastrointest Res. Basel, Karger, 2002, vol 25, pp 1–13
Roles of Extracellular Matrix and

Cytoskeleton in Intestinal Epithelial
Restitution
Jerrold R. Turner a, Marc D. Basson b,c
Department of a Pathology, The University of Chicago, Ill.; b Department of Surgery,
Wayne State University School of Medicine, Detroit, Mich., and c John D. Dingell VA
Medical Center, Detroit, Mich., USA
The intestinal mucosa maintains a barrier that limits exposure of the body
to potentially injurious luminal contents. In this role, the epithelium is subject
to attack by infectious, pharmacologic and other chemical agents from the
luminal (apical) surface. The epithelium may also be damaged by inflammatory
cells as they traverse intercellular junctions while migrating from the intersti-
tium to the intestinal lumen. Thus, epithelial injury may range from damage to
intercellular junctions to erosion and ulceration. Healing of mild epithelial
injury may only require reassembly of intercellular junctions. In contrast, more
severe injury requires spreading, migration and proliferation of epithelial cells
to adequately seal the wound and restore barrier function. In the latter case,
healing of mucosal wounds has been characterized as a coordinated process
(fig. 1) in which epithelial cells at the edge of the defect alter their shape and
spread across the wound. The observation that these processes of spreading and
migration can be regulated by matrix proteins likely reflects distinct differences
in matrix composition between denuded basement membrane and ulcerated
mucosa from a teleologic standpoint. Healing can also be accelerated by growth
factors in a manner that is largely independent of their mitogenic properties,
since part epithelial proliferation only contributes significantly to healing if the
wound cannot be resealed in the first 24 h. Ultimately, as migrating cells spread
and cover the damaged surface, they must also form intercellular junctions to
restore barrier function. Cell shape, migration and junction assembly depend on
Growth
Growth
factor
factors
receptors
Cytoskeletal
Signals Motility
motors
Integrins
Matrix (and non-
proteins integrin
receptors)
Fig. 1. Growth factors and matrix proteins are altered in mucosal injury. Multiple
mechanisms, including direct effects of cell wounding, release of chemotactic factors from
injured tissue, and migration of inflammatory cells into injured tissue with growth factor
secretion may lead to increased concentrations of growth factors at sites of injury. Wounding
both disturbs the architecture of the tissue matrix and may denude the basement membrane,
thus exposing the interstitial matrix below. Growth factors and matrix proteins each act on
intestinal epithelial cells via specific receptors that initiate intracellular signal transduction
cascades. These ultimately regulate cytoskeletal motors that provide the force necessary for
cell motility. However, it has now become clear that additional complexity and cross talk
resides in the interactions between growth factor- and matrix-initiated signal cascades and
the ability of such signals to feed back to and alter the expression, organization and activity
of the growth factor and matrix receptors themselves.
an intact cytoskeleton. While each of these events is critical to ultimate healing

of mucosal wounds, the focus of this chapter will be limited to the roles of the
extracellular matrix and cytoskeleton in migration and restoration of barrier
function.
In vivo Processes and in vitro Models
As noted above, migrating epithelial cells abandon their columnar shape

and assume a flattened squamoid appearance. At the simplest level, this shape
change allows cells to cover a larger surface area and, thus partially compensate
for cell loss following mucosal injury. In conjunction with the shape change,
these cells also exhibit decreased expression of differentiation markers. This
was initially interpreted as dedifferentiation. However, the process might be
better described as a phenotypic shift, or redifferentiation, towards a specialized
migratory phenotype that reorganizes membrane, cytoskeleton and matrix
receptors [1] and also displays altered matrix-dependent intracellular signaling
patterns [2].
Turner/Basson 2
One in vitro model of intestinal epithelial injury that has been well char-
acterized involves superficial injury of guinea pig ileal mucosal sheets mounted
in Üssing chambers [3]. Treatment of the mucosal surface with low concentra-
tions of nonionic detergent for 5 min caused denudation of the epithelium at
villous tips without destruction of the basement membrane or crypt epithelium.
As a result of this injury, resistance to passive ion flow decreased significantly
and flux of the extracellular markers mannitol and inulin increased 3- to 5-fold
immediately after injury [3]. The villus tips were again covered by columnar
absorptive cells within 2 h after injury, a time course that is incompatible with
enhanced cell proliferation. Analysis of the structural events occurring during
recovery showed that absorptive cells shouldering the foci of denudation
rapidly changed shape after injury: they became flattened and sent cell projec-
tions over the denuded basement membrane. By 60 min after injury, cells from
opposite shoulders of the denudation abutted, thus resealing the defect.
Paralleling these structural changes, transepithelial resistance, potential differ-
ence, and mannitol and inulin fluxes returned toward control values. These data
show that focal epithelial discontinuities in the small intestine may be rapidly
resealed. This resealing correlates with epithelial cell shape change and migra-
tion but occurs too rapidly for cell division to be significantly involved. Such
reparative processes may substantially limit the deleterious physiologic impact
of superficial forms of intestinal injury.
Further studies in the isolated mucosal sheet model using functional anti-
bodies to matrix proteins and receptors have suggested a role for specific
matrix proteins in the regulation of restitution [4, 5]. However, it is difficult to
do more fundamental mechanistic studies in such whole tissue models. Thus,
experimental models using cultured cell models have been used to further ana-
lyze migration. However, mechanistic studies of enterocyte migration have
often lagged behind studies of fibroblast or inflammatory cell migration. In
part, this is due to technical difficulties inherent in the use of well-differentiated
intestinal epithelial cell lines. We have focused on Caco-2 cells, derived from
a human colonic adenocarcinoma, as a useful model. Specifically, we have
used a Caco-2 BBe subclone that displays a microvillus brush border with a
full range of small intestinal hydrolases [6], tight junctions [7], polarized pro-
tein sorting [8], and physiologic ion and drug transport [9–13]. Caco-2 cells
resemble enterocytes morphologically and functionally and are frequently
used to model normal human small intestinal epithelium [14–18]. Our labora-
tories have studied this Caco-2 BBe subclone to elucidate mechanisms by
which matrix and cytoskeleton influence enterocyte biology [19, 20]. Others
have studied cell migration using human tumor-derived T84 [21] and HT29
[22] cell lines or nontransformed rat intestinal epithelial cell lines, such as
IEC-6 [23].
Extracellular Matrix and the Cytoskeleton 3

Direct and Independent Matrix Effects
The earliest phase of intestinal epithelial restitution involves rapid migra-

tion and spreading of epithelial cells to fill an experimental wound. Cell prolif-
eration is also an important component of the healing process, but lags behind
cell migration. Thus, the earliest phase of healing, restitution, occurs before pro-
liferation can contribute to the process, and many reductionist studies directed
toward understanding mucosal healing have therefore focused on epithelial
sheet migration.
For instance, using a ‘fence’ assay, in which Caco-2 cell migration outward
from a defined starting area is measured over 6 days, the effects of various sub-
strates and growth factors was determined [20]. Although initial attachment and
spreading of Caco-2 cells on surfaces coated with collagen types I and IV and
laminin were similar, cell proliferation and migration on collagen I-coated
surfaces was significantly greater than on the other substrates [20]. These data
suggest that disruption of the basement membrane and direct exposure of the
epithelial cell basal surface to collagen I, a dominant collagen in the intersti-
tium beneath the basement membrane, may trigger a maximal signal for migra-
tion that is not achieved on basement membrane-derived matrix components
such as laminin.
Although restitution occurs prior to and independently of cell prolifera-
tion, even this early phase of wound healing can be accelerated by mitogens
such as epidermal growth factor [20], hepatocyte growth factor [21], fibroblast
growth factors [23], and insulin-like growth factor-1 [22]. This subject is
covered in more detail elsewhere in this volume.
An interesting additional hypothesis is that the effects of growth factors and
matrix proteins are interactive rather than simply additive. That is, we [20, 24]
and others [25] have suggested that matrix proteins not only directly affect gut
epithelial cell motility but also influence the response to growth factors.
Consistent with this hypothesis, epidermal growth factor was able to further
stimulate Caco-2 migration over laminin in a fence migration assay, but migra-
tion over collagen I-coated surfaces did not increase [20]. This matrix-dependent
difference in the ability of epidermal growth factor to enhance Caco-2 cell
migration was not due to differences in cell proliferation, since epidermal
growth factor stimulated proliferation equivalently on both laminin and collagen I
and since the differential effects of epidermal growth factor on migration over
laminin and collagen I persisted when migration studies were performed in the
presence of mitomycin C to block cell proliferation. Instead, the effect appeared
to be caused by laminin- and EGF-dependent co-modulation of ␣2 integrin sub-
unit expression and organization at the lamellipodial edge of the migrating cells
[20]. The effects of transforming growth factor-␤1 on Caco-2 cell migration over
Turner/Basson 4
laminin and collagen I also diverged, similarly to those of epidermal growth fac-
tor [24]. Transforming growth factor-␤1 stimulated migration over laminin but
inhibited migration over collagen I [24], although the mechanism for this effect
appeared independent of the ␣2 integrin subunit. Thus, differential regulation of
migration over various matrices is not as simple as maximal stimulation by col-
lagen I and graded stimulation by laminin. Just as growth factors alter the
expression and organization of matrix receptors, so it seems also likely that
matrix proteins alter the expression and function of some growth factor recep-
tors and their downstream signals (fig. 1) [25]. Indeed, in many cell types,
including epithelial cells, protection from apoptosis by tonic growth factor
signaling seems to require matrix adhesion [26, 27].
Integrins
Differences in migration rates over different matrixes are, in part, deter-

mined by cell-matrix interactions. In the intestine and elsewhere, integrins are
specific mediators of these interactions. In response to experimental wounds
created by mechanical injury of cultured monolayers, T84 intestinal epithelial
cells synthesize several laminin isoforms and deposit laminin 5 in the matrix
that extends into the wound [28]. This is accompanied by intracellular redistri-
bution of the laminin-binding ␣3b1 integrin and ␣6-containing integrins [28].
Moreover, blocking antibodies targeted against components of laminin 5, or
␣3-, ␤1- and ␣6-containing integrins markedly inhibited cell migration and
wound closure [28]. Thus, laminin and ␣3-, ␤1- and ␣6-containing integrins are
critical to wound healing in this T84 cell model [28].
In addition to laminin, migration over collagen types I and IV, the primary
collagens of the interstitium and basement membrane, respectively, is important
both biologically and experimentally important to our understanding of epithe-
lial wound healing. However, the laminin-binding ␣3-, ␤1- and ␣6-containing
integrins do not mediate cell attachment to collagen. This falls to the ␣2␤1 and
␣1␤1 integrins. When the ␣1-, ␣2- and ␤1-containing integrins were evaluated
in Caco-2 cells, a decrease in surface ␣1 integrin subunits was induced by epi-
dermal growth factor in cells cultured on collagen I [20]. In contrast, surface
pools of ␣1 integrin subunits increased following treatment with epidermal
growth factor when cells were cultured on laminin [20]. Additionally, although
quantitative analysis did not reveal significant changes in the distribution of ␣2
integrin subunits between surface and intracellular pools, morphological analy-
sis demonstrated marked differences in ␣2 integrin subunit organization at the
leading edge of cells migrating over collagen I or laminin. Antibody blockade
of ␣2 integrin prevented epidermal growth factor-mediated stimulation of

migration over laminin, but did not inhibit migration over laminin in the
absence of epidermal growth factor or over collagen I in either the presence or
absence of epidermal growth factor. Thus, epidermal growth factor appears to
exert a matrix-specific effect on enterocyte migration by modulating both
expression and subcellular distribution of ␣2 integrin [20].
Signal Transduction Processes
When bound to extracellular ligands, integrin receptors cluster and activate

intracellular signaling cascades. The best characterized of these is activation of
the tyrosine kinase focal adhesion kinase (FAK), which co-localizes with clus-
tered integrins and cytoskeletal proteins to form focal adhesion complexes.
Although the biochemistry of FAK activation is incompletely understood, it
does appear to require the cytoplasmic domain of a ␤ integrin subunit, focal
adhesion assembly and association with stress fibers. We have recently shown
in the Caco-2 cell model that activation of specific ␤1 heterodimers with func-
tional antibodies directed against individual ␣ integrin subunits differentially
affects Caco-2 cell proliferation, differentiation and signaling [29]. However, it
remains unclear whether these observations reflect true differences in the activ-
ity of integrins with different ␣ subunits or simply differences in their cellular
localization and linkages to downstream signaling molecules within the cell.
Once activated, FAK autophosphorylation occurs and FAK also associates
with and phosphorylates other molecules including adapter proteins, e.g. pax-
illin and Shc, and src kinases, e.g. c-src and Fyn. In turn, c-src phosphorylates
FAK, forming an SH2-binding site for Grb2. This complex series of phospho-
rylation events initiates a series of downstream signaling events that appear to
vary among cell types, but which eventually lead to activation of the mitogen-
activated protein kinase (MAPK) cascades [30].
When tyrosine kinase activation was compared in cultures of confluent
contact-inhibited Caco-2 cells and subconfluent migrating Caco-2 cells with
lamellipodia, total tyrosine kinase activity was increased in migrating cells and
these increases were proportionally greater in the detergent-insoluble (mem-
brane and cytoskeletal) fraction than in the detergent-soluble (cytosolic) fraction
[31]. Interestingly, although the proportion of intracellular FAK and paxillin that
is phosphorylated increased in cells migrating across matrix proteins, total pro-
tein pools of FAK and paxillin decreased substantially. Thus, the actual amount
of phosphorylated FAK and paxillin in migrating Caco-2 cells was less than that
present in confluent contact-inhibited Caco-2 cells. Similarly, ERK1 and ERK2
were activated while total protein content was changed following adhesion to
collagens I or IV or laminin. The patterns of activation and expression of these
Turner/Basson 6
signal proteins in migrating cells compared with contact-inhibited confluent
cells varied in a matrix-dependent manner [2]. As might be expected, the reor-
ganization of these proteins associated with Caco-2 cell motility is also matrix-
dependent [32]. The pattern of activation observed in Caco-2 cells migrating
over collagen is consistent with recent descriptions of ERK activation during gut
mucosal wound healing in vivo [33]. In addition, immunofluorescent micro-
scopic evaluation suggests that motility is also associated with a reorganization
of intracellular FAK away from the lamellipodial edge. Thus, Caco-2 motility
seems to be associated with regulation of FAK at three different levels – phos-
phorylation, total protein content and subcellular distribution. The biological rel-
evance of this FAK and ERK activation is also apparent, as expression of a
dominant negative FAK construct decreased Caco-2 motility and inhibited
ERK2 activation [2]. Additionally, pharmacological ERK inhibition with
PD98059 inhibited Caco-2 motility. Thus, the FAK-ERK pathway appears to be
critically involved in regulating Caco-2 cell migration [2, 30]. However, the
complexities of spatial changes in FAK activation remain poorly understood.
Actomyosin Contraction
Although intracellular signaling processes and membrane receptors that

mediate cell-matrix interactions are critical to cell migration, the movement of
cells over a surface ultimately requires the generation of force. Numerous studies
in a broad range of cell types have demonstrated the essential role of actomyosin
contraction in this process. In epithelial cells, actomyosin contraction is princi-
pally mediated by conventional type II myosin. Each myosin II molecule is com-
posed of two heavy chains, two essential light chains and two regulatory light
chains (MLC). Phosphorylation of MLC at serine 19 by myosin light chain kinase
triggers a conformational shift in myosin that leads to actomyosin contraction.
Thus, MLC phosphorylation is a biochemical marker of actomyosin contraction.
MLC phosphorylation is also a likely target of at least some of the signal trans-
duction pathways described above. For example, myosin light chain kinase can be
phosphorylated by MAP kinases ERK1 and ERK2 [34]. MAP kinase phospho-
rylated myosin light chain kinase is more sensitive to calmodulin activation [34].
Thus, via activation of myosin light chain kinase, MAP kinase activation can
increase MLC phosphorylation and, potentially, increase cell migration rates.
MLC phosphorylation has been studied in motile fibroblasts using a poly-
clonal antibody that specifically recognizes MLC phosphorylated at serine 19
[35]. Phosphorylated MLC was distributed in a polarized pattern related to
the direction of cell movement, with increased MLC phosphorylation near
membrane ruffles at the leading edge and in a trailing region that included the

nucleus [35]. This suggests that actomyosin contraction at both poles of migrat-
ing fibroblasts is involved in cell migration [35]. Similarly, phosphorylated
MLC was also concentrated in actomyosin bundles at the leading edge of
migrating sheets of renal epithelial cells [35]. These observations suggest that
actomyosin filaments at the leading edge may be involved in force production
during epithelial cell migration. This is consistent with the observation that
polyamine-depleted cultures of IEC-6 intestinal epithelial cells have a markedly
disrupted distribution of myosin II and reduced migration over experimental
wounds [36]. However, that interpretation contrasts sharply with previous con-
clusions that the primary role of myosin II in fibroblast cell migration is at the
rear of the migrating cells [37]. How can these disparate conclusions then be
resolved? Do the discrepancies reflect cell-specific differences, or is the role of
myosin II in cell migration still incompletely understood? Better understanding
of these issues will be critical to our understanding of epithelial cell migration
and wound healing in general.
To further evaluate the role of myosin II in intestinal epithelial cell migra-
tion, we sought to determine whether enhanced MLC phosphorylation, and
therefore actomyosin contraction, is sufficient to accelerate intestinal epithelial
cell migration. We established clonal populations of Caco-2 cells that express a
truncated MLC kinase catalytic domain construct [38] under the control of the
tetracycline-regulated inducible expression system [39]. The tMLCK construct
(tMLCK) lacks the regulatory/inhibitory domain of MLC kinase [38]. Thus, the
kinase is active when the protein is expressed and does not require Ca2+ and
calmodulin for activation [38]. In preliminary studies we showed that washout
of doxycycline led to increased MLC kinase activity and MLC phosphorylation
in cell lysates [40]. We have used the fence migration assay described above to
directly evaluate the effect of tMLCK expression on Caco-2 cell migration over
collagen I. To avoid the potential that tMLCK expression might alter cell
attachment, Caco-2 cells were allowed to attach within the fence overnight in
the presence of doxycycline (to prevent tMLCK expression). After cell attach-
ment, the fence was removed and the cells transferred to media with or without
doxycycline to repress or induce tMLCK expression, respectively. tMLCK
expression reduced the surface area covered by Caco-2 migration by 10 ⫾ 3%
(p ⬍ 0.02, fig. 2). This change is quantitatively similar to the TGF-␣-induced
inhibition of cell migration over collagen I [24]. Although this quantitative sim-
ilarity between the effects of tMLCK expression and TGF-␣ treatment is likely
coincidental, it is interesting to note that available data do support a role for
actomyosin contraction as an intermediate in the effects of TGF-␣ on endothe-
lial monolayer integrity [41].
Immunofluorescent microscopy was used to evaluate actin filaments at the
leading edge of Caco-2 cells with or without tMLCK expression after 1 day of
Turner/Basson 8
1.0
(normalized to control ⫾ SEM)
Area of migration after 4 days
0.8
0.6
0.4
0.2
0.0
Doxycycline ⫹ ⫺ ⫹ ⫺
ML-7 ⫺ ⫺ ⫹ ⫹
Fig. 2. Effects of tMLCK expression and MLC kinase inhibition on Caco-2 migration.
Cell migation was allowed to proceed for 4 days after the induction (or repression) of
tMLCK expression and addition (or not) of 10 ␮M ML-7. Migration area for each sample
was then determined. Mean area ⫾ SEM (n ⫽ 7) is shown. For the migration in the absence
or presence of doxycycline (both without ML-7), the difference in migration was significant
(p ⬍ 0.02). Differences in migration in the absence or presence of doxycycline (both with
ML-7) were not statistically significant.
Doxycycline ⫹ ⫺ ⫹ ⫺
ML-7 ⫺ ⫺ ⫹ ⫹
Fig. 3. Effects of tMLCK expression and MLC kinase inhibition on actin distribution
in migrating Caco-2 cells. Fluorescent photomicrographs of actin distribution at the leading
edge of migrating cells 1 day after the initiation of migration, tMLCK expression (by
washout of doxycycline) and addition of 10 ␮M ML-7. Arrows indicate lamellipodia.
migration. In the absence of tMLCK expression, lamellipodia were easily iden-

tified, as were long stress fibers parallel to the leading edge (fig. 3). In sharp
contrast, lamellipodia were almost entirely absent in cells expressing tMLCK.
Moreover, the number of stress fibers parallel to the leading edge was markedly

decreased. These observations suggest that regulated actomyosin contraction
and relaxation is necessary for extension of lamellopodia. If true, this would be
at odds with other systems, for example Dictyostelium, where myosin II is
specifically absent from protrusive structures and myosin I is considered to be
of greater import [42]. However, it would be consistent with the idea that
myosin II is critically involved in the assembly of actin filaments and lamel-
lopodia at the leading edge of migrating epithelial cells [43].
To further evaluate the role of tMLCK expression in Caco-2 cell migra-
tion, we evaluated migration and actin assembly in cells with or without
tMLCK expression that were also treated with the MLC kinase inhibitor ML-7.
ML-7 inhibited cell migration over collagen I both in the absence and presence
of tMLCK expression (fig. 2). However, migration was inhibited to a greater
extent in cells without tMLCK expression, such that migration of cells with or
without tMLCK expression was comparable when ML-7 was present (fig. 2).
Thus, differences in cell migration induced by tMLCK expression are abolished
when both tMLCK and the endogenous MLC kinase are inhibited. Addition of
ML-7 to cultures without tMLCK expression not only limited forward progress
of the leading edge, but also led to the development of large gaps within the
surface over which the cells have migrated.
The morphology of the migrating monolayers also supports the idea that
MLC kinase inhibition with ML-7 may partially reverse the effects of tMLCK
expression. Lamellipodia at the leading edge of migrating cells were poorly
seen in cells with tMLCK expression, but addition of ML-7 to cells expressing
tMLCK restored lamellipodia and, to some degree, stress fibers assembly at the
leading edge. This result supports the hypothesis that the effects of tMLCK
expression on cell migration are at least partly due to excessive and inappro-
priate MLC phosphorylation.
Thus, actomyosin contraction, mediated by myosin II, appears to be critical to
migration of intestinal epithelial cells. The result that both inhibitors of MLC
kinase, e.g. ML-7, and activators of MLC phosphorylation, e.g. tMLCK, suggest
that neither is sufficient to activate cell migration. In contrast, it appears that a
complete cycle of actomyosin contraction and relaxation is necessary for cell
migration, and lamellipodium formation, to occur. Thus, for example, tMLCK
expression may prevent cell migration by inducing a rigor-like contraction of the
cytoskeleton and thereby preventing force generation and leading edge assembly.
Conclusions
Intestinal epithelial biology is profoundly affected by complex interactions

between the extracellular matrix and the actin cytoskeleton. In this chapter we
Turner/Basson 10
have focused on the roles of these effectors, both in concert and separately, in
cell motility and mucosal wound healing. Although the mechanisms by which
these processes occur remain incompletely understood, it is clear that integrin-
mediated cell-matrix adhesion is a central event. In addition to providing the
mechanical coupling between the cell and the substratum, binding of integrins
to extracellular matrix components may activate signaling cascades necessary
for cell migration to proceed. Examples include activation of focal adhesion
kinase and MAPK. However, differing subcellular localizations of integrin-
matrix binding interactions during motility are likely to also provide more com-
plex spatiotemporal signaling than is presently appreciated. It is also likely that
bidirectional interactions between myosin II-triggered actomyosin contraction
and integrin function and distribution remain to be defined. Thus, the interac-
tions between extracellular mediators, transmembrane proteins and cytoskeletal
components are likely to jointly influence cell phenotype during restitution.
This represents an important area for further study.
Acknowledgements
This work was supported in part by the National Institutes of Health (DK56121 and
DK61931 to J.R.T., and DK47051 and DK60771 to M.D.B.), the Children’s Research Center
of Michigan (J.R.T.) and the RAG and MERIT funding by the Department of Veterans
Affairs (M.D.B.).
References
1 Basson MD, Rashid Z, Turowski GA, West AB, Emenaker NJ, Sgambati SA, Hong F, Perdikis DM,
Datta S, Madri JA: Restitution at the cellular level: Regulation of the migrating phenotype. Yale
J Biol Med 1996;69:119–129.
2 Yu CF, Sanders MA, Basson MD: Human Caco-2 motility redistributes FAK and paxillin and acti-
vates p38 MAPK in a matrix-dependent manner. Am J Physiol Gastrointest Liver Physiol
2000;278:G952–G966.
3 Moore R, Carlson S, Madara JL: Rapid barrier restitution in an in vitro model of intestinal epithe-
lial injury. Lab Invest 1989;60:237–244.
4 Miller MA, Bunnett NW, Debas HT: Laminin mediates the restitution of rat gastric mucosa in
vitro. Exp Physiol 1994;79:647–659.
5 Moore R, Madri J, Carlson S, Madara JL: Collagens facilitate epithelial migration in restitution of
native guinea pig intestinal epithelium. Gastroenterology 1992;102:119–130.
6 Peterson MD, Mooseker MS: Characterization of the enterocyte-like brush border cytoskeleton of
the C2BBe clones of the human intestinal cell line, Caco-2. J Cell Sci 1992;102:581–600.
7 Turner JR: ‘Putting the squeeze’ on the tight junction: Understanding cytoskeletal regulation.
Semin Cell Dev Biol 2000;11:301–308.
8 Matter K, Brauchbar M, Bucher K, Hauri HP: Sorting of endogenous plasma membrane proteins
occurs from two sites in cultured human intestinal epithelial cells (Caco-2). Cell 1990;60:
429–437.

9 Riley SA, Warhurst G, Crowe PT, Turnberg LA: Active hexose transport across cultured human
Caco-2 cells: Characterization and influence of culture conditions. Biochim Biophys Acta 1991;
1066:175–182.
10 Turner JR, Lencer WI, Carlson S, Madara JL: Carboxy-terminal vesicular stomatitis virus G
protein-tagged intestinal Na+-dependent glucose cotransporter (SGLT1): Maintenance of surface
expression and global transport function with selective perturbation of transport kinetics and
polarized expression. J Biol Chem 1996;271:7738–7744.
11 Gonda T, Maouyo D, Rees SE, Montrose MH: Regulation of intracellular pH gradients by identi-
fied Na/H exchanger isoforms and a short-chain fatty acid. Am J Physiol 1999;276:G259–G270.
12 Fricker G, Drewe J: Enteral absorption of octreotide: Modulation of intestinal permeability by
distinct carbohydrates. J Pharmacol Exp Ther 1995;274:826–832.
13 Janecki AJ, Montrose MH, Zimniak P, Zweibaum A, Tse CM, Khurana S, Donowitz M:
Subcellular redistribution is involved in acute regulation of the brush border Na+/H+ exchanger
isoform 3 in human colon adenocarcinoma cell line Caco-2. Protein kinase C-mediated inhibition
of the exchanger. J Biol Chem 1998;273:8790–8798.
14 Peterson MD, Bement WM, Mooseker MS: An in vitro model for the analysis of intestinal brush
border assembly. II. Changes in expression and localization of brush border proteins during cell
contact-induced brush border assembly in Caco-2BBe cells. J Cell Sci 1993;105:461–472.
15 Beaulieu JF, Quaroni A: Clonal analysis of sucrase-isomaltase expression in the human colon
adenocarcinoma Caco-2 cells. Biochem J 1991;280:599–608.
16 Bement WM, Forscher P, Mooseker MS: A novel cytoskeletal structure involved in purse string
wound closure and cell polarity maintenance. J Cell Biol 1993;121:565–578.
17 Hauri HP, Matter K: Protein traffic in intestinal epithelial cells. Semin Cell Biol 1991;2:355–364.
18 Zweibaum A, Hauri HP, Sterchi E, Chantret I, Haffen K, Bamat J, Sordat B: Immunohistological
evidence, obtained with monoclonal antibodies, of small intestinal brush border hydrolases in
human colon cancers and foetal colons. Int J Cancer 1984;34:591–598.
19 Turner JR, Rill BK, Carlson SL, Carnes D, Kerner R, Mrsny RJ, Madara JL: Physiological regu-
lation of epithelial tight junctions is associated with myosin light-chain phosphorylation. Am J
Physiol 1997;273:C1378–C1385.
20 Basson MD, Modlin IM, Madri JA: Human enterocyte (Caco-2) migration is modulated in
vitro by extracellular matrix composition and epidermal growth factor. J Clin Invest 1992;
90:15–23.
21 Nusrat A, Parkos CA, Bacarra AE, Godowski PJ, Delp-Archer C, Rosen EM, Madara JL:
Hepatocyte growth factor/scatter factor effects on epithelia. Regulation of intercellular junctions
in transformed and nontransformed cell lines, basolateral polarization of c-met receptor in trans-
formed and natural intestinal epithelia, and induction of rapid wound repair in a transformed
model epithelium. J Clin Invest 1994;93:2056–2065.
22 Andre F, Rigot V, Thimonier J, Montixi C, Parat F, Pommier G, Marvaldi J, Luis J: Integrins and
E-cadherin cooperate with IGF-I to induce migration of epithelial colonic cells. Int J Cancer
1999;83:497–505.
23 Dignass AU, Tsunekawa S, Podolsky DK: Fibroblast growth factors modulate intestinal epithelial
cell growth and migration. Gastroenterology 1994;106:1254–1262.
24 Basson MD, Modlin IM, Flynn SD, Jena BP, Madri JA: Independent modulation of enterocyte
migration and proliferation by growth factors, matrix proteins, and pharmacologic agents in an in
vitro model of mucosal healing. Surgery 1992;112:299–307.
25 Wolpert SI, Wong MD, Wang JY, Bass BL: Epithelial-matrix interactions: Laminin downregulates
enterocyte EGF receptor and IGF-I receptor expression. J Surg Res 1996;63:345–348.
26 Frisch SM, Vuori K, Ruoslahti E, Chan-Hui PY: Control of adhesion-dependent cell survival by
focal adhesion kinase. J Cell Biol 1996;134:793–799.
27 Danilkovitch A, Donley S, Skeel A, Leonard EJ: Two independent signaling pathways mediate the
antiapoptotic action of macrophage-stimulating protein on epithelial cells. Mol Cell Biol 2000;
20:2218–2227.
28 Lotz MM, Nusrat A, Madara JL, Ezzell R, Wewer UM, Mercurio AM: Intestinal epithelial
restitution. Involvement of specific laminin isoforms and integrin laminin receptors in wound
closure of a transformed model epithelium. Am J Pathol 1997;150:747–760.
Turner/Basson 12
29 Basson MD, Emenaker NJ, Sanders MA: Alpha integrin subunits regulate human (Caco-2) intesti-
nal epithelial proliferation and phenotype. Cell Physiol Biochem 2000;10:27–36.
30 Sanders MA, Basson MD: Collagen IV-dependent ERK activation in human Caco-2 intestinal
epithelial cells requires focal adhesion kinase. J Biol Chem 2000;275:38040–38047.
31 Liu YW, Sanders MA, Basson MD: Human Caco-2 intestinal epithelial motility is associated with
tyrosine kinase and cytoskeletal focal adhesion kinase signals. J Surg Res 1998;77:112–118.
32 Yu CF, Basson MD: Matrix-specific FAK and MAPK reorganization during Caco-2 cell motility.
Microsc Res Tech 2000;51:191–203.
33 Pai R, Ohta M, Itani RM, Sarfeh IJ, Tarnawski AS: Induction of mitogen-activated protein kinase
signal transduction pathway during gastric ulcer healing in rats. Gastroenterology 1998;114:
706–713.
34 Klemke RL, Cai S, Giannini AL, Gallagher PJ, de Lanerolle P, Cheresh DA: Regulation of cell
motility by mitogen-activated protein kinase. J Cell Biol 1997;137:481–492.
35 Matsumura F, Ono S, Yamakita Y, Totsukawa G, Yamashiro S: Specific localization of serine 19
phosphorylated myosin II during cell locomotion and mitosis of cultured cells. J Cell Biol 1998;
140:119–129.
36 Wang JY, McCormack SA, Johnson LR: Role of nonmuscle myosin II in polyamine-dependent
intestinal epithelial cell migration. Am J Physiol 1996;270:G355–G362.
37 Conrad PA, Giuliano KA, Fisher G, Collins K, Matsudaira PT, Taylor DL: Relative distribution
of actin, myosin I and myosin II during the wound healing response of fibroblasts. J Cell Biol
1993;120:1381–1391.
38 Guerriero V Jr, Russo MA, Olson NJ, Putkey JA, Means AR: Domain organization of chicken
gizzard myosin light chain kinase deduced from a cloned cDNA. Biochemistry 1986;25:
8372–8381.
39 Gossen M, Bujard H: Tight control of gene expression in mammalian cells by tetracycline-
responsive promoters. Proc Natl Acad Sci USA 1992;89:5547–5551.
40 Turner JR, Guerriero V Jr, Black ED, Haelewyn K: Regulated expression of the myosin light chain
kinase catalytic domain increases paracellular permeability and alters tight junction structure.
Gastroenterology 2000;118:A432.
41 Hurst VI, Goldberg PL, Minnear FL, Heimark RL, Vincent PA: Rearrangement of adherens junc-
tions by transforming growth factor-␤1: Role of contraction. Am J Physiol 1999;276:L582–L595.
42 Jay PY, Elson EL: Surface particle transport mechanism independent of myosin II in
Dictyostelium. Nature 1992;356:438–440.
43 Small JV, Herzog M, Anderson K: Actin filament organization in the fish keratocyte lamel-
lipodium. J Cell Biol 1995;129:1275–1286.
Marc D. Basson, MD, PhD, Chief, Surgical Service (11S),

John D. Dingell VA Medical Center, 4646 John R Detroit, MI 48201–1932 (USA)
Tel. +1 313 576 3598, Fax +1 313 576 1002, E-Mail marc.basson@med.va.gov

Cytokines in Restitution
James Yooa, Margaret M. Lotza, Jeffrey B. Matthewsb
a
Beth Israel Deaconess Medical Center, Department of Surgery,
Boston, Mass. and bUniversity of Cincinnati Medical Center,
Department of Surgery, Cincinnati, Ohio, USA
Cytokines play a critical role in the response to mucosal injury, regulat-

ing a complex set of processes that are involved in restoring normal mucosal
architecture. An intact mucosa is essential to the fundamental physiologic
processes performed by the gastrointestinal tract, including nutrient absorp-
tion, electrolyte and water homeostasis, immune regulation, and barrier func-
tion. The epithelial lining of the gut selectively regulates the movement of
substances from the lumen to the bloodstream and vice versa. The junctional
complex between epithelial cells allows for this selectivity, preventing harm-
ful luminal contents including bacteria, toxins and noxious chemicals from
gaining access to the internal environment. Mucosal injuries disrupt this bar-
rier and increase host susceptibility to infectious agents and other potentially
pathogenic substances.
Restoration of barrier function requires rapid mucosal repair, and is char-
acterized by an early and late phase. The early phase occurs in the absence
of cellular proliferation and was termed restitution by Silen and Ito [1] in
1985. This homeostatic process occurs over the course of minutes to hours to
re-establish epithelial continuity. It involves the sloughing of damaged cells fol-
lowed by epithelial cell flattening, spreading, and migration from the wound
margins to rapidly restore mucosal integrity. Cell division, proliferation, and
re-differentiation characterize the later phase of mucosal repair.
Restitution involves a coordinated series of events involving alterations in
the underlying extracellular matrix (ECM) as well as changes within the intesti-
nal epithelium to ultimately promote resealing by epithelial cells. These cells
undergo phenotypic changes as they temporarily switch functional states,
advancing over the epithelial defect to re-establish cellular continuity and
restore the gastrointestinal tract lining. Cytokines appear to influence, if not
directly regulate, the events involved in restitution, allowing for communication
between cell populations as they interact to coordinate the many steps of early
mucosal repair.
Cytokines Are Secreted at Sites of Mucosal Injury
Various lines of evidence suggest that restitution occurs on a routine basis

in the gastrointestinal tract. Depending on the severity of the damage, immune
and mesenchymal cells may or may not be stimulated to release cytokines.
Mucosal epithelial cells quickly restitute denuded villous tips as cells apoptose
during the last stage of crypt to villous migration. These microscopic defects
probably do not provoke much reaction by inflammatory cells or even the
underlying mesenchyme. In cases of mild injury, despite the fact that minimal
damage is noted in the basal lamina, myofibroblasts are neurally stimulated to
contract the villous, greatly diminishing the denuded surface area [2, 3]. Thus,
even in these superficial injuries, non-epithelial cells are activated, suggesting
the potential for lamina propria immune cells to release cytokines capable of
mediating restitution. Finally, conditions such as peptic ulcer disease or inflam-
matory bowel disease can involve deeper layers of the gut wall. In these cases,
there is an immediate reaction by inflammatory and immune cell populations,
along with intestinal epithelial cells, leading to the production of an array of
cytokines and peptides that modulate host responses to injury but which also
mediate mucosal healing (fig. 1, 2).
A number of different cell populations participate in restitution, leading to
changes in the epithelial cell phenotype as well as alterations in the ECM to
promote cell migration. The release of cytokines and growth factors in response
to mucosal injury occurs by mesenchymal cells such as fibroblasts,
macrophages, neutrophils, lymphocytes, myofibroblasts and epithelial cells.
Mesenchymal cells are known to secrete keratinocyte growth factor (KGF),
a member of the fibroblast growth factor (FGF) family that is involved in cell
migration as well as proliferation. Macrophages and mononuclear cells secrete
many pro-inflammatory cytokines, such as interleukin (IL)-1, IL-2, tumor
necrosis factor (TNF-␣) and transforming growth factor (TGF), which have all
been shown in some model systems to contribute to restitution. Fibroblasts are
involved in the production of FGF, while epithelial goblet cells secrete intesti-
nal trefoil factor (ITF), which stimulates cell migration through several mech-
anisms. Myofibroblasts are located immediately subjacent to the epithelium in
the lamina propria and, in addition to their role in wound contraction, are
involved in the secretion of hepatocyte growth factor (HGF), which also has
Cytokines in Restitution 15
Injury
ITF ITF
ITF
IL-1
TNF
IFN-␥
FGF IL-1
IL-8 IL-2 IL-8
KGF TNF
TGF-␤ TGF
TNF IFN-␥
IL-2
Fig 1. Cytokines (IL-1, TNF-␣ and IFN-␥) are released at sites of mucosal damage.
In response to tissue injury, many different cell populations release cytokines that are
involved in both immune regulation/inflammation and mucosal healing. Fibroblasts
secrete FGF and KGF. Macrophages and mononuclear cells secrete IL-1, IL-2, TNF-␣, and
TGF. Epithelial goblet cells are involved in the secretion of ITF, while myofibroblasts
secrete HGF.
TNF-␣ IL-17
IL-1␤
ITF
TNF-␣
EGF [Zonula occludens]
EGFR
Rho HGF
[Cytoskeleton] [Zonula adherens]

IL-1␣
IFN-␥
Integrin
[ECM]
TGF-␣ TGF-␤
EGF
IFN-␥ IL-1␤
Fig. 2. Cytokines regulate restitution by leading to changes in cell structure, cell-cell

contacts, and cell-ECM interactions. Changes in cell structure involve alterations in the actin
cytoskeleton. Changes in cell-cell contacts involve effects on components of the zonula
occludens (occludin, claudin, ZO-1, ZO-2, ZO-3) and components of the zonula adherens
(cadherin, ␣-, ␤- and ␥-catenin), which associate with the underlying cytoskeleton. Changes
in cell-ECM matrix dynamics involve the regulation of integrins.
Yoo/Lotz/Matthews 16
pro-migratory effects. The various cytokines and growth factors secreted from
these cells lead to functional and phenotypic changes within the epithelial cell
and invoke changes to the ECM, allowing for the initiation of cell migration.
Cytokines Stimulate Migration
The contributory role of cytokines in restitution has been demonstrated in

multiple in vitro and in vivo model systems. Well-established wounding proto-
cols have been utilized to simulate mucosal injury using either intestinal epithe-
lial monolayers or various animal models. Many cytokines and peptides have
been shown through these model systems to enhance epithelial cell migration
following injury. While little is known regarding the mechanism whereby a spe-
cific cytokine affects restitution, in many instances the effect may be indirect,
working through the effects of growth factors. Key growth factors and
cytokines that appear to facilitate cell migration include HGF, platelet-derived
growth factor (PDGF), epidermal growth factor (EGF), KGF, insulin-
like growth factor (IGF), TGF-␣, TGF-␤, IL-1␣, IL-1␤, IL-2, IL-4, IL-8,
IL-15, IFN-␥, TNF-␣, and ITF. While trefoil peptides regulate epithelial cell
migration through apically located epithelial cell surface receptors, cytokines
and growth factors mediate their effects through basolateral receptors.
Subsequent cell signaling events, which include pathways involving protein
kinase C (PKC), the Rho family of GTPases, phosphoinositol-3-kinase (PI3K),
phospholipase C (PLC) and mitogen-activated protein kinase (MAPK), lead to
a phenotypic shift of the epithelial cells. Their primary function changes from
being cells specialized in vectorial transport and the maintenance of barrier
function to becoming migratory cells whose main function is to seal the
mucosal defect, re-establish cell-cell contacts and tight junctions, and thereby
restore epithelial continuity.
IL-2
IL-2 is a key cytokine involved in the initiation of the immune response,

and is found in abundance at sites of inflammation and mucosal injury. It is
secreted mainly by T cells, leading to the activation of other T cells as well as
cell populations involved in the inflammatory response. In addition to its fun-
damental role in immune responses, IL-2 has also been found to stimulate
epithelial cell restitution. Functional IL-2 receptors have been found on multi-
ple intestinal epithelial cell lines, including IEC-6, IEC-17, HT-29, Caco2, and
T84 cell lines [4, 5]. Exogenous IL-2 stimulates epithelial cell restitution in
IEC-6 cells in the absence of proliferation. Its ability to stimulate cell migration
appears to be secondary to the induction of TGF-␤1 expression, a growth factor
with known pro-migratory effects [4]. IL-2 leads to an elevation in both TGF-␤1
mRNA and in the amount of bioactive peptide. The ability of IL-2 to stimulate
restitution can be blocked by either functional antibody to TGF-␤1 or to the
IL-2 receptor ␤ chain. This suggests that IL-2, and other pro-inflammatory
cytokines that are released in response to mucosal injury, not only mediate the
immune response but also regulate elements of mucosal repair.
IL-1␤ and IFN-␥
IL-1␤ and IFN-␥ are found at sites of mucosal injury or inflammation.

Similar to IL-2, they have also been shown to enhance restitution through
an indirect mechanism involving growth factors. IL-1␤ and IFN-␥ along
with TGF-␣, EGF, and the aforementioned IL-2, all enhance restitution and
have been shown to increase TGF-␤ production in the IEC-6 cell line [4, 6].
TGF-␤ is an important mediator of the restitution process, and appears to be
necessary for the initiation of epithelial cell motility following mucosal injury
[6–10]. It is secreted by intestinal epithelial cells and mononuclear cells in
an inactive form bonded to a pro-peptide, requiring post-translational modifi-
cation (cleavage of the non-covalent bond) for activation. Therefore, its pro-
duction and activity can be regulated by cytokines at multiple levels. It is
constitutively produced, which may explain how epithelial cells can respond so
rapidly to injury, and it is upregulated after injury (as evidenced by elevated
mRNA levels in wounded monolayers [6]). This upregulation occurs at least
in part through the effects of the cytokines IL-1␤ and IFN-␥, and the growth
factors TGF-␣ and EGF, which increase the amount of bioactive peptide
available [6, 7].
TGF-␤ appears to mediate the motogenic effects of a number of cytokines.
The importance of TGF-␤ in initiating migration is supported by the fact that
anti-TGF-␤1 antibody prevents cell migration following wounding in IEC-6
cells [7]. Further evidence for the key role of TGF-␤1 on epithelial cell migra-
tion comes from studies showing that anti-TGF-␤1 antibody also blocks the
effect of other pro-migratory cytokines (TGF-␣, EGF, IL-1␤, IL-2, IFN-␥) on
restitution. Taken together, these findings suggest that TGF-␤1 may be the final
common pathway for these cytokines, or at least a necessary co-factor, in their
pro-migratory function. In the IEC-6 and T84 cell lines, TGF-␤3, released by
subepithelial myofibroblasts, also enhances restitution [10]. Myofibroblast-
conditioned media was found to stimulate migration post-wounding, while
inhibiting cellular proliferation. Antibody to TGF blocked this effect, while
antibody to the TGF-␤1 and TGF-␤2 subtypes had no effect, suggesting that, in
these cell lines, the TGF-␤3 subtype was the likely contributing factor in
promoting migration [10].
The pro-migratory effect of IL-1␤ was also demonstrated in a canine
gastric cell line. Interestingly, in this model system the motogenic effects of
IL-1␤ appear to be linked to a different growth factor, EGF [11]. Mab-528, a
monoclonal antibody to the EGF receptor (EGF-R), completely blocked IL-1␤-
stimulated migration, suggesting that the downstream cell signaling pathways
associated with EGF are required for IL-1␤-induced cell migration. Multiple
studies using different cell lines and in vivo model systems (MSIE, IEC-6,
rabbit duodenum, canine oxyntic mucosa cells) have demonstrated a role for
EGF in restitution [11–14]. In the gastrointestinal tract, EGF is produced by the
salivary glands, the duodenum (Brunner’s glands) and the pancreas, and it is
found both luminally and systemically [14]. Using the immortalized murine
small intestinal cell line (MSIE) and a well-established epithelial wounding
protocol, a study by Polk [12] demonstrated that the growth factor EGF
promotes cell migration through a mechanism involving PLC and PKC.
IL-1␣ and TNF-␣
Both IL-1␣ and TNF-␣ may indirectly stimulate restitution through the
increased production of HGF, also known as scatter factor. HGF is a hepato-
trophic factor known to promote liver regeneration and has been shown to
stimulate cell migration in gastric epithelial cells [15, 16]. A study by Takahashi
et al. [15] demonstrated that prostaglandins (PGE1, PGE2), IL-1␣ and TNF-␣ all
stimulate HGF protein synthesis in human gastric fibroblasts, though PGE1 and
PGE2 had the most profound effects. They speculated that the protective effect
of prostaglandins on NSAID-induced gastric mucosal injury might be, in part,
mediated by their ability to stimulate the release of HGF by gastric fibroblasts.
PGE1, a cytoprotective agent that is known to accelerate restitution, increases
mRNA expression of HGF by fibroblasts as assessed by Northern analysis, and
its protective effects are inhibited by antibody to HGF. This suggests that the
ability of PGE1 to stimulate cell migration may be due, in part, to the release of
HGF. IL-1␣ and TNF-␣ are likely to have similar effects mediated through HGF.
In another study by Takahashi [16], using an in vitro round wound restitution
model on gastric epithelial cell monolayers, he demonstrated that HGF produced
by gastric fibroblasts enhanced restitution. Fibroblast-conditioned media stimu-
lated restitution, and this effect was inhibited by the addition of anti-HGF anti-
body. Similarly, a study by Nursat et al. [17] demonstrated pro-migratory effects
of HGF on the T84 human intestinal cell line using a similar round wound
restitution model. In addition, IL-1 and TNF increase the production of IL-8,
which has also been shown to have pro-migratory effects (see below).
KGF
KGF-2 (FGF-10) is a member of the FGF family. It is secreted by fibro-

blasts and targets epithelial cells, which bear its receptor. In a study by Han et al.
[18] using Lewis rats, they demonstrated that KGF-2 attenuates indomethacin-
induced epithelial damage. In parallel wounding studies using Caco2 cells,
they found that KGF-2 stimulates epithelial cell migration to the same extent
as TGF-␤. It also increases COX-2 mRNA levels in Caco2 cells, potentially
leading to the production of cytoprotective prostaglandins. Elevated levels of
PGE2 were seen in both Caco2 supernatants as well as in the small intestinal
tissue from Lewis rats, a finding that supports this hypothesis.
IL-8
In response to mucosal injury, IL-8, like many other cytokines, appears to

be involved in the dual processes of inflammation and subsequent mucosal
repair. Intestinal epithelial crypt cells produce IL-8 in response to injury or in
the presence of pro-inflammatory cytokines such as TNF-␣ and IL-1. It is
secreted from the basolateral surface where it acts as a leukocyte chemoattrac-
tant. Neutrophils localize to sites of injury and then secrete a host of other pro-
inflammatory cytokines that may also contribute to restitution. Many have been
discussed already and include TGF-␤, TNF-␣, IL-2, and IFN-␥. In fact, using
an in vitro co-culture model with IEC-6 cells, wounded monolayers showed
enhanced restitution in the presence of peripheral blood mononuclear cells
(PBMC) [19]. This effect could be blocked with neutralizing antibody to
IL-2 and IFN-␥, suggesting that these are important cytokines in the PBMC-
mediated regulation of restitution.
In addition to its role in neutrophil chemotaxis, IL-8 has also been shown
to promote cell motility through a more direct mechanism. This was demon-
strated using a circular wound model in the human colon cancer cell line
LIM1215 [20]. Exogenous IL-8 stimulated migration in this cell line in a dose-
dependent fashion. TNF-␣ which upregulates the production of IL-8, also stim-
ulated migration. Interestingly, the motogenic effects of TNF-␣ were blocked in
the presence of IL-8 antibody. In this model system the pro-migratory effects
of TNF-␣ may be secondary to subsequent IL-8 production or to a signaling
event downstream of IL-8.
ITF
ITF is a member of the trefoil peptide family (which also includes pS2 and
spasmolytic polypeptide (SP)) and is an important regulator of restitution.
While pS2 and SP are found mainly in the stomach, ITF is found predominantly
in the small and large intestine. It is secreted by epithelial goblet cells and
remains in the lumen within the overlying mucinous layer in contact with the
apical side of the intestinal epithelium. It works synergistically with mucinous
glycoproteins in exerting its effects on epithelia. ITF has been shown to be
upregulated in areas of mucosal injury such as at the edges of mucosal ulcera-
tion and it promotes restitution through several mechanisms [21, 22]. Its main
functional contribution appears to be in the initiation of migration following
mucosal injury. In IEC-6 cells, ITF-induced migration is associated with ERK
phosphorylation and can be blocked by the MEK inhibitor PD98059, implicat-
ing this signaling cascade in this process [23].
Cytokines Promote Other Pro-Migratory Factors
Cytokines also lead to the production of other factors that may be involved
in regulating restitution. There appears to be some cross talk between various
cytokines and the end products of the lipoxygenase and cyclooxygenase path-
ways. A study by Zushi et al. [21] demonstrated that in the IEC-6 and Caco2
cell lines the accelerating effects of EGF, HGF and TGF-␤ on epithelial cell
restitution are, in part, mediated by the production of prostaglandins. Their
wounding model demonstrated that the enhancing effects on resealing speed
were reduced by roughly 50–60% in the presence of piroxicam, which inhibits
prostaglandin synthesis. In the IEC-6 cell line, lysophosphatidic acid (LPA), a
naturally occurring phospholipid that is released by growth factor-stimulated
fibroblasts, also enhances cell migration and inhibits cell proliferation through
a TGF-␤ independent pathway [24]. The effects of LPA were confirmed in a
Sprague-Dawley rat model, showing decreased trinitrobenzene-induced colitis
with exogenous administration of LPA.
Cytokines Regulate Early and Late Phases of Mucosal Repair
Cytokines may regulate the process of mucosal healing by coordinating the

timing of the early and late phases. The early phase is characterized by rapid
epithelial cell migration in the absence of proliferation to re-establish epithelial
continuity, while the late phase involves cell division and proliferation to fill in
the defect and restore mucosal architecture. Cytokines are released immediately
at sites of injury, and are at least partially responsible for the rapid response that
occurs. The appropriate timing of these processes is important. While restitu-
tion begins as soon as 15 min after mucosal injury, cell proliferation generally
begins 12 h after injury, and continues for 1–2 days [16, 25]. Some cytokines
involved in the promotion of restitution are also anti-proliferative, and this
may account for the delay in the onset of cell proliferation. This may serve
to regulate the timing of these two distinct but overlapping phases of wound
healing allowing for an orderly, coordinated repair response. Cytokines with
anti-proliferative properties include TGF-␤. Multiple model systems have
demonstrated that TGF-␤ promotes epithelial cell migration while at the same
time inhibiting cell proliferation [10].
Cytokines Lead to Changes in Cell Structure
Following mucosal injury, epithelial cells abutting the damage rapidly

spread into the denuded area. Epithelial migration requires dynamic changes in
epithelial cell structure. Cytokines regulate various functional and phenotypic
changes that occur in epithelial cells that are necessary for the initiation of
migration. Within the epithelial cell there are changes in the cytoskeleton
involving a rearrangement of actin fibers. Cytokine-stimulated growth factors
have been shown to alter actin stress fibers and focal adhesion molecules nec-
essary for motility [13]. There are also changes in cell surface proteins involved
in cell-cell contact that regulate tight junction formation. Additionally, the
epithelial cell interacts dynamically with its underlying ECM, as it undergoes
reversible interactions with matrix components allowing it to advance over this
scaffolding.
In culture systems that reproduce mucosal injury, such as the T84 and
Caco2 cell lines, the integrity of the undamaged epithelium is maintained and
cooperative cell flattening heals the defect. Cells surrounding the damage flatten
and those immediately adjacent to the wound extend lamellipodia, broad, thin
cytoplasmic protrusions, to cover the mucosal defect [26]. Lamellipodia appear
to generate the force required to induce cell flattening. Actin filaments inside
lamellipodia engage integrins, cell surface proteins that adhere to underlying
matrix proteins. Adhesive strength between integrins and the stationary matrix
generates the traction, transmitted to the cell cytoplasm via actin filaments, to
stretch the epithelial cells over the matrix.
Cytokines appear to influence the cytoskeletal events that occur as the
epithelial cell begins to spread and migrate. Santos et al. [13] showed that EGF
has important regulatory effects on actin filaments leading to increased stress
fiber formation. The resultant cytoskeletal rearrangements are necessary for
the initiation of motility in IEC-6 cells. Alterations in actin fibers lead to cell
flattening and the formation of lamellipodia and membrane ruffling, which are
early phenotypic changes of migratory cells. Santos found that this effect of
EGF is mediated through Rho GTPases, and that specific inhibition of Rho pro-
teins inhibited motility following wounding [13]. In fact, the EGF-R is directly
linked to the cytoskeleton [14], which would explain how EGF affects actin fiber
rearrangements. Other studies have demonstrated a role for prostaglandins and
leukotrienes in the EGF-induced changes that occur in the cytoskeleton [27].
Cytokines Alter Cell-Cell Interactions
In an intact monolayer, epithelial cells form a selectively permeable barrier

through the formation of tight junctions and also adhere to each other via
adherens junctions. These cell-cell contacts stabilize the epithelium and are
necessary for its normal functioning. As mentioned previously, gastrointestinal
epithelial cells have not been observed to detach from each other and migrate
as single cells during restitution, suggesting that cell-cell contacts are not bro-
ken during resealing. However, the dynamics of the resealing process, during
which epithelial cells alter their morphology and spread, dictate that the inter-
face between individual cells is not rigid. Therefore, during restitution cell-cell
contacts most likely are breaking and reforming continuously. Tight junctional
complexes are regulated, in part, by various cytokines. Claudins, a family of at
least 24 different membrane proteins that are integral components of tight junc-
tions, are regulated by IL-17 in T84 cells [28]. Similarly, TNF alters epithelial
permeability in the HT29/B6 cell line through alterations in tight junction
structure [29]. Cytokines that are released at sites of mucosal injury regulate
cell-cell contact to allow for the initiation of cell migration. Once the epithe-
lial cells have migrated to fully cover the mucosal defect, the physical barrier is
re-established.
ITF has been shown to alter cell surface proteins involved in tight junction
regulation [22, 30–32]. Cadherins are a group of calcium-dependent surface
proteins that mediate intercellular adhesion, and their expression is regulated by
ITF. Their cytoplasmic domains are in contact with members of the catenin
family, which establish communication with the cytoskeleton. E-cadherin is a
120-kD peptide located at the adherens junction, and is the key component
involved in cell-cell adhesion. In HT-29 cells, ITF leads to loss of E-cadherin
from the membrane surface and its redistribution to the cytoplasm [32]. This
redistribution not only disrupts the cell-cell contact formed by neighboring
cadherin molecules, but it also disrupts the cadherin-catenin complex, which
links the adherens junction to the cytoskeleton. These changes de-stabilize
cell-cell contacts and promote migration.
EGF may inhibit cadherin function through a similar mechanism. The
EGF-R, a receptor tyrosine kinase, has been shown to be involved in phospho-
rylation of the cadherin-catenin complex, leading to disruption of the adherens
junction [31]. EGF and ITF both increase the association of EGF-R with
␤-catenin. In HT-29 cells, ITF leads to phosphorylation of the EGF-R and
␤-catenin, leading to decreased surface expression of E-cadherin [31]. This
downregulation of surface E-cadherin expression has also been shown to occur
in response to HGF [32], which phosphorylates ␣- and ␥-catenin. EGF has
also been shown to phosphorylate the tyrosine residue of ␤- and ␥-catenin in
Caco2 cells. This leads to the disruption of cell-cell contacts through the loss of
E-cadherin from junctional complexes, resulting to cell scattering [32].
In the Caco2 cell line, Th1 cytokines appear to regulate the cadherin-catenin
complex as well [33]. TNF-␣, IL-1, and IFN-␥ all lead to a significant reduc-
tion in E-cadherin expression, while TNF-␣ also reduces ␤-catenin expression,
though to a lesser degree. TNF-␣ led to a reversible decrease in both membrane
and cytoplasmic E-cadherin as well as in the membrane-associated ␤-catenin,
though cytoplasmic ␤-catenin was unaffected. ␤-Catenin, a 92-kD protein,
plays an important role not only in the regulation of cell-cell adhesion, but also
in downstream cell signaling and in communication with the cytoskeleton.
Cytokines Alter Cell-Matrix Interactions
During restitution, gut epithelial cells require an intact basal lamina over
which they can extend themselves, making the interaction between gut epithe-
lial cells and the underlying matrix a key target for the mediating actions of
cytokines. Superficial injuries leave an intact basement membrane, and are
characterized by a more rapid restitution response compared to deeper defects,
where the basement membrane has been disrupted. This suggests that the
condition of the underlying matrix is important in regulating cell migration. In
fact, the ECM composition can be affected by cytokines [34, 35]. There is
also a change in the interactions between epithelial cells and the underlying
matrix, as they extend and retract lamellipodia over the denuded basement
membrane.
Components of the ECM influence cell migration. The basement membrane
is composed of collagen type IV, fibronectin, glycosaminoglycans and laminin,
some of which have the potential to be bound to growth factors. Earlier work
showed that antibodies to collagen IV and fibronectin but not laminin 1 were
able to inhibit restitution in vivo, indicating that they are important substrates for
gut epithelial migration [36]. More recent evidence indicates that the laminin
family of matrix proteins is also important in restitution. In vitro healing assays
suggest that laminin 5 is particularly conducive to wound resealing [37].
Cytokines may also mediate restitution by regulating matrix production by
either the epithelial cells or the underlying lamina propria cells. Both the
epithelium and the underlying mesenchymal cells synthesize the basal lamina
[38]. Thus it is important to examine the effects of growth factors on both cell
types. In a study by Goke et al. [34], collagen type IV, fibronectin and laminin
1 were found to be upregulated by TGF-␤ in an IEC-6 cell line following
wounding. In the absence of TGF-␤, these proteins were downregulated. In
addition, fibronectin and laminin distribution was altered in response to wound-
ing, an affect that was also inhibited by the presence of TGF-␤. The effect of
TGF-␤ on matrix composition has been demonstrated in other model systems
as well, including in vitro guinea pig gastric mucosa [8]. TGF-␤ is also able to
upregulate collagen and fibronectin synthesis by fibroblasts [39, 40]. These
results suggest that TGF-␤ may promote cell restitution in part by sustaining the
ECM composition and structure to create a substrate that is optimal for cell
migration during restitution.
Integrins are the main regulators of cell-matrix interactions and are impor-
tant targets of cytokines. They are transmembrane heterodimers that communi-
cate with components of the ECM through interactions with their ␣ and ␤
chains. Integrins are involved in matrix adhesion and the regulation of matrix
formation and degradation. They also serve as points of traction for cell migra-
tion. Integrins associate with the cytoskeleton through ␣-actinin, talin and
vinculin, and they localize to focal adhesions along with these cytoskeletal
elements in areas of close contact with the ECM. Through this contact, epithe-
lial cells are able to monitor changes that may be occurring in their surround-
ing environment. It appears likely that cell restitution involves integrin
reorganization, and that integrins are an important target of cytokine-mediated
signaling.
Evidence from other fields has shown that cytokines operate on integrins.
In many instances, this may be exerted via their effects on specific growth
factors, particularly TGF-␤ and EGF. For instance, in epidermal wound healing,
TGF-␤ induces the expression of avb6 integrin on keratinocytes, allowing
them to migrate on plasma fibronectin newly introduced by clot formation [41].
EGF upregulates the cell surface receptor for urokinase in embryonic kidney
cells which in turn is able to inhibit ␤1 integrin function, reducing adhesion to
fibronectin [42]. These types of regulatory events could very well occur during
restitution, especially in the cases where the epithelium needs to reseal after a
deeper injury. In these circumstances, clot formation would briefly present the
epithelial cells with new matrix molecules with which to interact transiently.
Several studies show that intestinal cells respond to cytokines by altering
integrin interactions. In Caco2 cells, both EGF and TGF-␣ have been shown to
regulate integrin function [32, 43]. These growth factors increase migration of
Caco2 cells on laminin 1 and collagen. EGF affects integrin ␣2 subunit expres-
sion and localization [43]. The integrins ␣2␤1 and ␣3␤1, which are receptors for
laminin and collagen, are upregulated by EGF and TGF-␣, which stimulate
Caco2 cell migration in this system. Functional antibody to the ␣2 and ␤1 inte-
grin chains inhibits EGF and TGF-␣ stimulated Caco2 cell migration, impli-
cating these integrins as key targets of cytokine regulation [43]. In the HD3
colon cell line, TGF-␤1 upregulates the ␣2␤1 integrin protein in precursor cells.
It appears to do this by mediating formation of the ␤1 integrin chain through a
process involving ras proteins [44]. Ras stimulates conversion of the ␤1 integrin
precursor to its mature active form through ␤1-chain glycosylation, a process
that can be accelerated with exogenous TGF-␤1 and inhibited with antibody to
this factor. This post-translational modification also affects surface expression
of this collagen receptor.
Altered cell interactions with the ECM involve the careful regulation of
proteases such as urokinase (u-PA) and the matrix metalloproteinases, as well
as the tissue inhibitors of matrix metalloproteinases (TIMP). The receptor for
urokinase (u-PAR) has been shown to co-localize with integrins and alter their
function [25]. EGF may upregulate the urokinase receptor, and may therefore
stimulate migration through this mechanism. EGF is known to produce ECM
proteases, to stimulate matrix formation, and to lead to changes within the
epithelial cell cytoskeleton, effects that may also contribute to a migratory
phenotype [13].
References
1 Silen W, Ito S: Mechanisms for rapid re-epithelialization of the gastric mucosal surface. Annu Rev
Physiol 1985;47:217–229.
2 Moore R, Carlson S, Madara JL: Villus contraction aids repair of intestinal epithelium after injury.
Am J Physiol 1989;257:G274–G283.
3 Moore R, Carlson S, Madara JL: Rapid barrier restitution in an in vitro model of intestinal epithe-
lial injury. Lab Invest 1989;60:237–244.
4 Dignass AU, Podolsky DK: Interleukin-2 modulates intestinal epithelial cell function in vitro. Exp
Cell Res 1996;225:422–429.
5 Ciacci C, Mahida YR, Dignass A, Koizumi M, Podolsky DK: Functional interleukin-2 receptors
on intestinal epithelial cells. J Clin Invest 1993;92:527–532.
6 Dignass AU, Podolsky DK: Cytokine modulation of intestinal epithelial cell restitution: Central
role of transforming growth factor ␤. Gastroenterology 1993;105:1323–1332.
7 Podolsky DK: Healing the epithelium: Solving the problem from two sides. J Gastroenterol
1997;32:122–126.
8 Yanaka A, Muto H, Fukutomi H, Ito S, Silen W: Role of transforming growth factor-␤ in the resti-
tution of injured guinea pig gastric mucosa in vitro. Am J Physiol 1996;271:G75–G85.
9 Ciacci C, Lind SE, Podolsky DK: Transforming growth factor ␤ regulation of migration in
wounded rat intestinal epithelial monolayers. Gastroenterology 1993;105:93–101.
10 McKaig BC, Makh SS, Hawkey CJ, Podolsky DK, Mahida YR: Normal human colonic subep-
ithelial myofibroblasts enhance epithelial migration (restitution) via TGF-␤3. Am J Physiol
1999;276:G1087–G1093.
11 Kato K, Chen MC, Nguyen M, Lehmann FS, Podolsky DK, Soll AH: Effects of growth factors
and trefoil peptides on migration and replication in primary oxyntic cultures. Am J Physiol
1999;276:G1105–G1116.
12 Polk DB: Epidermal growth factor receptor-stimulated intestinal epithelial cell migration requires
phospholipase C activity. Gastroenterology 1998;114:493–502.
13 Santos MF, McCormack SA, Guo Z, Okolicany J, Zheng Y, Johnson LR, Tigyi G: Rho proteins
play a critical role in cell migration during the early phase of mucosal restitution. J Clin Invest
1997;100:216–225.
14 Riegler M, Sedivy R, Sogukoglu T, Cosnetini E, Bischof G, Teleky B, Feil W, Schiessel R,
Hamilton G, Wenzl E: Effect of growth factors on epithelial restitution of human colonic mucosa
in vitro. Scand J Gastroenterol 1997;32:925–932.
15 Takahashi M, Ota S, Hata Y, Mikami Y, Azuma N, Nakamura T, Terano A, Omata M: Hepatocyte
growth factor as a key to modulate anti-ulcer action of prostaglandins in stomach. J Clin Invest
1996;98:2604–2611.
16 Takahashi M, Ota S, Shimada T, Hamada E, Kawabe T, Okudaira T, Matsumara M, Kaneko N,
Terano A, Nakamura T, Omata M: Hepatocyte growth factor is the most potent endogenous stim-
ulant of rabbit gastric epithelial proliferation and migration in primary culture. J Clin Invest
1995;95:1994–2003.
17 Nusrat A, Parkos CA, Bacarra AE, Godowski PJ, Delp-Archer C, Rosen EM, Madara JL:
Hepatocyte growth factor/scatter factor effects on epithelia. Regulation of intercellular junctions
in transformed and nontransformed cell lines, basolateral polarization of c-met receptor in trans-
formed and natural intestinal epithelia, and induction of rapid wound repair in a transformed
model epithelium. J Clin Invest 1994;93:2056–2065.
18 Han DS, Li F, Holt L, Connolly K, Hubert M, Miceli R, Okoye Z, Santiago G, Windle K, Wong
E, Sartor RB: Keratinocyte growth factor-2 (FGF-10) promotes healing of experimental small
intestinal ulceration in rats. Am J Physiol 2000;279:G1011–G1022.
19 Cario EC, Becker A, Sturm A, Goebell H, Dignass AU: Peripheral blood mononuclear cells
promote intestinal epithelial restitution in vitro through an interleukin-2/inteferon-␥-dependent
pathway. Scand J Gastroenterol 1999;34:1132–1138.
20 Wilson AJ, Byron K, Gibson PR: Interleukin-8 stimulates the migration of human colonic epithe-
lial cells in vitro. Clin Sci 1999;97:385–390.
21 Zushi S, Shinomura Y, Kiyohara T, Minami T, Sugimachi M, Higashimoto Y, Kanayama S,
Matsuzawa Y: Role of prostaglandins in intestinal epithelial restitution stimulated by growth
factors. Am J Physiol 1996;270:G757–G762.
22 Podolsky DK: Mucosal immunity and inflammation V. Innate mechanisms of mucosal defense and
repair: The best offense is a good defense. Am J Physiol 1999;277:G495–G499.
23 Kinoshita K, Taupin DR, Itoh H, Podolsky DK: Distinct pathways of cell migration and
antiapoptotic response to epithelial injury: Structure-function analysis of human intestinal trefoil
factor. Mol Cell Biol 2000;20:4680–4690.
24 Sturm A, Sudermann T, Schulte KM, Goebell H, Dignass AU: Modulation of intestinal epithelial
wound healing in vitro and in vivo by lysophosphatidic acid. Gastroenterology 1999;117:368–377.
25 Wilson AJ, Gibson PR: Epithelial migration in the colon: Filling in the gaps. Clin Sci 1997;93:
97–108.
26 Lotz MM, Rabinovitz I, Mercurio AM: Intestinal restitution: Progression of actin cytoskeleton
rearragements and integrin function in a model of epithelial wound healing. Am J Pathol 2000;
156:985–996.
27 Peppelenbosch MP, Tertoolen LG, Hage WJ, de Laat SW: Epidermal growth factor-induced actin
remodeling is regulated by 5-lipoxygenase and cyclooxygenase products. Cell 1993;74:565–575.
28 Kinugasa T, Sakaguchi T, Gu X, Reinecker HC: Claudins regulate the intestinal barrier in response
to immune mediators. Gastroenterology 2000;118:1001–1011.
29 Gitter AH, Bendfeldt K, Schmitz H, Schulzke JD, Bentzel CJ, Fromm M: Epithelial barrier defects
in HT-29/B6 colonic cell monolayers induced by tumor necrosis factor-␣. Ann NY Acad Sci
2000;915:193–203.
30 Poulsom R, Begos DE, Modlin IM: Molecular aspects of restitution: Functions of trefoil peptides.
Yale J Biol Med 1996;69:137–146.
31 Liu D, El-Hariry I, Karayiannakis J, Wilding J, Chinery R, Kmiot W, McCrea PD, Gullick WJ,
Pignatelli M: Phosphorylation of ␤-catenin and epidermal growth factor receptor by intestinal
trefoil factor. Lab Invest 1997;77:557–563.
32 Pignatelli M: Modulation of cell adhesion during epithelial restitution in the gastrointestinal tract.
Yale J Biol Med 1996;69:131–135.
33 Perry I, Tselepis C, Hoyland J, Iqbal T, Sanders SA, Cooper BT, Jankowski JAZ: Reduced
cadherin/catenin complex expression in celiac disease can be reproduced in vitro by cytokine
stimulation. Lab Invest 1999;79:1489–1499.
34 Goke M, Zuk A, Podolsky DK: Regulation and function of extracellular matrix in intestinal
epithelial restitution in vitro. Am J Physiol 1996;34:G729–G740.
35 Basson MD, Modlin IM, Flynn SD, Jena BP, Madri JA: Independent modulation of enterocyte
migration and proliferation by growth factors, matrix proteins and pharmacologic agents in an in
vitro model of mucosal healing. Surgery 1992;112:299–308.
36 Moore R, Madri J, Carson S, Madara JL: Collagens facilitate epithelial migration in restitution of
native guinea pig intestinal epithelium. Gastroenterology 1992;102:119–130.
37 Lotz MM, Nusrat A, Madara JL, Ezzell R, Wewer UM, Mercurio AM: Intestinal epithelial resti-
tution: Involvement of specific laminin isoforms and integrin laminin receptors in wound closure
of a transformed model epithelium. Am J Pathol 1997;150:747–759.
38 Simon-Assman P, Bouziges F, Haffen AK, Kedinger M: Epithelial-mesenchymal interactions in
the production of basement membrane components in the gut. Development 1988;102:339–347.
39 Raghow R, Postlethwaite AE, Keski-Oja J, Moses HL, Kang AH: Transforming growth factor-␤
increases steady-state levels of type I procollagen and fibronectin messenger RNAs posttran-
scriptionally in cultured human dermal fibroblasts. J Clin Invest 1987;79:1285–1288.
40 Varga J, Rosenbloom J, Jimenez SA: Transforming growth factor ␤ causes a persistent increase
in steady-state amounts of type I and type III collagen and fibronectin mRNAs in normal human
dermal fibroblasts. Biochem J 1987;247:597–604.
41 Zambruno G, Marchisio PC, Marconi A, Vaschieri C, Melchiori A, Giannetti A, De Luca M:
Transforming growth factor-␤1 modulates ␤1 and ␤5 integrin receptors and induces the de novo
expression of the avb6 heterodimer in normal human keratinocytes: Implications for wound heal-
ing. J Cell Biol1995;129:853–865.
42 Sakurai H, Tsukamoto T, Kjelsberg CA, Cantley LG, Nigam SK: EGF receptor ligands are a large
fraction of in vitro branching morphogens secreted by embryonic kidney. Am J Physiol 1997;
273:F463–F472.
43 Basson MD: Role of integrins in enterocyte migration. Clin Exp Pharmacol Physiol 1998;
25:280–285.
44 Bellis SL, Newman E, Friedman EA: Steps in integrin ␤1-chain glycosylation mediated by
TGF-␤1 signaling through ras. J Cell Physiol 1999;181:33–44.
Jeffrey B. Matthews, MD, FACS, 231 Albert B. Sabin Way,

PO Box 670558, Cincinnati, OH 45267–0558 (USA)
Tel. ⫹1 513 558 5333, Fax ⫹0 513 558 2585, E-Mail Jeffrey.matthews@uc.edu
Ca2ⴙ Signaling in Epithelial Restitution

Jaladanki N. Rao a,b, Jian-Ying Wang a–c
Departments of a Surgery and c Pathology, University of Maryland School
of Medicine and b Baltimore Veterans Affairs Medical Center,
Baltimore, Md., USA
Calcium ion (Ca2) is a life and death signal in eukaryotic cells and
performs an especially large number of cellular functions. Ca2 acts as an
intracellular messenger, relaying information within the cells to coordinate the
Ca2-dependent processes including cell motility, proliferation, differentiation
and apoptosis [1–6]. To do all of these cellular functions, Ca2 signals need to
be flexible yet precisely regulated. At the cellular level, cytosolic free Ca2
([Ca2]cyt) is derived from two sources – external and internal. Ca2 can enter
from outside the cell by passing through channels that span the external barrier,
plasma membrane and be also released from internal Ca2 stores (endoplasmic
and/or sarcoplasmic reticulums) [6, 7]. It is well known that [Ca2]cyt exerts its
functions by regulating activity of a wide variety of the downstream target
signals such as protein kinases, phosphatases, cytoskeleton and cytoskeleton-
binding proteins, membrane channels, and other biochemical substances [5, 6].
The restoration of normal gastrointestinal mucosal integrity – successful
repair of wounds and ulcers – requires epithelial cell decisions that regulate
signaling networks controlling gene expression, cell survival, migration and
proliferation [3, 4, 8–11]. Early mucosal restitution is a primary repair modal-
ity in the gastrointestinal tract and occurs by sloughing off the damaged epithe-
lial cells and migration of remaining viable cells from areas adjacent to, or just
beneath, the injured surface to cover the wounded area [12, 13]. This early
mucosal re-epithelialization is able to rapidly reseal the superficial wounds and
does not require cell proliferation. The process of cell migration during restitu-
tion begins with an initial protrusion or extension of the plasma membrane at
the front leading edge of the cell mainly driven by polymerization of a network
of cytoskeletal elements and is stabilized through the formation of adhesive
complex [14, 15].
There is increasing evidence indicating that Ca2 plays an important role in
the regulation of gastrointestinal epithelial restitution following superficial injury
[16–20]. An increase in [Ca2]cyt concentration stimulates epithelial cell migra-
tion [18, 19], while removal of extracellular Ca2 ([Ca2]o) and/or depletion of
intracellularly stored Ca2 inhibit cell motility during restitution [19, 20]. In this
chapter, we will summarize the findings regarding role of Ca2 in gastrointesti-
nal mucosal epithelial migration during restitution and also discuss mechanisms
involved in the regulation of Ca2 homeostasis and its downstream signaling.
Ca2ⴙ in Restitution in vivo
Restitution of gastrointestinal mucosa was originally studied in vivo and

widely reviewed by Silen and co-workers [17, 21, 23]. In frog fundic mucosa
mounted in Ussing chambers, the surface epithelium is restored within 4 h after
exposure to 1 mol/1 NaCl for 10 min. This rapid process is completely inde-
pendent of cell division because autoradiography of [3H]thymidine incorpora-
tion shows no increase in labeling within 4 h of hypertonic NaCl damage. This
chemical form of mucosal damage is widely used as a specific model for stud-
ies responsible for mechanisms of mucosal restitution, since there is no involve-
ment of DNA synthesis in this system.
Using this model, several studies examined the effect of Ca2 on restitution
in vivo and have demonstrated that severe depletion of tissue Ca2 decreases the
formation of intercellular junctions, inhibits cell migration and delays early
mucosal restitution [16, 21]. Addition of 2 mM Ca2 to the bath buffer not only
recovers the formation of junctional complexes and mucosal electrical activity
but also increases cell migration over the denuded areas. In another set of stud-
ies in frog gastric mucosa conducted by Logsdon and Machen [22], simple
washing of the tissues with Ca2-free media following superficial injury lowers
the tissue calcium content and slightly inhibits the cell migration, but does not
completely prevent recovery of mucosal electrophysiological activities.
However, more stringent depletion of Ca2 by treatment with EDTA and wash-
ing with Ca2-free solutions almost totally inhibits migration of epithelial cells
over the bare basal lamina and recovery of mucosal electrical activity following
injury with hypertonic NaCl. The role of Ca2 in maintaining the integrity of the
gastric mucosal barrier was also observed in the mucosa of the rabbit by demon-
strating that chelation of Ca2 with treatment with EDTA disrupts function of
the gastric mucosal barrier as measured by mucosal permeability.
In other tissues, Grzesiak and Pierschbacher [27] demonstrated that shifts
in the concentrations of Ca2 in early porcine and rat wound fluid activate the
cell migratory response during cutaneous wound healing. The level of Ca2 in
Rao/Wang 30
the wound fluid drops when cell migration into the wound site is initiated. As
wound healing progresses, Ca2 concentration in the wound fluid begins to
return to normal plasma levels.
There is also indirect evidence indicating that Ca2 is absolutely required
for gut mucosal restitution. Transglutaminases are a family of Ca2-dependent
enzymes, which catalyze cross-linking reactions between polypeptide chains of
membrane and cytoskeletal proteins to form cross-linking proteins [24, 25].
The most important feature of transglutaminases is that its enzyme activity
absolutely requires Ca2 [26]. Wang and Johnson [24, 25] reported that stress-
induced mucosal erosions in the stomach and duodenum are associated with
a significant increase in transglutaminase activity. Induction of the enzyme
and subsequent formation of cross-linking proteins are necessary for normal
mucosal repair because inhibition of the enzyme activity delays early epithelial
restitution. In addition, the polymerization of -tubulin in microtubules is a
Ca2-dependent process and has also been shown to play a role in early mucosal
repair following the damage of rat gastric mucosa induced by exposure to 3.4 M
hypertonic NaCl [15].
Taken together, these direct and indirect findings from an in vivo system
clearly show that Ca2 is essential for the early rapid mucosal restitution fol-
lowing superficial wounds in the gastrointestinal tract. The Ca2 signal plays a
critical role in the activation and maintenance of cell migratory phenotypes of
gastrointestinal epithelial cells involved in the restitution process.
Ca2ⴙ in Restitution in vitro
Gastrointestinal epithelial restitution is a complex process, for it necessi-

tates that columnar epithelial cells flatten, spread, migrate and ultimately
repolarize. The mechanisms by which epithelial cells change to the flattened
phenotype and become motile during restitution remain to be understood. Since
the limitation to study such issues in natural mucosae, due to their complex
geometry, heterogeneity and finite in vitro life span, an in vitro model for gas-
trointestinal epithelial restitution is absolutely needed to characterize migration
events during this process in detail.
Using cultured IEC-6 cells, derived from rat small intestinal crypts, we have
elucidated the role of Ca2 in cell migration in the in vitro system that mimics
the early cell division-independent stages of epithelial restitution. Cells are
grown in the standard DMEM medium to achieve a confluence and the mono-
layer is scraped with a razor blade to initiate migration. After wounding caused
by removal of approximately half of the cell layer, healing is allowed to occur
over the denuded areas and numbers of migrating cells are counted at 6 or 8 h.
Ca2 Signaling in Epithelial Restitution 31

Cell migration over the denuded area increases with time after wounding in
IEC-6 cells. To determine the involvement of cell proliferation in this model, the
rate of DNA synthesis is also measured by [3H]thymidine incorporation tech-
nique. There are no significant changes in DNA synthesis within 8 h after
wounding, indicating that cell division does not participate in this process [11].
To examine the possible role of Ca2 in the process of epithelial restitution
in vitro, [Ca2]cyt concentration is decreased by removal of [Ca2]o. After
wounding the monolayer of IEC-6 cells, the Ca2-free medium is immediately
added to the cultures and numbers of migrating cells across the wound edge are
measured at various times. As shown in figure 1, exposure to the Ca2-free
medium significantly decreases the rate of cell migration compared to control
cells. The rate of cell migration is decreased by ~80% when cells are exposed
to the Ca2-free medium for 6 h. In addition, depolarization of membrane
potential (Em) by treatment with 4-aminopyridine (4-AP) has been shown to
reduce [Ca2]cyt concentration through decreasing the driving force for Ca2
influx in intestinal epithelial cells [3, 4]. Exposure of IEC-6 cells to 4-AP also
inhibits cell migration after wounding. 4-AP inhibits cell migration in dose-
dependent manner. Neither the removal of [Ca2]o nor the application of 4-AP
alters the cell viability. These results clearly indicate that intestinal epithelial
cells require Ca2 for the stimulation of migration after wounding in an in vitro
system.
Recently, a stable differentiated intestinal epithelial cell line (IEC-Cdx2L1)
has been shown to provide a better in vitro model system for restitution. Early
mucosal restitution in vivo is the function of differentiated intestinal epithelial
cells, which are localized in the surface of the mucosa, rather than of the undif-
ferentiated stem cells, which are within the proliferative zone of the crypts.
However, IEC-6 cells are undifferentiated intestinal epithelial cells that are iso-
lated from the crypts [3, 4, 9, 10]. Because IEC-Cdx2L1 cells are associated with
a significant differentiated phenotype, the in vitro model using the differentiated
intestinal epithelial cells highly mimics the restitution observed in vivo. We have
demonstrated that these differentiated intestinal epithelial cells migrate over the
wounded edge much faster than undifferentiated parental IEC-6 cells [11]. It has
been shown that increased migration of differentiated intestinal epithelial cells
after wounding is partially due to the increase in [Ca2]cyt through enhancing the
Ca2 driving force for Ca2 influx during restitution [55].
In addition, some evidence has been reported demonstrating that
prostaglandins enhance early mucosal repair at least partially through increas-
ing [Ca2]cyt concentration. Treatment with prostaglandin E2 stimulates villous
contraction of the small intestinal mucosa during restitution following injury
and the effect of prostaglandin E2 is mediated by Ca2 [28–30]. Prostaglandin
I2 increases [Ca2]cyt and protects intestinal epithelial barrier function in porcine
Rao/Wang 32
140
Control
120
Cell migration (cells/mm)
100
80
60
40
Ca2-free
20 *
0 * *
0 2 4 6
a Time after wounding (h)
120 Control
Cell migration (cells/mm)
100
4-AP
* 1.0 mM
80
60 * * 2.5 mM
40 * * 5.0 mM
20
*
0
0 2 4 6
b Time after wounding (h)
Fig. 1. Effects of removal of extracellular Ca2 (a) and 4-aminopyridine (4-AP; b) on

cell migration in IEC-6 cells. Cells are grown in the standard DMEM medium to achieve a
confluence and the monolayer is scraped with a razor blade to initiate migration. The Ca2-
free medium and 4-AP at different concentrations (1–5 mM) were given immediately after
wounding, and cell migration was assayed at various times after treatment. Values are means
SE from 6 dishes. *p 0.05 vs. controls.
ileum [31]. Lysophosphatidic acid (LPA) stimulates intestinal epithelial cell

migration after wounding via cytoskeletal activation and remodeling mediated
by [Ca2]cyt [32, 33]. Exposure of smooth muscle cells to arachidonic acid has
also been shown to enhance Ca2 influx and stimulate cell migration [35].
The role of Ca2 in cell migration has been reported in various other types
of cells including vascular endothelial cells [20]. For example, activation of
Ca2/calmodulin-dependent protein kinase II (CaM Kinase II) plays an important

role in endothelial cell migration and the inhibition of CaM Kinase II by decreas-
ing [Ca2]cyt diminishes cell migration [20]. Increasing [Ca2]cyt by treatment
with the calcium ionophore, ionomycin, activates CaM Kinase II activity and
enhances the process of endothelial cell migration. One study indicates that treat-
ment with exogenous ATP induces the Ca2 signals in gastric microvascular
endothelial cells and promotes cell migration [34].
Regulation of Ca2ⴙ Homeostasis
Ca2 homeostasis is a dynamic process, involving the mechanisms of Ca2

entry, release, and sequestration of cellular events. The mechanisms involved in
this complex process are cell type-dependent and are controlled by numerous
factors. Excitable cells such as neurons and muscle cells highly express L-type
voltage-dependent Ca2 channels (VDCC) that are activated by membrane
depolarization, leading to an increase in [Ca2]cyt. In contrast, nonexcitable
cells including intestinal epithelial cells do not express VDCC and membrane
depolarization decreases [Ca2]cyt concentration through reducing the driving
force for Ca2 influx.
[Ca2]cyt undergoes rapid and often substantial fluctuations in response to
various extracellular messengers binding to their cognate receptors on target
cells. Since [Ca2]o concentration is maintained stably at 1.6–18 mM, ~10,000-
to 20,000-fold higher than the resting [Ca2]cyt (50–150 nM) under physiologi-
cal conditions, [Ca2]cyt is primarily controlled by Ca2 influx in a variety of
cell types [6, 7]. Although the mechanisms responsible for Ca2 homeostasis
in excitable cells have been extensively studied, nonexcitable cells have not
received as much attention as excitable cells until recently. In this part of the
chapter, we will focus on the regulation of Ca2 homeostasis in nonexcitable
intestinal epithelial cells.
Ca2 Driving Force and Membrane Potential (Em)

Ca2 influx in nonexcitable cells depends on the Ca2 driving force or the
electrochemical gradient across the plasma membrane [3, 4, 36]. While the
chemical gradient, the ratio of [Ca2]o to [Ca2]cyt, and the Ca2 equilibrium
potential, ECa {ECa = 12.5 ln ([Ca2]o/[Ca2]cyt) = 117–131 mV at 25 °C} are
constant, the Ca2 driving force is mainly determined by the electrical gradient,
the difference between Em and ECa (Em – ECa). In other words, Em is a major
determinant of the driving force for Ca2 influx in nonexcitable cells. Em is pri-
marily determined by the K permeability and K gradient across the plasma
membrane [37, 40, 41]. Since K gradient is maintained by Na,K-ATPase,
the K permeability is directly related to the activity and number of membrane
Rao/Wang 34
K channels. Voltage-gated K channels (Kv) have been shown to be a major
determinant of resting Em in many types of cells [37]. When K channel closes
or the number of total K channels decreases, Em becomes less negative (i.e.,
depolarization). When K channel opens or the number of total K channels
rises, Em becomes more negative (i.e., hyperpolarization). Therefore, inhibition
of K channel gene expression decreases the number of K channels and atten-
uates K channel activity. The subsequent membrane depolarization decreases
the Ca2 driving force and thus inhibits Ca2 influx. Since Ca2 entry is a major
source for [Ca2]cyt, inhibition of Ca2 influx would reduce [Ca2]cyt in cells
lacking VDCC [36, 37].
By controlling the Ca2 driving force, Em is an important regulator of
2
[Ca ]cyt concentration in nonexcitable cells. Membrane depolarization
decreases the Ca2 driving force and inhibits Ca2 influx. In contrast, mem-
brane hyperpolarization increases the Ca2 driving force and enhances Ca2
influx. Therefore, in nonexcitable cells that do not express VDCC, Ca2 influx
is decreased by membrane depolarization but increased by membrane hyperpo-
larization [4, 36–39]. Nonetheless, in excitable cells, VDCC that are opened by
membrane depolarization are the major pathway for Ca2 influx [7, 36]. In con-
trast to the voltage-independent pathway for Ca2 influx in nonexcitable cells,
membrane depolarization opens VDCC and thus increases [Ca2]cyt in excitable
cells [37].
We have recently demonstrated that intestinal epithelial cells (IEC-6 line)
do not express VDCC and that the [Ca2]cyt concentration is primarily regulated
by Ca2 influx through the Ca2 driving force [3]. Exposure of IEC-6 cells to
-difluoromethylornithine (DFMO), a specific inhibitor for polyamine biosyn-
thesis, not only completely depletes cellular polyamines (putrescine, spermi-
dine and spermine) but also significantly decreases Kv1.1 mRNA and protein
expression, which is associated with a reduction of whole cell K currents,
membrane depolarization and a decrease in [Ca2]cyt. In these depolarized
intestinal epithelial cells following polyamine depletion, cyclopiazonic acid
(CPA)- and ionomycin-induced transient increases in [Ca2]cyt are significantly
lower compared with those observed in control cells [3, 4]. This reduced
response of polyamine-deficient cells to CPA or ionomycin is apparently due to
the decreased Ca2 driving force as a result of membrane depolarization. Cell
migration in polyamine-deficient cells is reduced by 80% after wounding.
Restoration of Kv channel activity by exogenous natural polyamine spermidine
prevents depolarized Em, returns [Ca2]cyt level to near normal, and promotes
cell migration in polyamine-deficient cells. These results clearly indicate that
Kv channel activity plays a critical role in the regulation of Ca2 influx during
restitution and that Kv channel expression absolutely requires polyamines in
intestinal epithelial cells. Increased Kv channel activity results in membrane

hyperpolarization and raises [Ca2]cyt by enhancing the driving force for Ca2
influx and thus stimulates cell migration.
Regulation of [Ca2]cyt and Transient Receptor Potential Channels

As pointed out above, the Ca2 driving force in intestinal epithelial cells is
the major determinant of transmembrane influx of Ca2, but Ca2 channels in
this process are still unknown. In endothelial and epithelial cells, passive Ca2
leakage, receptor-operated Ca2 channels (ROC), nonselective cation channels
(NSOC) and store-operated channels (SOC) all contribute to Ca2 influx [7, 42].
Nonexcitable cells such as intestinal epithelial cells lack VDCC but have devel-
oped the Ca2 entry mechanism that is coupled with the depletion of intracellu-
lar Ca2 stores to activate transient receptor potential channels (TRPC). Thus the
capacitative Ca2 entry via TRPC channels may be a major source of intracellu-
lar Ca2 in mammalian cells [41, 43]. TRPC-1 and TRPC-5 that encodes Ca2
permeable channels are detected in the intestinal epithelial cells [3]. These
results indicate that the capacitative Ca2 entry through TRPC-1 and TRPC-5
channels may serve as an important source for the agonist-mediated rise in
[Ca2]cyt during restitution in intestinal epithelial cells. In support of this possi-
bility, we have recently demonstrated that mRNA and protein levels of TRPC-1
in differentiated intestinal epithelial cells are significantly higher than those
observed in undifferentiated parental cells. The higher levels of TRPC-1 in
differentiated epithelial cells are associated with a significant increase in basal
[Ca2]cyt concentration [55]. Clearly, studies to elucidate the exact role of TRPC
in the regulation of Ca2 influx in gastrointestinal epithelial cells are badly
needed.
Downstream Targets of Ca2ⴙ
Although Ca2 regulates almost everything that we do, different cell types
and cellular functions select combinations of Ca2 signals with the precise
parameters to fit their processes. A specific role of Ca2 in the regulation of
various biological functions completely depend on downstream targets of ele-
vated [Ca2]cyt. The research has been reviewed elsewhere, and the following
discussion will focus on just a few points of current interest. The coordinated
movement of intestinal epithelial cells during restitution is a complex process
that is controlled by the cytoskeleton. Changes in both the distribution and for-
mation of the cytoskeleton alter the adhesion, spreading, and motility of cells.
There is increasing evidence indicating that elevated [Ca2]cyt activates Rho
guanine nucleotide triphosphate (GTP)-binding proteins that are key regulators
of the cytoskeletal reorganization [44, 45].
Rao/Wang 36
Rho Proteins
The Rho family, including Rho, Rac and Cdc42, is a member of the Ras
superfamily of small guanosine triphosphatases (GTPases) and functions as
molecular switches by cycling between an active GTP-bound state and an inac-
tive GDP-bound state [44, 46, 50]. Activation of Rho proteins, through GDP-GTP
exchange, is stimulated by guanine nucleotide exchange factors (GEFs), whereas
inactivation of the proteins is promoted by GTPase-activating proteins (GAPs)
[48–50]. Activated Rho proteins interact with cellular target proteins or effectors
to regulate a signal transduction pathway linking surface receptors to the forma-
tion of actomyosin stress fibers and focal adhesions [46, 50]. The transformation
of RhoA from its inactive GDP-bound form to its active GTP-bound form acti-
vates Rho kinase that results in the formation of actomyosin stress fibers by ini-
tiating myosin light chain phosphorylation [52]. On the other hand, activation of
Rac promotes de novo actin polymerization at the cell periphery to form lamel-
lipodial extensions and membrane ruffles, and activation of Cdc42 results in actin
polymerization to form filopodia or microspikes [44, 46–48, 51, 54].
It has been recently reported that Ca2-activated RhoA activity plays a crit-
ical role in regulation of cell migration after wounding in intestinal epithelial
cells [4, 47]. Decreased [Ca2]cyt concentration, either by reducing the Ca2 driv-
ing force for Ca2 influx via membrane depolarization by polyamine depletion
or removal of [Ca2]o from the culture medium inhibits RhoA expression and
activity. Increasing [Ca2]cyt by the treatment with Ca2 ionophore ionomycin
stimulates RhoA activity in intestinal epithelial cells (fig. 2). Exposure to 1 M
ionomycin for 4 and 6 h not only increases RhoA protein level in normal cells
but also significantly overcomes the inhibitory effect of polyamine depletion on
RhoA expression. RhoA protein levels in response to ionomycin are correlated
to the elevation of [Ca2]cyt and the responses in polyamine-deficient cells are
smaller than those observed in normal cells because reduced driving force for
Ca2 influx. Elevation of [Ca2]cyt increases RhoA activity partially through
alteration of RhoA protein synthesis and stability in intestinal epithelial cells
[4]. The rate of newly synthesized RhoA protein is stimulated by increasing
[Ca2]cyt but inhibited after exposure to the Ca2-free medium. On the other
hand, the stability of RhoA protein also is regulated by [Ca2]cyt in intestinal
epithelial cells. Although elevation of [Ca2]cyt slightly increases the stability of
RhoA protein, removal of [Ca2]o dramatically destabilizes RhoA protein and
accelerates its degradation.
Furthermore, decreased Rho activity by treatment with Clostridium
botulinum exoenzyme C3 transferase (C3) blocks formation of actin-myosin
stress fibers and inhibits intestinal epithelial cell migration after wounding
[4, 47]. At present it is not clear whether other members of the mammalian Rho
family, including RhoB, RhoC, RhoD, RhoE, and RhoG, Rac1 and Rac2, and

200 Control DFMO
b
175
[Ca2]cyt (nM) 150
125
a
100 c
75
Ionomycin
50 Ionomycin 2min
a
a. Control b. DFMO
Ionomycin Ionomycin
0 4 6 0 4 6 Time (h)
kD
21 Rho-A
43 -Actin
b
Fig. 2. Effect of Ca2 ionophore ionomycin on [Ca2]cyt and RhoA expression in normal
and polyamine-deficient cells. a Representative records of [Ca2]cyt measured in peripheral
areas of cells before, during and after application of 1 M ionomycin. Cells were grown in the
DMEM medium with or without 5 mM -difluoromethylornithine (DFMO) for 4 days.
Ionomycin at the concentration of 1 M was added into the media and [Ca2]cyt was continu-
ously monitored for 10 min after the administration of ionomycin. b Western immunoblots of
RhoA protein. After cells were exposed to ionomycin at the dose of 1 µM for 4 and 6 h, RhoA
protein expression (~21 kD) was examined by using the specific anti-RhoA antibody.
Cdc42, are regulated by [Ca2]cyt alterations and are involved in the regulation
of intestinal epithelial cell migration. These results indicate that increasing
[Ca2]cyt activates RhoA activity that increases the formation of actomyosin
stress fibers and stimulates intestinal epithelial migration during the early phase
of mucosal restitution.
Rao/Wang 38
Effector of Rho
To understand the exact mechanisms through which Rho GTPases regulate
the cytoskeletal rearrangement and other associated activities, enormous cellu-
lar effectors (targets) have been reported [48, 56]. Using yeast two-hybrid selec-
tion and affinity purification techniques, more than 20 candidate targets have
been identified so far, which represent a wide variety of enzymatic activities
and protein-protein interaction domains.
It has been shown that the Ser-Thr kinase p160ROCK interacts with Rho in
a GTP-dependent manner [53, 57]. This kinase is an excellent candidate for
mediating Rho-induced changes to actin-myosin cytoskeleton because it mimics
the function of Rho when overexpressed or constitutively activated. Both myosin
light chain (MLC) and MLC phosphatase, which are known to regulate the
assembly of actin-myosin filament bundle [46, 52], are substrates of the Ser-Thr
kinase p160ROCK. Whether p160ROCK is the only downstream target of Rho
required for inducing the formation of stress fibers remains to be demonstrated.
Treatment with cytochalasin D blocks assembly of stress fibers, suggesting that
some actin polymerization might be required. However, there are some results
indicating that the actin-myosin filaments induced by p160ROCK are not cor-
rectly organized nor are they contractile as they are when induced by Rho [56].
Although no direct targets of Rho are identified yet, the ERM proteins,
including ezrin, radixin, and moesin, are emerging as key regulators of the
actin-myosin cytoskeleton. The interaction of ERM with a transmembrane pro-
tein, CD44, through their NH-2-termini has been shown to be regulated by Rho,
and their COOH-terminal ends interact with filamentous actin [58]. Moreover,
ERM proteins are essential for both Rho- and Rac-induced cytoskeletal organi-
zation. It is possible that ERM proteins behave as regulatable scaffold proteins
that anchor actin-myosin filaments to the membrane and that this is a prerequi-
site for Rho and Rac to induce stress fibers and lamellipodia, respectively.
Summary and Conclusion
The results summarized in this chapter clearly indicate that a rise in [Ca2]cyt
is an important stimulus for epithelial cell migration after superficial wounds in
the gastrointestinal mucosa. Decreasing [Ca2]cyt suppresses intestinal epithelial
cell motility but increasing [Ca2]cyt stimulates cell migration and enhances
mucosal restitution after wounding. Since intestinal epithelial cells do not express
VDCC, Ca2 influx is primarily controlled by Em (the driving for Ca2 influx). Em
is dependent on K channel activity. Enhanced K channel protein expression
induces the number of functional Kv channels, increases the K currents, causes
membrane hyperpolarization, increase the driving force for Ca2 influx, and raises

[Ca2]cyt. The resultant increase in [Ca2]cyt activates RhoA activity and others
that drive the assembly of actin-myosin microfilaments, thereby stimulating cell
migration and intestinal epithelial restitution.
Although important discoveries have been made over the past decade,
many critical issues regarding the role of Ca2 in mucosal restitution remains
to be demonstrated. Studies to examine the molecular mechanism responsible
for K channel gene expression, Ca2 channels of intestinal epithelial cells, and
Ca2 sensitive proteins and their downstream targets are needed to fully eluci-
date the biological function of Ca2 signal in the regulation of mucosal restitu-
tion. It can be predicted with some confidence that the cellular and molecular
analysis of the signal pathways controlled by Ca2 will lead to a better under-
standing of mechanisms of gastrointestinal epithelial restitution.
References
1 Glenney JR Jr, Bretscher A, Weber K: Calcium control of the intestinal microvillus cytoskeleton:
Its implications for the regulation of microfilament organizations. Proc Natl Acad Sci USA 1980;
77:6458–6462.
2 Tansey MG, Word RA, Hidaka H, Singer HA, Schworer CM, Kamm KE, Stull JT: Phosphorylation
of myosin light chain kinase by the multifunctional calmodulin-dependent protein kinase II in
smooth muscle cells. J Biol Chem 1992;267:12511–12516.
3 Wang JY, Wang J, Golovina VA, Li L, Platoshyn O, Yuan JX: Role of K channel expression in
polyamine-dependent intestinal epithelial cell migration. Am J Physiol 2000;278:C303–C314.
4 Rao JN, Li L, Golovina VA, Platoshyn O, Strauch ED, Yuan JX, Wang JY: Ca2-RhoA signaling
pathway required for polyamine-dependent intestinal epithelial cell migration. Am J Physiol 2001;
280:C993–C1007.
5 Niki I, Yokokura H, Sudo T, Kato M, Hidaka H: Ca2 signaling and intracellular Ca2 binding pro-
teins. J Biochem 1996;120:685–698.
6 Bootman MD, Berridge MJ. The elemental principles of calcium signaling. Cell 1995;83:
675–678.
7 Parekh AB, Penner R: Store depletion and calcium influx. Physiol Rev 1997;77:901–930.
8 Lauffenburger DA, Horwitz AF: Cell migration: A physically integrated molecular process. Cell
1996;84:359–369.
10 Wang JY, Viar MJ, Li J, Shi HJ, McCormack SA, Johnson LR: Polyamines are necessary for
normal expression of the transforming growth factor- gene during cell migration. Am J Physiol
1997;272:G713–G720.
11 Rao JN, Li J, Li L, Bass BL, Wang JY: Differentiated intestinal epithelial cells exhibit increased
migration through polyamines and myosin II. Am J Physiol 1999;277:G1149–G1158.
12 Rutten MJ, Ito S: Morphology and electrophysiology of guinea pig gastric mucosal repair in vitro.
Am J Physiol 1983;244:G171–G182.
13 Nusrat A, Delp C, Madara JL: Intestinal epithelial restitution: Characterization of cell culture
model and mapping of cytoskeletal elements in migrating cells. J Clin Invest 1992;89:
1501–1511.
14 Horwitz AR, Parsons JT: Cell migration – movin’ on. Science 1999;286:1102–1103.
15 Banan A, McCormack SA, Johnson LR: Polyamines are required for microtubule formation
during gastric mucosal healing. Am J Physiol 1998;274:G879–G885.
Rao/Wang 40
16 Critchlow J, Magee D, Ito S, Takeuchi K, Silen W: Requirements for restitution of the surface
epithelium of frog stomach after mucosal injury. Gastroenterology 1985;88:237–249.
17 Paimela H, Goddard PJ, Silen W: Present views on restitution of gastrointestinal epithelium. Dig
Dis Sci 1995;40:2495–2496.
18 Clapham DE: Calcium signaling. Cell 1995;80:259–268.
19 Bilato C, Pauly RR, Melillo G, Monticone R, Gorelick-Feldman D, Gluzband YA, Sollott SJ,
Ziman B, Lakatta EG, Crow MT: Intracellular signaling pathways required for rat vascular smooth
muscle cell migration. Interactions between basic fibroblast growth factor and platelet-derived
growth factor. J Clin Invest 1995;96:1905–1915.
20 Pauly RR, Bilato C, Sollott SJ, Monticone R, Kelly PT, Lakatta EG, Crow MT: Role of calcium/
calmodulin-dependent protein kinase II in the regulation of vascular smooth muscle cell migra-
tion. Circulation 1995;91:1107–1115.
21 Silen W, Ito S: Mechanisms for rapid re-epithelialization of the gastric mucosal surface. Annu Rev
Physiol 1985;47:217–219.
22 Logsdon CD, Machen TE: Involvement of extracellular calcium in gastric stimulation. Am
J Physiol 1981;241:6365–6375.
23 Chung RSK, Sum PT, Goldman H, Field M, Silen W: Effects of chelation of calcium on the gastric
mucosal barrier. Gastroenterology 1970;59:200–207.
24 Wang JY, Johnson LR: Role of transglutaminase and protein cross-linking in the repair of mucosal
stress erosions. Am J Physiol 1992;262:G818–G825.
25 Wang JY, Viar MJ, Johnson LR: Transglutaminase in response to hypertonic NaCl-induced gastric
mucosal injury in rats. Gastroenterology 1993;104:65–74.
26 Lorand L, Conrad SM: Transglutaminases. Mol Cell Biochem 1984;58:9–35.
27 Grzesiak JJ, Pierschbacher MD: Shifts in the concentrations of magnesium and calcium in early
porcine and rat wound fluids activate the cell migratory response. J Clin Invest 1995;95:227–233.
28 Yanaka A, Muto H, Fukutami H, Ito S, Silen W: Role of transforming growth factor- in the resti-
tution of injured guinea pig gastric mucosa in vitro. Am J Physiol 1996;271:G75–G85.
29 Paimela H, Goddard PJ, Carter K, Khakee R, McNeil PL, Ito S, Silen W: Restitution of frog
gastric mucosa in vitro: Effect of basic fibroblast growth factor. Gastroenterology 1993;104:
1337–1345.
30 Erickson RA, Tarnawski A, Dines G, Stachura J: 16,16-Dimethyl-prostaglandin E2 induces villous
contraction in rats without affecting intestinal restitution. Gastroenterology 1990;99:708–716.
31 Blikslager AT, Roberts MC, Rhoads JM, Argenzio RA: Prostaglandin I2 and E2 have a synergistic
role in rescuing epithelial barrier function in porcine ileum. J Clin Invest 1997;100:1928–1933.
32 Hines OJ, Ryder N, Chu J, McFadden D: Lysophosphatidic acid stimulates intestinal restitution
via cytoskeletal activation and remodeling. J Surg Res 2000;92:23–28.
33 Szabo S: Mechanisms of gastric mucosal injury and protection. J Clin Gastroenterol
2000;13:S21–S34.
34 Ehring GR, Szabo IL, Jones MK, Sarfeh IJ, Tarnawski A: ATP-induced Ca2-signaling enhances
rat gastric microvascular endothelial cell migration. J Physiol Pharmacol 2000;51:799–811.
35 Toyoda KT, Morita I, Murota S: Arachidonic acid pretreatment enhances smooth muscle cell
migration via increased Ca2 influx. Prostaglandins Leukot Essent Fatty Acids 1998;58:25–31.
36 Brown EM, Vassilev PM, Hebert SC: Calcium as extracellular messengers. Cell 1995;83:679–682.
37 Fleischmann BK, Washabau RJ, Kotlikoff MI: Control of resting membrane potential by delayed
rectifier potassium currents in ferret airway smooth muscle cells. J Physiol (Lond) 1993;469:
625–638.
38 Himmel HA, Whorton AR, Strauss HC: Intracellular calcium, currents and stimulus-response
coupling in endothelial cells. Hypertension 1993;21:112–127.
39 Wang JY, Johnson LR: Luminal polyamines stimulate repair of gastric mucosal stress ulcers. Am
J Physiol 1990;259:G584–G592.
40 Nelson MT, Quayle JM: Physiological roles and properties of potassium channels in arterial
smooth muscle. Am J Physiol 1995;268:C799–C822.
41 Moore TM, Brough GH, Babal P, Kelly JJ, Li M, Stevens T: Store-operated calcium entry
promotes shape change in pulmonary endothelial cells expressing Trp1. Am J Physiol 1998;275:
L574–L582.

42 Nilius B, Viana F, Droogmans G: Ion channels in vascular endothelium. Annu Rev Physiol 1997;
59:145–170.
43 Zhu X, Jiang M, Peyton M, Boulay G, Hurst R, Stefane E, Birnbaumer L: Trp, a novel mammalian
gene family essential for agonist-activated capacitative Ca2 entry. Cell 1996;85:661–671.
44 Nobes CD, Hall A: Rho, Rac and Cdc42 GTPases regulate the assembly of multimolecular focal
complexes associated with actin stress fibers, lamellipodia and filopodia. Cell 1995;81:53–62.
45 Kiosses WB, Daniels RH, Otey C, Bokoch GM, Schwartz MA: A role for P21-activated kinase in
endothelial cell migration. J Cell Biol 1999;147:831–843.
46 Amano M, Chihara K, Kimura K, Fukata Y, Nakamura N, Matsuura Y, Kaibuchi K: Formation of
actin stress fibers and focal adhesions enhanced by Rho-kinase. Science 1997;275:1308–1311.
47 Santos MF, McCormack SA, Guo Z, Okolicany J, Zheng Y, Johnson LR: Rho proteins play a crit-
ical role in cell migration during early phase of mucosal restitution. J Clin Invest 1997;100:
216–225.
48 Hall A: Rho GTPase and the actin cytoskeleton. Science 1998;279:509–514.
49 Mackay JG, Hall A: Rho GTPase. J Biol Chem 1998;273:20685–20688.
50 Narumiya S: The small GTPase Rho: Cellular functions and signal transduction. J Biochem
1996;120:215–228.
51 Burridge K: Cross talk between Rac and Rho. Science 1999;283:2028–2029.
52 De Lanerolle P, Paur RJ: Myosin phosphorylation/dephosphorylation and regulation of airway
smooth muscle contractility. Am J Physiol 1991;261:L1–L14.
53 Leung T, Chen XQ, Manser E, Lim L: The P160 RhoA-binding kinase ROK is a member of
a kinase family and is involved in the reorganization of the cytoskeleton. Mol Cell Biol 1996;
16:5313–5327.
54 Tapon N, Hall A: Rho, Rac and Cdc42 GTPases regulate the organization of the actin cytoskele-
ton. Curr Opin Cell Biol 1997;9:86–92.
55 Rao JN, Platoshyn O, Li L, Guo X, Golovina VA, Yuan JXJ, Wang JY: Activation of K channels
and increased migration of differentiated intestinal epithelial cells after wounding. Am J Physiol
2002;282:C885–C898.
56 Aelst LV, D’Souza-Schorey C: Rho GTPases and signaling networks. Gene Dev 1997;11:
2295–2322.
57 Kimura K, Ito M, Amano M, Chihara K, Fukata Y, Nakafuku M, Yamamori B, Feng J, Nakano T,
Okawa K, Iwamatsu A, Kaibuchi K: Regulation of myosin phosphatase by Rho and Rho-associ-
ated kinase (Rho-kinase). Science 1996;273:245–248.
58 Pestonjamasp K, Amieva MR, Strassel CP, Nauseef WM, Furthmayr H, Luna EJ: Moesin, ezrin
and p205 are actin-binding proteins associated with neutrophil plasma membranes. Mol Biol Cell
1995;6:247–259.
Jaladanki N. Rao, PhD, Department of Surgery, Baltimore VA Medical Center,

10 North Greene Street, Baltimore, MD 21201 (USA)
Tel. 1 410 605 7000/ext. 6430, Fax 1 410 605 7919, E-Mail jrao@umaryland.edu
Rao/Wang 42
Polyamines in Intestinal Epithelial

Restitution
S.A. McCormack, Ramesh M. Ray, L.R. Johnson
Department of Physiology, University of Tennessee Health Science Center,
Memphis, Tenn., USA
Gastrointestinal epithelial damage can result from infection (ulcer), chemi-

cal agents (alcohol, drugs), or mechanical forces (stretching) and requires
prompt repair to preserve the epithelial barrier to luminal antigens. Although
repair is a continuous process, it can be divided into two stages depending on
whether the primary action is one of cell spreading and migration or of prolif-
eration. During the early restitution period (conservatively up to 12 h), viable
cells bordering the lesion extend lamellipodia and migrate over the damaged
area. During the late period, migration continues and cell proliferation becomes
a major factor in repair [1]. Although many processes are active during the early
stage and continue into the late stage, some are unique to the early stage, for
instance, the immediate rapid secretion of the trefoil peptides. Cells that are
deficient in polyamines fail to migrate normally during early restitution [2–4]
and throughout the whole process of repair. Cell proliferation is slowed to a
basal level as well [5].
The physiological polyamines spermidine, spermine and putrescine
(a diamine) are found in every cell of virtually all living species. In spite of their
early discovery in 1677 by Leewenhoek [6], little progress toward understand-
ing their biological functions was made before 1970. Today it is well known that
polyamines play essential roles in embryological development, cell prolifera-
tion, attachment, spreading, migration, gene expression and transcription, cell
cycle regulation, and probably in other areas not yet investigated. Polyamines
are involved in the processes that maintain the dynamic turnover and integrity
of the gastrointestinal epithelium in which many individual cells have a lifetime
of only 3–4 days. The following summary of polyamine biosynthesis, regu-
lation, and function concentrates on their participation in the early stage of
Ornithine
DFMO
1
Methylthio CO 2
Adenosine
H2N⫹ N⫹H2 Putrescine
7
Decarboxylated (CH 2 ) 3 NH 2
S-Adenosylmethionine 2 5
HN ⫹
4 H2N⫹ N⫹H2 Spermidine
CO 2
S-Adenosylmethionine (CH 2 ) 3 NH 2
3 6
ATP-Mg ⫹⫹ HN ⫹
PP i ⫹ P i H2N⫹ N⫹H2 Spermine
N⫹H
L-Methionine
1 Ornithine Decarboxylase (ODC)
2 Spermidine Synthase
3 Spermine Synthase
4 S-Adenosylmethionine Decarboxylase (SAMDC)
5 Polyamine Oxidase
6 Polyamine Oxidase
7 Acetyldiamine Transferase
Fig. 1. The biosynthesis of the major physiologic polyamines and their precursor
putrescine. Putrescine has 4 CH groups, spermidine 7, and spermine 10. They carry 2, 3 and 4
positive charges that reflect the strength of their binding ability. Enzymes involved in the
reactions are numbered.
gastrointestinal epithelial repair (restitution), a brief period that is independent

of cell proliferation but vital to rapid resealing of the epithelial barrier.
Polyamine Biosynthesis
Putrescine, spermidine and spermine are organic, aliphatic, cationic

amines. They are produced in a cyclic reaction sequence starting from ornithine
or methionine that results in the synthesis and interconversion of putrescine,
spermidine and spermine (fig. 1). The key regulatory enzymes are ornithine
decarboxylase (ODC) and S-adenosylmethionine decarboxylase (SAMDC).
ODC decarboxylates ornithine to produce putrescine and is the primary rate-
limiting enzyme of polyamine biosynthesis. SAMDC decarboxylates S-adenosyl
methionine to produce propylamine groups that are added to synthesize
spermidine and spermine. SAMDC is considered the second rate-limiting
enzyme in polyamine biosynthesis. In addition to biosynthesis from ornithine,
putrescine can be synthesized from spermidine and spermine by the enzymatic
removal of their propylamine groups. The enzymes involved in these intercon-
versions are spermidine synthase, spermine synthase and polyamine oxidase,
McCormack/Ray/Johnson 44
an enzyme that oxidatively splits the monoacetyl derivatives of spermidine
and spermine. These enzymes are usually present in excess and are not rate
limiting under physiological conditions. In nonproliferating cells, most of the
putrescine comes from spermidine rather than from ODC. ODC activity is low
in nonproliferating cells and primarily provides putrescine moieties lost by
cells to secretion and degradation. In rapidly dividing cells, such as embryonal
cells, gut mucosal cells and tumor cells, putrescine is synthesized almost
entirely by ODC [7].
The ubiquitous presence of polyamines in cells makes the study of
polyamine actions difficult. Therefore, one must often resort to investigating
them from the standpoint of an artificial deficiency produced by the use of an
inhibitor or by transfection. Because tumors often have very high polyamine
levels due to their defective regulatory mechanisms and because polyamines
stimulate tumor growth, a variety of drugs that inhibit enzymes at various
stages of polyamine biosynthesis has been synthesized in the search for effec-
tive anti-cancer treatments. One of the most useful of these drugs for experi-
mental use is dl-␣-difluoromethylornithine (DFMO). DFMO irreversibly
inhibits ODC, blocking the production of putrescine from ornithine, and effec-
tively inhibiting growth and migration in gastrointestinal cells. Significantly,
if a polyamine is added to supplement DFMO in the treated cells, growth and
migration can be maintained at control levels [2]. Another polyamine synthesis
inhibitor, diethylglyoxal bis(guanylhydrazone) (DEGBG), can be used to
inhibit SAMDC to block the production of spermine and spermidine from
S-adenosyl-L-methionine [8].
Polyamine deficiency induced by polyamine synthesis inhibitors reveals
the role of polyamines in maintaining the cytoskeleton and regulating ionic and
osmotic intracellular control. Despite the near specificity of the two inhibitors
(especially DFMO), some nonspecific effects have been reported. For example,
the expression of genes for two major elements of the cytoskeleton, ␤-actin and
␣-tubulin, are reduced in mouse splenocytes [9]. Also, a persistent lowering of
intracellular pH (pHI) homeostasis is induced in L1210 leukemia cells due to a
lowering of the pHI setpoint of the Na⫹/H⫹ exchanger [10].
Polyamine Regulation
Although polyamines are necessary for normal gastrointestinal function,

increased intracellular polyamines in the absence of a stimulus for growth or
migration may cause apoptosis [11–13], lead to cancer [7, 14], or have signifi-
cant toxicity [15]. Extracellular polyamines originating from food, intestinal
flora and sloughed degraded cells are constantly available at many times
PAs in Intestinal Restitution 45

intracellular levels to epithelial cells lining the intestinal lumen. In addition,
polyamines are available from the bloodstream. Therefore, tight regulation of
intracellular polyamine levels is important and is achieved by endogenous
inhibitors and by adjusting inward and outward transport.
Antizymes are endogenous inhibitors of ODC induced by the excess pro-
duction of polyamines [16]. Antizymes bind to and inactivate ODC and, at the
same time, inhibit polyamine uptake [16–18] and stimulate polyamine excre-
tion and degradation by the 26S proteosome [19, 20]. Five antizymes of varying
abundance, species and tissue specificity have been described [16, 21–23].
Cells use a variety of transport mechanisms for import as well as export of
polyamines. In mammalian cells, polyamine transport is specific, saturable and
energy-dependent. In some cells, transport may be carrier-mediated and require
RNA and protein synthesis [24]. Many cells have a common transporter for
polyamines as well as separate transporters for spermidine and spermine
[25–27]. Inward transport activity is increased under growth-promoting condi-
tions and decreased when there is no need for exogenous polyamines. Using
IEC-6 cells (a noncancerous intestinal epithelial cell line) grown on cell culture
inserts, we found that inward transport of putrescine occurs preferentially from
the apical side, reflecting their position in the intestinal mucosa [28]. In LoVo
cells (a human colon cancer cell line), inward transport is greater from the basal
side than the apical side and implies increased transport from the bloodstream
[29]. In both cell lines, transport is Na⫹-independent [28, 29]. Because the nor-
mal milieu of the epithelial cells lining the intestinal lumen has a high concen-
tration of polyamines, outward transport is more active than inward transport
unless there is a stimulus for growth or repair.
Several pathological states that damage the gastrointestinal epithelium
to the extent of requiring restitution are associated with polyamine excess or
deficiency, i.e. cancer, bacterial toxicity, aging and ischemic damage. The up-
regulation of ODC mRNA expression (followed by an increase in polyamine
levels) is an early event in the development and progression of Barrett’s-associated
adenocarcinoma of the esophagus. The increased polyamine levels may be
useful in detecting this disease in an occult stage [30]. The small GTPase Rho,
which regulates a variety of cell functions, also serves as a specific substrate for
bacterial toxins. The intracellular polyamines can be cross-linked with Rho by
the Bordetella dermonecrotizing toxin causing them to play an unwanted role
in the toxin’s potency [31]. During aging in man, susceptibility to gastric
mucosal damage increases while mucosal repair ability declines. At the same
time, polyamines, prostaglandins and activated epidermal growth factor recep-
tors (EGFR) also decrease. The lowered levels of prostaglandins, growth fac-
tors and polyamines may be the cause of the aging effects on mucosal injury
and repair [32, 33]. In addition, polyamine deficiency reduces the tyrosine
phosphorylation of EGFR by 50% and alters its recycling as well [34]. In rats,
the administration of arginine or ornithine, both precursors of polyamines, did
not prevent intestinal ischemic damage but did accelerate morphological repair,
enhance cell proliferation, and increase polyamine content [35]. These and
other outright disease/polyamine associations deserve further investigation for
both diagnostic and therapeutic possibilities.
What Is the Route of Signals between Wounding and Restitution?
Signals for repair originate from the wound by a paracrine route, probably
as cytokines, growth hormones and changes in Ca2⫹ levels. The signals stimu-
late two responses, migration and proliferation. Migration begins immediately,
but proliferation takes place more slowly because the cells in the original con-
fluent monolayer were quiescent before the wound, and time is required for
DNA replication before dividing. The signals begin a cascade that ultimately
results in many coordinated processes. These include increased actin polymeri-
zation and stress fiber formation, the formation of new focal adhesions and the
de-adhesion of others, cell polarization, the synthesis of new plasma membrane
for the protrusion of lamellipodia and filopodia, increased actin-myosin motor
assembly, and finally proliferation.
Tabor and Tabor [5] had shown the effect of polyamine deficiency on cell
growth in 1984. Others had investigated the effects of polyamines on various
aspects of cell migration, i.e., human sperm motility [36], metastasis in rat
breast cancer [37], migration of cells over tooth root surfaces [38] and cell
attachment to fibronectin [39]. We began experiments on gastrointestinal
mucosal healing with in vivo experiments in rats in 1989. We found that
polyamine deficiency seriously inhibited would healing in the gastrointestinal
tract caused by stress [40], corticosterone administration [41] or hypertonic
saline [42, 43] and were eager to find at what point the polyamines were
required.
The restitution period, because it encompasses signaling events that may be
brief yet set a cascade of others in motion, called for an experimental model in
culture. In general, restitution in vitro occurs with the same coordinated reorga-
nization of the actin cytoskeleton as in vivo. Namely, myosin and actin are trans-
ported to the cell cortex, lamellipodia extend from surviving cells at the wound
edges, ruffles form on the outer edges of the lamellipodia, transverse stress
fibers contract, the tail retracts, lifts and detaches from the substrate [44–46].
Therefore, we developed an in vitro model of restitution using the IEC-6 cell
line, a well-characterized intestinal crypt cell line which originated from normal
rats by Quaroni [47] in 1979. IEC-6 cells were plated on a solubilized basement

a b
Fig. 2. The migrating edge of IEC-6

cells. The cells are shown at 3 h of migration
following 4 days of no treatment (a), DFMO
(b) or DFMO plus putrescine (c). F-actin is
stained with Texas Red phalloidin. The
abundance of F-actin stress fibers is obvious
in the control and DFMO/putrescine groups,
as are cell polarization and lamellipodia.
The DFMO group shows the scarcity of
interior stress fibers along with the heavy
cell cortex and lack of lamellipodia typical
c of polyamine-depleted cells.
membrane matrix (Matrigel) and allowed to grow to confluency. The monolayer

was wounded by scraping an area with a razor blade to remove cells. Migration
was assessed by counting the cells that migrated over the scraped area after a
fixed period, usually 6 h [48]. In this model, migration by IEC-6 cells after
4 days of treatment with the polyamine synthesis inhibitor DFMO was reduced
by ~80% and polyamine levels were reduced concomitantly. Putrescine, sper-
midine or spermine added with DFMO maintained migration within control
levels, showing that the reduced migration was due to the absence of polyamines
and not to DFMO per se. Microscopic observation of fixed DFMO-treated cells
revealed a drastic change in cytoskeletal architecture (fig. 2). Actin is present in
cells in two forms, filamentous actin (F-actin), a polymerized form in helical
conformation, and monomeric actin (G-actin), an unpolymerized form of single
monomers. In the polyamine-depleted cells, F-actin is concentrated in a heavy
actin cortex, and stress fibers are short and sparse in the cell interior. Cells lose
their polarization, and lamellipodia are minimal or nonexistent. The distribution
of F- and G-actin is altered, but total intracellular levels of F-actin and G-actin
do not change significantly [49].
Microtubules modulate the functional changes in focal adhesions [50–52].
While microtubules are not necessary for lamellipodial extension because it
depends on actin polymerization, they do influence cell morphology, polariza-
tion and the direction of migration [53]. Microtubules are required for normal
cell migration, tail retraction and the modulation of cell-matrix adhesion. When
microtubules are stabilized with nocodazole or taxol, the cells show decreased
ruffling and lamellipodial formation [54]. In vitro, microtubule assembly is
enhanced by polyamines [55], and in vivo, they are required for microtubule
formation [43].
Polyamine-depleted cells also show changes in cytoskeletal proteins, espe-
cially those associated with or bound to actin, for example, tropomyosin and
nonmuscle myosin II. Tropomyosin is a dimeric rod-shaped molecule that lies
along the groove of actin filaments where it can either increase or decrease
the stability and Ca2⫹ sensitivity of stress fibers. Sensitivity to fluctuations in
intracellular Ca2⫹ is important because changes in Ca2⫹ levels stimulate rapid
reorganization of the cytoskeleton [56]. Tropomyosin can modulate sensitivity
to Ca2⫹ levels by competing with myosin II and other actin-binding proteins
binding to actin. This competition affects the rigidity of stress fibers, cell motil-
ity and reorganization of the cytoskeleton [57]. In polyamine-depleted IEC-6
cells, tropomyosin associates with the remaining short stress fibers and the
heavy actin cortex. In addition, a lower molecular mass tropomyosin isoform
(~25 kD), thought to associate primarily with short stress fibers, appears [58].
The maintenance and dynamics of the cytoskeleton are important func-
tions of nonmuscle myosin II in the cell. These activities control cell shape [59],
response to changing Ca2⫹ levels [60] and transport and distribution of
organelles and endocytosed molecules [61]. Polyamine deficiency decreases
nonmuscle myosin II protein by 75% and changes its distribution so that,
instead of binding to the actin stress fibers that remain, myosin II concentrates
in patches scattered through the cytoplasm and on the cell cortex. Obviously,
these changes contribute to the severe reduction of migratory ability in
polyamine deficient cells. Normal levels of myosin II can be maintained by
supplying putrescine or spermidine along with DFMO [62]. Similar results as
well as a reduction of myosin II mRNA have been found in IEC-6 cells that
have been transfected to cause differentiation [3].
The intracellular Ca2⫹ concentration is a key element in the induction of
some intracellular signaling events. An increase in free cytosolic Ca2⫹ is a

stimulus for cell contraction [63] and migration [64, 65] in restitution. In gas-
trointestinal epithelial cells, polyamine-dependent K⫹ (Kv) channels determine
the membrane potential and driving force for Ca2⫹ and indirectly regulate
cytosolic free Ca2⫹ concentration. Polyamines stimulate the expression and
activity of Kv channels. Since intestinal epithelial cells do not express voltage-
dependent Ca2⫹ channels, activation of the Kv channels by polyamines causes
membrane hyperpolarization that increases the driving force for Ca2⫹ influx,
raises free Ca2⫹ concentration in the cytosol and stimulates migration. Poly-
amine deficiency, on the other hand, results in hypopolarization, fewer
Kv channels, less activation of Kv channels and reduced cell migration. As
with other effects of polyamine deficiency, supplementation of DFMO with
putrescine, spermidine or spermine reverses these effects [4]. Subsequently,
these investigators have shown that the formation of myosin II stress fibers is
partially initiated by Ca2⫹ influx correlated with RhoA levels. The activity of
RhoA was not measured [66].
Attachment to the extracellular matrix (ECM) is the first step in migration,
and IEC-6 cells require polyamines for attachment to be successful [67]. The
migration of intestinal epithelial cells also requires RhoA as shown by the inhi-
bition of migration by the C3 toxin of Clostridium difficile which ribosylates
RhoA [68, 69]. RhoA requires polyamines for its activity, synthesis and nor-
mal functioning, but RhoA activity alone is not sufficient for migration in
polyamine-depleted cells. We found that polyamine depletion inhibits migra-
tion to the same extent whether in wild-type IEC-6 cells, those transfected
with empty vector, or those stably transfected with constitutively active RhoA
(*V14). Therefore, we believe that polyamines must be essential either for
the activation of RhoA or for a step downstream from RhoA, or both [unpubl.
data].
Signals from the wound, probably cytokines and growth factors, stimulate
signaling and structural molecules to aggregate around small groups of ligand-
bound integrins to produce nascent focal adhesions at the leading edge of the
cell. These early adhesions are the source of strong propulsive forces [70].
Stress fibers increase as a result of actin polymerization and myosin light chain
kinase (MLCK) phosphorylation of myosin II. The ligand-bound integrins relay
signals from the ECM which convey its composition and density, determining
factors in the success and rate of attachment and migration. Attachment acti-
vates focal adhesion kinase (FAK) which then activates paxillin, ␣-actinin and
other molecules in the focal adhesions that participate in attachment to the
cytoskeleton, spreading and migration.
Fusion proteins such as green fluorescent protein (GFP) have made it pos-
sible to study the locations of the various molecules involved in migration,
either in fixed or live cells. The activated molecules enter and exit the leading
edge in a hierarchical manner that insures, at first, attachment and force for
propulsion, then stabilization, and finally detachment. Using fusion proteins,
Laukaitis et al. [71] have investigated the dynamics of ␣5 integrin, paxillin and
␣-actinin during the formation and disassembly of focal adhesions in migrating
cells. They have found that all three are present in new protrusions and have
shown, for each, the sequence of their first appearance, route and fate through-
out the stages of focal adhesion development. To our knowledge, no informa-
tion on the effect of polyamine deficiency on the sequence of these processes
is available.
We have found that integrin signaling is decreased significantly in
polyamine-deficient cells. Integrin signaling requires heterodimer configura-
tion for activity, and immunocytochemistry reveals that little co-localization of
the integrin heterodimers ␣2 and ␤1 occurs in DFMO-treated cells although
integrin protein is not decreased. Since the integrins activate FAK activity, pax-
illin phosphorylation is decreased as well. Cell attachment, focal adhesions,
actin stress fibers, spreading and migration are all reduced in these cells [72].
Integrin mediates adhesion to the ECM by signaling through the small
GPTases Rho, Rac1 and Cdc42 [73, 74]. Activated Rho, Cdc42 and Rac1 reg-
ulate cell motility directly and indirectly. Cdc42 and Rac1 modulate the forma-
tion of lamellipodia, new focal adhesions, ruffling at the leading edge [75–77],
and cell spreading [78]. RhoA stimulates actin polymerization, the formation of
stress fibers and the continued development of nascent focal adhesions [76, 77,
79]. As cells reach confluence, Rac1 inhibits motility in lamellipodia by
directly down-regulating Cdc42. Similarly, RhoA slows lamellipodial forma-
tion by down-regulating both Cdc42 and Rac1 [80].
What is the mechanism of lamellipodia extension? In epithelial cells, phos-
phatidylinositol 4,5-bisphosphate (PIP3), Cdc42, the neuronal Wiskott-Aldrich
syndrome protein (N-WASP) and the Arp2/3 complex of 7 proteins (Arp2/3) are
involved. The signaling pathway leads from PIP3 to actin monomer nucleation
through Cdc42, N-WASP and ARP2/3. PIP3 indirectly activates Cdc42 which,
with PIP3, activates N-WASP. N-WASP activates ARP2/3 which localizes on
filaments at the leading edge and nucleates actin monomers [81]. The nucleated
monomers rapidly polymerize into new filaments that branch from the sides of
the original filaments at a 70⬚ angle to form a dense dendritic array at the
migrating edge. Rapid polymerization of these filaments is considered to be
responsible for the extension of the lamellipodia [75, 82, 83].
Actin stress fibers are indispensable for migration because they provide a
stable form for the cell as well as the force for contraction. A new view of stress
fibers has recently been reported by Katoh et al. [84]. They have used inhibitors
of Rho kinase and MLCK to demonstrate that there are two types of stress fibers
with different functions in distinct areas of the cell. Peripheral stress fibers are

Wounding
ECM
Cdc42 N-WASP Arp2/3
Rac PAK Stress fiber
Signals disassembly
䉬
K ⫹, 䉬
Ca 2⫹ *Rho
*Integrins Rho-kinase Actin polymerization

MLCK
*FAK
*Focal adhesion *Stress fiber formation
*Paxillin
Vinculin
ERMs Spreading
*Migration
*Restitution
Fig. 3. A simplified model of the route to restitution showing areas where polyamines
are required. Abbreviations are as follows: extracellular matrix (ECM); focal adhesion kinase
(FAK); ezrin, radixin, moesin, actin-binding proteins of the focal adhesion plaque (ERMs);
myosin light chain kinase (MLCK); neuronal Wiskott-Aldrich syndrome protein (N-WASP);
paxillin kinase (PAK). *L.R. Johnson laboratory; 䉬J.Y. Wang laboratory.
used for the formation of lamellipodia and filopodia and for maintaining an
extended morphology. They depend on the activity of MLCK, a specific kinase
for the activation of myosin II. Central stress fibers and their associated focal
adhesions depend on the activity of Rho-kinase, although it is not specific for
them. Whether polyamines function similarly in these two types of stress fibers
is not known.
Where Do Polyamines Make a Difference?
Under conditions in which polyamine levels are deficient, processes required

for signaling, attachment, cytoskeletal remodeling, stress fiber formation,
spreading and migration cannot be carried out in a normal manner in intestinal
epithelial cells. Integrin ␣2/␤1 heterodimers, FAK, paxillin, RhoA and myosin II
are all reduced in amount or activity. These deficiencies and their effects
demonstrate the pervasive involvement of polyamines. The following model
shows some of these relationships and indicates steps that have been shown to
depend on an adequate supply of polyamines (fig. 3). It is immediately obvious
that many points of possible polyamine involvement remain for investigation.
References
1 Silen W: Mechanisms for rapid re-epithelialization of the gastric mucosal surface. Annu Rev
Physiol 1985;47:217–229.
2 McCormack SA, Viar MJ, Johnson LR: Polyamines are necessary for cell migration by a small
intestinal crypt cell line. Am J Physiol 1993;264:G367–G374.
3 Rao JN, Li J, Li L, Bass BL, Wang JY: Differentiated intestinal epithelial cells exhibit increased
migration through polyamines and myosin II. Am J Physiol 1999;277:G1149–G1158.
4 Wang JY, Wang J, Golovina VA, Li L, Platoshyn O, Yuan JX: Role of K(⫹) channel expression in
polyamine-dependent intestinal epithelial cell migration. Am J Physiol 2000;278:C303–C314.
5 Tabor CW, Tabor H: Polyamines, Annu Rev Biochem 1984;53:749–790.
6 Cohen S: A Guide to the Polyamines, ed 1. New York, Oxford University Press, 1998.
7 Seiler N, Atanassov CL, Raul F: Polyamine metabolism as target for cancer chemoprevention.
Int J Oncol 1998;13:993–1006.
8 Svensson F, Kockum I, Persson L: Diethylglyoxal bis(guanylhydrazone), a potent inhibitor of
mammalian S-adenosylmethionine decarboxylase. Effects on cell proliferation and polyamine
metabolism in L1210 cells. Mol Cell Biochem 1993;124:141–147.
9 Kaminska B, Kaczmarek L, Grzelakowska-Sztabert B: Inhibitors of polyamine biosynthesis affect
the expression of genes encoding cytoskeletal proteins. FEBS Lett 1992;304:198–200.
10 Poulin R, Pegg AE: Stable intracellular acidification upon polyamine depletion induced by
␣-difluoromethylornithine or N1,N12-bis(ethyl)spermine in L1210 leukaemia cells. Biochem J
1995;312:749–756.
11 Steffanelli C, Bonavita F, Stanic I, et al: Spermine triggers the activation of caspase-3 in a
cell-free model of apoptosis. FEBS Lett 1999;451:95–98.
12 Bratton D, Fadok V, Richter D, et al: Polyamine regulation of plasma membrane phospholipid
flip-flop during apoptosis. J Biol Chem 1999;274:28113–28120.
13 Hashimoto Y, Hibasami H, Tamaki S, et al: Induction of apoptotic cell death in human hepato-
cellular carcinoma SK-HEP-1 cells by a polyamine synthesis inhibitor, methylglyoxal bis
(cyclopentylamidiinohydrazone). Anticancer Drugs 1999;10:323–327.
14 Wallace HM: The physiological role of the polyamines. Eur J Clin Invest 2000;30:1–3.
15 Brunton VG, Grant MH, Wallace HM: Mechanisms of spermine toxicity in baby-hamster kidney
(BHK) cells. The role of amine oxidases and oxidative stress. Biochem J 1991;280:193–198.
16 Mitchell JL, Judd GG, Bareyal-Leyser A, Ling SY: Feedback repression of polyamine transport is
mediated by antizyme in mammalian tissue culture cells. Biochem J 1994;299:19–22.
17 He Y, Suzuki T, Kashiwagi K, Igarashi K: Antizyme delays the restoration by spermine of growth
of polyamine-deficient cells through its negative regulation of polyamine transport. Biochem
Biophys Res Commun 1994;203:608–614.
18 Sakata K, Kashiwagi K, Igarashi K: Properties of a polyamine transporter regulated by antizyme.
Biochem J 2000;347:297–303.
19 Almrud J, Oliveira M, Kern A, Grishin N, Phillips M, Backert M: Crystal structure of human
ornithine decarboxylase at 2.1 Å resolution: Structural insights to antizyme binding. J Mol Biol
2000;295:7–16.
20 Murakami Y, Matsufuji S, Hayaski S, Tanahashi N, Tanaka K: Degradation of ornithine decar-
boxylase by the 26S proteasome. Biochem Biophys Res Commun 2000;267:1–6.
21 Ivanov I, Gesteland R, Atkins J: A second mammalian antizyme: Conservation of programmed
ribosomal frameshifting. Genomics 1998;52:119–129.
22 Ivano I, Rohrwasser A, Terreros D, Gesteland RF, Atkins J: Discovery of a spermatogenesis
stage-specific ornithine decarboxylase antizyme: Antizyme 3. Proc Natl Acad Sci USA 2000;
97:4808–4813.
23 Saito T, Hascilowicz T, Ohkido I, et al: Two zebrafish (Danio rerio) antizymes with different
expression and activities. Biochem J 2000;345:99–106.
24 Seiler N, Dezeure F: Polyamine transport in mammalian cells. Int J Biochem 1990;22:211–218.
25 Minchin RF, Raso A, Martin RL, Ilett KJ: Evidence for the existence of distinct transporters
for the polyamines putrescine and spermidine in B16 melanoma cells. Eur J Biochem 1991;200:
457–462.

26 Toursarkissian B, Endean ED, Aziz SM: Characterization of polyamine transport in rat aortic
smooth muscle cells. J Surg Res 1994;57:401– 407.
27 Aziz SM, Olson JW, Gillespie MN: Multiple polyamine transport pathways in cultured pulmonary
artery smooth muscle cells: Regulation by hypoxia. Am J Respir Cell Mol Biol 1994;10:160–166.
28 McCormack SA, Johnson LR: Putrescine uptake and release by a normal rat small intestine crypt
cell line, IEC-6. Exp Cell Res 1991;193:241–252.
29 McCormack SA, Johnson LR: Putrescine uptake and release by colon cancer cells. Am J Physiol
1989;256:G868–G877.
30 Brabender J, Lord R, Danenberg K, et al: Upregulation of ornithine decarboxylase mRNA expres-
sion in Barrett’s esophagus and Barrett’s-associated adenocarcinoma. J Gastrointest Surg 2001;
5:174–182.
31 Masuda M, Betancourt L, Matsuzawa T, et al: Activation of rho through a cross-link with
polyamines catalyzed by Bordetella dermonecrotizing toxin. EMBO J 2000;19:521–530.
32 Sakamoto K, Fujiyama Y, Bamba T: Altered polyamine biosynthesis with aging after massive
proximal small bowel resection in rat. J Gastroenterol 1996;31:338–346.
33 Majumdar AP, Fligiel SE, Jaszewski R: Gastric mucosal injury and repair: Effect of aging. Histol
Histopathol 1997;12:491–501.
34 McCormack SA, Blanner PM, Zimmerman BJ, et al: Polyamine deficiency alters EGF distribu-
tion and signaling effectiveness in IEC-6 cells. Am J Physiol 1998;274:C192–C205.
35 Raul F, Galluser M, Schleiffer R, Gosse F, Hasselmann M, Seiler N: Beneficial effects of
L-arginine on intestinal epithelial restitution after ischemic damage in rats. Digestion 1995;
56:400–405.
36 Singer R, Sagiv M, Levinsky H, Maayan R, Segenreich E, Allalouf D: The influence of seminal
plasma and polyaminic substances on the motility of isolated human sperm. Int J Fertil 1989;
34:224–230.
37 Manni A, Wright C: Polyamines as mediators of estrogen action on the growth of experimental
breast cancer in rats. J Natl Cancer Inst 1984;73:511–514.
38 Lowenberg B, Thibault J, Lawrence C, Sodek J: The influence of chemically-induced modifica-
tions of root surfaces on cell migration, attachment, and orientation. J Dent Res 1986;65:
1010–1015.
39 Iwata M, Nakano E: Characterization of sea-urchin fibronectin. Biochem J 1983;215:205–208.
40 Wang JY, Johnson RL: Luminal polyamines stimulate repair of gastric mucosal stress ulcers.
Am J Physiol 1990;259:G584–G592.
41 Wang JY, Johnson LR: Gastric and duodenal mucosal ornithine decarboxylase and damage after
corticosterone. Am J Physiol 1990;258:G942–G950.
42 Banan A, Wang JY, McCormack S, Johnson L: Relationship between polyamines, actin distri-
bution, and gastric mucosal healing in rats. Am J Physiol 1996;271:G893–G903.
43 Banan A, McCormack S, Johnson L: Polyamines are required for microtubule formation during
gastric mucosal healing. Am J Physiol 1998;274:G879–G885.
44 Conrad P, Giuliano K, Fisher G, Collins K, Matsudaira P, Taylor D: Relative distribution of actin,
myosin I, and myosin II during the wound healing response of fibroblasts. J Cell Biol 1993;120:
1381–1391.
45 Dignass AU: Mechanisms and modulation of intestinal epithelial repair. Inflamm Bowel Dis
2001;7:68–77.
46 Lauffenburger DA, Horwitz AF: Cell migration: A physically integrated molecular process. Cell
1996;84:359–369.
47 Quaroni A, Wands J, Trelstad RL, Isselbacher KJ: Epithelioid cell cultures from rat small intes-
tine. Characterization by morphologic and immunologic criteria. J Cell Biol 1979;80:248–265.
48 McCormack SA, Viar MJ, Johnson LR: Migration of IEC-6 cells: A model for mucosal healing.
Am J Physiol 1992;263:G426–G435.
49 McCormack SA, Ray RM, Blanner PM, Johnson LR: Polyamine depletion alters the relationship
of F-actin, G-actin, and thymosin ␤4 in migrating IEC-6 cells. Am J Physiol 1999;276:
C459–C468.
50 Kaverina I, Krylyshkina O, Small JV: Microtubule targeting of substrate contacts promotes their
relaxation and dissociation. J Cell Biol 1999;146:1033–1044.
51 Elbaum M, Chausovsky A, Levy E, Shtutman M, Bershadsky A: Microtubule involvement in
regulating cell contractility and adhesion-dependent signalling: A possible mechanism for polar-
ization of cell motility. Biochem Soc Symp 1999;65:147–172.
52 Pletjushkina OJ, Belkin AM, Ivanova OJ, Oliver T, Vasiliev JM, Jacobson K: Maturation of cell-
substratum focal adhesions induced by depolymerization of microtubules is mediated by increased
cortical tension. Cell Adhes Commun 1998;5:121–135.
53 Kaverina I, Krylyshkina O, Gimona M, Beningo K, Wang YL, Small JV: Enforced polarisation
and locomotion of fibroblasts lacking microtubules. Curr Biol 2000;10:739–742.
54 Ballestrem C, Wehrle-Haller B, Hinz B, Imhof B: Actin-dependent lamellipodia formation and
microtubule-dependent tail retraction control-directed cell migration. Mol Biol Cell 2000;
11:2999–3012.
55 Wolff J: Promotion of microtubule assembly by oligocations: Cooperativity between charged
groups: Biochemistry 1998;37:10722–10729.
56 Smillie LB: Structure and functions of tropomyosins from muscle and non-muscle sources. Trends
Biochem Sci 1979;4:151–155.
57 Lehman W, Hatch V, Korman V, et al: Tropomyosin and actin isoform modulate the localization of
tropomyosin strands on actin filaments. J Mol Biol 2000;302:593–606.
58 McCormack SA, Wang JY, Johnson LR: Polyamine deficiency causes reorganization of F-actin
and tropomyosin in IEC-6 cells. Am J Physiol 1994;267:C715–C722.
59 Xu XS, Lee E, Chen T, Kuczmarski E, Chisholm RL, Knecht DA: During multicellular migration,
myosin II serves a structural role independent of its motor function. Dev Biol 2001;232:255–264.
60 Eddy R, Pierini L, Matsumura F, Maxfield F: Ca2⫹-dependent myosin II activation is required for
uropod retraction during neutrophil migration. J Cell Sci 2000;113:1287–1298.
61 Wu X, Jung G, Hammer JA 3rd: Functions of unconventional myosins. Curr Opin Cell Biol
2000;12:42–51.
63 Clapham D: Calcium signaling. Cell 1995;80:259–268.
64 Bilato C, Pauly R, Melillo G, et al: Intracellular signaling pathways required for rat vascular
smooth muscle cell migration. Interactions between basic fibroblast growth factor and platelet-
derived growth factor. J Clin Invest 1995;96:1905–1915.
65 Ehring GR, Szabo IL, Jones MK, Sarfeh IJ, Tarnawski AS: ATP-induced Ca2⫹-signaling enhances
rat gastric microvascular endothelial cell migration. J Physiol Pharmacol 2000;51: 799–811.
66 Rao JN, Li L, Golovina VA, et al: Ca2⫹-RhoA signaling pathway required for polyamine-
dependent intestinal epithelial cell migration. Am J Physiol Cell Physiol 2001;280:C993–C1007.
67 Santos MF, Viar MJ, McCormack SA, Johnson LR: Polyamines are important for attachment of
IEC-6 cells to extracellular matrix. Am J Physiol 1997;273:G175–G183.
68 Chardin P, Boquet P, Madaule P, Popoff M, Rubin E, Gill D: The mammalian G protein rhoC is
ADP-ribosylated by Clostridium botulinum exoenzyme C3 and affects actin microfilaments in
Vero cells. EMBO J 1989;8:1087–1092.
69 Santos MF, McCormack SA, Guo Z, et al: Rho proteins play a critical role in cell migration
during the early phase of mucosal restitution. J Clin Invest 1997;100:216–225.
70 Beningo K, Dembo M, Kaverina I, Small J, Wang Y: Nascent focal adhesions are responsible for
the generation of strong propulsive forces in migrating fibroblasts. J Cell Biol 2001;153:
881–887.
71 Laukaitis CM, Webb DJ, Donais K, Horwitz AF: Differential dynamics of ␣5 integrin, paxillin,
and ␣-actinin during formation and disassembly of adhesions in migrating cells. J Cell Biol 2001;
153:1427–1440.
72 Ray MR, Viar MJ, McCormack SA, Johnson LR: Focal adhesion kinase signaling is decreased in
polyamine depleted IEC-6 cells. Am J Physiol 2001;281:C475–C485.
73 Clark E, King W, Brugge J, Symons M, Hynes R: Integrin-mediated signals regulated by members
of the rho family of GTPases. J Cell Biol 1998;142:573–586.
74 Del Pozo M, Price L, Alderson N, Ren X, Schwartz M: Adhesion to the extracellular
matrix regulates the coupling of the small GTPase Rac to its effector PAK. EMBO J 2000;19:
2008–2014.

75 Borisy G, Svitkina T: Actin machinery: Pushing the envelope. Curr Opin Cell Biol 2000;12:
104–112.
76 Nobes CD, Hall A: Rho, rac and cdc42 GTPases regulate the assembly of multimolecular focal
complexes associated with actin stress fibers, lamellipodia and filopodia. Cell 1995;81:53–62.
77 Nobes CD, Hall A: Rho GTPases control polarity, protrusion, and adhesion during cell movement.
J Cell Biol 1999;144:1235–1244.
78 Price LS, Leng J, Schwartz MA, Bokoch GM: Activation of Rac and Cdc42 by integrins mediates
cell spreading. Mol Biol Cell 1998;9:1863–1871.
79 Ridley AJ, Hall A: The small GTP-binding protein rho regulates the assembly of focal adhesions
and actin stress fibers in response to growth factors. Cell 1992;70:389–399.
80 Cox E, Sastry S, Huttenlocher A: Integrin-mediated adhesion regulates cell polarity and mem-
brane protrusion through the Rho family of GTPases. Mol Biol Cell 2001;12:265–277.
81 Rohatgi R, Ho HY, Kirschner MW: Mechanism of N-WASP activation by CDC42 and phos-
phatidylinositol 4,5-bisphosphate. J Cell Biol 2000;150:1299–1310.
82 Bailly M, Ichetovkin I, Grant W, et al: The F-actin side binding activity of the Arp2/3 complex is
essential for actin nucleation and lamellipod extension. Curr Biol 2001;11:620–625.
83 Welch MD, DePace AH, Verma S, Iwamatsu A, Mitchison TJ: The human Arp2/3 complex is
composed of evolutionarily conserved subunits and is localized to cellular regions of dynamic
actin filament assembly. J Cell Biol 1997;138:375–384.
84 Katoh K, Kano Y, Amano M, Kaibuchi K, Fujiwara K: Stress fiber organization regulated by
MLCK and Rho-kinase in cultured human fibroblasts. Am J Physiol Cell Physiol 2001;280:
C1669–C1679.
Leonard R. Johnson, PhD, Department of Physiology, College of Medicine,

The University of Tennessee Health Science Center, 894 Union Avenue,
Room 426 Nash Building, Memphis, TN 38163 (USA)
Tel. ⫹1 901 448 7088, Fax ⫹1 901 448 7126, E-Mail ljohn@physiol.utmem.edu
Epithelial Restitution and

Physical Stress
Taro Osada a, Sumio Watanabe b, Nobuhiro Sato a
a
Department of Gastroenterology, Juntendo University School of Medicine,
Tokyo and b First Department of Internal Medicine Akita University School of
Medicine, Akita, Japan
The wound repair process of peptic ulcer involves epithelial and mes-
enchymal restoration. Especially, the repair of gastric mucosal lesions has been
investigated with regard to gastric mucosal blood flow, prostaglandins, growth
factors, the mucus layer, and gastric acid secretion in vivo and in vitro [1–5].
Cells such as gastric epithelial cells, vascular endothelial cells, fibroblasts and
smooth muscle cells continuously receive repetitive physical stress by adaptive
relaxation relating to food storage, peristalsis and fasting contraction such as
the interdigestive migrating complex. This repetitive physical stress might play
an important role on the process of gastric wound repair. The effect of mechan-
ical strain on vascular smooth muscle cells and vascular endothelial cells in the
cardiovascular system in order to reveal the mechanism of hypertension and
atherosclerosis has been well studied. Pedel et al. [6] examined the effect of
mechanical strain on vascular smooth muscle cells of the internal mammary
artery and saphenous vein and found that mechanical strain increased cell num-
ber and DNA synthesis in the saphenous vein but not in the internal mammary
artery. Moreover, mechanical strain induces growth of neonatal rat smooth
muscle cells and increases the autocrine response to platelet-derived growth
factor (PDGF) [7], and increases extracellular matrix (ECM) accumulation
(e.g. collagen and fibronectin) induced by transforming growth factor ␤1 (TGF␤1)
[8]. Mechanical strain also increases the expression of vascular endothelial
growth factor (VEGF) [9]. VEGF is a potent endothelial cell specific mitogen
that induces marked increases in vascular endothelial permeability. In vascular
endothelial cells, mechanical strain increases the production of endothelin-1
[10], nitric oxide [11] and prostacyclin [12]. It has been thought that VEGF may
act in a paracrine loop to regulate endothelial growth and permeability [9].
Fig. 1. Phase-contrast microphotographs showing the process of wound restoration in
primary gastric epithelial cells. In the control cultures (not subjected to strain), epithelial
restoration was completed at 48 h after wounding.
This review focuses on the effect of mechanical strain to gastric epithelial restora-
tion with original wound repair model in vitro compared with the response of
cells isolated from other organs.
Gastric Wound Repair Model
Wound repair of gastric lesions frequently has been described using in vivo
systems. However, these in vivo experimental models pose difficulty in quanti-
tatively investigating the precise role of modulating factors and the capacity for
repair of gastric epithelial cells. We have created an original gastric epithelial
wound repair model in vitro according to methods described previously [13].
We used gastric epithelial cells isolated from Japanese white rabbits and a
gastric epithelial cell line isolated from normal Wistar rat (RGM-1). Briefly,
after making complete monolayers of these cells, a round artificial wound of
constant size was made using a pencil-type mixer with a rotating silicone tip. It
is best to use a silicone tip for this purpose, as it does not damage the surface
material of the culture dishes. The process of gastric epithelial wound repair was
monitored. In this gastric epithelial wound repair model, just after wounding by
mechanical cell denudation, the cells located at the wound edge began to migrate
toward the center of the wound until the cell-free area completely closed.
In control cultures of the gastric epithelial cells isolated from Japanese
white rabbits, the size of the cell-free area closed completely within 48 h after
wounding (fig. 1). Just after wounding, the cells on the edge of wound began
to form lamellipodia, which ruffled and moved to the center of the wound.
Lamellipodia disappeared after complete restoration of the monolayer.
Proliferating cells were detected by monoclonal anti-5⬘-bromodeoxyuridine
Osada/Watanabe/Sato 58
Fig. 2. Microphotographs showing BrdU-positive proliferative cells during restoration
after wounding in primary gastric epithelial cells. They were detected around the margin of
the wound. Their number was highest in the 24–36 h group and then decreased. Fewer BrdU-
positive cells were detected until initial 24 h. W: artificial wound.
(BrdU) antibody. Gastric epithelial cell proliferation was negligible in controls

during the first 24 h after wounding, peaked at 36 h and returned to baseline by
48 h. BrdU-positive cells were detected around the wound, but barely detected
in cells located ⬎1,000 ␮m from the wound edge during the wound repair [14]
(fig. 2). In the RGM-1 cells wound repair model, the speed of wound repair
was slower than that of primary gastric epithelial cells isolated from Japanese
rabbits. The cell-free area of RGM-1 cells closed completely about 72 h after
wounding (fig. 3). However, the number of BrdU-positive cells were detected
more than that of primary gastric epithelial cells during restoration (fig. 4). The
role of cell migration was much more important in gastric epithelial restoration
than cell proliferation. These results suggest that gastric epithelial cell restora-
tion consists of two steps, early stage cell migration, followed by both prolifer-
ation and migration. The restoration process of RGM-1 cells was similar to that
of primary gastric epithelial restoration.
Application of Strain to Gastric Epithelial Cells
Detailed study of the direct effect of mechanical force on gastric epithelial

restoration has until recently not be done by the lack of suitable research tech-
nology. A new system, the Flexercell Strain Unit (Flexcell Corp.) (fig. 5), was
developed and mainly utilized for cardiovascular cell biology. With this system,
mechanical strain was applied to adherent cells to the silicone rubber membrane.
This system utilizes vacuum pressure regulated by a solenoid valve to deform
Epithelial Restitution and Physical Stress 59

a
Fig. 3. Phase-contrast microphotographs showing the process of wound restoration in

RGM-1 cells. In the control cultures (a) (not subjected to strain), mucosal restoration was
completed at 72 h after wounding. In the 10% strain groups (b), the process of restoration
was retarded by mechanical strain.
a b c
d e f
Fig. 4. Microphotographs showing BrdU-positive proliferative cells during restoration

after wounding. In the control (a–c) and strain (d–f ) groups, they were detected around the
margin of the wound. Their number was highest in the 24–48 h group (b and e) and then
decreased in the 48–72 h group (c and f ). Fewer BrdU-positive cells were detected in the
10% strain groups than in controls, but the difference was not statistically significant until
24 h (a and d). W: artificial wound.
Rest Stretching
35mm 35mm
Flexible membrane
a Vacuum
Cell stretching
Cell
Fig. 5. Schema of cell stretching with the Flexercell Strain Unit. Precise vacuum level
is applied to the system. The culture plate bottoms are deformed by a known percentage,
which is translated to the culture cells, resulting strain (a). The force on attached cells is
primarily in one axis, almost radial strain (b).
the silicone rubber membrane on which the cells have been cultured. The fre-
quency, duration and magnitude of the applied strain can be varied by computer.
The frequency of mechanical strain corresponds to each organic rhythm, cardio-
vascular materials are adjusted to pulse regularis and respiratory cells are
matched with respiratory rate. The frequency of gastric peristaltic movement is
3 cycles/min in humans [15] and 4–6 cycles/min in small animals [16, 17]. This
system can be studied in a biomechanically active environment.
Effect of Strain on Gastric Epithelial Restoration
Using the original wound repair model and the Flexercell Strain Unit, we
can analyze the effect of mechanical strain to gastric epithelial restoration,

which includes early stage cell migration (restitution) followed by both prolif-
eration and restitution.
In controls of RGM-1 cells (not subjected to mechanical strain), the size
of the cell-free area decreased gradually and closed completely within 72 h after
wounding (fig. 3a). In the strain groups, wound repair was slower than in con-
trols (fig. 3b). Wound repair was inhibited by mechanical strain in a strength-
dependent manner [18]. The difference in the speed of restoration was caused
by a difference in the speed of cell migration. Therefore, mechanical strain
mainly inhibited gastric epithelial cell migration.
Proliferating cells were detected by BrdU labeling index. This experiment
of RGM-1 cells compared three groups 0–24, 24–48 and 48–72 h after wound-
ing. In the process of wound repair, the number of BrdU-positive cells increased
gradually, peaked in the 24–48 h group (fig. 4b, e), and then decreased (fig. 4c, f)
in both the control and strain groups. Detection of BrdU-positive cells decreased
in a strain strength-dependent manner around the wound [18]. In this study,
BrdU-positive cells were hardly detected in cells located far from the wound
edge during the wound repair in both control and strain groups. Thus, these
studies showed that mechanical strain inhibited gastric epithelial proliferation
around the wound. Therefore, mechanical strain inhibited both gastric epithelial
cell migration and proliferation around the wound.
Effect of Strain on the Cells Isolated from Other Organs
We showed the effect of mechanical strain on cultured primary gastric

smooth muscle cell restoration with a similar wound repair model [19]. Gastric
smooth muscle cells were isolated from Japanese male white rabbits. The
mechanical strain applied to gastric smooth muscle cells was equal in strength
and duration to gastric epithelial cells (RGM-1). In controls of gastric smooth
muscle cells, the size of the cell-free area almost closed at 48 h after wounding.
In the strain group, wound repair was slower than in controls, a similar result
to gastric epithelial cells. The wound repair of gastric smooth muscle cells was
also inhibited by the restoration in a strain strength-dependent manner.
Proliferating cells around the wound were also evaluated by BrdU labeling
index, but the number of proliferating cells was equal to controls in the strain
groups. Therefore, cell migration but not proliferation of primary gastric
smooth muscle cells as well as gastric epithelial cells was inhibited by mechan-
ical strain throughout the process of wound healing.
Recently, some studies have shown that mechanical strain affects other
original epithelial cells (table 1). Savla and Waters [20] examined the effect of
mechanical strain, cyclic elongation and compression, on airway epithelium
Table 1. Effect of mechanical strain on various cells in vitro
Cell type Type of mechanical strain Migration Proliferation Mechanism

and experimental design
Gastric epithelial 5 and 10% elongation ↓ ↓ Cytoskeletal dysfunction

cells (RGM-1) (average) Inhibition of integrin
5 cycles/min aggregation (?)
Wound repair model
Gastric smooth 5 and 10% elongation ↓ → Cytoskeletal dysfunction
muscle cells (average)
(primary culture) 5 cycles/min
Wound repair model
Intestinal epithelial 10% elongation (average) Not done ↑ Integrin-mediated
cells (Caco-2) 10 cycles/min tyrosine kanase and
PKC activation
Airway epithelial 9% elongation (average) ↓ ↓ Not evaluated
cells (CTE, NHBE) 10–30 cycles/min
Wound repair model
Pulmonary epithelial 20% elongation Not done ↑ Integrin-mediated
cells (H441) (maximum) tyrosine kanase and
60 cycles/min PKC activation
and assessed the repair of wound in vitro. Both cyclic elongation and com-
pression significantly retarded the repair of wound closure by inhibiting the
cell migration, spread and proliferation. Their results were similar to the
response of mechanical strain to gastric epithelial restoration. On the other
hand, pulmonary epithelial (H441) cells, in the human adenocarcinoma cell
line, are accelerated cell proliferation by cyclic mechanical strain in a
strength-dependent manner [21]. Basson et al. [22, 23] examined the effect
of mechanical strain on human intestinal epithelial Caco-2 cell monolayers.
They assessed the cell proliferation and the expression of the brush border
enzymes alkaline phosphatase and dipeptidyl dipeptidase as marker of cell
differentiation. Mechanical strain promoted proliferation in a strain strength-
dependent manner, and modulated differentiation, selectively stimulating
dipeptidyl dipeptidase while inhibiting alkaline phosphatase, via tyrosine
kinase activity. Caco-2 cells are a well-differentiated human intestinal epithe-
lial cell line derived from a human colon cancer cells [24]. Therefore, while
the degree of cell differentiation and experimental design is different,
mechanical strain may produce diverse effects on epithelial cells derived from
stomach and intestine.

a b
Fig. 6. Fluorescence microphotographs showing actin staining at the margin of the

wound. In the control (not subjected to strain), widely spread lamellipodia and many filopo-
dia were observed (a) at the margin of the wound. In the strain group, few lamellipodia and
filopodia were observed (b) at the same place. W: artificial wound.
Mechanism of Mechanical Strain Retards Gastric

Epithelial Restoration
Effect of Strain on Gastric Epithelial Cytoskeletal

System during the Restoration
The actin cytoskeleton maintains cellular shape and plays an important role
in cell motility and cell division. Actin stress fibers consist of long bundles of
filaments that traverse the cell and are linked to the ECM through integrins
and focal adhesion complexes, while a highly compact meshwork of actin
filaments can be found at the leading edge of motile cells in lamellipodia and
ruffles. In addition, short bundles of actin filaments are often found protruding
from the cell surface, particularly in motile cells, to produce microspikes and
filopodia. Our results using this wound repair model with primary cultured gas-
tric epithelial cells suggested that during the restoration cytochalasin B, which
blocks the formation of F-actin, and wortmannin, a myosin light chain kinase
inhibitor, inhibited the wound repair effected by cell migration and proliferation
[13, 25]. Actin was detected by rhodamine phalloidin staining [26]. Polymerized
actin was stained in the lamellipodia and filopodia of the gastric epithelial cells
located on the wound margin in controls (fig. 6a). On the other hand, in cells
treated with mechanical strain, actin was rarely detected in the lamellipodia at
the same place (fig. 6b). The shape of control cells was more elongated, and
the gap between cells was wider compared to cells subjected to mechanical
strain [18]. In the gastric smooth muscle cells, actin could be detected on the
stress fibers directed toward the center of the wound throughout the cytoplasm
in controls. However, in cells treated with mechanical strain, the direction of
stress fiber was not consistent. Actin-containing stress fibers might be disrupted
by mechanical strain [19]. Ethanol also retards gastric epithelial restoration due
to cytoskeletal dysfunction at the wound margin [27]. The cells treated with 3%
ethanol showed narrowed lamellipodia. As mentioned above, the cytoskeletal
system plays an important role in gastric epithelial wound repair. Especially, it
is important for the cytoskeletal re-assembly and modeling at the cells located
in the wound margin during cell migration and proliferation.
Effect of Strain on Gastric Epithelial Adhesion Plaque to

ECM and Signal Transduction during Restoration
Physical stress within the gastrointestinal tract influences the ECM, as the
ligand for integrin modulates gastric mucosal wound repair. Previously we
found that the gastric epithelial wound-healing rate was different for each type
of ECM [28]. Adherence of cells to the ECM is mediated through integrins.
Integrins cluster while adhering to the ECM in the presence of growth factors
and recruit proteins such as vinculin, talin and paxillin to form focal adhesion
complexes [29, 30] which regulate cell migration and proliferation [31, 32]. The
Rho family is a well-known member of the Ras superfamily of small guanosine
triphosphatases (GTPases) which exhibits both GDP/GTP binding and GTPase
activities. The Rho family regulates signal transduction from receptors in the
membrane to a variety of cellular events related to cell morphology [33], motility
[34], cytoskeletal systems [35] and tumor invasion [36]. Rho triggers the for-
mation of contractile stress fibers and focal adhesion complexes in Swiss 3T3
cells [35]. Rac, a member of the Rho family, induced lamellipodial protrusions
and focal complexes in the lamellipodium in the same cells [26]. We assessed
the distribution of focal adhesion plaque protein, vinculin and small GTPase
RhoA and Rac1 by immunofluorescence staining during the process of gastric
epithelial (RGM-1) cell restoration [18]. Intense vinculin, RhoA and Rac1 flu-
orescence was observed in the control cells located at the margin of the wound,
especially at the edge of the cells located around the wound. When the control
group and the strain group were compared, vinculin, RhoA and Rac1 were
much more intensely stained in control cells than in the strain group, especially
in cells located at the margin of the wound.
Focal adhesion plaque proteins play an important role in mediating signals
from the ECM. Integrin aggregation and specific tyrosine phosphorylation result
in focal accumulation of some signal transduction molecules, including RhoA,
Rac1, Ras, Raf, MEK, ERK and JNK besides FAK (focal adhesion kinase). This
signal transduction activates mitogen-activated protein kinase (MAPK) and
modulates cell proliferation and gene expression [37]. Vinculin, talin, and
␣-actinin, adhesion plaque proteins, require both integrin aggregation and ligand

Migration toward wound center
L⫹0.1L
MAPK( ) 1
2 Ras superfamily Nucleus
Wound Rho, Rac Cdc42( )

3
FAK ( )
Vinculin( )
Integrin
ECM
Stretch force
Fig. 7. Schematic illustration for cellular migration and proliferation under conditions
of strain. We speculate the mechanism that mechanical strain inhibits gastric epithelial
restoration via inhibition of integrin aggregation. L: length of adherence cell; L⫹0.1L: the
dimension of adherent cells is increased an average of 10% by mechanical strain. 䊊 1 : signal
transduction to the nucleus for cell division (proliferation); 䊊
2 : signal for cell movement via
cytoskeletal system (migration); 䊊 3 : lamellipodial formation.
occupancy for accumulation. Furthermore, if tyrosine phosphorylation pro-

ceeds, F-actin and associated cytoskeletal proteins accumulate [38]. This signal
transduction suggests cell movement via the cytoskeletal system (migration). In
our study, the control cells located around the wound were more intensely
stained for vinculin, RhoA, Rac1 and lamellipodial formation than cells remote
from the wound edge. These migrating and proliferating cells were activated by
integrin-mediated signal transduction. However, in the strain group, vinculin,
RhoA, Rac1 and lamellipodial formation were poorly expressed, compared with
the control in the same area. The reason might be that repetitive stretching of
basement membrane inhibited the integrin aggregation in only the migrating and
proliferating cells (fig. 7).
On the other hand, mechanical strain activated FAK-dependent and
integrin-modulated MAPK in Caco-2 cells [39] (table 1). This activation appeared
responsible for the mitogenic effects. In human pulmonary epithelial H441
cells, the tyrosine kinase inhibitor genistein blocked the strain-induced prolif-
eration. Mechanical strain induced pulmonary epithelial cell proliferation via
tyrosine kinase activity [21]. In this case, mechanical strain might affect prolif-
eration cells activated by integrin-mediated signal transduction. In vascular
smooth muscle cells and endothelial cells, integrins appear to function as
mechanotransducers and play an important role in transducing mechanical
strain stimulation into intracellular signals. This phenomenon may be responsi-
ble for the difference of the cell type and experimental design.
The mechanical strain-induced delay of gastric epithelial cellular migra-
tion and proliferation might support the phenomenon that the liquid diet accel-
erated the repair of gastric ulcers compared with the chow diet in vivo [2].
Therefore, although many factors affect gastric ulcer healing, repetitive
mechanical strain may directly influence retardation of gastric mucosal wound
repair in vivo. In the future, it will be necessary to study whether mechanical
strain affects gastric mesenchymal cells and the interaction between gastric
epithelial cells and mesenchymal cells with respect to gastric ulcer healing.
References
1 Lacy E, Ito S: Rapid epithelial restitution of rat gastric mucosa after ethanol injury. Lab Invest 1984;
51:573–583.
2 Newell AA: Gastric mucosal blood flow following damage by ethanol, acid or aspirin.
Gastroenterology 1970;58:311–320.
3 Kelly DG, Code CF: Physiological and morphological characteristics of progressive disruption of
the canine gastric mucosal barrier. Dig Dis Sci 1979;29:424–441.
4 Paimela H: Restitution of frog gastric mucosa in vitro: Effect of basic fibroblast growth factor.
5 Kamada T, Kawano S, Sato N, et al: Gastric mucosal blood distribution and its changes in the heal-
ing process of gastric ulcer. Gastroenterology 1983;83:1541–1546.
6 Pedel HG, Yang Z, Segresser LV, et al: Implications of pulsatile stretch on growth of saphenous
vein and mammary artery smooth muscle. Lancet 1992;340:878–879.
7 Emily W, Qing M, Krishnankutty S, et al: Mechanical strain induces growth of vascular smooth
muscle cells via autocrine action of PDGF. J Cell Biol 1993;123:741–747.
8 O’Callaghan CJ, Gallachar B, Williams B: Mechanical strain increases TGF mRNA expression and
matrix production by human vascular smooth muscle cells. J Hypertens 1996;14(suppl 1):S84.
9 Kimber PDK, O’Callaghan CO, Williams B: Chronic cyclic stretch increases vascular permeabil-
ity factor mRNA and peptide production by human vascular smooth muscle cells. J Hypertens
1996;14(suppl 1):S204.
10 Wang DL, Tang CC, Wung BS, et al: Cyclic strain increases endothelin-1 secretion and gene
expression in human endothelial cells. Biochem Biophys Res Commun 1993;195:1050–1056.
11 Awolesi MA, Sessa WC, Sumpio BE: Cyclic strain upregulates nitric oxide synthase in cultured
bovine aortic endothelial cells. J Clin Invest 1995;96:1449–1454.
12 Sumpio BE, Banes AJ: Prostacyclin synthetic activity in cultured aortic endothelial cells under-
going cyclic mechanical deformation. Surgery 1988;104:383–389.
13 Watanabe S, Hirose M, Yasuda T, Sato N: Role of actin and calmodulin in migration and prolifer-
ation of rabbit gastric mucosal cells in culture. J Gastroenterol Hepatol 1994;9:325–333.
14 Watanabe S, Wang XE, Hirose M, et al: Platelet-derived growth factor accelerates gastric epithe-
lial restoration in a rabbit cultured cell model. Gastroenterology 1996;110:775–779.
15 Banes AJ, Gilbert J, Taylor D, et al: A new vacuum-operated stress-providing instrument that
applies static or variable duration cycle tension or compression to cells in vitro. J Cell Sci 1985;
75:35–42.
16 Rosales OR, Sumpio BE: Changes in cyclic strain increase inositol triphosphate and diacylgly-
cerol in endothelial cells. Am J Physiol 1992;262:C956–C962.
17 Cooper JA, Blum JD, Pollard TD: Acanthamoeba castellaniii capping protein: Properties, mech-
anism of action, immunologic cross-reactivity, and localization. J Cell Biol 1984;99:217–225.

18 Osada T, Watanabe S, Tanaka H, et al: Effect of mechanical strain on gastric cellular migration and
proliferation during mucosal healing: Role of Rho-dependent and Rac-dependent cytoskeletal
reorganisation. Gut 1999;45:508–515.
19 Tanaka H, Hirose M, Osada T, et al: Implications of mechanical stretch on wound repair of gas-
tric smooth muscle cells in vitro. Dig Dis Sci 2000;45:2470–2477.
20 Savla U, Waters CM: Mechanical strain inhibits repair of airway epithelium in vitro. Am J Physiol
1998;274:883–892.
21 Chess PR, Tola L, Finkelstein JN: Mechanical strain-induced proliferation and signaling in pul-
monary epithelial H441 cells. Am J Physiol Lung Cell Mol Physiol 2000;279:L43–L51.
22 Basson MD, Li GD, Hong F, et al: Amplitude-dependent modulation of brush border enzymes and
proliferation by cyclic strain in human intestinal Caco-2 monolayers. J Cell Physiol 1996;168:
476–488.
23 Han O, Sumpio BE, Basson MD: Mechanical strain rapidly redistributes tyrosine phosphorylated
proteins in human intestinal Caco-2 cells. Biochem Biophys Res Commun 1998;250:668–673.
24 Peterson MD, Mooseker MS: Characterization of the enterocyte-like brush border cytoskeleton of
the C2BBe clones of the human intestinal cell line, Caco-2. J Cell Sci 1992;102:581–600.
25 Watanabe S, Wang XE, Hirose M, et al: Effect of myosin light chain kinase inhibitor wortmannin
on the wound repair of cultured gastric mucosal cells. Biochem Biophys Res Commun 1994;
199:799–806.
26 Hall A: Rho GTPases and the actin cytoskeleton. Science 1998;279:509–514.
27 Murai T, Watanabe S, Hirose M, et al: Ethanol retards gastric epithelial restoration in monolayer
cultures. Dig Dis Sci 1996;41:2062–2069.
28 Mikami H, Watanabe S, Hirose M, et al: Role of extracellular matrix in wound repair by cultured
gastric mucosal cells. Biochem Biophys Res Commun 1994;202:285–292.
29 Burridge K, Fath K: Focal contacts: Transmembrane links between the extracellular matrix and the
cytoskeleton. Bioessays 1989;10:104–108.
30 Burridge K, Fath K, Kelly T, et al: Focal adhesions: Transmembrane junctions between the extra-
cellular matrix and the cytoskeleton. Annu Rev Cell Biol 1988;4:487–525.
31 Huttenlocher A, Sandborg RR, Horwitz AF: Adhesion in cell migration. Curr Opin Cell Biol
1995;7:697–706.
32 Clark EA, Brugge JS: Integrins and signal transduction pathways: The road taken. Science
1995;268:233–239.
33 Paterson HF, Self AJ, Garrett MD, Just I, Aktories K, Hall A: Microinjection of recombinant p21rho
induces rapid changes in cell morphology. J Cell Biol 1990;111:1001–1007.
34 Takaishi K, Kikuchi A, Kuroda S, et al: Involvement of rho p21 and its inhibitory GDP/GTP
exchange protein (rho GDI) in cell motility. Mol Cell Biol 1993;3:72–79.
35 Kishi K, Sasaki T, Kuroda S, et al: Regulation of cytoplasmic division of Xenopus embryo by rho
p21 and its inhibitory GDP/GTP exchange protein. J Cell Biol 1990;120:1187–1195.
36 Yoshioka K, Matsumura F, Akedo H, et al: Small GTP-binding protein Rho stimulates the acto-
myosin system, leading to invasion of tumor cells. J Biol Chem 1998;273:5146–5154.
37 Chen Q, Kinch MS, Lin TH, et al: Integrin-mediated cell adhesion activates mitogen-activated
protein kinases. J Biol Chem 1994;269:26602–26605.
38 Miyamoto S, Teramoto H, Coso OM, et al: Integrin function: Molecular hierarchies of cytoskele-
tal and signal molecules. J Cell Biol 1995;131:791–805.
39 Li W, Duzgun A, Sumpio BE, et al: Integrin and FAK-mediated MAPK activation is required for
cyclic strain mitogenic effects in Caco-2 cells. Am J Physiol Gastrointest Liver Physiol 2001;280:
G75–G87.
Prof. Nobuhiro Sato, Department of Gastroenterology,

Juntendo University School of Medicine, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113-8421 (Japan)
Tel. ⫹81 3 3813 3111/ext 3608, Fax ⫹81 3 3813 8862, E-Mail nsato@med.juntendo.ac.jp
The Diacylglycerol/Protein Kinase C

Pathway in Gastrointestinal Mucosal
Injury and Defense
Thomas A. Miller, Maria J. Redlak, Leah M. Coy, Mohiuddin M. Taher
Department of Surgery, Medical College of Virginia Campus of Virginia,
Commonwealth University, Richmond, Va., USA
The remarkable ability of prostaglandins (PGs), independent of their known

antisecretory properties, to prevent injury to the gastrointestinal epithelium by
such diverse noxious substances as ethanol, non-steroidal anti-inflammatory
drugs (NSAIDs), boiling water, bile, concentrated acid and concentrated alkali
was first reported by Robert and co-workers [1, 2] in the late 1970s. Despite
over two decades of intensive research, resulting in literally hundreds of publi-
cations, the precise mechanism(s) underlying this ‘cytoprotective’ property of
PGs has remained illusive. Although in vivo models of gastrointestinal injury
have shown that maintenance of mucosal blood flow, stimulation of mucous/
bicarbonate secretion, and enhanced epithelial renewal are all important to the
cytoprotective process elicited by PGs [3], the observation that gastrointestinal
injury under in vitro conditions can also be prevented or markedly attenuated
by PGs and other protective substances, such as growth factors, indicates that
the events responsible for injury result in cellular and molecular perturbations
that are somehow prevented by these protective agents [4, 5]. This being the
case, many laboratories attempting to unravel the processes responsible for
injury and how protective substances render the gastric and intestinal lining
more resistant to damage have focused their investigative efforts on the molecu-
lar events responsible for these actions.
Emerging data suggest that the diacylglycerol/protein kinase C (DAG/
PKC) pathway may play an important role in gastrointestinal injury and defense.
A variety of extracellular and/or endogenous signals are known to induce the
formation of DAG which then binds to a regulatory domain of PKC [6]. Upon
activation, PKC can modulate a range of diverse cellular processes through
phosphorylation of specific target proteins [6]. Current knowledge concerning
PKC indicates that it consists of a family of serine-threonine protein kinases
that may be involved in the regulation of such functions as cell growth, cell dif-
ferentiation, apoptosis and calcium homeostasis. At least 12 isoforms of PKC
have been described and differ from one another in terms of tissue distribution,
biochemical characteristics and subcellular distributions [7, 8]. The PKC family
may be divided into three main groups based on their structures and activator
requirements. Classical PKCs (␣, ␤1, ␤2 and ␥) require calcium and DAG
for optimal activation. Novel PKCs (␦, ⑀, ␮, ␩ and ␪) also require DAG
but are independent of calcium. Finally, atypical PKCs (␨, ␭ and ␶) do not
require DAG or calcium for kinase activation. Accordingly, activation of spe-
cific PKCs may have differential effects on a given cell resulting in injury and
eventual cell death or programming the cell to resist the injurious environment
to which it is exposed. The purpose of this discussion is to review relevant find-
ings concerning the role that PKC might play in gastrointestinal injury and
protection.
Gastric Mucosal Injury and Defense
The first demonstration that PG-induced protection against gastric

mucosal injury may involve a DAG/PKC pathway was reported by Konda et al.
[9] in 1990. Using a guinea pig-derived gastric chief cell preparation, pretreat-
ment with PGE2 or PGE1 significantly reduced not only ethanol but also tauro-
cholic acid-induced injury as measured by LDH release. PGs equipotently
stimulated increases in cellular DAG and cyclic AMP, and the rank order of the
potency was equal to their ability to reduce cellular injury. Pretreatment of chief
cells with OAG (a synthetic DAG and activator of PKC) or TPA (an activator of
PKC) also reduced injury to these cells while 4␣-PDD (an inactive phorbol
ester that does not activate PKC) failed to elicit this effect. In addition, when
chief cells were pretreated with PGE2 in the presence of H7, a PKC inhibitor,
the protective action of PG was reversed. Further, simultaneous pretreatment of
chief cells with OAG plus the calcium ionophore A23187 significantly reversed
the protective action of OAG. Because of this finding, these investigators con-
sidered the possibility that OAG may exert its protective action by increasing
calcium efflux from cells. Accordingly, additional studies were conducted in
which cells were prelabeled with 45Ca and then incubated with OAG or PGE2.
When measuring changes in cellular calcium following such incubation, both
agents were shown to significantly stimulate calcium efflux from the chief
cells. Interestingly, agents known to alter the cyclic AMP signal transduction
pathway, such as dibutyryl cyclic AMP or forskolin, had no effect on gastric
Miller/Redlak/Coy/Taher 70
injury and further were unable to influence calcium efflux. It was concluded
from these studies that PGs possess direct protective action against ethanol or
taurocholic acid-induced injury in guinea pig chief cells, presumably through
activation of the DAG/PKC signaling pathway.
Recent studies in our laboratory [10] using a human gastric carcinoma cell
line, known as AGS cells, extended the initial observations of Konda et al. and
paralleled their findings. When incubated with 10 mM aspirin (ASA) cellular
viability (as measured by trypan blue uptake) in AGS cells was markedly
decreased when compared with control. Both epidermal growth factor (EGF)
and transforming growth factor-␣ (TGF-␣) prevented the damaging effects of
ASA. The calcium ionophore, A23187, similar to ASA, caused a significant
reduction in cell viability, and when given in combination with either growth
factor prior to ASA exposure, obviated the protective effects of these agents.
Since these results suggested that changes in intracellular calcium by the
ionophore could override the protective action of these growth factors, further
studies were undertaken to determine whether EGF and/or TGF-␣ might pre-
vent intracellular calcium accumulation by inducing the efflux of calcium from
cells. Accordingly, cells previously loaded with 45Ca were noted to have sub-
stantially less radioactivity when treated with these growth factors than
occurred under control conditions, supporting the conclusion that both EGF
and TGF-␣ induced calcium efflux from cells. Additional studies were under-
taken to determine what role the DAG/PKC pathway might play in these find-
ings. In AGS cells challenged with ASA and pretreated with the PKC inhibitor,
chelerythrine, not only were the protective effects of EGF and TGF-␣ prevented
but, in addition, the ability of these growth factors to induce calcium efflux
was also obviated. Further, stimulators of PKC activity, such as OAG and TPA,
also induced calcium efflux from gastric cells to the same extent as EGF and
TGF-␣. Of further note, both growth factors elicited increased PKC activity in
this cell line (not dissimilar to OAG) when compared to control conditions, and
these effects were also prevented by chelerythrine. Finally, OAG and TPA were
equally effective in preventing ASA damage to that observed with EGF and
TGF-␣. These results were interpreted as indicating that a potential mechanism
by which EGF and TGF-␣ orchestrated their protective properties against ASA
damage was through the induction of calcium efflux via activation of a DAG/
PKC pathway.
In the aforementioned studies, only total PKC activity was considered so that
specific isoforms that may have been responsible for injury and/or protection
were not identified. Similarly, indices of damage (i.e. LDH release and
trypan blue uptake) employed have traditionally been considered to be markers of
gastric necrosis so that the role apoptosis may have played in these experimental
findings was not considered. Both of these issues were addressed in a series of
Protein Kinase C and Mucosal Defense 71

studies by Zhu et al. [11, 12]. Also using AGS cells as the experimental cell line,
these investigators observed that both ASA and indomethacin inhibited cell
growth and induced apoptosis in AGS cells in a dose- and time-dependent man-
ner, without altering the cell cycle [11]. Further experiments with indomethacin
demonstrated that this damaging agent increased the oncogene expression of both
c-myc mRNA and protein, but had no effect on the levels of p53 and p21waf1/cip1.
Down-regulation of c-myc by its antisense oligonucleotides reduced apoptosis by
indomethacin. Activation of PKC by TPA inhibited indomethacin-induced apop-
tosis as well as the overexpression of c-myc and increased expression of
p21waf1/cip1. It was concluded that NSAID-induced gastric apoptosis may be medi-
ated by up-regulation of c-myc proto-oncogene whereas PKC activation elicits its
protective action by abrogating this effect [11].
In a follow-up study, the specific PKC isoforms that might be involved in
these responses were evaluated [12]. Ten of the 12 known PKC isoforms were
found to be constitutively expressed in this cell line including all of the classi-
cal PKC isoforms, namely ␣, ␤1, ␤2 and ␥. Exposure of cells to indomethacin
decreased the expression of PKC-␤1, increased that of PKC-␤2, ␩ and ⑀, but did
not alter the expression of PKC ␣, ␥, ␨, ␦, ␫ or µ. Overexpression of PKC-␤1
by transfection with its complimentary DNA rendered AGS cells relatively
resistant to indomethacin-induced apoptosis, suggesting that this isoform acted
as a survival mediator in response to indomethacin exposure. Because over-
expression of PKC-␤2 did not show a similar protective effect, it would appear
that the ␤1 and ␤2 isoforms have divergent effects on apoptosis even though it
is known that they are derived from a single gene and only differ by 50 amino
acids at the carboxyl end of the protein [13]. Further, in attempting to determine
the signaling pathways from the specific PKC isoform to the downstream
gene(s) activated, it was noted that overexpression of PKC-␤1 increased
p21waf1/cip1 expression in both the absence and presence of indomethacin. This
suggested that the PKC-␤1 isoform may be a part of a signal transduction cas-
cade that modulates expression of p21waf1/cip1. Further support of this conclusion
was that inhibition of PKC-␤1-mediated overexpression of p21waf1/cip1 by its
antisense cDNA could partially inhibit the antiapoptotic effect of PKC-␤1. The
authors concluded that it is likely that PKC-␤1 suppresses indomethacin-
induced apoptosis partly by modulating the expression of p21waf1/cip1.
Intestinal Injury and Protection
Evidence supporting a role for the DAG-PKC pathway in the mediation of

intestinal defense mechanisms has also emerged. Using a human colonic cell
line (Caco-2), Banan et al. [14] noted that the damaging effects of ethanol
(i.e. 7.5 and 10%), as measured by trypan blue uptake, were significantly
overcome when cells were preincubated with PGE2. It was further observed that
PG could prevent the corresponding disruption in microtubule stability elicited
by ethanol at the same time that it significantly increased PKC activation in
these cells. Further studies demonstrated that pretreatment of Caco-2 cells
with OAG or TPA prior to exposure to 7.5 or 10% ethanol significantly reduced
cell injury and increased microtubule stability as confirmed by confocal
microscopy. In contrast, pretreatment of Caco-2 cells with 4␣-PDD (an inactive
phorbol ester) failed to prevent cell injury and disruption of the microtubule
cytoskeleton. Preincubation with staurosporine (a PKC inhibitor) abolished the
protective effects of PG in cells exposed to these ethanol concentrations.
Incubation of Caco-2 cells with the calcium ionophore, A23187, similar to 10%
ethanol, caused significant reduction in cell viability and microtubule stability.
In addition, preincubation with A23187 in combination with PG or OAG prior
to subsequent exposure to ethanol significantly abolished the protective effects
of PG or OAG pretreatment. Of further note, pretreatment with OAG, TPA or
PG resulted in significant increases in 45Ca efflux from cells which correlated
with increased stability of the microtubule cytoskeleton and maintenance of
cellular integrity. These data suggested that PG possessed direct protective
effects against ethanol injury in this colonic cell line, and may act by stabiliz-
ing the microtubule cytoskeleton through a DAG/PKC signal transduction
pathway and its consequent stimulation of calcium extrusion from cells.
In another study using Caco-2 cells, Banan et al. [15] evaluated the ability
of the growth factor, EGF, to protect microtubules and barrier integrity against
oxidative injury. In this study, monolayers of this cell line were pretreated with
EGF, PKC modulators or calcium modulators before exposure to the oxidants,
hydrogen peroxide (H2O2) or hypochlorous acid (HOCl). These oxidants were
noted to disrupt microtubule stability and barrier integrity, both of which were
protected from this damage by EGF pretreatment. Further, EGF caused a rapid
distribution of the PKC isoforms, ␣, ␤1 and ␨ to cell membranes, enhancing
PKC activity of membrane fractions while reducing PKC activity of cytosolic
fractions. EGF was also observed to enhance calcium extrusion from cells and
prevented the oxidant-induced sustained rises in intracellular calcium. PKC
inhibitors abolished and PKC activators mimicked EGF protection. Oxidant-
induced damage was mimicked by and potentiated by a calcium ionophore
(A23187), exacerbated when the cellular bathing solution contained high con-
centrations of calcium, and prevented by removal of calcium from the bathing
solution by chelation or calcium channel antagonists. Finally, PKC activators
mimicked EGF in terms of also enhancing calcium efflux from cells and stabi-
lizing intracellular calcium concentration; membrane calcium-ATPase pump
inhibitors which prevented calcium extrusion from cells obviated the protection by

both EGF or PKC activators. These experimental findings were consistent with
the conclusion that EGF protection of microtubules and the intestinal epithelial
barrier requires activation of a PKC signal transduction pathway through which
normalization of intracellular calcium equilibrium is orchestrated.
Further study by these investigators explored the role that the ␤1 isoform
of PKC may play in mediating these effects [16]. Using a similar model of
oxidant-induced microtubule instability and intestinal barrier disruption, Caco-2
cells were transfected to stably over- or underexpress the PKC-␤1 isoform. It
was found that transfected cells that overexpressed PKC-␤1 are severalfold
more sensitive to protection by EGF as well as the PKC activator OAG. This
overexpression synergized with the addition of EGF or OAG to enhance the sta-
bility of polymerized tubulin which thereby maintained microtubule cytoskeletal
stability and also stabilized monolayer barrier integrity. This protection
required activation through the translocation of PKC-␤1 from the cytosolic to
the particulate fraction. In contrast, cells which were transfected with anti-sense
to PKC-␤1, resulting in underexpression of this isoform to 10% of normal levels,
were rendered severalfold less sensitive to the protective effects of EGF and
OAG. Under this circumstance, EGF and OAG were unable to enhance tubulin
assembly, stabilize the microtubule cytoskeleton and maintain monolayer
barrier integrity. These data provided convincing evidence that EGF-induced
protection against oxidant injury in this intestinal cell line requires activation of
the PKC-␤1 isoform.
Since intracellular activation of DAG is a major pathway through which
PKC activation ultimately occurs, Banan et al. [17] hypothesized that phospho-
lipase C (PLC) might be the initiator of this cascade of events by growth factors
since activation of PLC is known to induce the production of DAG and there-
from activation of PKC. Again using the Caco-2 cell line, these investigators
observed that transfected cells in which PLC-␥1 activity was severely attenuated
were no longer protected by EGF against oxidant damage to tubulin assembly,
microtubule stability and barrier integrity. These transfection experiments were
performed because pharmacologic inhibitors of PLC are specific only to total
PLC activity and not to any one PLC isoform. Moreover, PLCz, the negative
dominant fragment of PLC-␥, obtained from the Z region of PLC-␥1, is known
to contain the domains necessary for activation/phosphorylation of PLC-␥1 by
the EGF receptor. Further, blockade of this receptor with specific tyrosine
kinase inhibitors also blocked the protective effects of EGF against oxidant-
induced tubulin disassembly, disruption of the microtubule cytoskeleton and
barrier leakiness. It was concluded from this study that protection by
EGF occurs through activation of the EGF receptor which then activates
PLC-␥1 and initiates formation of DAG and subsequent activation of the PKC-␤1
isoform.
Although the aforementioned studies strongly implicate a role for the
DAG/PKC pathway in mediating intestinal defense, evidence also exists sug-
gesting that PKC may be involved in various injurious processes in the gut. As
examples, PKC has been noted to be elevated in colonic mucosal samples
obtained from patients with inflammatory bowel disease [18], and activation of
PKC by luminal instillation of phorbol ester has been demonstrated to induce
ileal and colonic inflammation in experimental animal models [19–21]. PKC
activity has also been shown to be elevated in mucosa taken from animals in
which experimental colitis was induced via instillation of 2,4,6-trinitrobenzene
sulfonic acid (TNBS) [22]. Finally, elevated PKC activity has been reported
to increase intestinal endothelial and epithelial permeability under various
experimental conditions [23, 24].
To gain better insight into the role of the PKC pathway in intestinal injury,
Brown et al. [25] induced experimental colitis in rats via the intrarectal instil-
lation of TNBS or the PKC activator PMA. Both PKC activity and mucosal
injury increased significantly within 4 h of the TNBS treatment. Mucosal dam-
age became maximal at 1 day and declined after 7 days whereas the PKC activ-
ity was maximal at 7 days and declined by 14 days. The ␤, ␦ and ⑀ isoforms of
PKC were all increased in response to TNBS whereas the ␣ isoform decreased.
Both staurosporine (a nonspecific inhibitor) and GF-109203X (a specific
inhibitor of PKC) reduced TNBS-induced changes in mucosal PKC activity
concomitant with the degree of mucosal damage. Further, activation of mucosal
PKC activity with PMA also initiated mucosal damage which was correspond-
ingly inhibited by pretreatment with a PKC antagonist. It was concluded that
increases in PKC activity likely play a causative role in TNBS-induced colitis,
and that the elevated ␤, ␦ and ⑀ isoforms of PKC appear to be responsible for
the TNBS-induced inflammation. In contrast, the observation that PKC-␣ was
depressed by this inflammatory process, and did not return to normal levels
until the process had subsided, suggested that this isoform may play a role in
the restitution process.
In another study by this group of investigators [26], the role of PKC
isoforms in tumor necrosis factor (TNF)-␣-mediated cytotoxicity and apoptosis
in intestinal cells was studied using a rat epithelial cell line, known as IEC-18.
The genesis of this study stems from earlier work in which TNF-␣ had been
shown to cause significant small intestinal injury in rodents [27] and had been
detected as a possible mediator of colonic mucosal damage in patients with
Crohn’s disease [28]. Accordingly, cells were incubated with TNF-␣ in the pres-
ence or absence of the transcription inhibitor actinomycin D (AMD). As noted
previously [29], the extent of cell damage was enhanced when AMD was added
to the incubation medium, suggesting that new protein synthesis plays a role in
the cytotoxic action of TNF-␣. Interestingly, TNF-␣ induced the translocation

of the ␣, ␦ and ⑀ isoforms of PKC from the cytosol to the membrane fraction
of the intestinal cells, suggesting that TNF-␣ may require the induction of these
isoforms to mediate its cytotoxic effects. This conclusion was further strength-
ened upon pretreatment of the cells with the PKC-⑀ translocation inhibitor,
PKC-⑀V1–2, following which it was noted that the cytotoxic and apoptotic
effects of TNF were markedly reduced. Of further note, cells incubated with
PMA also displayed an increase in cell injury although the extent of this cyto-
toxicity and apoptosis was not enhanced by AMD. Nonetheless, the PMA-
induced cell damage was reduced by rottlerin, a PKC-␦ inhibitor. Additional
studies indicated that caspase-3, an enzyme implicated in the mediation of
apoptosis, was activated in cells in response to either TNF-␣ or PMA stimula-
tion and its effects on this activity were reduced by selective inhibition of
PKC-⑀ and -␦, respectively. As observed by this investigative team in an earlier
report [30], these same isoforms appeared to be involved in nitric oxide-
induced damage to rat isolated colonic mucosal cells. Taken together, data from
both of these studies suggested that activation of specific and selective PKC
isoforms may play a role in the induction of various types of intestinal epithelial
cell injury.
Critical Overview of the Role of the DAG/PKC Pathway in

Gastrointestinal Injury and Defense
The studies reviewed in this report provide compelling evidence for a role
for the DAG/PKC pathway in both injury and defense mechanisms involving
the gastrointestinal epithelium. Depending upon the experimental circum-
stance, it seems clear that activation of PKC can both mediate processes lead-
ing to cellular injury and ultimate death as well as trigger other processes
that enable cells to resist injury and maintain cellular integrity. In the few
studies in which specific PKC isoforms have been examined, information
to date would suggest that some isoforms target processes leading to injury
while others initiate protective processes. In the study by Zhu et al. [12] exam-
ining indomethacin-induced gastric injury, the PKC-␤1 isoform appeared to
play a protective role, while the PKC-␤2 isoform was involved with injury.
Interestingly, Banan et al. [16] observed in Caco-2 cells that EGF-induced
protection against oxidant injury was also associated with activation of the
PKC-␤1 isoform. Such findings evoke the interesting question of whether the
␤1 isoform is always an initiator of processes leading to protection or whether
in other cell types or under different experimental conditions it would act in
ways that eventually would result in cellular injury. Clearly, much needs to be
learned about these isoforms, the circumstances under which they are activated,
and the processes through which their effects are mediated. When one realizes
that most, if not all, of the 12 isoforms currently identified exist in a given cell,
it raises the interesting possibility that each isoform may play a strategically dif-
ferent role in carrying out a particular intracellular function emphasizing the
complexity alone of this enzyme system, not to mention the multiplicity of
other enzymes that exist in a cell.
Of further note, once a particular PKC isoform is activated, the down-
stream processes that ultimately lead to injury or protection are also uncertain.
In studies showing an association of PKC activation with protection, at least
two protective processes have been identified. These include cytoskeletal stabi-
lization and maintenance of intracellular calcium homeostasis. Thus, activation
of PKC via various growth factors or PGs has shown that depolymerization
of the microtubule cytoskeleton can be prevented. Similarly, these protective
agents have also been shown to induce extrusion of calcium from the cell, pre-
sumably through PKC, since activators of this enzyme, such as OAG and TPA,
can also initiate calcium extrusion. Whether the apparent effects of PKC acti-
vation on cytoskeletal stabilization and calcium efflux are independent actions
of this enzyme or occur via the resultant intracellular calcium homeostasis
following efflux of excess calcium from the cell is uncertain. More likely, the
latter process is operational since it is known that elevated intracellular calcium
can result in disassembly of both the actin and microtubule cytoskeleton
[31, 32]. Further, the specific genes activated by PKC to initiate its downstream
effects in gastrointestinal epithelium are virtually unknown. Although Zhu et al.
[12] reported that the protective action of the PKC-␤1 isoform against
indomethacin-induced gastric injury was associated with an increase in the
expression of p21waf1/cip1, suggesting that this gene might play a key role in
mediating the protective process, whether the action of this gene was mediated
through enhanced calcium efflux and/or cytoskeletal stabilization or initiated
through some other pathway remains unknown. Obviously, these uncertainties
indicate the potential fruitfulness of this area of research.
Finally, the type of underlying injury in a given experimental situation
and its linkage with the PKC enzyme system requires further clarification.
Specifically, some of the research cited would suggest that injury follows a
necrotic pathway of death, whereas in other studies apoptosis seems to be
responsible for the resultant cell death. Are PKC isoforms that initiate injury
involved in both types of cell death or is one type more likely linked with the
PKC system? Similarly, can protector isoforms of PKC prevent both cellular
necrosis and apoptosis or is one mode of death more likely to be affected than
the other? The reason these considerations are important is because necrotic
cell death is characterized by ATP depletion, decreased mitochondrial function,
swelling of the cytoplasm, increased membrane permeability and fragmentation

of DNA into various random sizes [33]. In contrast, apoptosis is actually an
energy requiring program of cell death that is typified by shrinkage of the cyto-
plasm, condensation of the nucleus, and fragmentation of DNA into 200-base
pair multiple fragments [33]. Clearly, then, much more focused research is
required to distinguish between these two types of cell death and what role the
PKC pathway plays in their mediation and/or prevention.
Taking into account the experimental considerations enumerated above, a
paradigm is proposed as to how calcium homeostasis, cytoskeletal stability, the
PKC pathway and apoptosis/necrosis fit together. Figure 1 details these inter-
actions. Although not specifically discussed in this review, many damaging
insults to the gastrointestinal epithelium have been shown to alter intracellular
calcium concentration by enhancing the release of calcium from intracellular
stores and inducing the influx of calcium through calcium channels [10,
34–36]. Further, nuclear factor ␬B (NF-␬B) is an important modulator of
apoptosis and its activation limits this form of cell death through anti-apoptotic
gene transcription, whereas blockade of such activation enhances programmed
cell death [37]. Hence, these experimental observations were included in
preparing this figure.
Current knowledge suggests that the DAG/PKC pathway plays a major

role in both injury and defense mechanisms involving the gastrointestinal
epithelium. Specific isoforms appear to play a role in the mediation of gas-
trointestinal injury and ultimate cellular death while other isoforms appear to
trigger responses that enable the cell to maintain its health. At least two processes
appear to be operational in initiating protection via the PKC pathway by pro-
tective agents such as PGs and growth factors. These include the maintenance
of cytoskeletal stability and the provision of intracellular calcium homeostasis.
Other mechanisms are probably operational that have yet to be identified. While
cellular injury and death can ultimately occur through two distinctly different
pathways, namely necrosis and apoptosis, it is uncertain whether PKC’s role in
cell injury and death involves both pathways or is primarily mediated by one.
Similarly, prevention of injury by PKC activation can also occur by obviating
necrosis or apoptosis. Again, whether one or both of these pathways is more
likely controlled by PKC will require further clarification. Because of these
present uncertainties, determining the precise role of the DAG/PKC pathway
in mucosal injury and defense mechanisms involving the gastrointestinal
epithelium will most assuredly prove to be a fruitful area of investigation for
some time to come.
Damaging agent Ca2⫹
Protective agent
Inhibits Efflux
Ca2⫹
Membrane
PLC PKC isoform Inhibits

PKC isoform
Cytosol DAG
Ca2⫹ (Decreased)
NF-␬B
IP3 Ca2⫹
(Increased)
ER
Stabilizes
Nucleus
Destabilizes
Apoptosis
Cytoskeleton Protection
Necrosis
Fig. 1. The scheme details putative mechanisms whereby a damaging agent (solid
lines) induces injury and ultimate death via apoptosis/necrosis in gut epithelial cells.
According to this paradigm, a damaging agent induces increased intracellular calcium
accumulation by releasing calcium from the endoplasmic reticulum (ER) via IP3 generation
following phospholipase C (PLC) activation and causing extracellular calcium influx
through calcium channels. If unchecked, the resultant increase in intracellular calcium con-
centration disrupts cytoskeletal integrity and may ultimately induce apoptosis, either directly
by activating gene transcription mechanisms within the nucleus, or indirectly by activating a
PKC isoform (either alone or in combination with DAG) that then activates the nucleus to
induce apoptosis. If excessive intracellular calcium accumulation ensues, cell death may
result from necrosis. In contrast, a protective agent (broken lines) prevents the intracellular
calcium accumulation by possibly altering the influx of calcium through calcium channels,
preventing the generation of IP3 and eliciting extrusion of excess calcium from cells
via PKC activation. In addition, the protective agent may activate the same or another PKC
isoform that mediates translocation of NF-␬B from the cytosol to the nucleus and thereby
induces cell survival genes that prevent apoptosis and insures the cell’s health and longevity.
References
1 Robert A, Nezamis JE, Lancaster C, Hanchar AJ: Cytoprotection by prostaglandins in rats:

Prevention of gastric necrosis produced by alcohol, HCl, NaOH, hypertonic NaCl and thermal
injury. Gastroenterology 1979;77:433–443.
2 Robert A: Cytoprotection by prostaglandins. Gastroenterology 1979;77:761–767.

3 Miller TA, Smith GS, Baretto JC: Gastroduodenal defense: Role of epithelial factors; in
Goldie R (ed): Immunopharmacology of Epithelial Barriers. London, Academic Press, 1994,
pp 197–211.
4 Lichtenberger LM: Gastroduodenal mucosal defense; in Yamada T (ed): Current Opinion in
Gastroenterology. Philadelphia, Lippincott-Williams & Wilkins, 1999, pp 463–472.
5 Pai R, Tarnawski AS: Signal transduction cascades triggered by EGF receptor activation:
Relevance to gastric injury repair and healing. Dig Dis Sci 1998;43:14S–22S.
6 Newton A: Protein kinase C: Structure, function and regulation. J Biol Chem 1995;270:
28495–28498.
7 Hug H, Sarre TF: Protein kinase C isoenzymes: Divergence in signal transduction? Biochem J
1993;291:329–343.
8 Deacon EM, Pongracz J, Griffiths G, Lord JM: Isoenzymes of protein kinase C: Differential
involvement in apoptosis and pathogenesis. Mol Pathol 1997;50:124–131.
9 Konda Y, Nishisaki H, Nakano O, Matsuda K, Wada K, Nagao M, Matozaki T, Sakamoto C:
Prostaglandin protects isolated guinea pig chief cells against ethanol injury via an increase in
diacylglycerol. J Clin Invest 1990;86:1897–1903.
10 Miller TA, Kokoska ER, Smith GS, Banan A: Role of calcium homeostasis in gastric mucosal
injury and protection. Life Sci 2001; 69:3091–3102.
11 Zhu GH, Wong BCY, Eggo MC, Ching CK, Yuen ST, Chan EYT, Lai KC, Lam SK: Non-steroidal
anti-inflammatory drug-induced apoptosis in gastric cancer cells blocked by protein kinase C
activation through inhibition of c-myc. Br J Cancer 1999;79:393–400.
12 Zhu GH, Wong BCY, Slosberg ED, Eggo MC, Ching CK, Yuen ST, Lai KC, Soh JW, Weinstein
IB, Lam SK: Overexpression of protein kinase C-␤1 isoenzyme suppresses indomethacin-induced
apoptosis in gastric epithelial cells. Gastroenterology 2000;118:507–514.
13 Ono Y, Kikkawa U, Ogita K, Fujii T, Kurokawa T, Asaoka Y, Sekiguchi K, Ase K, Igarashi K,
Nishizuka Y: Expression and properties of two types of protein kinase C: Alternative splicing from
a single gene. Science 1987;236:1116–1120.
14 Banan A, Smith GS, Deshpande Y, Rieckenberg CL, Kokoska ER, Miller TA: Prostaglandin
protects human intestinal cells against ethanol injury by stabilizing microtubules: Role of protein
kinase C and enhanced calcium efflux. Dig Dis Sci 1999;44:697–707.
15 Banan A, Fields JZ, Zhang Y, Keshavarzian A: Key role of PKC and Ca2⫹ in EGF protection of
microtubules and intestinal barrier against oxidants. Am J Physiol 2001;280:G828–G843.
16 Banan A, Fields JZ, Talmage DA, Zhang Y, Keshavarzian A: PKC-␤1 mediates EGF protection of
microtubules and barrier of intestinal monolayers against oxidants. Am J Physiol 2001;281:
G833–G847.
17 Banan A, Fields JZ, Zhang Y, Keshavarzian A: Phospholipase C-␥ inhibition prevents EGF
protection of intestinal cytoskeleton and barrier against oxidants. Am J Physiol 2001;281:
G412–G423.
18 Sakanoue Y, Hatada T, Horai T, Shoji Y, Kusumoki M, Utsunomiya J: Protein kinase C activity of
colonic mucosa in ulcerative colitis. Scand J Gastroenterol 1992;27:275–280.
19 Berin MC, Buell MG: Phorbol myristate acetate ex vivo model of enhanced colonic epithelial
permeability. Reactive oxygen metabolite and protease independence. Dig Dis Sci 1995;40:
2268–2279.
20 Fretland DJ, Widomski DL, Levin S, Gaginella T: Colonic inflammation in the rabbit induced by
phorbol-12-myristate-13-acetate. Inflammation 1990;14:143–150.
21 Overdahl MC, Julian MW, Weisbrode SE, Dorinsky PM: Anti-CD18 antibody does not block ileal
injury induced by phorbol myristate acetate. Am J Respir Crit Care Med 1995;152:1331–1336.
22 Brown GR, Lindberg G, Meddings J, Silva M, Beutler B: Tumor necrosis factor inhibitor amelio-
rates murine intestinal graft-versus-host disease. Gastroenterology 1999;116:593–601.
23 Northover AM, Northover BJ: Stimulation of protein kinase C activity may increase microvascu-
lar permeability to colloidal carbon via ␣-bioenzyme. Inflammation 1994;18:481–487.
24 Perez M, Barber A, Pon Z: Modulation of intestinal paracellular permeability to intracellular
mediators and cytoskeleton. Can J Physiol Pharmacol 1997;75:287–292.
25 Brown JF, Chang Q, Soper BD, Tepperman BL: Protein kinase C mediates experimental colitis in
the rat. Am J Physiol 1999;276:G583–G590.
26 Chang Q, Tepperman BL: The role of protein kinase C isozymes in TNF-␣-induced cytotoxicity
to a rat intestinal epithelial cell line. Am J Physiol 2001;280:G572–G583.
27 Garside P, Bunce C, Tomlinson C, Nichols BL, Mowat AM: Analysis of enteropathy-induced by
tumor necrosis factor ␣. Cytokine 1993;5:24–30.
28 Beil WJ, Weller PF, Peppercorn MA, Galli SJ, Dvorak AM: Ultrastructural immunogold localiza-
tion of subcellular sites of TNF-␣ in colonic Crohn’s disease. J Leukoc Biol 1995;58:284–298.
29 Leist Gantner MF, Bohlinger I, Germann PG, Tiegs G, Wendel A: Murine hepatocyte apoptosis
induced in vitro and in vivo by TNF-␣ requires transcriptional arrest. J Immunol 1994;153:
1778–1788.
30 Tepperman BL, Chang Q, Soper BD: The involvement of protein kinase C in nitric oxide-induced
damage to rat isolated colonic mucosal cells. Br J Pharmacol 1999;128:1268–1274.
31 Hori M, Sato H, Kitakaze M, Iwai K, Takeda H, Inoue M, Kamada T: ␤-Adrenergic stimulation
disassembles microtubules in neonatal rat cultured cardiomyocytes through intracellular Ca2⫹
overload. Circ Res 1994;75:324–334.
32 Kakiuchi S, Sobue K: Ca2⫹- and calmodulin-dependent flip-flop mechanism in microtubule
assembly-disassembly. FEBS Lett 1981;132:141–143.
33 Que FG, Gores GJ: Cell death by apoptosis: Basic concepts and disease relevance for the
gastroenterologist. Gastroenterology 1996;110:1238–1243.
34 Kokoska ER, Smith GS, Deshpande Y, Rieckenberg CL, Miller TA: Adaptive cytoprotection
induced by ethanol in human intestinal cells: Role of prostaglandins and calcium homeostasis.
Ann Surg 1998;288:123–130.
35 Kokoska ER, Smith GS, Wolff AB, Deshpande Y, Rieckenberg CL, Banan A, Miller TA: Role of
calcium in adaptive cytoprotection and cell injury induced by deoxycholate in human gastric cells.
Am J Physiol 1998;275:G322–G330.
36 Kokoska ER, Smith GS, Deshpande Y, Wolff AB, Rieckenberg C, Miller TA: Calcium accentuates
injury induced by ethanol in human gastric cells. J Gastroint Surg 1999;3:308–318.
37 Baeuerle PA, Baltimore D: NF-␬B – ten years after. Cell 1996;87:13–20.
Thomas A. Miller, MD, Department of Surgery, Division of General Surgery,

MCV Campus of Virginia Commonwealth University,
PO Box 980519, Richmond, VA 23298–0519 (USA)
Tel. ⫹1 804 675 5112, Fax ⫹1 804 675 5390, E-Mail thomas.miller3@med.va.gov

Part 2: Mucosal Repair and Ulcer Healing
Expression of Early Primary Response

Genes in Healing of Gastrointestinal
Mucosal Injury
Jian-Ying Wang
Departments of Surgery and Pathology, University of Maryland School of
Medicine and Baltimore Veterans Affairs Medical Center,
Baltimore, Md., USA
Gastrointestinal mucosal injury occurs in circumstances commonly

encountered in daily life, from mild physical trauma during digestion to
localized damage from the ingestion of alcohol, aspirin, and/or non-steroidal
anti-inflammatory compounds, or from Helicobacter pylori infection. Acute
mucosal injury also occurs in critical illness including various surgical condi-
tions such as trauma, thermal injury, shock and sepsis. After injury, gastroin-
testinal mucosal tissue exhibits a spectrum of responses. In the acute response
to injury damaged cells are sloughed, and remaining viable cells from areas
adjacent to or just beneath the injured surface, migrate to cover the denuded
area. This early restitution is independent of cell proliferation and appears to be
an initial host response to prevent noxious agents from causing deeper tissue
damage. In contrast to this rapid repair process, deeper damage and chronic
ulcers manifest long-term complex responses that require de novo mRNA and
protein synthesis and cell replication. Altered gene regulation in response to
wounding or ulceration results in an increase in cell proliferation to replace lost
cells.
Over the last decade, considerable progress has been made in understand-
ing the roles of early primary response genes in events responsible for the
process of cell renewal during ulcer and wound healing in the gastrointestinal
mucosa and other tissues. Most of these early primary response genes belong
to the family of protooncogenes and are responsible for control of the cell cycle
[1, 2]. Because the expression of these early primary response genes is rapid
and transient following injury or when normal quiescent cells are exposed to
mitogenic substances, they have been thought to act as mediators linking short-
term signals, immediately after cell surface stimulation, to proliferation by
regulating the activation of specific genes. These early primary response genes
such as protooncogenes code for sequence-specific DNA-binding nuclear pro-
teins with a potential to influence directly the expression of specific genes at
the transcriptional level. Therefore, activation of early primary response gene
expression plays important roles in healing following wounding the gastroin-
testinal mucosa and other tissues. In this chapter we will provide an overview
of early primary response gene expression during wound healing and then
analyze in some detail the possible pathways by which damage induces the
activation of protooncogenes.
Gene Activation in Response to Wounding
Normal cells respond to wounding by altering rapidly the expression of

various genes whose products are central to cell migration and proliferation.
At the early response following injury, the increased synthesis of transcription
factors is critical to the modulation of expression of cell-type-specific or devel-
opmentally regulated genes [1]. This provides a pathway for controlling the
expression of a gene whose product is infrequently required under physiologi-
cal conditions. The process of transcription is a fundamental element in gene
expression and is an attractive control point for the regulation of gene activa-
tion. The region immediately upstream of the transcribed sequence contains
two types of elements: (1) sequences involved in the process of transcription
itself, and (2) sequences found in genes transcribed in a particular tissue or
in response to a specific signal [3, 4]. These tissue- and stimulus-specific
sequences are implicated in selective cellular responses to wounding and to
growth factors or cytokines. Transcription factors are regulated to become
active only in the appropriate cell type or in response to the appropriate stimu-
lus. One of the mechanisms responsible for the control of transcription factor
activity is the regulation of the synthesis of the factor itself. The other is the
regulation of the activity of preexisting factor.
Although low basal expression of the nuclear protooncogenes c-fos, c-jun
and c-myc is observed in most cells, their expression is rapidly and transiently
induced following wounding in vivo [5] as well as in vitro [6]. The products of
these protooncogenes are nuclear transcription factors and bind to specific
DNA sequences in the vicinity of target genes, which are then activated. The
c-Jun protein can bind DNA as either Jun-Jun homodimers or Jun-Fos hetero-
dimers. The Fos protein must form heterodimers with c-Jun proteins because
Fos protein cannot form stable Fos-Fos homodimers. The Jun-Fos heterodimers
Early Primary Response Genes in Healing of GI Mucosal Injury 83

are more stable and exhibit more DNA-binding activity and transactivation
capability compared with corresponding Jun-Jun dimers. These homo- or het-
erodimers interact with a DNA regulating element known as the AP-1 site [3, 4].
Binding sites for AP-1 are founded in the promoters of numerous inducible
genes and are regulatory elements for gene expression. The Myc protein
interacts with Max, a basic helix-loop-helix zipper protein, and the Myc-Max
heterodimeric complexes stimulate transcription of variety of inducible genes
and are involved in the process of activation of ‘second responsive genes’ in
generative tissues following injury [7]. Recently, cyclin-dependent kinase-4
(CDK4) and p21 are identified as target genes of c-myc and the c-myc activa-
tion stimulates expression of the CDK4 gene [53] but represses the p21
promoter activity [54], providing a direct link between c-myc and cell-cycle
regulation.
In addition to AP-1 and c-myc, there are additional cis-acting elements,
including the serum responsive element of SP1/SP2, the nuclear factor-␬B
(NF-␬B)-binding site, and the AP-2 activation site, that also may be in involved
in the process of wound healing. These elements are recognized by transcrip-
tion factors distinct from AP-1 and c-myc, but their activity is modulated by
exposure to phorbol esters or other protein kinase C activators.
Protooncogenes and Ulcer Healing
Cell Renewal and Ulceration

The cells of the gastrointestinal mucosa are among the most rapidly pro-
liferating ones in the body [8, 9]. Normal function of the mucosa depends on
a regulated rate of division of proliferating cells in the mucous neck region
[9, 10] in the stomach and the crypts [9, 10] in the small intestine. In addition
to being dependent on the typical metabolic hormones such as insulin and
growth hormone, the growth of the mucosa is affected by a plethora of agents
whose presence is triggered by or dependent upon food intake and the process-
ing of food by the digestive tract [9]. Cells lost into the lumen are replaced by
new cells. Factors affecting either cell loss or regeneration rapidly alter the
healthy mucosa, which is a balance between these processes. Theoretically
then, ulcers may develop if either the factors acting to damaging the mucosa are
increased or those acting to produce new cells are decreased. In other words, it
may be said that an ulcer develops when the rate of cell loss exceeds the rate of
cell renewal.
Numerous studies have elucidated the importance of cell renewal in the
ulcerogenic process. Vanamee et al. [11] have showed that growth hormone
reduces the incidence of gastric lesions caused by restraint stress in rats. It has
Wang 84
been also demonstrated that the susceptibility of rats to stress ulcers is directly
correlated with the rate of gastric mucosal DNA synthesis [12]. In a chronic
experimental ulcer model, pentagastrin has been shown to stimulate DNA syn-
thesis and accelerate the healing of acetic acid-induced gastric and duodenal
ulcers [13].
Increased Expression of Protooncogenes in Healing of

Gastric Mucosal Ulcers
Protooncogenes play an important role in cell proliferation, especially in
control of the cell cycle, differentiation and apoptosis [1, 2]. Expression of the
protooncogenes is enhanced during regeneration of damaged tissues. Makino
et al. [14] and others [15, 16] have reported that when growth is stimulated
in the normally quiescent adult rat liver by partial hepatectomy, steady-state
levels of mRNAs for protooncogenes c-fos and c-myc increase sequentially
during the pre-replicative phase which precedes DNA synthesis.
Using the stress ulcer model, we have tested the hypothesis that expression
of protooncogenes c-fos and c-myc is involved in healing of the damaged mucosa
in the gastrointestinal tract [5]. Stress caused visible lesions in the oxyntic gland
mucosa after 2 h. The severity and number of lesions increased with the duration
of stress. By 4 h the incidence of oxyntic lesions was 100%. These lesions
appeared as elongated bands ranging from 2 to 10 mm in length and 1 to 3 mm
in width, generally parallel to the long axis of the stomach. Microscopic exami-
nation of the gastric mucosa after 6 h of stress showed a discontinuous surface
with many cells sloughed off into the lumen. The gastric pits were greatly short-
ened and some were eliminated entirely [17, 18]. Exposure to stress also resulted
in the rapid appearance of c-fos and c-myc mRNA in the gastric oxyntic gland
mucosa (fig. 1). The increases in c-fos mRNA levels in rats stressed for 2 h were
3- to 4-fold the normal value. By 4 h of stress, the increased expression of c-fos
had disappeared. Baseline expression of c-myc was enhanced significantly after
the 6 h of stress and remained significantly elevated for 4 more hours. Maximal
increases in c-myc mRNA levels were 7–8 times normal. By 8 h after the priod
of stress, expression of c-myc had returned to normal levels. Increased levels of
c-fos and c-myc mRNAs in the gastric mucosa after stress were paralleled by
increases in c-Fos and c-Myc proteins. Immediately following 2 h of stress, the
c-Fos protein content was significantly elevated and represented approximately
3 times the control values. Mucosal c-Myc protein content reached its highest
level following a 6-hour period of stress and remained elevated for 4 h into the
recovery period. The level of c-Myc protein reached 6–7 times normal and
returned to control value 8 h following the period of stress.
The rate of DNA synthesis as measured by [3H]thymidine incorporation in
the gastric mucosa was significantly decreased during stress and then began to

8 *
*
c-myc
Relative mRNA level 6

c-fos
4 *
a ⫺6 ⫺4 ⫺2 0 2 4 6 8 10 12
* *
*
300
DNA synthesis
200
100 *
*
⫺6 ⫺4 ⫺2 0 2 4 6 8 10 12
b Hours after stress
Fig. 1. Expression of protooncogenes c-fos and c-myc and DNA synthesis in oxyntic
gland mucosa in rats stressed for different times or in those allowed to recover for 4, 8 or 12 h
after a 6-hour period of stress. a Relative mRNA levels for c-fos and c-myc from quantita-
tive analysis of Northern blots by densitometry. Values were means from three separate
experiments with a SE of ~10%. Levels of mRNA species were corrected for RNA loading
as measured by densitometry of glyceraldehyde-3-phosphate dehydrogenase (GAPDH).
b DNA synthesis as measured by [3H]thymidine incorporation. Values are means ⫾ SE from
5 or 6 rats/groups. *p ⬍ 0.05 compared with control group.
increase with time following the 6-hour stress period. The change in the expres-
sion of c-fos and c-myc mRNAs and proteins preceded an increased rate of
DNA synthesis (fig. 1). The maximal increase in DNA synthesis, occurring
between 4 and 12 h after the period of stress, was twice the basal level. By 24 h,
although the rate of DNA synthesis remained significantly elevated, it was no
higher than that observed 4 and 12 h after stress. These results clearly indicate
Wang 86
that protooncogenes are involved in healing of gastric mucosal stress ulcers and
suggest that increased expression of protooncogenes is necessary for the early
modulation of cell division after damage.
In the support of findings in vivo, we have demonstrated that normal
growth of cultured intestinal crypt cells is associated with a significant increase
in expression of c-myc and c-jun protooncogenes [19]. After initial plating,
steady-state levels of c-myc and c-jun mRNA significantly increased from day
2 and peaked on day 4, and then returned to the basal expression, which was
maintained after the cells decreased their growth at confluence. Consistent
with the induction of the gene expression, cells began to divide on day 2. Cell
numbers increased significantly for 4 days and then entered a plateau phase on
day 6. Inhibition of the expression of c-myc and c-jun by decreasing the rate of
gene transcription results in the inhibition of cell proliferation [2, 3].
On the other hand, it is possible that other nuclear transcription factors
may also play a role in the process of gut mucosal repair following injury. One
recent report shows that NF-␬B is involved in gastric ulcer healing in rats
[46]. NF-␬B was activated in ulcerated tissue but not in normal mucosa, and
the level of the activation was decreased with ulcer healing, suggesting that
activated NF-␬B might up-regulate the expression of healing-promoting
factors responsible for gastric ulcer healing. However, contrary evidence has
been reported, indicating that NF-␬B has harmful action in the gastric
mucosa. For example, NF-␬B has been shown to play a pivotal role in the
pathogenesis of gastric mucosal inflammation and injury caused by H. pylori
infection [47].
Cells Expressing Protooncogenes in Healing of the Damaged Mucosa

Gastric glands consist of various types of cells, including mucous cells,
mucous neck cells, parietal cells, chief cells and endocrine cells. Following
mucosal injury, all of these cells have the potential to be restored by DNA
synthesis and cell division. The process of this regeneration has been analyzed
morphologically [20] and by microautoradiography with [3H]thymidine incor-
poration [21]. These studies have indicated that following injury the prolifera-
tive zone gets broader toward the bottom of glands around the lesions, and that
not only mucous neck cells but also other cells take part in the regeneration of
the gastric mucosa. These findings suggest the possibility that mature cells
making up gastric oxyntic glands may express the protooncogenes related to
regeneration after mucosal injury.
Using in situ hybridization and immunocytochemical techniques, Ito et al.
[22] examined changes in expression of the protooncogenes at different levels
of the mucosa during gastric regeneration after injury caused by indomethacin.
The administration of indomethacin consistently induced mucosal injury

which was detected both macroscopically and microscopically. Most of the
hemorrhagic lesions were observed in the fundus from 3 h after treatment.
Silver grains, corresponding to the expression of the c-myc gene, were aggre-
gated over the gastric mucosal cells, and these cells were scattered not only in
the isthmus and neck regions of glands but also at the bottom of the glands
around the lesions. Cells expressing c-myc mRNA were microscopically iden-
tified as mucous neck, parietal and chief cells. Also, expression of c-myc
mRNA was observed over enterochromaffin-like cells, which were recognized
by immunostaining with anti-histidine decarboxylase antibody. Expression of
c-myc was localized to nuclei and reached a maximum at 3 h, while expression
of the c-ha-ras gene was localized to the cytoplasm with a peak between 6 and
12 h after treatment. In addition, the distribution of cells immunostained for
c-Myc and c-Ha-Ras proteins was roughly coincident with that of cells
expressing the corresponding mRNA. On the other hand, the distribution of
cells in the S-phase roughly coincided with that of cells in which expression
of the protooncogenes was detected. These results suggest that various types
of gastric mucosal cells participate in the sequential regulated expression of
protooncogenes during healing of the damaged and may have a self-replicative
potential.
Protooncogenes in Wound Healing of Other Tissues
Protooncogene Expression in Corneal Healing

Early restitution also is an important primary repair modality in corneal
epithelial cells. This rapid process occurs as a consequence of epithelial migra-
tion into the defect, a process not requiring epithelial cell proliferation [23].
Transformation of epithelial cells to a motile form is the first step in wound
healing of the corneal epithelium. Okada et al. [24] examined the expression of
protooncogenes c-fos, c-jun, Jun-b and Jun-D in injured corneal epithelium
using in situ hybridization and demonstrated that c-fos and c-jun mRNAs were
detected in the corneal epithelium surrounding the wound 15 min after injury.
These signals reached peaks 30 –60 min after wounding but were no longer
evident at 120 min. Immunoreactivities for c-Fos and c-Jun proteins were also
detected in the same area 60 –120 min after wounding. Jun-B mRNA was not
detected around the defect until 60 min after wounding, later than the other
protooncogenes, and reached a peak after 90 min. Jun-D mRNA was detected
in normal tissue and was not altered by wounding. Li and Tseng [25] reported
that increased expression of c-fos, c-jun and c-myc was involved in the
transcriptional regulation of interleukn-1␣ and its receptor genes in injured
corneal epithelial cells. These results indicate that transcriptional activation
Wang 88
of epithelial cells is initiated in the early phase following injury and that
increased expression of protooncogenes plays a role in the healing of corneal
epithelium.
Endothelial Cells in Response to Injury

Endothelial cells occupy the interface between the blood and the sur-
rounding vascular tissue and respond to a variety of both physiological and
pathophysiological stimuli. Although originally envisioned as passive, inert
vascular lining cells, endothelial cells have been shown to be involved in
the regulation of vascular tone, coagulation, fibrinolysis, cell growth and differ-
entiation. When endothelial cells are injured, two mechanical stimuli, pressure
and shear stress, induce acute responses including ion channel activation and
the formation of prostaglandins, cytokines and nitric oxide [26]. However, these
same stimuli also selectively alter the expression of various genes resulting in
changes in endothelial morphology and cytoskeletal proteins, synthesis and
release of growth factors, and endothelial cell growth [27, 28].
Attempts to identify the role of protooncogene expression in healing
following injury in endothelial cells have not been fruitful. Injured endothelial
cells release various cytokines such as interleukins, interferon-␥, and tumor
necrosis factor. Colotta et al. [29] reported that exposure of human umbilical
vein endothelial cells (HUVEC) to interleukin (IL)-1 and tumor necrosis factor,
but not to interferon-␥ induced transient c-fos expression. They observed c-fos
expression within 1 h of exposure to cytokine which became undetectable after
4–7 h. Pretreated with cycloheximide caused c-fos expression to be superinduced
by IL-1 and tumor necrosis factor, indicating that the constitutive synthesis of
protein inhibits c-fos expression.
IL-4 interacts with other cytokines in the regulation of endothelial acti-
vation following injury. IL-4 receptors have been demonstrated on a variety
of cell types including nonhemopoietic cells and trigger many different
responses. In microvascular endothelial cells cultured from macaque
lymph node and large vessel endothelial cells derived from human umbilical
vein, IL-4 acts synergistically with IL-1 to promote endothelial cell activa-
tion, lymphocyte adhesion, and expression of the vascular cell adhesion
molecule-1 (VCAM-1) gene during injury or the inflammatory response [30].
IL-4 selectively enhances HUVEC adhesiveness for T cells but not neu-
trophils. IL-4 also significantly inhibits the constitutive expression of inter-
cellular adhesion molecule-1 (ICAM-1) and prevents the increased
expression of ICAM-1 in HUVEC exposed to IL-1, tumor necrosis factor, or
interferon-␥ [31]. These results suggest that IL-4 may play a critical role in
the transition of an acute injury or inflammatory state to a chronic state
involving an immune component.

Mechanisms of Increased Expression of Protooncogenes
The exact signal transduction pathways leading to increased expression of

protooncogenes following injury are unclear at the present time, but are thought
to result both from transcriptional and posttranscriptional mechanisms.
Transcriptional mechanisms could be due to alterations in the intracellular
location, activity, or generation of transactivating factors as well as numerous
intracellular mediators that directly or indirectly bind cis-acting elements in
the promoters of protooncogenes and influence the rate of transcription either
positively or negatively. Posttranscriptional mechanisms include regulation of
mRNA stability and translation.
Regulation of Transcription by Polyamines during Ulcer Healing

The regulation of protooncogene transcription in regenerative tissues
unquestionably is cell-type- and stimuli-dependent. We have gathered consid-
erable evidence indicating that intracellular polyamines are involved in the
regulation of transcription of protooncogenes during healing following stress
ulcers in the gastrointestinal mucosa.
Polyamines and Mucosal Healing

The polyamines, spermidine, spermine and their precursor putrescine, are
organic cations of low molecular weight, ubiquitous in eukaryotic cells, and are
intimately involved in, and required for, cell proliferation and differentiation
[32–34]. At physiological pH, putrescine, spermidine and spermine possess
two, three, and four positive charges, respectively. Together with magnesium
ions, they account for majority of the intracellular cationic charges [32, 35]. In
contrast to magnesium, intracellular polyamine levels are highly regulated by
the cell according to the state of growth. Their levels completely depend on the
activation or inhibition of the rate-limiting enzyme ornithine decarboxylase
(ODC). Polyamines are believed to exert their effects by binding to negatively
charged macromolecules such as DNA, RNA and proteins, and influencing the
chromatin structure and sequence-specific DNA-protein interactions, which
alter the level of gene transcription. It has been shown that entry of cells into
the cell cycle is accompanied by large increases in newly synthesized
polyamines by ODC, which occur in concert with increases in the expression
of a number of early response genes [32, 34]. Polyamine synthesis usually
precedes DNA synthesis, and the depletion of intracellular polyamines by treat-
ment with ␣-difluoromethylornithine (DFMO), a specific inhibitor of ODC,
results in a decrease in cell proliferation [8, 36].
A series of observations from our previous studies has demonstrated
that polyamines, either synthesized endogenously or supplied luminally, are
Wang 90
absolutely required for the process of cell division during healing of gastric and
duodenal mucosal stress ulcers [17, 18] and for normal mucosal growth [36].
Stress significantly increased ODC activity, which remained markedly elevated
over that of corresponding controls for 12 h after the period of stress in both
gastric and duodenal mucosa. Increases in mucosal content of putrescine, sper-
midine and spermine paralleled the changes in ODC activity and peaked 4 h
after stress. Administration of DFMO not only inhibited the increases of ODC
activity and polyamine levels but, also, almost totally prevented healing in both
tissues. In addition, oral administration of polyamines immediately after stress
increased the normal rate of healing and prevented the inhibition of repair
caused by DFMO [17]. Spermidine or spermine accelerated healing better than
putrescine. The delayed recovery of DNA synthesis and contents of DNA and
RNA after stress in the DFMO-treated animals was also significantly prevented
by exogenous polyamines.
We have further demonstrated that increased expression of c-fos and c-myc
protooncogenes in regenerative gastric mucosa after stress is regulated by
cellular polyamines [5]. Exposure of rats to stress results in a rapid increase in
the activity of ODC, which is associated with an increase in c-myc gene expres-
sion in the gastric mucosa. The significant increase in ODC activity occurred at
4 h of stress and peaked 4 h after a 6-hour period of stress. Baseline expression
of c-myc mRNA and protein was enhanced dramatically after 6 h of stress and
remained significantly elevated for 8 h. By 12 h after the period of stress, the
expression of c-myc has returned to near normal levels.
Administration of DFMO (500 mg/kg) in stressed animals not only prevented
the marked increases in ODC (fig. 2a) and polyamines, including putrescine,
spermidine and spermine, but also inhibited the induced expression of the c-myc
gene in the gastric mucosa (fig. 2b). The c-myc mRNA and protein levels were
decreased by ~70% immediately after the 6-hour stress period and totally absent
following a 4-hour recovery period from the 6-hour stress in DFMO-treated rats.
In cultured IEC-6 cells, we also demonstrated that polyamines stimulated
normal cell growth in association with their ability to regulate expression of the
protooncogenes [19]. Depletion of cellular polyamines by DFMO prevented
increased expression of c-myc and c-jun during log-phase growth and also
significantly reduced steady-state levels of c-myc and c-jun mRNAs during the
plateau phase. Treatment with DFMO also totally prevented the increased
expression of c-fos when 5% dialyzed fetal bovine serum was given after
24-hour serum deprivation. The remarkable parallelism that exists between
increased intracellular polyamines and induced expression of protooncogenes
during healing or normal cell growth led us to test the possibility that
polyamines are involved in the regulation of transcription of the protoonco-
genes in intestinal epithelial cells.

6
*
*
(pmol/h/mg protein)
ODC activity
4
⫹ DFMO
2
⫹
Stress
0
⫺6 0 4
a Hours after stress
a. mRNA level O
O M
M DF
DF ry⫹
⫹ ve
l ss ss
t ro t re t re eco
n s s r
Co 6h 6h 4h
c-Myc
GAPDH
b. Protein level
62 kD c-Myc
Fig. 2. Changes in ornithine decarboxylase (ODC) activity and c-myc expression in

oxyntic gland mucosa from stressed rats treated with ␣-difluoromethylornithine (DFMO, a
specific inhibitor for polyamine biosynthesis). DFMO was administered intraperitoneally
60 min before stress and repeated at 8-hour intervals after stress. Measurements were taken
either immediately after 6 h of stress or 4 h following the stress. a ODC activity. Data are
means ⫾ SE from 6 rats/per group. *, ⫹p ⬍ 0.05 compared with control rats and rats stressed
but treated without DFMO, respectively. b Levels of c-myc mRNA and protein as measured
by Northern or Western blot analysis. Three experiments are performed that showed similar
results.
Wang 92
Polyamines in Protooncogene Transcription
In order to determine the role of intracellular polyamines in the regulation
of protooncogene transcription, we carried out three experiments in IEC-6 cells,
a line derived from rat small intestinal crypt cells [37]. In the first experiment,
depletion of cellular and nuclear polyamines by treatment with DFMO for 4 or
6 days significantly decreased the steady-state levels of c-myc and c-jun
mRNA. The changes in c-myc and c-jun transcription paralleled those of their
respective cytoplasmic mRNA levels. The rate of c-myc transcription decreased
by 55% on day 4, and 60% on day 6 in DFMO-treated cells. The c-jun tran-
scription in DFMO-treated cells decreased by 75% on day 4 and 85% on day 6.
These low rates of c-myc and c-jun transcription in cells treated with DFMO
returned toward normal levels after administration of exogenous spermidine
(5 ␮M). The transcription of c-myc and c-jun in cells grown in the presence of
DFMO plus spermidine for 4 and 6 days was indistinguishable from that of
control cells. Because the decreased rate of transcription of c-myc and c-jun in
the DFMO-treated cells is completely overcome by exogenous spermidine,
inhibition of transcription of c-myc and c-jun mRNAs in the DFMO-treated
cells must be related to polyamine depletion rather than secondary to an effect
of DFMO unrelated to the inhibition of polyamine biosynthesis.
In the second experiment, we determined whether polyamine depletion
prevented the increased transcription of c-myc and c-jun after exposure of
quiescent cells to 5% dFBS, a known stimulator of these two genes [19, 37].
IEC-6 cells were grown in the presence or absence of DFMO for 4 days, and
serum was deprived for 24 h before experiments. Transcription rates of c-myc
and c-jun were measured 3 h after administration of 5% dFBS. As can be seen
in figure 3, 5% dFBS stimulated transcription rates significantly in normal
quiescent cells. The increased rates of c-myc and c-jun were 8 and 3.5 times the
control levels, respectively. These increases were prevented significantly by
polyamine depletion. In polyamine-deficient cells, the rate of c-myc transcrip-
tion slightly increased after exposure to 5% dFBS and was twice the control
level. The increased transcription of c-jun by 5% dFBS was completely
prevented in the DFMO-treated cells.
In the third experiment, we examined the effect of addition of spermidine
to nuclei isolated from control and DFMO-treated cells on the transcription
rates of c-myc and c-jun [37]. Cells were initially treated with or without DFMO
for 6 days, and then nuclei were isolated. When the transcription rates of c-myc
and c-jun were examined before and after spermidine addition to isolated nuclei,
significant differences between control and DFMO-treated cells were observed.
There was no significant alteration in the transcription rate of c-myc or c-jun
gene in the nuclei from control IEC-6 cells (without DFMO) after the addition
of spermidine at different concentrations. In contrast, spermidine addition to

Normal cells DFMO-treated cells
S
B
B
l
l
tro
tro
dF
dF
on
on
5%
5%
C
C
c-myc
c-jun
pSV2neo
GAPDH
Fig. 3. Serum-induced c-myc and c-jun transcription in control and polyamine-deficient

quiescent intestinal epithelial cells (IEC-6 cell line). Cells were grown in the presence or
absence of 5 mM ␣-difluoromethylornithine (DFMO) for 4 days and then serum deprived
for 24 h before experiments. The rate of transcription was measured 3 h after exposure to 5%
dialyzed fetal bovine serum (dFBS). Equal amounts of [␣-32P]UTP-labeled RNA
(5 ⫻ 106 cpm) were hybridized to filter containing immobilized plasmids of c-myc, c-jun,
GAPDH (severed as positive control) and pSV2neo (as negative control). Three experiments
were performed that showed similar results.
polyamine-deficient nuclei from DFMO-treated cells resulted in a marked

increase in the transcription rates of c-myc and c-jun genes. Administration of
spermidine resulted in 2- to 2.5-fold increase in c-myc and c-jun transcrip-
tion without altering the transcription of GAPDH gene. The increase in c-myc
and c-jun transcription in response to spermidine addition to nuclei was
concentration-dependent with a maximum increase observed at a concentration
of 0.5 mM. The effect of spermidine addition on the transcription rates of c-myc
and c-jun in isolated nuclei from DFMO-treated cells was not accounted for by
replacement of positive changes, since increasing MgCl2 concentration by
0.5–2.0 mM over the standard conditions of 5 mM MgCl2 in the assay system
had no effect on the transcription of any specific gene. These results are consis-
tent with data from other investigators [38, 52], who have demonstrated that
spermidine induces an increase in protooncogenes and other transcription
factors in nuclei from certain types of cells.
Taken together, these data clearly show that intracellular polyamines play
a critical role in the regulation of transcription of protooncogenes in intestinal
epithelial cells. In stress-induced gastric ulcers, mucosal polyamine levels
Wang 94
dramatically increased in response to damage [17, 18]. Because increased
cellular polyamines alter the rate of protooncogene transcription, it is likely that
increases in expression of the protooncogenes during healing of gastric mucosal
stress ulcers are primarily caused by the activation of gene transcription.
Possible Mechanisms of Action of the Polyamines

The molecular mechanisms by which polyamines modulate transcription of
protooncogenes in the gastrointestinal mucosa are unclear. Several studies have
suggested that changes in chromatin structure by polyamines may be correlated
with levels of transcription of c-myc and other genes [39]. In cell-free systems,
polyamines have been shown to affect B-to-Z conformational changes in DNA
as well as changes in chromatin and nucleosomal structure [39, 40]. Because the
addition of similar concentrations of simple divalent cations such as Mg2⫹ has
no specific effect on DNA structure, the alterations achieved by polyamines in
these cell-free systems is not the result of simply altering the ionic environment.
These findings are consistent with our previous results [18, 36] and those of
others [32, 34] showing that exogenous polyamines specifically reverse the
inhibition of cell proliferation following polyamine depletion.
In addition to effects on DNA structure, polyamines have been shown to
alter sequence-specific DNA protein-binding activities, which may affect the
regulation of initiation, elongation and termination during transcription.
Intracellular polyamine levels may provide an ionic environment that could be
modulated to alter the binding or release of transcriptional regulatory factors
[32, 35, 41]. Some of these effects are polyamine-specific, while others are due
to the general cationic nature of these compounds [35, 42]. Panagiotidis et al.
[43] reported that at physiological concentrations, polyamines specifically
enhanced the binding of several proteins including upstream stimulatory factor
(USF), transcription factor E3 (TFE3), immunoglobulin/enhancer-binding pro-
tein (Ig/EBP), nuclear factor-IL-6 (NF-IL6), and Yin-Yang-binding protein-1
(YY1) to DNA, but inhibited others such as octamer-binding protein-1 (Oct-1).
Polyamines facilitate formation of complexes involving binding of more than
one protein on a DNA fragment but do not influence DNA-protein contacts. The
decreased rate of c-myc and c-jun transcription that we observed in polyamine-
deficient IEC-6 cells may result from abnormalities in the interaction between
transcription factors and their cognate DNA sequences. Clearly, further study is
necessary to examine whether polyamines play a specific role in the interaction
or synthesis of key transcriptional regulatory factors in intestinal epithelial cells.
Posttranscriptional Regulation
Messenger RNAs that must be rapidly produced and rapidly cleared are
often regulated posttranscriptionally. Posttranscriptional regulation can be

simple, e.g., phosphorylation or dephosphorylation, or complex and circuitous,
e.g., regulation of the stability of cytoplasmic mRNAs. The mRNAs coding for
cellular structural proteins such as actin are remarkably stable, with half-lives
exceeding 17 h. In contrast, protooncogenes and many other inducible genes
have unstable mRNAs with half-lives of 10–60 min. The interaction of cis ele-
ments with proteins regulates the stability of mRNA. Polyadenylation of the
3⬘ end of mRNA transcripts has been shown to protect mRNA from rapid
turnover, and removal of the poly A tail seems to precede mRNA degradation.
The short-lived c-fos and c-myc mRNAs show a rapid poly A shortening,
preceding their complete degradation [44]. Poly A mRNA complexes with
cytoplasmic binding protein, a 70-kD poly A-binding protein. It is believed that
internal sequences in labile mRNAs interrupt or destabilize the binding of poly A-
binding protein to the poly A tail, which significantly increases susceptibility of
these mRNAs to deadenylation and decay [44, 45].
Because intracellular polyamine levels markedly increase during healing
following injury [17, 18] and these compounds have been shown to modulate
the transcription of protooncogenes in the gastrointestinal epithelial cells [38],
polyamines may regulate posttranscription of the protooncogenes by altering
protein binding. To test the possibility that polyamines alter the stability of
c-myc and c-jun mRNAs, we examined the half-lives of c-myc and c-jun
mRNAs in control and polyamine-deficient IEC-6 cells [37]. We have demon-
strated that the mRNA levels for c-myc and c-jun declined rapidly after the
administration of actinomycin D. In control cells, the mRNA levels for c-myc
decreased with a half-life of 30 min. The mRNA for c-myc in the polyamine-
deficient cells decreased at a same rate, with a half-life of 34 min. These slight
differences were not statistically significant. In general, the rate of degradation
of c-jun mRNA in IEC-6 cells was similar to that observed for c-myc. There was
no significant difference in the stability of mRNAs for c-jun between control
and the polyamine-deficient cells. The half-lives of c-jun mRNA in control
cells and cells treated with DFMO for 6 days were approximately 30 and
35 min, respectively. These results clearly show that intracellular polyamines
play no role in the regulation of posttranscription of the protooncogennes c-myc
and c-jun in intestinal epithelial cells.
It is interesting that cellular polyamines have been shown to modulate
expression of growth-inhibiting genes such as p53 and TGF-␤/TGF-␤ receptor
through posttranscriptional regulation in intestinal epithelial cells [48–51].
Polyamines are repressors for expression of growth-inhibiting genes and deple-
tion of cellular polyamines significantly increases mRNA levels for p53, TGF␤
and TGF␤ receptor in intestinal epithelial cells. These inhibitory effects of
polyamines result primarily from destabilization of the mRNA rather than
from alteration of gene transcription [49, 51]. However, the exact role of
Wang 96
polyamine-modulated expression of growth-inhibiting genes in the process of
gut mucosal healing remains to be demonstrated.
There are several additional posttranscriptional mechanisms by which
cells modulate gene expression by regulating mRNA processing, transport
and translation. However, no data are available regarding the regulation of
protooncogene expression during wound healing through those processes.
Therefore, the significance and role of mRNA processing, transport and trans-
lation in the activation of protooncogene expression in response to wounding
are still unknown.
Conclusions
The experimental data summarized in this chapter support the hypothesis

that increased expression of protooncogenes is involved in the process of ulcer
and wound healing in the gastrointestinal mucosa and other tissues. The repair
process following ulceration or wounding is associated with significant
increases in the expression of protooncogenes. In the gastric mucosa, the
changes in the expression of c-fos and c-myc precede the induction of DNA
synthesis after stress. Inhibition of expression of the protooncogenes by
decreasing cellular polyamine levels significantly decreased mucosal healing.
Although the wide variety of growth regulators has complicated our under-
standing of the overall regulation of healing and normal mucosal growth,
increased protooncogene expression is involved in the regulation of epithelial
cell division by both humoral and local stimulants. Therefore, an understanding
of the control of protooncogene expression provides novel information regard-
ing the mechanisms regulating mucosal growth during adaptive responses
as well as in the course of normal physiology. In corneal epithelium and
endothelial cells, a series of studies regarding expression of protooncogenes in
response to injury has not been fruitful, but available results also suggest that
protooncogene products are implicated in wound healing in these two tissues.
The activation of protooncogene expression is a rapid and transient process
in response to damage. The induced expression of protooncogenes in damaged
tissues is highly regulated and returns to normal levels following repair.
However, there are many situations where these acute alterations become fixed
and are associated either with the development or progression of chronic
disease. Sequential protooncogene expression is a major mechanism involved
in neoplastic diseases in the gastrointestinal mucosa and other tissues. In order
to understand the mechanism by which these acute transient changes become
permanent, it is necessary to fully elucidate the process responsible for the
regulation of protooncogenes during healing. Further studies of the control of

gene expression and acute changes in response to injury in the gastrointestinal
mucosa and endothelial cells will improve our understanding of the patho-
genesis of neoplastic diseases as well as the processes involved in the repair of
normal mucosa.
References
1 McMahon SB, Monroe G: Role of primary response genes in generating cellular responses to
growth factors. FASEB J 1992;6:2707–2715.
2 Facchini LM, Penn LZ: The molecular role of Myc in growth and transformation: Recent discov-
eries lead to new insights. FASEB J 1998;12:633–651.
3 Angel P, Karin M: The role of Jun, Fos and the AP-1 complex in cell-proliferation and transfor-
mation. Biochim Biophys Acta 1991;1072:129–157.
4 Abate C, Curran T: Encounters with fos and jun on the road to AP-1. Semin Cancer Biol 1990;
1:19–26.
5 Wang JY, Johnson LR: Expression of proto-oncogenes c-fos and c-myc in healing of gastric
mucosal stress ulcers. Am J Physiol 1994;266:G878–G886.
6 Verrier B, Muller D, Bravo R, Muller R: Wounding a fibroblast monolayer results in the rapid
induction of the c-fos proto-oncogene. EMBO J 1986;5:913–917.
7 Gu W, Bhatia K, Magrath IT, Dang CV, Dallon-Favern R: Binding and suppression of the myc
transcriptional activation domain by p107. Science 1994;264:251–254.
8 McCormack SA, Johnson LR: Role of polyamines in gastrointestinal mucosal growth. Am
J Physiol 1991;260:G795–G806.
9 Johnson LR: Regulation of gastrointestinal mucosal growth. Physiol Rev 1988;68:456–502.
10 Madara JL, Trier JS: The functional morphology of the mucosa of the small intestine; in
Johnson LR (ed): Physiology of the Gastrointestinal Tract, ed 3. New York, Raven Press, 1994,
pp 1577–1622.
11 Vanamee P, Winawer SJ, Sherlock P, Somenberg M, Lipkin M: Decreased incidence of restraint-
stress induced gastric erosions in rats treated with bovine growth hormone. Proc Soc Exp Biol
Med 1970;135:259–262.
12 Takeuchi K, Johnson LR: Pentagastrin protects against stress ulceration in rats. Gastroenterology
1979;76:327–334.
13 Takeuchi K, Johnson LR: Effect of cell proliferation on healing of gastric and duodenal ulcers in
rats. Digestion 1986;33:92–100.
14 Makino R, Hagashi K, Sugimura T: c-myc transcript is induced in rat liver at a very early stage of
regeneration or by cycloheximide treatment. Nature 1984;310:697–698.
15 Fausto N: Messenger RNA in regenerating liver: Implications for the understanding of regulated
growth. Mol Cell Biochem 1984;59:131–147.
16 Thompson NH, Mead JE, Brown L, Goyette M, Shank PR, Fausto N: Sequential proto-oncogene
expression during rat liver regeneration. Cancer Res 1986;46:3111–3117.
17 Wang JY, Johnson LR: Luminal polyamines stimulate repair of gastric mucosal stress ulcers. Am
J Physiol 1990;259:G584–G592.
18 Wang JY, Johnson LR: Polyamines and ornithine decarboxylase during repair of duodenal mucosa
after stress damage in rats. Gastroenterology 1991;100:333–343.
19 Wang JY, McCormack SA, Viar MJ, Wang H, Tzen CC, Scott RE, Johnson LR: Decreased expres-
sion of proto-oncogenes c-fos, c-myc and c-jun following polyamine depletion in IEC-6 cells. Am
J Physiol 1993;265:G331–G338.
20 Hunt TE: Regeneration of the gastric mucosa in rat. Anat Rec 1961;131:193–212.
21 Myhre E: Regeneration of the fundic mucosa in rats. Arch Pathol 1960;70:476–485.
22 Ito T, Hayashi N, Sasaki Y, Morita Y, Kawano S, Fusamoto H, Sato N, Tohyama M, Kamada T:
Sequential protooncogene expression during regeneration in rat stomach. Gastroenterology 1990;
98:1525–1531.
Wang 98
23 Silen W, Ito S: Mechanism for rapid epithelialization of the gastric mucosal surface. Annu Rev
Physiol 1985;47:217–229.
24 Okada Y, Saika S, Hashizume N, Kobata S, Yamanaka O, Ohnishi Y, Senba E: Expression of fos
family and jun family protooncogenes during corneal epithelial wound healing. Curr Eye Res
1996;15:824–832.
25 Li DQ, Tseng SC: Differential regulation of cytokine and receptor transcript expression in human
corneal and limbal fibroblasts by epidermal growth factor, transforming growth factor-␣,
platelet-derived growth factor B and interleukin-1␤. Invest Ophthalmol Vis Sci 1996;37:
2068–2080.
26 Gerritsen ME, Bloor CM: Endothelial cell gene expression in response to injury. FASEB J 1993;
7:523–532.
27 Ando J, Nomura H, Kimaya A: Effect of fluid shear stress on the migration and proliferation of
cultured endothelial cells. Microvasc Res 1987;33:62–70.
28 Levesque MJ, Nerem RM, Sprague EA: Vascular endothelial cell proliferation in culture and the
influence of flow. Biomaterials 1990;11:702–707.
29 Colotta F, Lampugnanii MG, Polentaretti N, Dejana E, Montovani A: Interleukin-1 induces c-fos
protooncogene expression in cultured human endothelial cells. Biochem Biophys Res Commun
1988;152:1104–1110.
30 Masinovsky B, Urdal D, Gallatin WM: IL-4 acts synergistically with IL-2␤ to promote lymphocyte
adhesion to microvascular endothelium by induction of vascular adhesion molecule-1. J Immunol
1990;145:2886–2895.
31 Thornhill MH, Haskard DO: IL-4 regulates endothelial cell activation by IL-1, tumor necrosis
factor or IFN-␥. J Immunol 1990;145:865–872.
32 Pegg AE: Polyamine metabolism and its importance in neoplastic growth and as a target for
chemotherapy. Cancer Res 1988;48:759–774.
33 Bachrach U, Wang YC, Tabib A: Polyamines: New cues in cellular signal transduction. News
Physiol Sci 2001;16:106–109.
34 Tabor CW, Tabor H: Polyamines. Annu Rev Biochem 1984;53:749–790.
35 Janne J, Alhonen L, Leinonen P: Polyamines: From molecular biology to clinical applications.
Ann Med 1991;23:241–259.
36 Wang JY, McCormack SA, Viar MJ, Johnson LR: Stimulation of proximal small intestinal
mucosal growth by luminal polyamines. Am J Physiol 1991;261:G504–G511.
37 Patel AR, Wang JY: Polyamines modulate transcription but not posttranscription of c-myc and
c-jun in IEC-6 cells. Am J Physiol 1997;273:C1020–C1029.
38 Celano P, Baylin SB, Caserro RA: Polyamines differentially modulate the transcription of growth-
associated genes in human colon carcinoma cells. J Biol Chem 1989;264:8922–8927.
39 Gross DS, Garrard WT: Nuclease hypersensitive sites in chromatin. Annu Rev Biochem 1988;
57:159–197.
40 Elgin SCR: The formation and function of DNase I hypersensitive sites in the process of gene
activation. J Biol Chem 1988;263:19259–19262.
41 Moore JJ, Lungren DW, Moore RM, Andersen B: Polyamines control human chorionic
gonadotropin production in the JEG-3 choriocarcinoma cell. J Biol Chem 1988;263:12765–12769.
42 Watanabe T, Sherman M, Shafman T, Iwata T, Kufe D: Effects of ornithine decarboxylase
inhibition on c-myc expression during murine erythroleukemia cell proliferation and differentia-
tion. J Cell Physiol 1986;127:480–484.
43 Panagiotidis CA, Artandi S, Calame K, Silverstein SJ: Polyamines alter sequence-specific DNA-
protein interactions. Nucleic Acids Res 1995;23:1800–1809.
44 Raghow R: Regulation of messenger RNA turnover in eukaryotes. Trends Biochem Sci 1987;12:
358–360.
45 Malter JS: Identification of an AUUUA-specific messenger RNA binding protein. Science
1989;246:664–666.
46 Takahashi S, Fujita T, Yamamoto A: Role of nuclear factor-␬B in gastric ulcer healing in rats. Am
J Physiol 2001;280:G1296–G1304.
47 Keates S, Hitti YS, Upton M, Kelly CP: Helicobacter pylori infection activates NF-␬B in gastric
epithelial cells. Gastroenterology 1997;113:1099–1109.

48 Li L, Li J, Rao JN, Li M, Bass BL, Wang JY: Inhibition of polyamine synthesis induces p53 gene
expression but not apoptosis. Am J Physiol 1999;276:C946–C954.
49 Li L, Rao JN, Guo X, Liu L, Santora R, Bass BL, Wang JY: Polyamine depletion stabilizes p53
resulting in inhibition of normal intestinal epithelial cell proliferation. Am J Physiol 2001;
281:C941–C953.
50 Patel AR, Li J, Bass BL, Wang JY: Expression of the transforming growth factor-␤ gene during
growth inhibition following polyamine depletion. Am J Physiol 1998;275:C590–C589.
51 Rao JN, Li L, Bass BL, Wang JY: Expression of the TGF-␤ receptor gene and sensitivity to growth
inhibition following polyamine depletion. Am J Physiol 2000;279:C1034 –C1044.
52 Shah N, Thomas T, Shirahata A, Sigal LH, Thomas TJ: Activation of nuclear factor ␬B by
polyamines in breast cancer cells. Biochemistry 1999;38:14763–14774.
53 Hermeking H, Rago C, Schuhmacher M, Li Q, Barrett JF, Obaya AJ, Connell BC, Mateyak MK,
Tam W, Kohlhuber F, Dang CV, Sedivy JM, Eick D, Vogelstein B, Kinzler KW: Identification of
CDK4 as a target of c-myc. Proc Natl Acad Sci USA 2000;97:2229–2234.
54 Gartel AL, Ye X, Goufman E, Shianov P, Hay N, Najmabadi F, Tyner AL: Myc represses the
p21(waf1/cip1) promoter and interacts with Sp1/Sp3. Proc Natl Acad Sci USA 2001;98:
4510–4515.
Dr. Jian-Ying Wang, Department of Surgery, Baltimore VA Medical Center,

10 North Greene Street, Baltimore, MD 21201 (USA)
Tel. ⫹1 410 605 7000/ext 5678, Fax ⫹1 410 605 7919, E-Mail Jwang@smail.umaryland.edu
Wang 100
Role of Angiogenesis and Angiogenic

Growth Factors in Mucosal Repair and
Ulcer Healing
Andrzej S. Tarnawski a,b, Michael K. Jones a,b, Dolgar Baatar b, Rama Pai b
a
VA Medical Center, Long Beach, Calif. and the bUniversity of California,
Irvine, Calif., USA
Gastrointestinal mucosal structural integrity, viability and function are

critically dependent on blood flow through mucosal microvessels (capillaries,
collecting venules) which is essential for delivery and oxygen and nutrients
[1, 2]. In the gastrointestinal mucosa there is a dense network of blood
microvessels [3, 4] (fig. 1–3). At the level of muscularis mucosae, submucosal
arteries branch into capillary vessels which transverse the lamina propria, in
proximity to gastric glands, to the base of the surface epithelial cells where they
converge into collecting venules [3–5]. The endothelial cells lining microves-
sels generate potent vasodilators such as nitric oxide (NO) and prostacyclin
which protect the gastric mucosa against injury and oppose the mucosal dam-
aging action of vasoconstrictors such as leukotriene C4, thromboxane A2 and
endothelin.
When the gastric mucosa is exposed to injurious factors or acid, a marked
and rapid increase in mucosal blood flow occurs [6] that serves to remove the
back-diffusing HCl and/or noxious agents and to dilute them. This hyperemic
response appears to be essential for mucosal defense since its abolishment via
mechanical restriction of blood flow leads to the development of hemorrhagic
necrosis. Holzer [7] demonstrated that this hyperemic response is mediated by
sensory afferent nerves. These nerves, having endings localized just beneath the
surface epithelium, sense the presence of injurious factors or acid in the gastric
lumen [8]. Activation of these sensory nerves leads to the release of neurotrans-
mitters such as calcitonin gene-related peptide (CGRP) and substance P at the
nerve terminal, located within or close to the large submucosal vessels [7–9].
Release of CGRP from sensory afferent neurons produces direct vasodilatation
a b
Fig. 1. a Vascular capillary casts of normal gastric mucosa in rat. Vasculature was
filled with Mercox resin (Dainippon Inc., Co, Ltd, Tokyo, Japan). Thirty minutes later, the
stomach was removed, cut into small pieces and immersed in 20% NaOH for 6 h to dissolve
tissue external to the cast. Desiccated specimens were coated with gold/palladium (60:40)
and examined under a scanning electron microscope 5500, Hitachi, Ltd, Tokyo, Japan) at
10 kV [reprinted from 3, with permission]. b Transmission electron micrograph of normal
human gastric mucosa. A capillary vessel with erythrocytes (E) in the lumen about 10–15 ␮m
below the surface epithelial cells. This capillary represents the most superficial capillary
vessel. MG ⫽ Mucosal granules, N ⫽ nucleus of endothelial cell. ⫻1,800. [reprinted from
5, with permission].
and in addition stimulates local generation of NO. NO causes relaxation of adja-

cent vascular smooth muscle cells – vasodilatation and increased blood flow [9].
Interference with any aspect of the sensory innervation interferes with the
hyperemic response and, thus, with the resistance of the gastric mucosa to injury.
Ablation of the sensory afferent nerves (by chronic, large dose treatment with
capsaicin) abolishes the hyperemic response and greatly increases the suscepti-
bility of gastric mucosa to damage [6, 7, 9]. Similar to the hyperemic response
to irritants, increased mucosal blood flow in conjunction with acid secretion is
mediated, at least in part, by NO [9, 10].
NO is generated from the terminal guanidino-nitrogen atoms of L-arginine
by the enzyme NO synthase (NOS) [11]. The NOS enzyme exists in two forms:
a constitutive moiety (cNOS) that is calcium-dependent and an inducible moiety
Tarnawski/Jones/Baatar/Pai 102
J
2 3
Fig. 2. Transmission electron micrograph of human gastric mucosal capillary. The

structure of the capillary wall and endothelial cell cytoplasm is normal with a characteristic
fenestration (arrows) and presence of endothelial vesicles. BM ⫽ Basement membrane,
E ⫽ erythrocytes in the capillary lumen, J ⫽ junction between two neighboring endothelial
cells, CF ⫽ collagen fibers. ⫻17,400 [reprinted from 5, with permission].
Fig. 3. Electron micrograph of the gastric mucosa 5 min after alcohol administration.
The continuity of the capillary wall is broken and erythrocytes (E), together with coagulated
plasma (P), are leaking into edematous lamina propria. ⫻17,400 [reprinted from 5, with
permission].
(iNOS) that is calcium-independent. In the gastric mucosa, NO plays a major role

in mucosal defense by modulating the mucosal circulation [12]. Endogenous and
exogenous NO protect the gastric mucosa against injury by ethanol and endothe-
lin 1, whereas inhibition of NOS (resulting in reduced NO generation) increases
gastric mucosal injury [13].
Disruption of the mucosal defense permits ulcerogenic agents and aggres-
sive factors to penetrate into the mucosa initiating release of proinflammatory
and vasoactive mediators (serotonin, leukotriene C4, platelet-activating factor,
endothelin) and directly digesting cellular and connective tissue components of
the mucosa. The chain of events culminates in formation of mucosal erosions
or, if submucosal vessels are involved, in ulcerations [4, 14].
Gastrointestinal microvessels are also major targets of injury produced by
a variety of necrotizing and/or ulcerogenic agents including nonsteroidal anti-
inflammatory drugs (NSAIDs), ethanol, bile acids and others [5, 15, 16]. Injury
Angiogenesis and Mucosal Repair 103

of the endothelial cells leads to microvessel damage (rupture, thrombi forma-
tion) with resulting blood flow stasis and cessation of oxygen and nutrient
delivery. Microvascular damage occurs early during mucosal injury, preceding
necrosis of glandular cells, and adds an ischemic component to the direct toxic
injury of the cells [15, 16]. Vascular changes (e.g., constriction of veins) pro-
duced by the release of vasoactive, proinflammatory mediators from damaged
mast cells, macrophages and endothelial cells, further impair the mucosal
microcirculation and ultimately result in mucosal necrosis. When the microvas-
culature is damaged, endothelial cells lining microvessels in the periphery of
injured areas initiate repair and reconstruction of the microvascular network
through the angiogenic process.
The vascular and microvascular changes are the earliest events critical for
the development of experimental gastric ulcers [14]. These vascular changes
(thrombi, vascular constriction) cause mucosal ischemia, free radical formation
and cessation of nutrient delivery, all resulting in ischemic necrosis of the
mucosa and muscularis mucosae. The difference between an erosion and an
ulcer is that the former is confined to the mucosa, while an ulcer penetrates
through muscularis mucosae.
Repair of Mucosal Injury
Following acute gastric mucosal necrosis such as deep erosions or ulcers,

all mucosal components, including microvessels, are destroyed within the focal
lesions [17, 18]. Healing of such deep mucosal lesions requires reconstruction
of the surface epithelium, glandular epithelial structures, restoration of the lam-
ina propria and, most importantly, reconstruction of the mucosal microvascular
network essential for delivery of oxygen and nutrients to the healing site [4, 17,
18]. While repair of the surface epithelium through restitution has been well
characterized and extensively studied [6, 19, 20], the repair of deep mucosal
injury, e.g., restoration of connective tissue and sensory nerves, is less under-
stood. Our studies demonstrated that the repair of gastric mucosal erosions fol-
lowing ethanol injury, and restoration of the gastric mucosal microvascular
network, require angiogenesis [17, 21].
Angiogenesis is a fundamental process essential for reproduction, embry-
onic development, postnatal growth and reparative processes such as wound
healing [22]. In certain situations (e.g., wound or focal tissue necrosis), the rest-
ing phenotype of endothelial cells is changed to an angiogenic phenotype [22].
As a result, the microvascular endothelial cells from preserved microvessels at
the wound edge migrate, proliferate and attempt to re-establish a microvascular
network through the process of angiogenesis [17, 18, 21].
Angiogenesis has been studied both in vitro and in vivo in tissue other
than gastrointestinal mucosa. In vitro studies have demonstrated that microvas-
cular endothelial cells, when grown under appropriate conditions, possess the
ability to self-associate into three-dimensional structures resembling a capil-
lary network [23, 24]. Further studies have indicated that endothelial cell pro-
liferation and formation of microvascular structures are triggered by certain
extracellular matrix components (collagens I and III), phorbol esters, basic
fibroblast growth factor (bFGF) and vascular endothelial growth factor
(VEGF) [22–24].
Our studies demonstrated that, following acute alcohol injury, mucosal
microvessels in the gastric mucosa-bordering necrosis, at the edge of non-
necrotic gastric mucosa, undergo angiogenesis: basement membrane dissolu-
tion, endothelial cell budding into the extravascular space, tube formation and,
ultimately, reconstruction of the capillary network [21]. Gastric mucosal angio-
genesis is strongly stimulated by prostacyclin and human recombinant bFGF,
e.g., 5 ␮m bFGF mutein enhanced angiogenesis by 800%, wild-type bFGF
10 ␮g/kg by 500% [25, 27]. Preliminary qualitative assessment indicates that
enhancement of angiogenesis results in a faster and more complete (‘good qual-
ity’) mucosal restoration [26]. In contrast, inhibition of gastric angiogenesis by
indomethacin (0% angiogenic response at 24 h after ethanol injury), resulted in
delayed and less complete (‘poor quality’) mucosal repair [27].
We have also demonstrated that ethanol-induced gastric mucosal injury
triggers activation of genes encoding bFGF and its receptors and increased
bFGF protein expression in the mucosa-bordering necrosis, especially in sprout-
ing microvascular endothelial tubes [28].
In a separate study, we demonstrated that acute gastric mucosal injury by
ethanol triggers a 4- to 6-fold overexpression of VEGF mRNA and proteins in
the mucosa-bordering necrosis where the angiogenic process takes place [17]
(fig. 4–6). VEGF mRNA expression was increased ~630, 553 and 385% (all
p ⬍ 0.001) at 3, 6 and 24 h respectively after injury [17]. VEGF165 protein
expression was also significantly increased in injured gastric mucosa-bordering
necrosis ~365, 318 and 185% respectively at 3, 6 and 24 h [17]. Quantitative
assessment of angiogenesis 24 h following alcohol injury demonstrated that
~9% of microvessels in the mucosa-bordering necrosis display endothelial
sprouting reflecting angiogenesis [17]. Since VEGF is the most potent angio-
genic growth factor specifically acting on endothelial cells (because only these
cells possess VEGF receptors), these data strongly suggested that angiogenesis
in injured gastric mucosa was triggered by activation of the VEGF gene [17].
To determine the role of endogenous VEGF in angiogenesis and repair of
injured gastric mucosa, we injected anti-VEGF neutralizing antibody intra-
venously concurrently with intragastric administration of ethanol. Anti-VEGF

erosion
a b
Fig. 4. Photomicrographs of gastric mucosa showing immunofluorescence staining for

vimentin. a In normal mucosa, vimentin shows a regular pattern of distribution in endothe-
lial cells lining microvessels. ⫻400. b In mucosal erosion resulting from ethanol-induced
injury, there is virtually no vimentin fluorescence, reflecting destruction of mucosal
microvessels (arrows). ⫻400. c Gastric mucosa 24 h after ethanol administration. Migrating
tubes of endothelial cells (arrows) are clearly identified, reflecting angiogenesis. In migrat-
ing tubes, vimentin was found to be co-localized with expression of factor VIII-related
antigen. ⫻1,000 [reprinted from 17, with permission].
antibody significantly reduced, by 3-fold, the angiogenic response to alcohol

injury and delayed healing [17]. These data clearly demonstrate the essential
role of local VEGF overexpression in the stimulation of angiogenesis and in
mucosal injury healing. This study also demonstrated that, in rat endothelial
cells, the expression of VEGF is mediated by the Ras pathway and that the
1 2 3 4 1 2 3 4
Competitor
VEGF
a b
1 2 3 4 1 2 3 4
Competitor
VEGF
c d
p⬍0.001
0.8
p⬍0.001
Total RNA (amol/␮g)
0.6
p⬍0.001
0.4
0.2
0
Control 3h 6h 24h
e Water-treated ETOH-treated
Fig. 5. Quantification of VEGF mRNA expression in gastric mucosa at 3, 6 and 24 h

after ethanol treatment by competitive RT-PCR. Competitive RT-PCR was performed to pro-
vide quantification of VEGF mRNA expression in gastric mucosa of control animals (a) and
of animals at 3 h (b), 6 h (c) and 24 h (d) after intragastric administration of 100% ethanol.
(e) Quantitative data for competitive RT-PCR shown in a–d are given here in graph form.
Log of ratio of VEGF target intensity to competitor intensity was plotted against log of com-
petitor amount used. Amount of VEGF target cDNA was calculated by determining the
x-intercept for the point on the curve at which the ratio of target to competitor equaled 1.
Total mRNA content of each sample was normalized using ␤-actin as internal control. Values
are means ⫾ SD [reprinted from 17, with permission].
increase in Ras expression caused by mucosal injury likely upregulates VEGF

expression leading to the angiogenic response [17].
Another important mechanism for VEGF gene activation in response to
mucosal injury is hypoxia. Deep mucosal injury is associated with ischemia and
the resulting hypoxia. Mammalian cells are able to sense oxygen levels and
activate a number of genes in response to hypoxia. Hypoxia-inducible factor
(HIF)-1, a transcription factor, is an important component of the hypoxia
signaling pathway [29]. HIF-1 is a heterodimer composed of two members
of the basic helix-loop-helix transcription factor super family: HIF-1␣ and
HIF-1␤/ARNT. Hypoxia stabilizes the HIF-1␣ protein thereby facilitating its

VEGF
Controls 3h 6h 24 h
Water-treated ETOH-treated
a
p⬍0.001
1.2 p ⬍0.001
Relative density
0.8
p⬍ 0.002
0.4
0
Control 3h 6h 24 h
b Water-treated ETOH-treated
Fig. 6. Western blot analysis of VEGF protein expression in gastric mucosa after
ethanol injury compared with gastric mucosa of controls. a VEGF protein expression. Level
of VEGF protein from gastric mucosa samples of control animals is shown together with
that of gastric mucosa samples from animals at each time point after intragastric administra-
tion of 100% ethanol. b Quantitative data of VEGF protein expression (shown in a) obtained
by densitometric scanning using values of peak area. Values are means ⫾ SD [reprinted from
17, with permission].
accumulation within the cell [29, 30]. Following exposure to hypoxia, HIF-1␣
forms a heterodimer with the nuclear translocator HIF-1␤/ARNT. The HIF-1
complex is then translocated to the nucleus where it binds to the promoter ele-
ment of several genes including those encoding NOS and VEGF and initiates
[31, 32] transcription.
To determine whether HIF-1␣ is activated following gastric mucosal
injury, we examined its expression and localization in mucosa injured by alco-
hol [33]. Following alcohol injury, gastric mucosa-bordering necrosis demon-
strated a significant increase in HIF-1␣ mRNA at 3 and 6 h (40 ⫾ 4%,
19 ⫾ 2%; p ⬍ 0.05) and protein (⬎300 ⫾ 16%; p ⬍ 0.02) at all time points
with a peak at 1–3 h [33]. HIF-1␣ signal was detected in regenerating mucosal
microvessels, where it co-localized with VEGF [33]. Since HIF-1␣ initiates
transcription of VEGF mRNA, HIF-1␣ activation by ethanol-induced injury is
likely responsible for activation of the VEGF gene and induction of angiogen-
esis in response to the ischemia associated with ethanol-induced injury.
Angiogenesis:
Signaling
• Basement membrane dissolution
pathways • Endothelial cell proliferation and migration
• Formation of new capillary vessels
Fibroblast proliferation
Activation of genes for:

• bFGF & FGF-R, 2, 4
• VEGF and flk-1/KDR • Delivery of oxygen and nutrients to the
• Angiopoietin 1 (Ang1) healing site
• Angiopoietin 2 (Ang2) • Reconstruction of microvessels in the
• Tie2 receptor ulcer scar
• Tie1 receptor (ligand not identified yet) • Reconstruction of connective tissue lamina
propria (cells and matrix)
Fig. 7. Diagrammatic representation of cellular and molecular events in granulation

tissue. Granulation tissue develops within 48–72 h after ulceration. It consists of connective
tissue cells, macrophages, fibroblasts and proliferating endothelial cells forming microves-
sels through angiogenesis. In granulation tissue, the activation of genes encoding for bFGF,
VEGF, angiopoietins and their receptors promotes angiogenesis – new capillary vessel for-
mation – essential for delivery of oxygen and nutrients to the healing site and reconstruction
of the microvessels and connective tissue cells within the ulcer scar.
Ulcer Healing, Granulation Tissue and Angiogenesis
The main sites for capillary regeneration (angiogenesis) during healing of

erosions are preserved microvessels in the mucosa-bordering necrosis (fig. 7)
[17, 21, 25–27]. However, when the injury penetrates through muscularis
mucosae (ulcer), the predominant site for angiogenesis leading to microvascu-
lar reconstruction within the scar is granulation tissue. Granulation tissues,
which develop at the ulcer base within 48–72 h after ulceration [14, 18, 34],
consist of connective tissue cells, macrophages, fibroblasts and proliferating
endothelial cells, forming microvessels through angiogenesis [14, 18, 34]. The
latter process gives a characteristic granular appearance to the surface of gran-
ulation tissue. Granulation tissue is an important component of the ulcer heal-
ing process because it supplies connective tissue cells and microvessels for the
restoration of the microvasculature within the ulcer scar [14, 18, 34].
Angiogenesis is a major component of wound healing and is essential for
healing chronic gastroduodenal ulcers. The growth of new microvessels
through angiogenesis is promoted by angiogenic growth factors such as bFGF,
VEGF, PDGF and angiopoietin, and possibly by other growth factors and
cytokines including IL-1 and TNF-␣ [34–36].

Experimental studies demonstrated that bFGF and PDGF accelerate heal-
ing of experimental duodenal ulcers in rats. Folkman et al. [37] and Szabo et al.
[38] demonstrated that a stable form of bFGF, administered orally to rats, stim-
ulated angiogenesis in granulation tissue and significantly accelerated duode-
nal ulcer healing. bFGF is a direct mitogen for vascular endothelial cells,
fibroblasts and smooth muscle cells. A recent study has shown that VEGF also
accelerates healing of experimental duodenal ulcers [39].
While bFGF promotes proliferation and migration of a variety of epithe-
lial and mesenchymal cells, VEGF is an endothelial cell-specific mitogen
because its receptors are almost exclusively localized to endothelial cells [40].
VEGF (purified, cloned and amino acid-sequenced by Ferrara [40]) is a funda-
mental regulator of vasculogenesis and angiogenesis. The loss of a single VEGF
allele results in defective vascularization and early embryonic death [40].
VEGF stimulates physiological angiogenesis and the pathological angiogenesis
involved in proliferative diabetic retinopathy, tumor growth and metastasis [40].
Human VEGF protein is produced as four isoforms from alternative splicing of
a single VEGF transcript [40]. VEGF165 is the predominant molecular isoform
produced by a variety of cells and this isoform is strongly overexpressed in
wounded rat gastric mucosa [17].
VEGF binds to two specific receptors: VEGF-R1 or flt-1 (fms-like tyro-
sine kinase) and VEGF-R2 or flk-1/KDR (fetal liver kinase-1/kinase domain
region), expressed almost exclusively on endothelial cells, and thereby initiates
phosphorylation of numerous cytosolic proteins involved in signal transduction
leading to endothelial cell proliferation [40].
The activation of the VEGF-R2 tyrosine kinase in response to VEGF is
necessary for VEGF-induced endothelial cell proliferation [40]. It has also
recently been demonstrated that MAPK (Erk1/Erk2) activation and JNK kinase
cross-activation are necessary for induction of endothelial cell proliferation
by VEGF [41]. VEGF production is stimulated by PDGF, TGF-␤1, bFGF,
cytokines, NO and E-series prostaglandins [40]. Hypoxia is one of the best
characterized stimuli for the induction of VEGF expression, in a variety of cells
and tissues, acting via a HIF-1-binding site of the VEGF promoter [40].
Our recent studies demonstrated that activation of the MAP (Erk-1, -2)
kinase signal transduction pathway is crucial for VEGF-induced stimulation of
angiogenesis in vitro and that NSAIDs interfere with angiogenesis, in part, by
inhibiting the MAPK/Erk pathway [24]. VEGF-induced angiogenesis accelerates
healing of acute gastric injury, duodenal ulcers and ischemic wounds and provides
therapeutic benefits in animal models of myocardial or limb ischemia [40, 42].
Recently, several new angiogenic factors have been identified (in addition
to aFGF, bFGF and VEGF), namely, the angiopoietins and their receptor Tie2,
neuropilin, ephrin/Eph, leptin and CXCR-4. Angiopoietin-1 and -2 (Ang1 and
Ang2) appear to be involved in angiogenic processes occurring subsequent to
the actions of VEGF [35, 43–47]. Ang1 and Ang2 are secreted proteins sharing
approximately 60% amino acid homology. A ‘fibrinogen-like domain’ repre-
sents the receptor-binding portion of angiopoietins [44]. Both Ang1 and Ang2
bind to the endothelial specific receptor Tie2 (tyrosine kinase-containing
immunoglobulin-like loop and epidermal growth factor-like domain) with sim-
ilar affinity, but only Ang1 induces autophosphorylation of Tie2 [45–47]. Ang2
binding to Tie2 competitively inhibits Ang1-induced receptor phosphorylation
and kinase activity; therefore, Ang2 serves as a natural inhibitor of Tie2 activa-
tion [46]. The importance of Ang1, Ang2 and their common receptor, Tie2, in
vasculogenesis and angiogenesis is best illustrated by gene knockout studies in
mice showing that loss of Ang1 or Tie2 gene expression, or Ang2 overexpres-
sion, results in embryonic lethality due to impaired vasculogenesis/angiogene-
sis [35, 36, 45, 46]. Hypoxia and VEGF have been shown to upregulate
expression of Ang2 in cultured endothelial cells [47]. Local expression of Ang2
blocks Ang1/Tie2 signaling and causes loosening of tight vascular structures
resulting in exposure of endothelial cells to other endothelial growth factors,
notably VEGF, if present [35, 46]. Subsequently, VEGF triggers endothelial cell
migration, proliferation, sprouting and tube formation. Our preliminary studies
demonstrated that gastric injury erosions or ulcers activate expression of Ang1,
Ang2 and Tie2 genes in the areas of active angiogenesis [48, 49].
Signaling Pathways of VEGF and Ang1
The extracellular regulated kinases (Erks), members of a larger group of

mitogen-activated protein kinases (or MAP kinases), are key intermediaries in
several signaling pathways that activate transcription factors and early response
genes ultimately leading to cell proliferation and/or differentiation. The
paradigmatic pathway for this involvement is the Ras-MAP kinase pathway.
Activation of this pathway is initiated by the binding of a mitogen (e.g., growth
factor) to its specific cellular receptor. This results in transduction of the signal
via sequential activation of the serine/threonine kinases, Raf, MAP kinase
(MEK) and Erk. The active Erk isozymes, Erk1 and Erk2, are then translocated
to the nucleus where they participate in the activation of transcription factors.
Our recent studies demonstrated that NSAIDs interfere with Erk activation and
their nuclear translocation in endothelial cells [24].
Phosphatidylinositol 3⬘-kinase (PI-3 K) is another important kinase that
has been shown to be activated by most mitogens and has been implicated as
a critical factor in the regulation of cell proliferation, survival and motility
[34, 50, 51]. This kinase is a heterodimer composed of an 85-kD (p85)

adaptor/activator subunit and a 110-kD (p110) catalytic subunit. PI-3 K phos-
phorylates the D-3 position of the inositol ring of phosphoinositides, which in
turn act as second messengers in the activation of other kinases, notably PDK1
(3-phosphoinositide-dependent protein kinase 1) [reviewed in 34, 50, 51].
Inactive p110 is associated with p85 by a region between two Src homology
2 (SH2) domains on p85. The p85/p110 heterodimer binds to tyrosine phos-
phorylated receptors via the two SH2 domains and, in this manner, is recruited
to the inner cell membrane. Phosphorylation of the p85 subunit by the receptor
kinase then leads to activation of the p110 catalytic subunit [reviewed in 34, 50,
51]. VEGF has been shown to activate PI-3 K, suggesting an important role for
this kinase in VEGF-mediated signaling. Recent evidence also indicates that
Ang1 activates PI-3 K and its downstream effector, Akt [52].
Both VEGF and Ang1 are endothelial cell survival factors and both prevent
endothelial cell apoptosis [52, 54]. The anti-apoptotic effect of VEGF requires
activation of Flk-1/KDR [54] while that of Ang1 requires activation of the
Tie2 receptor [52]. In both cases, inhibition of apoptosis is mediated via the PI
3-K/Akt signaling pathway [52, 54]. In the case of VEGF, inhibition of apopto-
sis has been shown to be mediated by PI 3-K/Akt-dependent expression of
the FLICE-inhibitory protein, FLIP, leading to inhibition of Fas-mediated
apoptosis [55].
Prostaglandins (PGs) stimulate angiogenesis but the precise mechanisms
of their pro-angiogenic action remain unexplained. We investigated whether
prostaglandin E2 (PGE2) can induce VEGF expression in rat gastric microvascu-
lar endothelial cells (RGMEC) and the signaling pathway(s) involved [53]. We
demonstrated that PGE2 significantly increased ErK2 and JNK1 activation and
VEGF mRNA and protein expression. Incubation of RGMEC with PD 98059
(MEK kinase inhibitor) significantly reduced PGE2-induced Erk2 activity,
VEGF mRNA and protein expression [53]. Furthermore, PD 98059 treatment
almost completely abolished JNK1 activation. Our data suggest that PGE2-
stimulates VEGF expression in RGMEC via transactivation of JNK1 by Erk2.
NSAIDs Inhibit Angiogenesis in Ethanol-Injured Gastric Mucosa,

Ulcer Granulation Tissue and Endothelial Cells in vitro
We have demonstrated that NSAIDs inhibit angiogenesis in vivo in gastric

mucosa wounded by ethanol [27]. Quantitative assessment of angiogenesis in
the mucosa-bordering alcohol-induced gastric erosions demonstrated that, in the
placebo-treated group, 9% and 8 ⫾ 5% of microvessels at 24 and 48 h, respec-
tively, are undergoing angiogenesis (reflected by sprouting of microvascular
tubes) vs. only 5 ⫾ 1% in indomethacin-treated rats [27]. Experimental studies
demonstrated that NSAIDs delay experimental gastric ulcer healing, in part, by
significantly inhibiting (⬎45%) angiogenesis in granulation tissue [56, 57].
Our recent study [24] demonstrated that: (1) NSAIDs, both Cox2-selective
(NS-398) and nonselective (indomethacin), inhibit angiogenesis in a variety of
endothelial cells in vitro (rat aortic, human umbilical, human microvascular)
through a direct action on endothelial cells; (2) inhibition of angiogenesis by
NSAIDs is associated with and strongly correlated (r ⫽ 0.947) with the inhibi-
tion of MAP (Erk2) kinase activity, and Erk2 translocation to the nucleus, and
is independent of PKC. This study also demonstrated that PGE2, alone or in
combination with PGI2, partly reversed inhibition of angiogenesis caused by
NS-398 but not that caused by indomethacin. This study clearly indicates that
NSAIDs can inhibit important enzymes (e.g., Erk kinase) of signal transduction
pathways, in addition to Cox-1 and Cox-2 enzymes. Since endothelial cell
structure, function and properties markedly differ between organs and tissues,
it is not certain whether NSAIDs will inhibit angiogenesis in gastric mucosal
microvascular endothelial cells and whether this action will be mediated
through MAPK, or perhaps by other signaling pathway(s).
A subsequent study in vitro demonstrated that NSAIDs (indomethacin)
inhibit endothelial cell proliferation by suppressing cell cycle proteins and PRB
phosphorylation [58]. Since endothelial cell proliferation is essential for angio-
genesis, this study provides a new molecular mechanism for the antiangiogenic
action of NSAIDs.
Angiogenesis is also important for growth of some gastrointestinal tumors,
e.g., colon cancer, because the rapidly growing cell population requires nutrient
and oxygen delivery. This important topic, however, is beyond the scope of this
chapter.
Acknowledgements
This work was supported in part by a Merit Review Award and a Research
Enhancement Program Award to A.S. Tarnawski from the Medical Research Service of the
Department of Veterans Affairs.
References
1 Guth PH, Leung FW: Physiology of the gastric circulation; in Johnson L (ed): Physiology of the
Gastrointestinal Tract. New York, Raven Press, 1987, vol 7, pp 1031–1054.
2 Crissinger KD, Granger N: Gastrointestinal blood flow; in Yamada T (ed): Textbook of
Gastroenterology, ed 2. Philadelphia, Lippincott, 1995, pp 518–545.
3 Ichikawa Y, Tarnawski A, Sarfeh IJ, Ischikawa T, Shimada H: Distorted microangioarchitecture
and impaired angiogenesis in gastric mucosa of portal hypertensive rat. Gastroenterology 1994;
106:706–708.

4 Tarnawski A: Cellular and molecular mechanisms of mucosal defense and repair; in Bioregulation
and Its Disorders in Gastrointestinal Tract. Tokyo, Blackwell Science, 1998, pp 3–17.
5 Tarnawski A, Stachura J, Gergely H, et al: Microvascular endothelium: a major target for alcohol
injury of the human gastric mucosa: Histochemical and ultrastructural study. J Clin Gastroenterol
1988;10(suppl 1):S53.
6 Wallace JL, Granger DN: The cellular and molecular basis of gastric mucosal defense. FASEB
J 1996;10:731–740.
7 Holzer P: Sensory nerves and neuropeptides in gastroenterology. From basic science to clinical
perspectives. Adv Exp Med Biol 1991;298:3–17.
8 Tarnawski A, Sarfeh IJ, Lu SY, Santos AM: Quality of ulcer healing: Evidence for impaired
restoration of sensory CGRP nerves in the scar of experimental gastric ulcer. Eur J Gastroenterol
Hepatol 1993;5:S81–S85.
9 Peskar BM, Plate S, Stroff T: Gastroprotective effects of gut peptides: Role of afferent neurons,
calcitonin gene-related peptide and nitric oxide; in Cheli R, Iaquinto G, Szabo S (eds):
Gastroduodenal Mucosal Damage: Problems of Protection and Healing. Milano, Medserve, 1997,
pp 70–83.
10 Li DS, Raybould HE, Quintero E, Guth PH: Calcitonin gene-related peptide mediates the gastric
hyperemic response to acid back-diffusion. Gastroenterology 1992;102:1124–1128.
11 Palmer RMJ, Ashton DS, Moncada S: Vascular endothelial cells synthesize nitric oxide from
L-arginine. Nature 1988;333:664–666.
12 Whittle BJR: Neuronal and endothelium-derived mediators in the modulation of the gastric micro-
circulation: Integrity in the balance. Br J Pharmacol 1993;110:3–17.
13 Masuda E, Kawano S, Nagano K, Tsuji S, Takei Y, Tsujii M, Oshita M, Michida T, Kobayashi I,
Nakama A, Fusamoto H, Kamada T: Endogenous nitric oxide modulates ethanol-induced gastric
mucosal injury in rats. Gastroenterology 1995;108:58–64.
14 Tarnawski A, Hollander D, Stachura J, Krause WJ, Eltorei M, Dabros W, Gergely H: Vascular and
microvascular changes – Key factors in the development of acetic acid-induced gastric ulcers in
rats. J Clin Gastroenterol 1999;2(suppl 1):S148–S157.
15 Szabo S, Trie JS, Brown A, Schnoor J: Early vascular injury and increased vascular permeability
in gastric mucosal injury caused by ethanol in the rat. Gastroenterology 1985;88:228–236.
16 Tarnawski A, Hollander D, Stachura J: Ultrastructural changes in the gastric mucosal microves-
sels after ethanol. Gastroenterol Clin Biol 1985;9:93–97.
17 Jones MK, Itani RM, Wang H, Tomikawa M, Sarfeh IJ, Szabo S, Tarnawski AS: Activation of
VEGF and Ras genes in gastric mucosa during angiogenic response to ethanol injury. Am J
Physiol 1999;276:G1345–G1355.
18 Tarnawski A, Hollander D, Stachura J, Gergely H, Krause WJ, Sarfeh IJ: Role of angiogenesis in
healing of experimental gastric ulcer; in Halter F, Garner A, Tytgat GNJ (eds): Mechanisms of
Peptic Ulcer Healing. Dordrecht, Kluwer Academic, 1991, pp 165–171.
19 Lacy ER, Ito S: Microscopic analysis of ethanol damage to rat gastric mucosa after treatment with
prostaglandin. Gastroenterology 1982;83:619–625.
20 Tarnawski A, Hollander D, Stachura J, Krause WJ, Gergely H: Prostaglandin protection of the gas-
tric mucosa against alcohol injury – A dynamic time-related process: the role of mucosal prolif-
erative zone. Gastroenterology 1985;88:334–352.
21 Tarnawski A, Santos AM, Ichikawa Y, Lu SY, Sarfeh IJ: Antacid talcid stimulates angiogenesis in
injured gastric mucosa. A new mechanism for its mucosal healing action. Eur J Gastroenterol
Hepatol 1993;5:S125–S132.
22 Folkman J, Shin Y: Angiogenesis. J Biol Chem 1992;267:10931–10934.
23 Montesano R, Pepper M, Orci L: Angiogenesis in vitro: Morphogenetic and invasive properties of
endothelial cells. News Int Physiol Sci 1990;75–79.
24 Jones MK, Wang H, Levin E, Itani RM, Sarfeh IJ, Tarnawski AS: Inhibition of angiogenesis by
NSAIDs. Insight into the mechanisms and implications for cancer growth and ulcer healing. Nat
Med 1999;5:1418–1423.
25 Tarnawski A, Stachura J, Sarfeh IJ, Sekhon S, Krause WJ, Gergely H: Prostacyclin, endothelial
cell growth factor and antacid stimulate angiogenesis in injured gastric mucosa. Gastroenterology
1991;100:A174.
26 Tarnawski A, Stachura J, Sarfeh IJ, Sekhon S, Gergely H: Angiogenesis in gastric mucosa injured
by ethanol. Effect of epidermal growth factor, b-fibroblast growth factor and sucralfate (abstract).
27 Tarnawski A, Hollander D, Stachura J, Sarfeh IJ, Gergely H, Krause WJ: Angiogenic response of
gastric mucosa to ethanol injury is abolished by indomethacin (abstract). Gastroenterology
1989;96:A505.
28 Tarnawski AS, Florkiewicz RZ, Santos A, Irwin FL, Krause WJ, Sarfeh IJ: Basic fibroblast growth
factor and angiogenesis in gastric mucosa injured by ethanol (abstract). Gastroenterology 1994;
106:A194.
29 Wenger RH, Gassmann M: Oxygen(es) and the hypoxia-inducible factor-1. Biol Chem 1997;
376:609–616.
30 Salceda S, Caro J: Hypoxia-inducible factor 1␣ (HIF-1␣) protein is rapidly degraded by the ubiq-
uitin-proteasome system under normoxic conditions. Its stabilization by hypoxia depends on
redox-induced changes. J Biol Chem 1997;272:22642–22647.
31 Melillo G, Musso T, Sica A, Taylor LS, Cox GW, Varesio L: A hypoxia-responsive element medi-
ates a novel pathway of activation of the inducible nitric oxide synthase promoter. J Exp Med
1995;182:1683–1693.
32 Sima DT, Adamis AP, Ferrara N, Yeo KT, Yeo TK, Allende R, Folkman J, D’Amore PA: Hypoxic
induction of endothelial cell growth factors in retinal cells: Identification and characterization of
vascular endothelial growth factor as the mitogen. Mol Med 1995;1:182–193.
33 Szabo IL, Kawanaka H, Jones MK, Pai R, Soreghan B, Baatar D, Tarnawski A: Activation of
hypoxia inducible factor-1␣ in gastric mucosa in response to ethanol injury. A trigger for angio-
genesis? Life Sci 2001 (in press).
34 Tarnawski A: Molecular mechanisms of ulcer healing. Drug News Perspect 2000;13:158–168.
35 Risau W: Mechanisms of angiogenesis. Nature 1997;386:671–673.
36 Folkman J, D’Amore PA: Blood vessels formation: What is its molecular basis? Cell 1996;
87:1153–1155.
37 Folkman J, Szabo S, Stovroff M, McNeil P, Li W, Shing Y: Duodenal ulcer. Discovery of a new
mechanism and development of angiogenic therapy that accelerates healing. Ann Surg 1991;
214:414–427.
38 Szabo S, Folkman J, Vattay P, et al: Accelerated healing of duodenal ulcers by oral adminis-
tration of a mutein of basic fibroblast growth factors in rats. Gastroenterology 1994;106:
1106–1111.
39 Szabo S, Folkman J, Vincze A, Sandor ZS, Gombos A: Modulation of vascular factors by
VEGF/VPF is sufficient for chronic ulcer healing and acute gastroprotection (abstract).
40 Ferrara N: Role of vascular endothelial growth factor in the regulation of angiogenesis. Kidney Int
1999;56:794–814.
41 Pedram A, Razandi M, Levin ER: Extracellular signal-regulated protein kinase/Jun kinase cross-
talk underlies vascular endothelial cell growth factor-induced endothelial cell proliferation. J Biol
Chem 1998;273:26722–26728.
42 Ferrara N, Alitalo K: Clinical applications of angiogenic growth factors and their inhibitors. Nat
Med 1999;5:1359–1364.
43 Asahara T, Chen D, Takashashi T, et al: Tie2 receptor ligands angiopoietin-1 and angiopoietin-2
modulate VEGF-induced postnatal neovascularization. Circ Res 1998;83:233–240.
44 Procopio WN, Pelavin PI, Lee WMF, Yeilding NM: Angioproietin-1 and -2 coiled coil domains
mediate distinct oligomerization patterns, but fibrinogen-like domains mediate ligand activity.
J Biol Chem 1999;274:30196–30201.
45 Suri C, Jones PF, Patan S, et al: Requisite role of angiopoietin-1, a ligand for the TIE2 receptor,
during embryonic angiogenesis. Cell 1996;87:1171–1180.
46 Maisonpierre PC, Suri C, Jones PF, et al: Angiopoietin-2, a natural antagonist for Tie2 that dis-
rupts in vivo angiogenesis. Science 1997;277:55–60.
47 Oh H, Takagi H, Suzuma K, Otani A, et al: Hypoxia and vascular endothelial growth factor selec-
tively up-regulate angiopoietin-2 in bovine microvascular endothelial cells. J Biol Chem 1997;
274:15723–15739.

48 Tarnawski A, Pai R, Jones MK, Szabo I, Kawanaka H, Sarfeh IJ: Gastric injury activates angiopoi-
etin, Tie2 and VEGF genes in mucosa-bordering necrosis. A key to angiogenesis? Gastroenterology
2000;118:A659.
49 Tarnawski A, Pai R, Jones MK, Szabo I, Sarfeh IJ: Gastric ulceration triggers activation of
angiopoietin-1, -2 and Tie2 important for angiogenesis. Gastroenterology 2000;118:A242.
50 Suzuma K, Naruse K, Suzuma I, Takahara N, Ueki K, Aiello LP, King GL: Vascular endothelial
growth factor induces expression of connective tissue growth factor via KDR, Flt1, and phos-
phatidylinositol 3-kinase-akt-dependent pathways in retinal vascular cells. J Biol Chem 2000;
275:49725–49731.
51 Jiang BH, Zheng JZ, Aoki M, Vogt PK: Phosphatidylinositol 3-kinase signaling mediates angio-
genesis and expression of vascular endothelial growth factor in endothelial cells. Proc Natl Acad
Sci USA 2000;97:1749–1753.
52 Kim I, Kim HG, So JN, Kim JH, Kwak HJ, Koh GY: Angiopoietin-1 regulates endothelial cell
survival through the phosphatidylinositol 3⬘-kinase/Akt signal transduction pathway. Circ Res
2000;86:24–29.
53 Pai R, Szabo IL, Soreghan BA, Atay S, Kawanaka H, Tarnawski AS: PGE2 stimulates VEGF
expression in endothelial cells via ERK2/JNK1 signaling pathways. BBRC 2001;286:923–928.
54 Gerber HP, McMurtrey A, Kowalski J, Yan M, Keyt BA, Dixit V, Ferrara N: Vascular endothelial
growth factor regulates endothelial cell survival through the phosphatidylinositol 3⬘-kinase/Akt
signal transduction pathway. Requirement for Flk-1/KDR activation. J Biol Chem 1998;273:
30336–30343.
55 Suhara T, Mano T, Oliveira BE, Walsh K: Phosphatidylinositol 3-kinase/Akt signaling controls
endothelial sensitivity to Fas-mediated apoptosis via regulation of FLICE-inhibitory protein
(FLIP). Circ Res 2001;89:13–19.
56 Tarnawski A, Stachura J, Douglass TG, Krause WJ, Gergely H, Sarfeh IJ: Indomethacin impairs
quality of experimental gastric ulcer healing: A quality histologic and ultrastructural analysis; in
Garner A, O’Brien PR (eds): Mechanisms of Injury, Protection and Repair of the Upper
Gastrointestinal Tract. Chichester, Wiley, 1991, pp 521–531.
57 Halter F, Tarnawski AS, Schmassmann A, Peskar BM: Cyclooxygenase 2 implications on mainte-
nance of gastric mucosal integrity ulcer healing: Controversial issues and perspectives. Gut 2001;
49:443–453.
58 Pai R, Szabo IL, Kawanaka H, Soregham BA, Jones MK, Tarnawski AS: Indomethacin inhibits
endothelial cell proliferation by suppressing cell cycle proteins and PRB phosphorylation. A key
to its anti-angiogenic action? Mol Cell Biol Res Commun 2001;4:111–116.
A. Tarnawski, MD, DSc, Gastroenterology (111G),

5901 E. Seventh Street, Long Beach, CA 90822 (USA)
Tel. ⫹1 562 494 5494, Fax ⫹1 562 961 8016, E-Mail atarnawski@yahoo.com
Role of Platelets in Gastric Ulcer

Healing: A Delivery System for
Growth Factors
Li Ma, John L. Wallace
Mucosal Inflammation Research Group, University of Calgary, Alberta, Canada
Gastric ulcers are deep, necrotic lesions that involve the full thickness of
the mucosa and penetrate through the muscularis mucosae. Ulcer healing is an
active and complicated process of filling the mucosal defect with proliferating
and migrating epithelial cells and connective components, so as to reconstruct
the mucosal architecture. This requires the concerted interaction of a variety
of tissues and cellular systems, including soluble mediators, formed blood
elements, extracellular matrix (ECM) and parenchymal cells.
Tissue damage leads to blood vessel disruption accompanied by extrava-
sation of blood constituents. Tissue repair is initiated with the aggregation of
platelets, formation of a fibrin clot, and the release of growth factors from the
activated coagulation pathways, injured cells, platelets and ECM, followed by
migration of inflammatory cells to the wound site. Thereafter, epithelial cells
migrate over the damage, angiogenesis is initiated, and fibroblasts deposit and
remodel the granulation tissue. Those processes are regulated by a complex
network of highly divergent factors, among them a broad spectrum of struc-
turally distinct regulatory peptides that have been identified within the gastric
mucosa and platelets. Epidermal growth factor (EGF), hepatocyte growth
factor (HGF), transforming growth factor (TGF)-␣, and insulin-like growth
factor (IGF) are mainly involved in the reconstitution of epithelial structure.
Platelet-derived growth factor (PDGF), basic fibroblast growth factor (bFGF),
vascular endothelial cell growth factor (VEGF) and TGF-␤ play major roles in
the reconstitution of connective tissue, including vessels and smooth muscle
cells, and the formation of ECM substrate for cell migration and differentiation.
The expression of these growth factors and their receptors is increased during
ulcer healing. Inhibition of their effects by neutralization with antibodies may
result in delayed ulcer healing, while administration of recombinant or natural
growth factors may improve ulcer repair. In addition, platelets release growth
factors during clotting at the wound. In this chapter, we review the regulatory
role of platelets in gastric ulcer healing and the underlying mechanisms.
Blood platelets are non-nucleated fragments of bone marrow-derived
megakaryocytes that circulate in the peripheral blood for 9–11 days. Platelets
possess at least three types of granules, termed ␣, dense and lysosomal.
Mitogenic activity has been localized to the platelet ␣-granules. Since platelets
contain only a rudimentary protein synthetic apparatus, they appear to be a
storage vehicle for active mitogens. Consequently, additional release of platelet
mitogens requires continuous deposition of new platelets. Platelets have been
shown to express a range of receptors on their surface, which can be stimulated
by a variety of circulating agonists (e.g. ADP, thrombin, collagen and arachidonic
acid) causing them to undergo a shape change, degranulation and aggregation.
Platelets are capable of releasing a variety of biologically active molecules upon
stimulation, and play important roles under physical and pathological conditions,
such as wound healing, atherosclerosis, myelofibrosis, connective tissue diseases
and neoplastic disorders. In the following sections, we provide an overview of
some platelet-associated growth factors and their roles in gastric ulcer healing.
VEGF
VEGF is a 34- to 42-kD dimeric peptide with remarkable specificity for

endothelial cells [10]. VEGF was first described as a potent endothelial perme-
ability factor and later was identified as an endothelial mitogen. VEGF is
virtually specific for angiogenesis, therefore playing a significant role in gastric
ulcer healing by delivering oxygen, nutrients and growth factors to the regen-
erated tissue. During initial clinical studies in which blood VEGF levels were
measured in normal volunteers, it was found that the serum VEGF levels were
consistently higher than those in the corresponding plasma. Therefore, it was
hypothesized that the difference of serum and plasma VEGF concentration was
due to the release of VEGF from platelets during activation of clotting and,
more specifically, that VEGF is secreted during platelet aggregation. An in vitro
study has confirmed that VEGF is released during platelet aggregation induced
by thrombin, collagen, epinephrine or adenosine diphosphate (ADP). It is a
general response of platelet aggregation, but not specific to a particular agonist
and does not occur in response to the agonist if aggregation is blocked [17].
Furthermore, it was found that the release of VEGF was accompanied by the
release of ␤-thromboglobulin, a marker of platelet ␣-granules [29], and it was
blocked by a degranulation inhibitor [12], suggesting that VEGF is stored in
Ma/Wallace 118
platelet ␣-granules. The presence of VEGF mRNA and protein in megakaryo-
cytes provides strong evidence that VEGF synthesis during thrombopoiesis is
the origin of platelet VEGF [19]. Therefore, platelets are also called a trans-
porter of VEGF and circulating levels of VEGF are dependent on the platelet
content in blood [13].
EGF
EGF is a 53-amino-acid peptide. It was first discovered as a factor from

mouse salivary glands, able to induce the premature eruption of teeth and
opening of eyelids in mice. Later it was found that EGF not only stimulates
epidermal growth, differentiation and repair, but is also a powerful mitogen for
fibroblasts. EGF is secreted from salivary glands, duodenum and pancreas.
EGF secreted into the lumen is subsequently carboxyterminally processed into
smaller, less active forms in the stomach and small intestine. Therefore, the
local production of EGF from an ulcer-associated cell lineage in the gastric
mucosa [32] or from aggregated platelets at the site of injury is critical for ulcer
healing. EGF accelerates ulcer healing by stimulating epithelial cell prolifera-
tion and migration.
EGF is another growth factor known to be present in the ␣-granules of
platelets. It was reported early in 1983 that EGF could be purified from
platelets [22]. The majority of EGF in platelets exist either in a covalently
bound form with some protein(s) or as a dimer intermolecularly cross-linked by
a S-S linkage. The latter could be transformed into a monomer through several
manipulations, including freeze-thawing or proteolytic digestion [24]. EGF is
released during platelet aggregation in response to different agonists with a
good correlation between rate of release of EGF and the extent of platelet
aggregation [20]. Furthermore, it has been demonstrated that the source of EGF
in platelets is the megakaryocyte and that this EGF is synthesized in the
megakaryocyte rather than being taken up from its environment [3]. The
platelet is the main source of serum EGF. Serum EGF levels correlated very
well with blood platelet counts. One platelet contains approximately 2.5 ⫻
10⫺18 g EGF. Serum EGF levels are reduced by 90% when platelet counts are
decreased through bone marrow ablation [14].
TGF-␤
TGF-␤1, a 25-kD dimeric peptide, is a platelet-derived cytokine involved

in both normal wound healing and scarring. In general, the release and activation
Role of Platelets in Gastric Ulcer Healing 119

of TGF-␤1 contribute to the reconstitution of connective tissue, including
vessels and smooth muscle cells, and the formation of ECM substrate for cell
migration and differentiation. Local application of TGF-␤1 leads to significant
acceleration of gastric ulcer healing by increasing the formation of granulation
tissue and cell migration [7].
TGF-␤1 has been purified from platelets, which contain ~100-fold more
TGF-␤1 than that of other non-neoplastic tissues [1]. TGF-␤1 is released as a
consequence of degranulation during platelet aggregation in a parallel manner
with the release of the ␣-granule marker ␤-thromboglobulin [2], suggesting that
TGF-␤1 is stored in platelet ␣-granules.
PDGF
As reviewed by Ross et al. [26], the discovery of PDGF came in 1974 when
it was observed that material released from platelets was the principal source of
mitogens present in whole blood serum, and was responsible for the growth of
many cells in culture that are serum-dependent. PDGF from human platelets is a
cationic glycoprotein of approximately 30 kD. It is stored in platelets and released
upon stimulation. There are approximately 0.06 ng/106 human platelets (~1,200
molecules/platelet). PDGF accounts for approximately 50% of the platelet-
derived mitogenic activity [26]. PDGF stimulates the proliferation of a variety of
cells in the tissue culture, including arterial smooth muscle cells, fibroblasts and
glial cells. PDGF together with bFGF and TGF-␤ play a major role in the recon-
stitution of connective tissue, including vessels and smooth muscle cells, and
provide the ECM substrate for cell migration and differentiation, therefore accel-
erating ulcer healing. The role of PDGF in wound repair has centered on the fact
that platelets, monocyte/macrophages, and possibly injured endothelial cells can
secrete PDGF together with other growth factors. PDGF could be important in
the initiation of the repair process because of its chemotactic properties for both
leukocytes and fibroblasts, while PDGF from macrophages could play a major
role in the continuing process of fibrogenesis [26].
Other Growth Factors
HGF was purified as a homogeneous material from rat platelets and was
found to stimulate DNA synthesis of adult rat hepatocytes in primary culture.
Platelet-derived HGF has a molecular mass of 82 kD [21]. HGF stimulates
epithelial cell proliferation and migration in vitro [30] and accelerates ulcer
healing in vivo [27].
Ma/Wallace 120
IGF was also found to be released from platelets stimulated with thrombin.
Disruption of platelets by nitrogen cavitation followed by separation of the
organelles by sucrose density gradient sedimentation showed that IGF and
mitogenic activity localized predominantly to fractions containing ␣-granules
rather than soluble cellular components, lysosomes, or dense granules [11]. IGF
has been shown to accelerate both epithelial and fibroblast wound healing, and
involves both cell proliferation and migration [31]. This suggests that IGF could
play a key role in gastric epithelial-mesenchymal interaction during the process
of gastric ulcer healing.
Platelet-derived endothelial cell growth factor (PD-ECGF, also known as
thymidine phosphorylase, TP) is a 45-kD single-chain polypeptide that has been
purified from human platelets. Most platelet-associated growth factors have
been reported to reside in ␣-granules, but PD-ECGF appears to be present in the
platelet cytoplasm [18]. PD-ECGF has been demonstrated to stimulate endothe-
lial cell growth and chemotaxis in vitro and promote angiogenesis in vivo [9],
suggesting its possible role in wound healing during clotting at the injury.
Other Platelet-Derived Substances Related to Angiogenesis
In addition to the growth factors released during the activation of clotting

cascades, platelet-derived lipid mediators are now also known to play a key
role in many aspects of the angiogenic response. The first indication of lipid
mediator involvement in angiogenesis was the discovery that lysophosphatidic
acid, phosphatidic acid and sphingosine 1-phosphate are high-affinity agonists
for G-coupled endothelial differentiation gene receptors. These lipids induce
many important endothelial cell responses associated with angiogenesis,
including liberation of endothelial cells from established monolayers, chemo-
tactic migration, proliferation, adherent junction assembly and morphogenesis
into capillary-like structures [6].
In addition to growth-promoting peptides (summarized in table 1),
platelets also contain some inhibitors of growth or angiogenesis, some of which
have been studied in the context of ulcer healing.
Endostatin
Endostatin, a 20-kD C-terminal fragment of collagen XVIII, is a potent,

endogenous angiogenesis inhibitor, which has been recently shown to be stored
in platelets [15]. Endostatin was first found in conditioned medium of mouse
hemangioendothelioma cells [23]. It has been demonstrated to be a potent

Table 1. Platelet-associated growth factors
Name Identified in platelets Roles in gastric ulcer healing

(reference)
VEGF 19 Stimulates endothelial cell

proliferation, migration and
angiogenesis
EGF 23 Stimulates epithelial cell
proliferation and migration
PDGF 26 Stimulates proliferation and
migration of smooth muscle cells
and fibroblasts, and synthesis of new
connective tissue
TGF-␤1 1 Inhibits proliferation of most cell
types, but induces the deposition of
extracellular matrix, and promotes
angiogenesis
HGF 20 Stimulates epithelial cell proliferation
IGF 11 Stimulates proliferation of epithelial
cells and fibroblasts, and increases
collagen deposition
PD-ECGF 18 Stimulates endothelial cell
proliferation and angiogenesis
endogenous specific angiogenic inhibitor. Endostatin inhibits endothelial cell

proliferation, migration and angiogenesis [23], and promotes endothelial cell
apoptosis [5]. Endostatin has been the subject of extensive studies because of its
ability to inhibit tumor growth by cutting of the blood supply. Release of endo-
statin from platelets differs from that of the growth factors discussed above, as
it does not occur as a consequence of degranulation of the platelets. We observed
that doses of thrombin and ADP that induced a similar degree of platelet aggre-
gation did not affect endostatin release to the same extent: thrombin caused
release while ADP did not [15]. Moreover, platelet endostatin release (at least in
rats) induced by thrombin was found to be mediated through protease-activated
receptor 4 (PAR-4) (fig. 1). Both thrombin and a PAR-4-activating peptide
(AYPGKF; AY-NH2) dose-dependently stimulated platelet aggregation [8] and
endostatin release. A selective PAR-4 antagonist (trans-cinnamoyl-YPGKF-
NH2; tcY-NH2) significantly inhibited platelet aggregation induced by thrombin
and AY-NH2, and also completely blocked the associated endostatin release
(fig. 1). Moreover, the release of endostatin, unlike the release of ␣-granule
Ma/Wallace 122
Saline
Thrombin (U ml⫺1)
AY-NH2 (␮M)
***
Endostatin release (ng ml⫺1)
20
***
15 **
###
10 ###
5
0 0 1 1 15 25 50 50 Agonist
⫺ ⫹ ⫺ ⫹ ⫺ ⫺ ⫺ ⫹ tcY-NH2
Fig. 1. Endostatin release from platelets in response to stimulation with thrombin or a

PAR-4 agonist (AY-NH2), and the effects of a selective PAR-4 antagonist (tcY-NH2; 400 ␮M).
Both thrombin and AY-NH2 induced significant release of endostatin, which was blocked by
prior exposure of the platelets to the PAR-4 antagonist. Values are expressed as means ⫾ SE
of 5 samples in each group. **p ⬍ 0.01, ***p ⬍ 0.001 vs. saline group; ###p ⬍ 0.001 vs.
corresponding thrombin or AY-NH2 group.
growth factors, occurs independently of platelet aggregation. ADP is involved in

thrombin-induced platelet aggregation by amplifying the signal. Apyrase, an
ADP scavenger, significantly inhibited thrombin-induced platelet aggregation,
but did not affect the associated endostatin release [16]. The separation of
platelet aggregation and endostatin release is consistent with the finding that
ticlopidine, an ADP receptor antagonist, significantly inhibited thrombin-induced
platelet aggregation, but dramatically increased platelet endostatin release [15].
The role of platelet-derived endostatin in gastric ulcer healing was demon-
strated by the fact that ticlopidine, which increases platelet endostatin content
and release, delays ulcer healing by depressing angiogenesis [15]. High levels of
endostatin were found in sera derived from ticlopidine-treated rats. Incubation of
those sera with human umbilical vein endothelial cells resulted in a significantly
decreased rate of proliferation, and an increased rate of apoptosis. These effects
were reversed by immunoneutralization of endostatin in the serum samples.
Thrombospondin
Thrombospondin (TSP) is a trimeric, 450-kD glycoprotein originally iden-

tified as a major component of platelet ␣-granules. Upon activation, TSP is

75
Ulcer area (mm2)

50
p⬍ 0.05
25
0
Normal rabbit Anti-platelet
serum serum
Fig. 2. Immunodepletion of circulating platelets significantly impairs gastric ulcer

healing. Rats were given anti-platelet (rabbit anti-rat) serum, or normal rabbit serum, intra-
venously on days 3 and 7 post-ulcer induction. On the 10th day, ulcer areas were measured.
Values are expressed as means ⫾ SE of 7 rats in each group. p ⬍ 0.05 vs. controls that were
treated intravenously with normal rabbit serum. The dotted line shows the mean ulcer area
on day 3 (i.e., prior to depletion of circulating platelets).
secreted from platelets and involved in the secondary secretion-dependent phase

of platelet aggregation. TSP is also synthesized by a variety of other cells and
incorporated into their ECM. TSP has been demonstrated to bind to integrins
of the ␣v␤3 and ␣3␤1 classes, and therefore, soluble secreted TPS inhibits
angiogenesis but matrix-bound TSP promotes microvessel formation [4].
Platelets and Ulcer Healing
Both angiogenesis and re-epithelialization are very important processes in

gastric ulcer healing. Although platelets contain both growth-promoting and
inhibiting substances, platelets appear to play an important role in the promo-
tion of wound healing. Angiogenesis (the growth of new blood vessels from the
pre-existing vessels) is a pivotal process of ulcer healing and regulated by pro-
and anti-angiogenic factors. While platelets carry both VEGF and endostatin,
immunodepletion of platelets resulted a significant inhibition of angiogenesis
and a delay of ulcer healing (fig. 2), suggesting that platelets primarily play a
beneficial role in ulcer healing. This is consistent with the finding that in rats
with gastric ulcers, the ratio of serum levels of the pro-angiogenic factor, VEGF
to the anti-angiogenic factor, endostatin, is increased. Thus, the pro-angiogenic
factors are predominant during tissue repair [15]. Platelets themselves have
also been shown to directly stimulate endothelial cell proliferation and capillary
tubular structure formation in vitro [25]. Furthermore, ␣-thrombin, a potent
activator of platelets, stimulates angiogenesis in the chick chorioallantoic
Ma/Wallace 124
membrane [28]. Of course, some of the growth factors contained within platelets,
like EGF, HGF and IGF, will also simulate proliferation and migration of other
cell types, including epithelial cells, which is important in re-establishment of
glandular architecture during ulcer healing.
Substances that affect the levels of pro- and anti-angiogenic factors in
platelets, or affect their release, might also affect gastric ulcer healing. We
found that ticlopidine, which inhibits platelet function via blocking the ADP
receptor, significantly delayed gastric ulcer healing by altering the platelet
content of pro- and anti-angiogenic factors. Ticlopidine inhibited platelet VEGF
release in parallel with the inhibition of platelet aggregation, but also dramati-
cally increased platelet endostatin content and release [15]. It was further found
that this effect occurred at the level of the bone marrow. This was demonstrated
through a complex series of experiments, which are summarized in figure 3.
Rats made thrombocytopenic through treatment with an anti-platelet antiserum
were treated with either vehicle or ticlopidine each day for 1 week. They were
then left for another week to allow for the release of new platelets from the bone
marrow. The platelets from these two groups of rats were then harvested and
were transfused into other groups of thrombocytopenic rats in which gastric
ulcers had been induced. It is important to note that in the ‘donor’ rats, the
exposure to ticlopidine occurred during a period when the rats had no circulat-
ing platelets. Thus, any effect of ticlopidine observed when platelets from those
rats were transfused into ‘recipient’ rats could be inferred to be due to effects
of ticlopidine on the bone marrow. What we observed was that infusion of
platelets derived from vehicle-treated rats into thrombocytopenic rats with
gastric ulcers resulted in a significant acceleration of ulcer healing relative to
thrombocytopenic rats that did not receive a transfusion. In contrast, transfusion
of platelets from rats that had been treated with ticlopidine resulted in a signif-
icant delay of gastric ulcer healing. Moreover, the platelet levels of endostatin
in the donor rats treated with ticlopidine were significantly elevated over con-
trol levels. Thus, ticlopidine produces effects at the level of the bone marrow
that results in significant changes in endostatin content of the platelets, and this
can have profound effects on healing.
Conclusion
It is becoming increasingly evident that the platelet performs a vast array

of functions in addition to their role in coagulation. Platelets represent one of
the largest sources of growth factors in the body, and it is through the release of
these growth factors, at least in part, that platelets are capable of profoundly
influencing healing. Of course, platelets are known to accumulate at sites

A B
Gastric Gastric
ulcer ulcer
Vehicle Ticlop
1 week 1 week
Anti-platelet
serum
Gastric Gastric
ulcer ulcer
Blood ⫹ ⫹
no no
platelets platelets
From A
From B
Washed
platelets
Fig. 3. Schematic diagram of the protocol for an experiment demonstrating that ticlo-
pidine exerts actions on the bone marrow, which result in platelet-dependent retardation of
gastric ulcer healing. The left side of the figure shows the treatment of ‘donor’ rats, while the
right side shows the treatment of ‘recipient’ rats. When washed platelets from rats treated
daily with ticlopidine for 1 week were transfused into thrombocytopenic rats with gastric
ulcers, the healing of those ulcers was significantly impaired relative to a control group
receiving platelets from vehicle-treated rats. Moreover, if the donor rats were thrombocy-
topenic (induced by treatment with anti-platelet serum) during the period of treatment with
ticlopidine, and then were left for 1 week prior to the harvesting of blood (to allow platelet
counts to recover), the transfusion still resulted in impairment of ulcer healing. Thus,
ticlopidine exerted actions on megakaryocytes in the bone marrow, which resulted in changes
in the platelets (i.e., increased endostatin content), which in turn resulted in impaired ulcer
healing in the recipient rats.
of injury, so this represents a very rational delivery system for these growth
factors to the sites at which they are required. We have found that platelets can
dramatically influence the rate of healing of gastric ulcers, and that certain
drugs (e.g., ticlopidine), by altering the content and release of growth factors in
platelets, can exert inhibitory effects on gastric ulcer healing. Of course, it is
Ma/Wallace 126
also possible that drugs may be identified that can influence platelets in the
opposite direction, so as to accelerate ulcer healing. A better understanding
of the mechanisms regulating growth factor content and release by platelets
will clearly be important if we are to exploit the ability of platelets to act as a
‘delivery system’ for growth factors.
References
1 Assoian RK, Komoriya A, Meyers CA, Miller DM, Sporn MB: Transforming growth factor-␤ in
human platelets. Identification of a major storage site, purification and characterization. J Biol
Chem 1983;258:7155–7160.
2 Assoian RK, Sporn MB: Type ␤ transforming growth factor in human platelets: Release during
platelet degranulation and action on vascular smooth muscle cells. J Cell Biol 1986;102:1217–1223.
3 Ben-Ezra J, Sheibani K, Hwang DL, Lev-Ran A: Megakaryocyte synthesis is the source of
epidermal growth factor in human platelets. Am J Pathol 1990;137:755–759.
4 Bornstein P: Thrombospondins as matricellular modulators of cell function. J Clin Invest 2001;
107:929–934.
5 Dhanabal M, Ramchandran R, Waterman MJ, Lu H, Knebelmann B, Segal M, Sukhatme VP:
Endostatin induces endothelial cell apoptosis. J Biol Chem 1999;274:11721–11726.
6 English D, Garcia JG, Brindley DN: Platelet-released phospholipids link haemostasis and
angiogenesis. Cardivasc Res 2001;46:588–599.
7 Ernst H, Konturek PC, Brzozowski T, Konturek SJ, Hahn EG: Subserosal application of
transforming growth factor-␤1 in rats chronic gastric ulcers: Effect on gastric ulcer healing and
blood flow. J Physiol Pharmacol 1996;47:443–454.
8 Hollenberg MD, Saifeddine M: Proteinase-activated receptor-4 (PAR-4): Activation and inhibition
of rat platelet aggregation by PAR-4-derived peptides. Can J Physiol Pharmacol 2001;79:
439–442.
9 Ishikawa F, Miyazono K, Hellman U, Drexler H, Wernstedt C, Hagiwara K, Usuki K, Takaku F,
Risau W, Heldin CH: Identification of angiogenic activity and the cloning and expression of
platelet-derived endothelial cell growth factor. Nature 1989;338:557–562.
10 Jakeman LB, Winer J, Bennett GL, Altar CA, Ferrara N: Binding sites for vascular endothelial
growth factor are localized on endothelial cells in adult rat tissues. J Clin Invest 1992;89:244–253.
11 Karey KP, Sirbasku DA: Human platelet-derived mitogens. II. Subcellular localization of insulin-
like growth factor I to the ␣-granule and release in response to thrombin. Blood 1989;74:
1093–1100.
12 Koehne P, Willam C, Strauss E, Schindler R, Eckardt KU, Buhrer C: Lack of hypoxic stimulation
of VEGF secretion from neutrophils and platelets. Am J Physiol Heart Circ Physiol 2000;279:
H817–H824.
13 Lee JK, Hong YJ, Han CJ, Hwang DY, Hong SI: Clinical usefulness of serum and plasma vascular
endothelial growth factor in cancer patients: Which is the optimal specimen? Int J Oncol 2000;
17:149–152.
14 Lev-Ran A, Hwang DL, Snyder DS: Human serum and plasma have different sources of epidermal
growth factor. Am J Physiol 1990;259:R545–R548.
15 Ma L, Elliott SN, Cirino G, Buret A, Ignarro LJ, Wallace JL: Platelets modulate gastric ulcer
healing: Role of endostatin and vascular endothelial growth factor release. Proc Natl Acad Sci
USA 2001;98:6470–6475.
16 Ma L, Hollenberg MD, Wallace JL: Thrombin-induced platelet endostatin release is blocked by a
proteinase-activated receptor-4 (PAR-4) antagonist. Br J Pharmacol 2001;134:701–704.
17 Maloney JP, Silliman CC, Ambruso DR, Wang J, Tuder RM, Voelkel NF: In vitro release of
vascular endothelial growth factor during platelet aggregation. Am J Physiol 1998;275:
H1054–H1061.

18 Miyazono K, Okabe T, Urabe A, Takaku F, Heldin CH: Purification and properties of an endothe-
lial cell growth factor from human platelets. J Biol Chem 1987;262:4098–4103.
19 Mohle R, Green D, Moore MA, Nachman RL, Rafii S: Constitutive production and thrombin-
induced release of vascular endothelial growth factor by human megakaryocytes and platelets.
Proc Natl Acad Sci USA 1997;94:663–668.
20 Nakamura T, Kasai K, Banba N, Ishikawa M, Shimoda S: Release of human epidermal growth
factor from platelets in accordance with aggregation in vitro. Endocrinol Jpn 1989;36:23–28.
21 Nakamura T, Nawa K, Ichihara A, Kaise N, Nishino T: Purification and subunit structure of
hepatocyte growth factor from rat platelets. FEBS Lett 1987;224:311–316.
22 Oka Y, Orth DN: Human plasma epidermal growth factor/␤-urogastrone is associated with blood
platelets. J Clin Invest 1983;72:249–259.
23 O’Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS, Flynn E, Birkhead JR, Olsen BR,
Folkman J: Endostatin: An endogenous inhibitor of angiogenesis and tumor growth. Cell 1997;
88:277–285.
24 Pesonen K, Viinikka L, Myllyla G, Kiuru J, Perheentupa J: Characterization of material with
epidermal growth factor immunoreactivity in human serum and platelets. J Clin Endocrinol Metab
1989;68:486–491.
25 Pipili-Synetos E, Papadimitriou E, Maragoudakis ME: Evidence that platelets promote tube
formation by endothelial cells on matrigel. Br J Pharmacol 1998;125:1252–1257.
26 Ross R, Raines EW, Bowen-Pope DF: The biology of platelet-derived growth factor. Cell 1986;
46:155–169.
27 Schmassmann A, Stettler C, Poulsom R, Tarasova N, Hirschi C, Flogerzi B, Matsumoto K,
Nakamura T, Halter F: Roles of hepatocyte growth factor and its receptor Met during gastric ulcer
healing in rats. Gastroenterology 1997;113:1858–1872.
28 Tsopanoglou NE, Pipili-Synetos E, Maragoudakis ME: Thrombin promotes angiogenesis by a
mechanism independent of fibrin formation. Am J Physiol 1993;264:C1302–C1307.
29 Wartiovaara U, Salven P, Mikkola H, Lassila R, Kaukonen J, Joukov V, Orpana A, Ristimaki A,
Heikinheimo M, Joensuu H, Alitalo K, Palotie A: Peripheral blood platelets express VEGF-C and
VEGF which are released during platelet activation. Thromb Haemost 1998;80:171–175.
30 Watanabe S, Hirose M, Wang XE, Kobayashi O, Nagahara A, Murai T, Iwazaki R, Miwa H,
Miyazaki A, Sato N: Epithelial-mesenchymal interaction in gastric mucosal restoration. J Gas-
troenterol 2000;35(suppl 12):65–68.
31 Watanabe S, Hirose M, Wang XE, Maehiro K, Murai T, Kobayashi O, Nagahara A, Sato N:
Hepatocyte growth factor accelerates the wound repair of cultured gastric mucosal cells. Biochem
32 Wright NA, Pike C, Elia G: Induction of a novel epidermal growth factor-secreting cell lineage by
mucosal ulceration in human gastrointestinal stem cells. Nature 1990;343:82–85.
Dr. J.L. Wallace, Department of Pharmacology, University of Calgary,

3330 Hospital Drive NW, Calgary, Alta T2N 4N1 (Canada)
Tel. +1 403 220 4539, Fax +1 403 270 3353, E-Mail wallacej@ucalgary.ca
Ma/Wallace 128
Intestinal Mucosal Function following

Ischemia/Reperfusion
Patrick Tsoa, Armin Wollinb
a
Department of Pathology and Laboratory Medicine, University of Cincinnati
College of Medicine, Cincinnati, Ohio, USA and
b
Department of Physiology, College of Medicine, University of Saskatchewan,
Saskatoon, Sask., Canada
It has been demonstrated both clinically and experimentally that ischemia

followed by reperfusion (I/R) is injurious to the gastrointestinal tract. Episodes
of I/R cause both mucosal [1–4] and vascular [5, 6] injury in the small intes-
tine. Mucosal injury is usually assessed either morphologically by light micro-
scopy or by measuring the permeability of the mucosal barrier to small probes,
such as 51Cr-EDTA or mannitol [7]. Other sugar probes that have been used to
determine gastrointestinal mucosal permeability are sucrose, lactulose and
sucralose [8]. Large solutes include the use of 125I-labeled albumin.
Morphological studies have demonstrated that intestinal ischemia prog-
resses gradually from subepithelial edema, occurring within 30 min following
total vascular occlusion, to the lifting of the epithelium from the villus base-
ment membrane about 1 h after the ischemia takes place. The villi are destroyed
after 2 h of ischemia [4]. The permeability of the intestinal mucosal barrier to
various probes that range in molecular weight from 300 to 70,000 depends on
the severity and duration of the ischemic episode [9]. Although the barrier func-
tion of the intestinal mucosa is clearly compromised by ischemia, little infor-
mation of the functional integrity of the mucosa is provided by this
permeability index. Moreover, even less information is known about the time
that is required or the mechanism for small intestinal mucosa recovery follow-
ing I/R-induced injury. Determining both the functional integrity and the recov-
ery time of the small intestine after I/R-induced injury is of interest to the basic
scientist and clinician alike.
This review will summarize our current understanding of intestinal
mucosal function and recovery after I/R-induced injury to the small intestine.
Lipid Absorption Is Potentially Useful for Determining Mucosal
Function following I/R-Induced Injury
An important role of the small intestine is to absorb the nutrients from the
lumen to the blood. Lipid absorption by the small intestine involves a series
of complex cellular processes, some of which are energy-dependent. Lipid
absorption involves protein, lipid and carbohydrate synthesis and their assem-
bly into triglyceride-rich lipoproteins, which are subsequently released by entero-
cytes via exocytosis. Readers interested in the mechanism and factors that
regulate lipid absorption are referred to the many reviews written on this topic
[10, 11]. Briefly, the digestion, absorption and transport of lipid by the small
intestinal cells (enterocytes) can be summarized as follows: (1) Lipid is
digested by pancreatic lipase in the intestinal lumen. (2) Micellar solubilization
of these digestion products by bile salts and their passage across the cell mem-
brane occurs either passively by diffusion or actively by transporters. (3) The
absorbed lipid is intracellularly transported into the endoplasmic reticulum.
(4) Triglyceride and phospholipid are re-synthesized in the endoplasmic retic-
ulum and apolipoproteins added. (5) Lipid droplets are transported to the Golgi
apparatus for terminal glycosylation of apolipoproteins. (6) Maturing intestinal
lipoproteins are packaged in the Golgi-derived vesicles. (7) Lipoproteins are
released from enterocytes by exocytosis.
Thus, with the complexity of the steps involved in intestinal fat absorption,
and because some are active, it is highly likely that the intestinal absorption of
lipid is susceptible to the deleterious effects of I/R.
Intestinal Lipid Absorption following I/R-Induced Injury
Using the well-established intestinal lymph fistula model, Fujimoto et al.

[12] demonstrated that intestinal lipid absorption is markedly compromised
after I/R-induced injury. In their experiments, rats were anesthetized and the
superior mesenteric artery (SMA) occluded for 10 min with a microbulldog
clamp. Lidocaine was applied directly on the SMA to ensure reperfusion.
Usually, the small intestine exhibits reactive hyperemia after ischemia and
becomes quite red – a hallmark that blood flow to the small intestine has
resumed. Resuming immediate blood flow to the small intestine after the clamp
is released is important because it allows the investigator to determine the exact
time of exposure to ischemia. The use of lidocaine further guarantees the
resumption of microvessel blood flow once the clamp is released. Next, the
intestinal lymph duct was cannulated, and a duodenal tube installed for steady
lipid infusion. Following surgery, the rats recovered overnight and received a
Tso/Wollin 130
100 24 h after I/R
24 h sham
48 h after I/R
48 h sham
75
% of hourly infused
50
25
0
Fasting 0 2 4 6 8
Hours of infusion
Fig. 1. Lymphatic radioactive lipid output expressed as % of radioactivity infused per

hour. The I/R rats had their SMA occluded for 10 min. The sham rats had their SMA isolated,
but not occluded. Four groups of rats, 5 animals/group, were studied. Values are expressed
as means SE. I/R, ischemia-reperfusion; sham, sham-operated controls.
continuous duodenal infusion of a glucose-saline solution to replenish fluid and

electrolyte loss. Rats, whose SMA had been isolated but not occluded, were
used as sham-operated controls.
Figure 1 shows that rats undergoing a 10-min occlusion of the SMA had
significantly less radioactive lipid (triolein with all three oleic acid mole-
cules labeled) transported into lymph than the sham-operated control rats
24 h following ischemia. Lymphatic lipid output reached a steady state
6 h after intraduodenal lipid infusion began. During the 7th and 8th h, the
I/R-injured rats transported about 40% of the infused radioactive lipid into
lymph, and the sham-operated control rats transported about 75%. Despite
the marked decrease in lymphatic lipid transport in I/R-injured rats, the mor-
phology of their small intestinal villi revealed no significant damage relative
to the sham-operated controls. This data therefore emphasizes the sensitivity
and importance of using a functional verses morphological assessment of
small intestine impairment after I/R-induced injury.
Fujimoto et al. [12] also examined the effect that length of ischemic
exposure had on lymphatic lipid transport. They found no impairment in the
small intestine’s ability to absorb and transport lipid into lymph as chylo-
microns within 7 min of ischemia. Additionally, they observed no difference
in intestinal lipid transport between a 10- and 20-min induced ischemic
episode. Furthermore, the transport of lipid in lymph was restored in the I/R-
injured rats to the same level that was observed in the sham-operated control
Intestinal Ischemia and Mucosal Function 131

rats 48 h after ischemia was induced. Fujimoto et al. [12] therefore concluded
that a short ischemic period (one between 7 and 10 min in length) followed by
reperfusion is sufficient to injure the small intestine thereby comprising its
ability to absorb and transport lipid into lymph, and that this deficiency is
fully restored 48 h after inducing ischemia. They further concluded that injury
to the small intestine is not exacerbated by 10 min of additional ischemic
exposure.
Mechanisms Responsible for Impaired Intestinal Lipid

Absorption following I/R-Induced Injury
Impaired intestinal lipid absorption can occur via several mechanisms.

One potential mechanism is impaired uptake of lipid digestion products by
enterocytes. Although lipid uptake has long been thought to occur through
passive diffusion, the idea that transporters are responsible for its uptake was
first suggested by Chow and Hollander [13]. Since these initial studies,
numerous transporters for the uptake of fatty acids, cholesterol and fat-
soluble vitamins have been proposed [14]. Because uptake by transporters
usually requires energy, it was speculated that the I/R-injured intestine is less
efficient in taking up lipid digestion products. However, Fujimoto et al. [12]
found the amount of radioactive lipid remaining in the intestinal lumen to
be the same in both I/R-injured rats and sham-operated controls (fig. 2).
Thus, the I/R-injured intestine is just as efficient in the uptake of lipid as the
uninjured intestine.
Reduced re-esterification of 2-monoglyceride and fatty acids to form
triglyceride was also speculated to be a possible cause for impaired intesti-
nal transport of lipid to lymph. However, as figure 2 demonstrates, the
radioactive lipid that remained in the intestinal mucosa was the same in
both I/R-injured and sham-operated control rats [12]. If re-esterification
of 2-monoglyceride and fatty acids to form triglyceride were impaired, a
high recovery of radioactive lipid in the mucosa would be observed in the
I/R-injured rats compared with the sham-operated controls. Therefore, Fujimoto
et al. [12] reasoned that it is unlikely that a defect in the re-esterification of
the absorbed lipids to form triglyceride is responsible for the compromised
lymphatic lipid transport in I/R-injured rats. This conclusion was supported
further by the distribution of the radioactive lipid in the intestinal mucosa.
More than 70% of the radioactive lipid was in the triglyceride of both groups
of rats, and no difference in the distribution of mucosal radioactive lipids
between the I/R-injured and sham-operated controls was observed.
Tso/Wollin 132
80
24 h after I/R
24 h sham
48 h after I/R
48 h sham
% of total infused dose
60
40
20
0
Lymph output Wall Luminal
Fig. 2. Recovery of radioactive lipid in the lymph, intestinal wall and lumen.
Following 8 h of continuous intraduodenal infusion of [3H]triolein emulsion, radioactive
lipid (as % of total infused dose) was measured in the lumen of the gastrointestinal tract and
the wall of the small intestine. The recovery of 3H-labeled lipid in lymph collected over an
8-h infusion period is also shown. **p 0.01 in I/R rats compared with sham rats. Four
groups of rats, 5 animals/group were studied. Values are expressed as means SE. I/R,
ischemia-reperfusion, sham, sham-operated controls.
Fujimoto et al. [12] then studied the portal transport of radioactive lipid
in I/R-injured and sham-operated control rats. They found a significant dif-
ference in the portal vein radioactive fatty acid concentration in the I/R-
injured rats compared with sham-operated controls. Considering blood flow,
Fujimoto et al. [12] therefore estimated that portal transport of lipid was
responsible for the observed difference in lymphatic transport of radioactive
lipid in the two groups of rats. Fujimoto et al. [12] speculated that the intesti-
nal mucosa becomes leaky in response to I/R injury, thereby allowing lipid
to access the interstitium and portal blood by circumventing the enterocytes.
This is a reasonable conclusion in view of findings from other investigators,
who have demonstrated that the tight junctions between enterocytes are sen-
sitive to I/R-induced injury, and thus the intestinal epithelium becomes more
sieve-like to 51Cr-EDTA and albumin [1, 3, 5, 15].
Fujimoto et al. [12] suggested that the increased radioactive fatty acid
concentration in the I/R-injured rats may be due to the immaturity of new
enterocytes that replace injured ones remaining on the intestinal villus to
efficiently absorb and transport lipids as chylomicrons thereby resulting in
more of the absorbed radioactive fatty acid to be transported in the portal
vein. This interesting idea is partly supported by the finding that the repair
of the small intestine after I/R injury is accomplished by the replacement of
injured cells [16].

Table 1. ODC activity in intestinal mucosa at 24 h following ischemia-
reperfusion induced injury
Duodenum Jejunum Ileum
I/R 4.6 1.6 46.8 8.4* 38.1 7.9*

Sham-operated 2.8 1.2 2.1 0.5 3.9 2.3
Nonoperated 5.2 2.3 4.1 1.3 4.3 2.1
Values are means SE and are expressed as pmol CO2 mg protein1 h1.
*p 0.01, compared with the other two control groups.
Mucosal Repair and Restoration of Intestinal Function

following I/R-Induced Injury
As previously noted, occlusion of the SMA for 10 min caused reduced

intestinal lymphatic lipid transport in rats 24 h following the onset of ischemia
and returned to a normal transport level at 48 h. Fujimoto et al. [17] examined
the role that ornithine decarboxylase (ODC) has on the repair of the small intes-
tine following I/R-induced injury in rats and found its activity markedly
elevated 24 h after I/R injury. Table 1 shows higher ODC activity (p 0.01)
in both the jejunum and ileum of the I/R-injured rats compared with sham-
operated and nonoperated controls. This does not necessarily imply, however, that
elevated ODC is involved in the repair of the small intestine after I/R-induced
injury and thus the restoration of intestinal lipid absorption. What is compelling
about ODC’s involvement in restoring intestinal function after I/R-induced
injury is that restoration of intestinal lymphatic lipid transported in I/R-injured
animals can be completely prevented with treatment of a suicide inhibitor of
ODC, -difluoromethylornithine (DFMO). Figure 3 shows that 2% DFMO
added to drinking water did not affect lymphatic lipid transport in I/R-injured
or sham-operated control rats 24 h following the onset of ischemia, but that it
did inhibit the restoration of normal lymphatic lipid transport in the I/R-injured
rats 48 h after ischemia was induced (p 0.01 in all groups). Fujimoto et al. [17]
therefore concluded that ODC activity in the intestinal mucosa of rats increases
markedly after a brief 10-min occlusion of the SMA and that this increase
in mucosal ODC activity appears to play a critical role in the restoration of
normal lymphatic lipid transport in the post-ischemic intestine.
Tutton [18] has demonstrated that exogenous administration of histamine
results in an increase in mitosis of rat jejunal epithelial cells. To further define
the role of histamine in the repair of the small intestinal mucosal function after
I/R-induced injury, Fujimoto et al. [19] measured the histamine output into
Tso/Wollin 134
100 24 h after I/R vehicle
24 h after sham vehicle
24 h after I/R 2% DFMO
24 h after sham 2% DFMO
Lymph radioactive lipid output (% of hourly infused)
50
0
0 2 4 6 8
100 48 h after I/R vehicle
48 h sham vehicle
48 h after I/R 2% DFMO
48 h sham 2% DFMO
50
0
0 2 4 6 8
Hours of infusion
Fig. 3. Lymphatic radioactive lipid output (% radioactive lipid infused/h). The I/R rats
had their SMA occluded for 10 min followed by reperfusion. The sham rats had their SMA
isolated, but not occluded. Rats in both groups received either vehicle only or 2% DFMO for
24 h (top) or 48 h (bottom) before receiving lipid infusion. Five rats in each group were studied.
Values are expressed as means SE. I/R, ischemia-reperfusion; sham, sham-operated
controls; DFMO, -difluoromethylornithine.
lymph in I/R-injured and sham-operated control rats. Figure 4 shows that

histamine output in lymph in the sham-operated control rats remained relatively
constant, at about 10 pmol/h during the entire 48-h experimental period. The
lymphatic histamine output increased in the I/R-injured rats 3 h following the
onset of ischemia. The differences in histamine output between the I/R-injured
and sham-operated control rats were significant (p 0.01) at 3, 6 and 24 h fol-
lowing induced ischemia. Lymphatic histamine output in the I/R-injured rats
returned to the level of the sham-operated controls at 48 h. When I/R-injured and
sham-operated control rats were given -fluoromethylhistidine (-FMH), an
inhibitor of histidine decarboxylase, increased lymphatic histamine output was
abolished in the I/R-injured rats. A potential source for increased histamine out-
put in the injured intestine is from mucosal mast cells. Mucosal mast cells also

60
Vehicle sham
Vehicle I/R
-FMH I/R
Histamine output
30
0
1 3 6 24 48
Hours after I/R
Fig. 4. Histamine output in intestinal lymph after I/R. Histamine output in I/R rats
treated with vehicle increased at the 3rd h compared with shams (p 0.01). This increase
was completely suppressed in rats treated with -FMH (p 0.01). Values are expressed
as mean SE. I/R, ischemia-reperfusion; sham, sham-operated controls; -FMH,
-fluoromethylhistidine.
secrete a specific protease marker called RMCP II [20, 21]. Consequently,

Fujimoto et al. [19] measured the RMCP II concentration in intestinal lymph and
found that it did not significantly change following I/R-induced injury. Thus,
Fujimoto et al. [19] concluded that increased lymphatic histamine in I/R-injured
rats is not caused by the degranulation of mucosal mast cells. The precise source
and reason for increased lymphatic histamine output is currently unknown.
Fujimoto et al. [19] next determined the role of histamine in the lymphatic
transport of dietary lipid following induced I/R injury. Figure 5 shows increased
lymph lipid output following lipid feeding in both the histamine and -FMH
administered sham-operated control groups. Steady output of lymph lipid was
reached at ~75% of the hourly infused radioactive lipid. Thus, it appears that
-FMH has no effect on lymphatic lipid transport. At 48 h following induced
ischemia, lymphatic lipid transport in the I/R-injured rats was restored to the
same level of efficiency as the sham-operated controls. This data concurs with
the previous finding that depressed lymphatic radioactive lipid transport caused
by I/R injury is restored to normalcy 48 h after inducing ischemia. Lymphatic
lipid transport was not restored to normal levels in the I/R-injured rats pretreated
with -FMH. Thus, lymphatic radioactive lipid output was significantly lower
and significantly different (p 0.01) for all time points in the -FMH adminis-
tered I/R-injured rats compared with the other three groups. These data clearly
support the important role of histidine decarboxylase and histamine in restoring
lymphatic lipid transport to normal output following I/R-induced injury.
Tso/Wollin 136
100
Vehicle 48 h sham
Vehicle 48 h after I/R
Lymph radioactive lipid output
-FMH 48 h sham
-FMH 48 h after I/R
(% of hourly infused)
75
50
25
0
0 2 4 6 8
Hours of infusion
Fig. 5. Lymphatic radioactive lipid output expressed as percentage of the radioactive

lipid infused per hour in rats 48 h following surgery. Values are expressed as means SE.
I/R, ischemia-reperfusion; sham, sham-operated controls; -FMH, -fluoromethylhistidine.
That ODC is involved in the repair of the gastrointestinal mucosa following

injury has been clearly demonstrated [17, 22, 23]. Its relationship with histamine
in the repair and restoration of gastrointestinal tract function following
I/R injury, however, was a question needing answering. Fujimoto et al. [19] found
that ODC was markedly increased following I/R injury and this increase was
markedly attenuated by -FMH treatment. This data therefore seems to suggest
that histamine mediates I/R-induced changes in ODC activity. The temporal
relationship between histidine decarboxylase and ODC activity following I/R
injury also suggests a role for histamine in ODC activity. Fujimoto et al. [24]
reported that following I/R-induced injury, histidine decarboxylase activity
increased in the 1st h, whereas ODC activity increased significantly at the
6th h. They also found that increased ODC activity following I/R injury was
attenuated by treatment with the H1 receptor antagonist [24]. The relationship
between histamine and ODC activity has also been demonstrated in the colon
and rat brain tissues [25, 26].
Intestinal Cell Apoptosis and Renewal following

I/R-Induced Injury
Herbst and Tso [16] examined the effect of I/R injury on intestinal mor-
phology, crypt cell proliferation, and migration of enterocytes along the villi.

Specifically, they were interested in determining whether normal lymphatic
lipid transport 48 h following I/R-induced injury was restored due to increased
cell proliferation and replacement of injured cells or by repair of injured
absorptive cells. Using light microscopy, Herbst and Tso [16] found that the
height of the duodenal and jejunal crypts and villi were similar in both the I/R-
injured and sham-operated control rats at 24 and 48 h following I/R injury.
Thus, decreased lymphatic lipid transport in I/R-injured rats 24 h following I/R-
induced injury is not caused by a reduction in the number of cells that absorb
ingested lipid. Using bromodeoxyuridine, Herbst and Tso [16] found that the
cell proliferation index and cell migration rate was significantly higher in the
crypts of the I/R-injured rats compared with the sham-operated controls. They
therefore concluded that the restoration of lymphatic lipid transport following
I/R injury is achieved by accelerated crypt cell proliferation and enterocyte
migration thereby resulting in new cells re-populating the villi and not by the
repair of injured enterocytes.
Noda et al. [27] examined apoptosis of enterocytes following I/R-induced
injury. They found that apoptosis markedly increased, and that the degree to
which it did was proportional to the length of the ischemic period. Noda et al. [27]
further demonstrated that the degree of apoptosis in the rat small intestine
was induced by ischemia, and this process was exacerbated by reperfusion.
The changes in apoptosis were not affected by the treatment of DFMO, thus
indicating that apoptosis is not affected by ODC activity.
Yoshida et al. [28] recently demonstrated that neither the H1 nor H2 recep-
tor antagonists effected the degree of apoptosis in the rat small intestine exposed
to extensive ischemia (60 min occlusion of the SMA). Interestingly, however,
they observed that treatment with aminoguanidine attenuated the increase in
mucosal apoptosis following I/R injury. Because aminoguanidine is an inhibitor
of diamine oxidase, one wonders if the decrease in apoptosis in the I/R-injured
intestine caused by aminoguanidine treatment is caused by the increased plasma
histamine level. This is unlikely because both the H1 and H2 receptor blocker
failed to show an effect on apoptosis. More likely, aminoguanidine’s effect is
probably caused by inducible nitric oxide synthase inhibition and not diamine
oxidase inhibition.
Protection of the Gastrointestinal Tract from

I/R Injury by Fish Oil
Bang and Dyerberg [29] reported that Greenland Eskimos consuming

a diet rich in fish oil have a significantly lower incidence of coronary heart
disease. Subsequent studies in both humans and animals have found that a
Tso/Wollin 138
fish oil-enriched diet lowers plasma triglyceride [30, 31] and reduces platelet
count and platelet aggregation [32]. Lee et al. [33] reported that healthy male
subjects ingesting fish oil (18 capsules of MaxEPA) had altered neutrophil
function and adherence of neutrophils to endothelial cell monolayers. The
neutrophils from these human subjects were shown to have significantly
reduced capacity to produce leukotriene B4, a result of the inhibition of the
5-lipoxygenase pathway in these neutrophils. Furthermore, the neutrophils from
these subjects had a markedly reduced chemotactic response to leukotriene B4.
Thus, Lee et al. [33] proposed that diets enriched with fish oil containing 3
fatty acids may have anti-inflammatory effects because of the effect these
fatty acids have on the 5-lipoxygenase pathway. A study conducted by Fisher
et al. [34] reported that neutrophils from healthy volunteers consuming a diet
enriched with fish oil for 6 weeks produced significantly less superoxide when
challenged by an inflammatory stimulus. Studies have demonstrated that neu-
trophils are involved in the injury of the small intestine caused by I/R [35].
Thus, it can be speculated that fish oil consumption may also protect the gas-
trointestinal tract from I/R-induced injury.
We fed male Sprague-Dawley rats a semi-purified diet supplemented
with 10% fish oil, 10% safflower oil, or 10% beef tallow for 4 weeks.
Following consumption of these diets, the rats’ SMA was occluded for
15 min and then reperfused. A radioactive lipid emulsion was infused
intraduodenally via mesenteric lymph fistula. Rats fed the 10% safflower oil
and beef tallow diets had significantly reduced lymphatic lipid output 24 h
following I/R-induced injury. However, the rats fed the 10% fish oil diet had
normal lymphatic lipid transport following I/R-induced injury. This prelimi-
nary data strongly suggests that chronic feeding of a diet enriched with
fish oil protects the intestine from I/R-induced dysfunction. Several ques-
tions worth addressing include (1) whether fish oil’s ability to protect
against I/R-induced injury is dose-dependent, (2) by which mechanism it
occurs, and (3) for what length of time it lasts. These questions require
further investigation.
Concluding Remarks
This review has focused on how the intestinal absorption of lipids has been
compromised by I/R-induced injury. However, it is not the authors’ intent to
imply that other functions of the gastrointestinal tract are not compromised by
I/R injury as well. For instance, Kles et al. [36] recently demonstrated that
occlusion of the SMA resulted in compromised intestinal absorption of glu-
cose. Intestinal lipid absorption is believed to be a good index of intestinal

mucosal function because it is probably one of the most complex functions
carried out by enterocytes. Restoration of small intestinal function following
I/R-induced injury is a fertile area of research for basic and clinical inves-
tigators. Organ preservation during intestinal transplant surgery and infor-
mation regarding the diseased state of the gastrointestinal tract due
to compromised splanchnic circulation is highly clinically relevant. Currently,
we know that circulation to the gastrointestinal tract can be greatly compro-
mised during exercise [37], and that some symptoms caused by exercise, such
as intestinal cramps, diarrhea and frequent bowel movements, are related to
decreased blood flow to the gastrointestinal tract [38, 39]. This is just the begin-
ning of the knowledge to be learned about the effect that I/R-induced injury has
on the human gastrointestinal tract.
Acknowledgement
The authors are extremely grateful to the generous support provided by the National
Institutes of Health DK54504, DK56910 and DK56863.
References
1 Schoenberg MH, Muhl E, Sellin D, Younes M, Schildberg FW, Haglund U: Posthypotensive gen-
eration of superoxide free radicals – Possible role in the pathogenesis of the intestinal mucosal
damage. Acta Chir Scand 1984;150:301–309.
2 Chiu CJ, McArdle AH, Brown R, Scott HJ, Gurd FN: Intestinal mucosal lesions in low-flow states.
A morphological, hemodynamic and metabolic reappraisal. Arch Surg 1970;101:478–483.
3 Crissinger KD, Granger DN: Mucosal injury induced by ischemia and reperfusion in the piglet
intestine: Influences of age and feeding. Gastroenterology 1989;97:920–926.
4 Haglund U, Lundgren O: Reactions within consecutive vascular sections of the small intestine of
the cat during prolonged hypotension. Acta Physiol Scand 1972;84:151–161.
5 Grogaard B, Parks DA, Granger DN, McCord JM, Forsberg JO: Effects of ischemia and oxygen
radicals on mucosal albumin clearance in intestine. Am J Physiol 1982;242:G448–G454.
6 Hernandez LA, Grisham MB, Granger DN: A role for iron in oxidant-mediated ischemic injury to
intestinal microvasculature. Am J Physiol 1987;253:G49–G53.
7 Krugliak P, Hollander D, Schlaepfer CC, Nuguyen H, Ma TY: Mechanisms and sites of mannitol
permeability of small and large intestine in the rat. Dig Dis Sci 1994;39:796–801.
8 Smecuol E, Bai JC, Sugai E, Vazquez H, Niveloni S, Pedreira S, Maurino E, Meddings J: Acute
gastrointestinal permeability responses to different non-steroidal anti-inflammatory drugs. Gut
2001;49:650–655.
9 Parks DA, Grogaard B, Granger DN: Comparison of partial and complete arterial occlusion mod-
els for studying intestinal ischemia. Surgery 1982;92:869–901.
10 Thomson ABR, Keelan M, Garg ML, Clandinin MT: Intestinal aspects of lipid absorption. Can J
Physiol Pharmacol 1989;67:179–191.
11 Nordskog BK, Phan CT, Nutting DF, Tso P: An examination of the factors affecting intestinal lym-
phatic transport of dietary lipids. Adv Drug Deliv Rev 2001;50:21–44.
12 Fujimoto K, Price VH, Granger DN, Specian R, Bergstedt S, Tso P: Effect of ischemia-reperfu-
sion on lipid digestion and absorption in rat intestine. Am J Physiol 1991;260:G595–G602.
Tso/Wollin 140
13 Chow SL, Hollander D: A dual, concentration-dependent absorption mechanism of linoleic acid
by rat jejunum in vitro. J Lipid Res 1979;20:349–356.
14 Abumrad NA, Sfeir Z, Connelly MA, Coburn C: Lipid transporters: Membrane transport systems
for cholesterol and fatty acids. Curr Opin Clin Nutr Metab Care 2000;3:255–262.
15 Kingham JGC, Whorwell PJ, Loehry CA: Small intestinal permeability: Effects of ischemia and
exposure to acetyl salicylate. Gut 1976;17:354–361.
16 Herbst JJ, Tso P: Repair after intestinal ischemia-reperfusion is accomplished by cell replacement.
17 Fujimoto K, Granger DN, Price VH, Tso P: Ornithine decarboxylase is involved in repair of small
intestine after ischemic-reperfusion in rats. Am J Physiol 1991;261:G523–G529.
18 Tutton PJM: The influence of histamine on epithelial cell proliferation in the jejunum of the rat.
Clin Exp Pharmacol Physiol 1976;3:369–373.
19 Fujimoto K, Imamura I, Granger DN, Wada H, Sakata T, Tso P: Histamine and histidine decar-
boxylase are correlated with mucosal repair in rat small intestine after ischemia-reperfusion.
J Clin Invest 1992;89:126–133.
20 Metcalfe DD: Mast cell mediators with emphasis on intestinal mast cells. Ann Allergy 1984;
53:563–574.
21 Befus D, Fujimake TD, Lee TD, Swieter M: Mast cells polymorphisms. Present concepts, future
directions. Dig Dis Sci 1988;33:16S–24S.
22 Johnson LR: Regulation of gastrointestinal mucosal growth. Physiol Rev 1988;68:456–502.
23 Wang JY, Johnson LR: Polyamines and ornithine decarboxylase during repair of duodenal mucosa
after stress in rats. Gastroenterology 1991;100:333–343.
24 Fujimoto K, Gotoh Y, Ogata S, Tsunada S, Ohyama T, Ootani A, Okamoto K, Sakata T:
Histaminergic control of mucosal repair in the small intestine. Obes Res 1995;5:795S–799S.
25 Sun Y, Li Y: Induction of ornithine decarboxylase and histidine decarboxylase activities in rat
colon mucosa after application of 12-O-tetra-decanoylphorbol-13-acetate, sodium deoxycholate
and indole. Cancer Lett 1988;39:77–84.
26 Morris G, Seidler FJ, Slotkin TA: Stimulation of ornithine decarboxylase by histamine or nor-
epinephrine in brain regions of the developing rat: Evidence for biogenic amine as trophic agents
in neonatal brain development. Life Sci 1983;32:1565–1571.
27 Noda T, Iwakiri R, Fujimoto K, Matsuo S, Aw TY: Programmed cell death induced by ischemia-
reperfusion in rat intestinal mucosa. Am J Physiol 1998;274:G270–G276.
28 Yoshida T, Iwakiri R, Noda T, Okamoto K, Kojima M, Fukuyama K, Fujimoto K: Histaminergic
effect on apoptosis of rat small intestinal mucosa after ischemia-reperfusion. Dig Dis Sci 2000;
45:1138–1144.
29 Bang HO, Dyerberg J: Plasma lipids and lipoproteins in Greenlandic west coast Eskimos. Acta
Med Scand 1972;192:85–94.
30 Goodnight SH Jr, Harris WS, Connor WE, Illingworth DR: Polyunsaturated fatty acids, hyper-
lipidemia and thrombosis. Arteriosclerosis 1982;2:87–113.
31 Wong SH, Nestle PJ, Trimble RP, Storer GB, Illman RJ, Topping DL: The adaptive effects of
dietary fish and safflower oil on lipid and lipoprotein metabolism in perfused rat liver. Biochim
Biophys Acta 1984;792:103–109.
32 Saynor R, Verel D, Gillot T: The long-term effect of dietary supplementation with fish lipid con-
centrate on serum lipids, bleeding time, platelets and angina. Atherosclerosis 1984;50:3–10.
33 Lee TH, Hoover RL, Williams JD, Sperling RI, Ravalese K, Spur BW, Robinson DR, Corey EJ,
Lewis RA, Austen KF: Effect of dietary enrichment with eicosapentaenoic and docosahexaehoic
acids on in vitro neutrophil and monocyte leukotriene generation and neutrophil function. N Engl
J Med 1985;312:1217–1224.
34 Fisher M, Upchurch KS, Levine PH, Johnson MH, Vaudreuil CH, Natale A, Hoogasian JJ: Effects
of dietary fish oil supplementation on polymorphonuclear leukocyte inflammatory potential.
Inflammation 1986;10:387–392.
35 Kurtel H, Fujimoto K, Zimmerman BJ, Granger DN, Tso P: Ischemia-reperfusion-induced
mucosal dysfunction: Role of neutrophils. Am J Physiol 1991;261:G490–G496.
36 Kles KA, Wallig MA, Tappenden KA: Luminal nutrients exacerbate intestinal hypoxia in the
hypoperfused jejunum. J Parenter Enteral Nutr 2001;25:246–253.

37 Peters HP, de Leeuw D, Lapham RC, Bol E, Mosterd WL, de Vries WR: Reproducibility of ultra-
sound blood flow measurement of the superior mesenteric artery before and after exercise. Int J
Sports Med 2001;22:245–249.
38 Brouns F, Beckers E: Is the gut an athletic organ? Digestion, absorption and exercise. Sport Med
1993;15:242–257.
39 Gil SM, Yazaki E, Evans DF: Aetiology of running-related gastrointestinal dysfunction. How far
is the finishing line? Sports Med 1998;26:365–378.
Patrick Tso, PhD, Department of Pathology and Laboratory Medicine,

University of Cincinnati College of Medicine,
231 Albert Sabin Way (ML 0529) Cincinnati, OH 45267–0529 (USA)
Tel. 1 513 558 2151, Fax 1 513 558 1006, E-Mail tsopp@email.uc.edu
Tso/Wollin 142
Helicobacter pylori Infection and

Gastroduodenal Mucosal Damage
and Healing
Harry Hua Xiang Xia, Benjamin Chun Yu Wong, Shiu Kum Lam
Department of Medicine, The University of Hong Kong,
Queen Mary Hospital, Hong Kong SAR, China
Helicobacter pylori is a bacterium colonizing the human stomach. In

developed countries 30–60% of the populations are infected with H. pylori,
while the rates increase to more than 60% in developing countries [1]. H. pylori
infection unequivocally leads to inflammation in the gastric mucosa, and is
responsible for up to 80% gastric ulcers and 90% of duodenal ulcer disease [2].
Moreover, there is now convincing evidence that H. pylori is causally linked to
gastric cancer [3, 4].
Studies have demonstrated that H. pylori infection not only causes mucosal
damage, but also delays mucosal healing. The aim of this chapter is to elucidate
the possible mechanisms. In addition, interaction between H. pylori infection
and non-steroidal anti-inflammatory drugs (NSAIDs) on the gastroduodenal
mucosal damage and healing is also discussed.
H. pylori-Induced Damage in Gastroduodenal Mucosa
H. pylori infection causes gastric mucosal damage by several mechanisms.

These include direct effect of virulence factors produced by H. pylori, inflam-
matory/immune responses, increase of acid secretion and induction of apopto-
sis of epithelial cells (fig. 1).
H. pylori Virulence Factors

H. pylori produces many molecules that possess toxic, antioxidant or meta-
bolic properties. While most virulence factors are shared by all H. pylori
strains, some of them are particularly expressed in certain strains.
H. pylori infection
Virulence factors produced by Inflammatory/immune responses and

H. pylori: urease, VacA, phospholipases, release of free oxygen radicals such
polyliposaccharide, alcohol as nitric oxide, and cytokines such as
dehydrogenase, proteases, etc. IL-8, TNF-␣, and IFN-␥, etc.
Acid secretion Cell apoptosis
Damage to gastroduodenal mucosa
Fig. 1. Proposed mechanisms by which H. pylori infection induces gastrointestinal

mucosal damage.
Urease is a unique enzyme that is produced by H. pylori in large amounts.

This enzyme plays a pivotal role in the colonization and pathogenesis, mainly
due to its ability to hydrolyse urea, which transudates from the bloodstream into
the gastric mucosa, to ammonia, which causes acute toxicity within the gastric
epithelial cells. It has been demonstrated that urease protects the organisms
from acidity and is essential for the organisms to adhere to gastric mucosa. This
enzyme is able to attract polymorphonuclear cells to infiltrate gastric mucosa,
and stimulate mononuclear phagocytes to produce inflammatory cytokines [5].
Furthermore, it may induce apoptosis of gastric epithelial cells and cause DNA
damage [5, 6].
A cytotoxin (VacA) that provokes the accumulation of vacuoles in the
cytoplasm of susceptible cells in vitro has been identified in some strains of
H. pylori. Animal studies have shown that infection with VacA⫹ strains pro-
duces more severe gastric pathology compared to VacA– strains [5]. Strains
with VacA are shown to be associated with severe gastritis, peptic ulcer disease,
precancerous lesions and gastric cancers in many studies, although this associ-
ation is not reproduced in other studies, particularly those carried out in Asian
countries [7]. One of the reasons for this discrepancy is the fact that so far the
presence of cytotoxic strains is indicated by detection of a cytotoxin-associated
Xia/Wong/Lam 144
gene product (CagA), but the correlation between VacA and CagA varies
between geographic regions. Moreover, approximately 50% of strains in western
countries whereas over 90% of strains in Asian countries are VacA-positive,
which may also affect results on the association [7, 8].
Phospholipases produced by H. pylori may also contribute to alteration
of the protective gastric mucosal barrier, thus favor ulcerogenesis. H. pylori
can produce different kinds of phospholipases: A1, A2 and C [5]. The putative
targets of these enzymes are the phospholipid-rich zone of gastric epithelial
cells and the phospholipidic components of the gastric mucus [5, 9]. These
enzymes attack phospholipids and generate lysolecithin and fatty acids, which
may possibly act as precursors of ulcerogenic components (like platelet-
activating factor), and as lipid mediators (like leukotrienes and prosta-
glandins) [5, 10]. As a result of the interaction of phospholipases with their
targets, the hydrophobicity of gastric mucus and mucosa in infected patients
may be impaired [5, 11].
Lipopolysaccharide (LPS) of gram-negative bacteria has been shown to
have an important role in the pathogenesis of infection. LPS of H. pylori has
low biological and immunological activity compared to other microorganisms
that induce higher cytokine, prostaglandin E2 and nitric oxide responses [5, 12].
This intrinsically low proinflammatory activity of H. pylori LPS could con-
tribute in part to the chronic course of H. pylori infection. H. pylori LPS may
alter the gastric integrity of gastric epithelial cells due to its ability to bind to
laminin, an extracellular matrix protein, and prevent it from interacting with its
integrin ligand [5]. H. pylori LPS has a negative effect on the synthesis of
sulfated gastric mucin and stimulates the gastric mucosal pepsinogen secre-
tion [5, 13]. Moreover, studies have shown that H. pylori LPS may be in part
responsible for the increase in acid production in some patients with H. pylori
infection, particularly those with duodenal ulceration [12]. Recent studies
have shown that H. pylori LPS induces epithelial apoptosis, probably by upreg-
ulation of gastric mucosal expression of endothelin-1 and TNF-␣ [14].
H. pylori produces alcohol dehydrogenase (ADH), which catalyses the
fermentation of glucose to ethanol and of ethanol to acetaldehyde, which may
accumulate in the gastric mucosa and cause damage. Acetaldehyde in vitro
reacts with phospholipids and proteins and forms adducts. In vivo, it can also
bind to various proteins and induce lipid peroxidation [5]. Moreover, chronic
exposure of rat gastric mucosa to this product yields a significant reduction in
epithelial cell proliferation [5].
Numerous studies have shown that H. pylori may produce proteases
that degrade proteins such as gastric mucins, transforming growth factor-␤
(TGF-␤) and PDGF [5, 15]. These proteins are involved in the mucosal defense
and repair. Their degradation, therefore, could result in mucosal damage.
Helicobacter pylori Infection and Gastroduodenal Mucosal Damage and Healing 145
Inflammatory/Immune Responses
The host’s response to H. pylori infection is characterized by infiltration of
plasma cells, lymphocytes, neutrophils and monocytes into the gastric mucosa
[16]. This inflammatory/immune response plays a major role in the induction
of gastric mucosal damage.
Activated neutrophils have many properties, which may contribute to
tissue damage. Neutrophil chemotaxis and activation can be induced directly by
products of H. pylori including a water-soluble neutrophil-activating protein.
Neutrophil infiltration and activation can be also indirectly through the inflam-
matory cytokine cascade [16]. The gastric epithelium secretes chemokines,
which have neutrophil attractant properties, such as IL-8 and GRO␣ in response
to bacterial infection [17]. In vivo production of these chemokines is increased
in the gastric mucosa with H. pylori infection, and in vitro, expression of these
chemokines is upregulated by proinflammatory cytokines such as tumor necro-
sis factor (TNF)-␣ and interleukin (IL)-1, which are produced from mono-
nuclear cells in the gastric mucosa in patients with H. pylori infection,
particularly the cytotoxic strains [16].
Following initial acute inflammation and associated changes in gastric per-
meability, continues exposure to antigen results in the generation of H. pylori
specific B- and T-cell responses [16, 18]. T-cell responses to H. pylori play an
important role in the induction of gastric mucosal damage. For activation of
T cells, interactions of B7-1 (CD80) or B7-2 (CD86) on antigen-presenting
cells with CD28 on T cells is required. Recent studies have suggested that gas-
tric epithelial cells could potentially function as antigen-presenting cells in
H. pylori infection. The infiltrating T cells in H. pylori-positive gastric mucosa
are predominately of the CD45RO⫹ phenotype. These cells frequently secrete
interferon (IFN)-␥ but not IL-4, indicating a Th1 phenotype in response to stim-
ulation with H. pylori antigens. A gastric Th1 response is more frequent in
patients with peptic ulcer disease [16].
Increase of Acid Secretion

H. pylori infection has been found to consistently elevate plasma
concentrations of gastrin (mainly gastrin 17) produced from G cells in the gas-
tric antrum. This occurs in the fasting state, after meals and during infusion of
gastrin-releasing peptide [19]. Eradication of H. pylori restores normal control
of gastrin release [19]. Function of G cells may be affected by products of
H. pylori itself, or by inflammatory factors released in H. pylori-associated gas-
tritis. On the other hand, expression of somatostatin in D cells is diminished in
H. pylori-infected mucosa, which would result in increased gastrin release and
acid secretion [19, 20]. IL-1 and TNF-␣ might be, in part, responsible for the
diminished expression of somatostatin [20]. Similarly, mucosal concentrations
Xia/Wong/Lam 146
of histamine are diminished in H. pylori-induced gastritis, probably due
to the suppression of enterochromaffin-like (ECL) cells by the H3 agonist
N-␣-methylhistamine produced by H. pylori or cytokines such as IL-1␤ [21].
Increased gastrin and diminished expression of somatostatin tend to increase
acid secretion. Moreover, some H. pylori products also stimulate parietal cells.
On the other hand, various substances produced by H. pylori and cytokines
TNF-␣ and IL-1␤ released during the infection inhibit parietal cells [19].
Therefore, the net effect of H. pylori infection on parietal cells and consequen-
tial acid secretion varies from time to time and from patient to patient, depend-
ing on the balance between stimulating and inhibiting factors. For example,
acid production is diminished at the early stage of infection and then returns
back to normal level. Whereas inhibiting factors may cause loss of parietal cells
and gastric atrophy, resulting in increased intragastric pH, and thus increased
risk of development of gastric cancer by allowing other bacteria persist and
produce carcinogens; stimulating factors may predispose to mucosal erosion or
ulceration by increasing acid secretion. Indeed, clinical studies have shown that
H. pylori infection is associated with elevated acid secretion in patients with
duodenal ulcer but decreased acid secretion in patients with gastric cancer [19].
Apoptosis of Epithelial Cells

Cell proliferation and apoptosis (a programmed cell death) are essential
events involved in the cellular turnover of epithelial tissue [4, 22, 23]. The lat-
ter is a physiological suicide mechanism that occurs during normal tissue
turnover and maintains homoeostasis of tissues. In the gastric epithelium, apop-
tosis plays an essential role in maintaining tissue integrity. Normally, the rate of
cell loss by apoptosis is matched by the rate of new cell production by prolif-
eration. However, this balance may be affected by H. pylori infection, leading
to various gastroduodenal diseases. It has been shown that H. pylori infection
induces apoptosis in the gastric epithelial cells, and subsequently results in an
increase in cell proliferation as a host response to apoptosis [4, 23, 24].
However, apoptosis induced by H. pylori infection that is not accompanied by
a matched increase in cell proliferation will result in loss of mucosal integrity
leading to gastric erosion and ulceration, or loss of gastric glands leading to
gastric atrophy [25]. On the other hand, apoptosis responding by an over-
matched increase in cell proliferation will result in hyperproliferation of gastric
mucosa, which is believed to increase the risk of the development of gastric
neoplasia [4, 23].
Both bacterial products and factors produced during H. pylori infection
may be involved in the induction of apoptosis [4]. It has been shown in rats
that intragastric application of H. pylori LPS to the gastric epithelium caused
a marked increase in apoptosis. Moreover, the number of apoptotic cells was
positively correlated with the grade of the induced acute inflammatory changes.
H. pylori-induced apoptosis may be accelerated by the urease of H. pylori,
which hydrolyses urea into ammonia and carbon dioxide [4, 26]. Ammonia can
easily penetrate cell membranes of gastric epithelial cells, and affect the intra-
cellular organelles, which may be a trigger for apoptosis. More importantly,
neutrophils in the lamina propria or infiltrating the epithelium activated during
H. pylori infection produce a large amount of reactive oxygen metabolites
including hypochlorous acid. The reaction product of ammonia and hypochlor-
ous acid, monochloramine, may be also involved in H. pylori-induced apopto-
sis through mitochondrial permeability transition, cytosolic caspases-3
activation and oxidative DNA damage [4]. H. pylori infection leads to the
expression of inducible nitric oxide synthase and sustained production of nitric
oxide by macrophages and neutrophils infiltrated in the gastric mucosa as part
of the host responses. These reactive nitrogen species may directly and indirectly
cause cell apoptosis. Moreover, cytokines produced by type 1 T-helper cells
(Th1), such as TNF-␣, IFN-␥, IL-2 and IL-1 markedly potentiate H. pylori-
induced apoptosis in gastric epithelial cells. There is evidence that TNF-␣,
IFN-␥, IL-2 and IL-1 increase apoptotic response of epithelial cells mediated
by the Fas/FasL signaling system [4].
Effects of H. pylori Infection on Healing of

Gastroduodenal Mucosa
Evidence on the effect of H. pylori on gastroduodenal mucosal healing is

accumulating although extensive studies are scarce. Animal studies have con-
sistently demonstrated that H. pylori infection delays gastric ulcer healing
[27–31]. H. pylori infection delays mucosal healing probably by affecting
the balance of cell apoptosis and proliferation and decreasing migration
of epithelial cells, blood flow and angiogenesis within the gastric mucosa
(fig. 2).
Effect on Growth Factors

It has been known that growth factors such as fibroblastic growth factor
(FGF), platelet-derived growth factor (PDGF), vascular endothelial growth
factor (VEGF), epithelial growth factor (EGF), transforming growth factor
(TGF), hepatocyte growth factor (HGF), and insulin-like growth factor 1 (IGF-1)
have a mitogenic activity. These factors and their receptors are increasingly pro-
duced following gastroduodenal ulceration, and play an important role in the
mucosal healing by increasing cell migration and proliferation, suppressing
apoptosis, enhancing blood flow, angiogenesis and production of granulation
Xia/Wong/Lam 148
H. pylori infection
Virulence factors produced by Inflammatory/immune responses and

H. pylori: urease, VacA, phospholipases, release of free oxygen radicals such
polyliposaccharide, alcohol as nitric oxide, and cytokines such as
dehydrogenase, proteases, etc. IL-8, TNF-␣ and IFN-␥, etc.
Growth factors: FGF,

PDGF, VEGF, EGF,
TGF, HGF and IGF-1
Cell apoptosis
Cell proliferation Cell migration Angiogenesis
Healing of gastroduodenal mucosa
Fig. 2. Proposed mechanisms by which H. pylori infection delays gastrointestinal

mucosal healing. ---䉴 = Stimulation or induction; ....䉴⫽ inhibition or suppression.
tissue, and inhibiting acid secretion [27, 32–38]. Animal experiments have
demonstrated that intragastric administration of extraneous growth factors
accelerates healing of chemically induced gastric and duodenal erosions and
ulcers [33, 39]. Recently, a multicenter, randomized, double-blind clinical trial
in patients with duodenal ulcers showed that the ulcers were healed in signifi-
cantly higher proportion of patients treated with human recombinant EGF
(50 ␮g/ml tid for 6 weeks), compared with those with placebo (59 vs. 26% at
4 weeks and 70 vs. 40% at 6 weeks of the treatment) [40].
Previous studies have demonstrated that H. pylori and its heparan sulfate
binding proteins (HSBP) bind to FGF with an extremely strong affinity as well
as heparan sulfate and heparin with high affinity while there was a binding of
H. pylori to IGF-1 and PDGF [41]. Heparin is required for the activation of
FGF receptors. Thus, H. pylori could efficiently interfere with growth factors
and their receptors, resulting in disturbance of the balance that controls the
renewal, maintenance and repair of the gastroduodenal mucosa [41]. In an in
vitro study, it was observed that H. pylori upregulated expression of EGF-
related peptides including amphiregulin and heparin-binding EGF-like growth
factor (HB-EGF), but not TGF-␣, in gastric epithelial cells. This effect was
independent of VacA, CagA and ammonia [27]. However, the proliferative
effect of these EGF-related peptides was overtaken by the inhibitory effect of
H. pylori on cell growth [27]. In addition, it has been observed that in vitro pro-
tease produced by H. pylori degrade 62% of PDGF and TGF-␤, although the
rate for EGF and basal FGF is less than 5 and 7%, respectively [15]. On the
other hand, clinical studies have shown that EGF, TGF and HGF are higher in
H. pylori-infected gastric mucosa than uninfected mucosa in patients with pep-
tic ulcers, and eradication of the infection results in a decrease in EGF and TGF
[34, 42–45].
Effect on Cell Proliferation

Studies have consistently shown that proliferation of the gastric epithelium
is significantly higher in patients with H. pylori infection than in normal con-
trols, and eradication of H. pylori infection leads to the reduction in cell prolif-
eration [4]. However, most in vitro studies have produced the opposite results;
proliferation of cell lines seems to decrease when incubated with H. pylori or
its supernatant, suggesting that H. pylori directly inhibits proliferation of gas-
tric epithelial cells, and that increased proliferation in gastric epithelium in vivo
is a secondary response to increased H. pylori-induced apoptosis, increased
levels of gastrin produced by G cells, or production of cytokines and oxygen
free radicals during the H. pylori inflammatory responses, all of which may
lead to increase in the growth factors mentioned above [4].
Effect on Mucosal Healing

However, the overall effect of H. pylori infection in patients with peptic
ulcers is that the infection impedes ulcer healing despite the increase of growth
factors and increased cell proliferation [27–31]. Indeed, in an animal study, 22
Japanese monkeys were inoculated with cytotoxic strains of H. pylori in the
stomach to induce gastritis while 22 other monkeys were not. In all monkeys
an active ulcer was produced at the angulus by injection of ammonia [44].
It was observed that healing of the ulcer was significantly delayed in monkeys
with H. pylori infection compared with controls when they were followed up at
weeks 4, 6 and 8 [44]. These observations are further confirmed in mice and
rats, which show that H. pylori infection markedly delays the healing of acetic
acid-induced or ischemia-reperfusion-induced gastric ulcers [28–31]. Moreover,
it is reported that eradication of the infection promotes ulcer healing [31].
Xia/Wong/Lam 150
Excess cell loss by apoptosis, decreased mucosal microcirculation in the ulcer
region, release of cytokines such as IL-1, and an impairment of the gastrin-
somatostatin regulation in H. pylori infection, as described above, are likely to
be the major factors overtaking the effect of growth factors and cell prolifera-
tion in vivo, and account for the delay of ulcer healing [20, 28–30]. Moreover,
some metabolic products during H. pylori infection such as ammonia, hydrogen
peroxide and monochloramine not only increase apoptosis, but are also
involved in the delay of ulcer healing [46, 47]. Recent studies have also
shown that VacA inhibits EGF-induced cell proliferation, alters cytoskeleton-
associated proteins and interferes with ulcer re-epithelialization [48]. In addi-
tion, there is evidence showing H. pylori may also inhibit migration of gastric
epithelial cells [46, 49].
Interaction between H. pylori Infection and NSAIDs
NSAIDs are an independent cause of gastroduodenal mucosal damage,

mainly by blocking the synthesis of prostaglandins from cyclooxygenase
(COX). Prostaglandins protect the gastric mucosa against various noxious
agents and thus maintain mucosal integrity. Studies on the interaction with
H. pylori in the gastroduodenal damage have produced conflicting results. It is
unclear whether they act indifferently, synergistically or antagonistically in the
development of gastroduodenal damage.
Association between H. pylori and NSAIDs

An inverse association between H. pylori infection and NSAID use has
been observed in some, but not all studies [51]. In patients with peptic ulcer dis-
ease, between 30 and 75% of H. pylori-negative patients are NSAID users,
whereas between 12 and 30% of H. pylori-positive patients are NSAID users
[50, 52]. The reason for such inverse association between H. pylori infection
and use of NSAIDs in peptic ulcer patients is unknown. Whether NSAIDs have
bacteriostatic or inhibitory activity against H. pylori, and thus occasionally
eliminate the organism with or without use of one or more antimicrobial agents,
requires further investigation.
Interaction on Gastroduodenal Ulceration and Bleeding

Whereas most studies have shown that H. pylori infection increases the
risk of peptic ulcers in NSAID users, others have failed to confirm this rela-
tionship [50, 53, 54]. Conflicting results have been reported regarding the ben-
efit of eradication of H. pylori infection in prevention of ulcer development of
relapse in long-term NSAID users. It has been reported that eradication of the
infection is associated with reduced risk of development of gastroduodenal
ulcers before and during administration of NSAIDs [53–56]. Moreover, an
Italian study of 278 NSAID users with ulcers showed that eradication of
H. pylori infection significantly reduced the ulcer relapse rate [57]. On the
other hand, a multicenter trial (the HELP study) of 85 NSAID users with pep-
tic ulcers, comparing triple therapy (omeprazole, clarithromycin and amoxi-
cillin) with maintenance of omeprazole, concluded that acid suppression, not
eradication of H. pylori infection, prevented ulcer relapse in established NSAID
users [58].
The role of H. pylori infection in the development of ulcer complications
such as bleeding has not been defined. Limited data have also produced
conflicting results. Whereas some studies have reported that H. pylori infection
is not related to ulcer bleeding in NSAID users, others observed a positive
(i.e. increased risk), or a negative (i.e. decreased risk) association between
H. pylori infection and ulcer bleeding in NSAID users [59]. The significance of
eradication of H. pylori infection in the prevention of ulcer bleeding has yet to
be evaluated.
Effect of H. pylori Infection on Mucosal Adaptation to Aspirin

Gastric mucosa shows enhanced resistance to injury after exposure to
repeated insults of noxious agents such as aspirin, alcohol, stress or H. pylori-
related gastrotoxins. This is called gastric adaptation [60, 61]. It has been
reported that gastric adaptation to aspirin injury involves enhanced cell prolif-
eration, presumably mediated by increased expression of TGF-␣ and COX-2
(an inducible isoform of COX) [43, 60]. However, H. pylori infection impairs
the ability of the gastric mucosa to adapt to continued aspirin challenge in indi-
viduals with or without peptic ulcer disease [60, 62]. However, eradication of
H. pylori restores this adaptation [62]. These findings may explain the obser-
vation that eradication of H. pylori infection before NSAID therapy lowers the
incidence of peptic ulcers [55].
Interaction on Mucosal Healing

It is conceivable that the presence of H. pylori infection may delay mucosal
healing in NSAID users, although clinical study specifically addressing this
issue is lacking. Similarly, the effect of NSAIDs on the healing of H. pylori-
induced ulcer is unknown. It is interesting to notice that H. pylori infection
induces expression in gastric fibroblasts of HGF, and COX-2 (and accordingly
prostaglandin E), which reflects the human defense response to H. pylori infec-
tion [43]. However, this defense mechanism is inhibited by non-aspirin
NSAIDs [43]. More extensive studies are required to clarify the interaction
between H. pylori infection and NSAIDs on mucosal healing.
Xia/Wong/Lam 152
Regimens of H. pylori Eradication
Eradication of H. pylori infection is recommended in patients with

H. pylori-associated peptic ulcer disease as clinical trials have concluded that
eradication of the infection not only results in the ulcer healing but also pre-
vents the relapse of ulcers and complications [2, 63]. The following is the
current regimens of treatment for H. pylori infection.
The combination of a bismuth compound (preferably colloidal bismuth
subcitrate, or CBS), metronidazole and either tetracycline (BMT) or amoxi-
cillin (BMA) for 2 weeks (so-called classical triple therapy) was recommended
in 1990 [50]. BMT appears to be superior to BMA, with mean eradication
rates of 84 and 73%, respectively. However, bismuth-based classical triple
therapy is cumbersome and causes considerable side effects. Subsequently, the
proton pump inhibitor (PPI)-based triple therapy was developed. The regimens
consist of a PPI and two of the following three antibiotics: amoxicillin,
a nitroimidazole (metronidazole or tinidazole), and clarithromycin [50]. An
international, randomized, double-blinded, placebo-controlled study, the
MACH1 study, demonstrated that the regimens with omeprazole plus metron-
idazole and clarithromycin (both 250 and 500 mg), or omeprazole plus amox-
icillin and clarithromycin (500 mg only), achieved eradication rates of over
90%, with minimal side effects. The eradication rates with omeprazole plus
amoxicillin and metronidazole or clarithromycin (250 mg) were 79 and 84%,
respectively [50]. Recently, 1-week triple therapies with ranitidine bismuth cit-
rate (RBC), clarithromycin and amoxicillin or a nitroimidazole (RBC-based
triple therapy) have been shown to achieve similar eradication rates to the
PPI-based triple therapy [50, 64]. It is emphasized that drug resistance which
may pre-exist or acquire during the treatment, should be assessed to guide
the choice of a substantial treatment regimen if the above triple therapy
fails. Quadruple therapy, which adds a PPI or H2 inhibitor to the bismuth-
based triple therapy, or RBC-based triple therapy, may be the choice for the
second-line treatment [50, 65].
Conclusions
H. pylori infection causes gastroduodenal mucosal damage, and delays

mucosal healing, by producing many virulence factors itself and promoting
inflammatory/immune responses and releasing a large amount of chemokines
and cytokines, which subsequently affect acid secretion, production of growth
factors, cell apoptosis and proliferation. The interaction between H. pylori
infection and NSAIDs in the mucosal damage and healing is still under
investigation although a synergism is likely. Currently, many treatment
regimens are quite effective in eradicating H. pylori infection, and thus aug-
menting mucosal healing.
References
1 Xia HHX, Talley NJ: Natural acquisition and spontaneous elimination of Helicobacter pylori
infection. Am J Gastroenterol 1997;92:1780–1787.
2 Consensus Development Panel on Helicobacter pylori in Peptic Ulcer Disease: Helicobacter
pylori in peptic ulcer disease. JAMA 1994;272:65–69.
3 International Agency for Research on Cancer: Schistosomes, Liver Flukes and Helicobacter
pylori. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. IARC 1994;161:
177–241.
4 Xia HHX, Wong B CY, Lam SK: Helicobacter pylori infection and gastric cancer. Asian J Surg
2001;24:217–221.
5 Figura N: Helicobacter pylori factors involved in the development of gastroduodenal mucosal
damage and ulceration. J Clin Gatsroenterol 1997;25(suppl 1):S149–S163.
6 Xia HHX, Talley NJ: Apoptosis in gastric epithelium induced by Helicobacter pylori infection:
Implications in gastric carcinogenesis. Am J Gastroenterol 2001;96:16–19.
7 Xia HHX, Gallagher C, English L, Hyde D, Talley NJ, Keane CT, O’Morain CA: Comparison
between McCoy cell line and HeLa cell line for detecting Helicobacter pylori cytotoxicity:
Clinical and pathological relevance. Ital J Gastroenterol Hepatol 1999;31:663–668.
8 Wong BCY, Yin Y, Berg DE, Xia HHX, Wang WH, Zhang JZ, Wong WM, Huang XR, Tang VSY,
Lam SK: Distribution of distinct VacA, CagA and IceA alleles in Helicobacter pylori in Hong
Kong. Helicobacter 2001;6:317–324.
9 Munch F, Bode G, Ditschumeit H, Malfertheiner P: Demonstration of a phospholipid-rich zone in
human gastric epithelium which is damaged by Helicobacter pylori. Gastroenterology
1993;105:1598–1704.
10 Langton SR, Cesareo SD: Helicobacter pylori associated phospholipase A2 activity: A factor in
peptic ulceration? J Clin Pathol 1992;45:221–224.
11 Goggin PM, Marrero JM, Spychal RT, Jackson PA, Corbishley CM, Northfield TC: Surface
hydrophobicity of gastric mucosa in Helicobacter pylori infection – Effect of clearance and erad-
ication. Gastroenterology 1992;103:1486–1490.
12 Moran AP, Wadstrom T: Pathogenesis of Helicobacter pylori. Curr Opin Gastroenterol
1998;14(suppl 1):9–14.
13 Slomiany A, Piotrowski J, Slomiany BL: Sucralfate counteracts the inhibition of gastric mucosal
mucin receptor by H. pylori lipopolysaccharide. Scand J Gastroenterol 1995;30(suppl 210):77–83.
14 Slomiany BL, Piotrowski J, Slomiany A: Up-regulation of endothelin-converting enzyme-1 in gas-
tric mucosal inflammatory responses to Helicobacter pylori lipopolysaccharide. Biochem
15 Piotroswski J, Slomiany A, Slomiany BL: Suppression of Helicobacter pylori protease activity
towards growth factors by sulglycotide. J Physiol Pharmacol 1997;48:345–351.
16 Shimoyama T, Crabtree JE: Bacterial factors and immune pathogenesis in Helicobacter pylori
infection. Gut 1998;43(suppl 1):S2–S5.
17 Shimizu T, Kusugami K, Ina K, Imada A, Nishio Y, Hosokawa T, Ohsuga M, Shimada M, Noshiro
M, Kaneko H, Ando T: Helicobacter pylori-associated gastric ulcer exhibits enhanced mucosal
chemokine activity at the ulcer site. Digestion 2000;62:87–94.
18 Fukuda Y, Bamba H, Okui M, Tamura K, Tanida N, Satomi M, Shimoyama K, Nishigami T:
Helicobacter pylori infection increases mucosal permeability of the stomach and intestine.
Digestion 2001;63(suppl 1):93–96.
19 Calam J, Gibbons A, Healey ZV, Bliss P, Arebi N: How does Helicobacter pylori cause mucosal
damage? Its effect on acid and gastrin physiology. Gastroenterology 1997;113:S43–S49.
Xia/Wong/Lam 154
20 Brzozowski T, Konturek PC, Konturek SJ, Kwiecien S, Pajdo R, Karczewska E, Stachura J, Hahn
EG: Water extracts of Helicobacter pylori delay healing of chronic gastric ulcer in rats: Role of
cytokines and gastrin-somatostatin link. Digestion 1999;60:22–33.
21 Konturek PC, Brzozowski T, Karczewska E, Duda A, Bielanski W, Hahn EG, Konturek SJ: Water
extracts of Helicobacter pylori suppress the expression of histidine decarboxylase and reduce
histamine content in the rat gastric mucosa. Digestion 2000;62:100–109.
22 Staunton MJ, Gaffney EF: Apoptosis. Basic concepts and potential significance in human cancer.
Arch Pathol Lab Med 1998;122:310–319.
23 Shirin H, Moss SF: Helicobacter pylori induced apoptosis. Gut 1998;43:592–594.
24 Zhu GH, Yang XL, Lai KC, Ching CK, Wong BCY, Yuen ST, Ho J, Lam SK: Nonsteroidal anti-
inflammatory drugs could reverse Helicobacter pylori-induced apoptosis and proliferation in
gastric epithelial cells. Dig Dis Sci 1998;43:1957–1963.
25 Leung WK, To KF, Chan FK, Lee TL, Chung SC, Sung JJ: Interaction of Helicobacter pylori erad-
ication and non-steroidal anti-inflammatory drugs on gastric epithelial apoptosis and prolifera-
tion: Implications on ulcerogenesis. Aliment Pharmacol Ther 2000;14:879–885.
26 Fan X, Gunasena H, Cheng Z, Espejo R, Crowe SE, Ernst PB, Reyes VE: Helicobacter pylori
urease binds to class II MHC on gastric epithelial cells and induces their apoptosis. J Immunol
2000;165:1918–1924.
27 Romano M, Ricci V, Di Popolo A, Sommi P, Del Vecchio Blanco C, Bruni CB, Ventura U, Cover
TL, Blaser MJ, Coffey RJ, Zarrilli R: Helicobacter pylori upregulates expression of epidermal
growth factor-related peptides, but inhibits their proliferative effect in MKN-28 gastric mucosal
cells. J Clin Invest 1998;101:1604–1613.
28 Konturek PC, Brzozowski T, Konturek SJ, Stachura J, Karczewska E, Pajdo R, Ghiara P, Hahn EG:
Mouse model of Helicobacter pylori infection: Studies of gastric function and ulcer healing.
Aliment Pharmacol Ther 1999;13:333–346.
29 Konturek SJ, Brzozowski T, Konturek PC, Kwiecien S, Karczewska E, Drozdowicz D, Stachura J,
Hahn EG: Helicobacter pylori infection delays healing of ischaemia-reperfusion induced gastric
ulcerations: New animal model for studying pathogenesis and therapy of H. pylori infection. Eur
J Gastroenterol Hepatol 2000;12:1299–1313.
30 Li H, Kalies I, Mellgard B, Helander HF: A rat model of chronic Helicobacter pylori infection.
Studies of epithelial cell turnover and gastric ulcer healing. Scand J Gastroenterol 1998;33:
370–378.
31 Koga T, Ito K, Kawada H, Tabata K: Delayed healing of chronic gastric ulcer after Helicobacter
pylori infection in mice. J Pharm Pharmacol 1998;50:943–948.
32 Slomiany BL, Piotrowski J, Slomiany A: Synchronized induction in ␤FGF and Cdk2 during
gastric ulcer healing. Biochem Mol Biol Int 1996;40:339–345.
33 Szabo S, Khomenko T, Gombos Z, Deng XM, Jadus MR, Yoshida M: Transcription factors and
growth factors in ulcer healing. Aliment Pharmacol Ther 2000;14(suppl 1):33–43.
34 Hori K, Shiota G, Kawasaki H: Expression of hepatocyte growth factor and c-met receptor in
gastric mucosa during gastric ulcer healing. Scand J Gastroenterol 2000;35:23–31.
35 Ma L, Wang HY, Chow JYC, Cho CH: Cigarette smoking increases apoptosis in the gastric
mucosa: Role of epidermal growth factor. Digestion 1999;60:461–468.
36 Tarnawski AS, Jones MK: The role of epidermal growth factor (EGF) and its receptor in mucosal
protection, adaptation to injury and ulcer healing: Involvement of EGF-R signal transduction
pathways. J Clin Gastroenterol 1998;27(suppl 1):S12–S20.
37 Scheiman JM, Meise KS, Greenson JK, Coffey RJ: Transforming growth factor-␣ levels in human
proximal gastrointestinal epithelium. Effect of mucosal injury and acid inhibition. Dig Dis Sci
1997;42:333–341.
38 Slomiany BL, Piotrowski J, Slomiany A: Role of basal fibroblast growth factor in the suppression
of apoptotic caspase-3 during chronic gastric ulcer healing. J Physiol Pharmacol 1998; 49:
489–500.
39 Pohle T, Shihin M, Domschke W, Konturek JW: Effect of basal fibroblast growth factor on gastric
ulcer healing and its own mRNA expression. Aliment Pharmacol Ther 1999;13:1543–1551.
40 Palomino A, Hernandez-Bernal F, Haedo W, Franco S, Mas JA, Fernandez JA, Soto G, Alonso A,
Gonzalez T, Lopez-Saura P: A multicenter, randomized, double-blind clinical trial examining the
effect of oral human recombinant epidermal growth factor on the healing of duodenal ulcers.
Scand J Gastroenterol 2000;35:1016–1022.
41 Ascencio F, Hansson HA, Larm O, Wadstrom T: Helicobacter pylori interacts with heparin and
heparin-dependent growth factors. FEMS Immunol Med Microbiol 1995;12:265–272.
42 Konturek PC, Bobrzynski A, Konturek SJ, Bielanski W, Faller G, Kirchner T, Hahn EG: Epidermal
growth factor and transforming growth factor-␣ in duodenal ulcer and non-ulcer dyspepsia
patients before and after Helicobacter pylori eradication. Scand J Gastroenterol 1998;33:143–151.
43 Takahashi M, Katayama Y, Takada H, Kuwayama H, Terano A: The effect of NSAIDs and a COX-
2-specific inhibitor on Helicobacter pylori-induced PGE2 and HGF in human gastric fibroblasts.
Aliment Pharmacol Ther 2000;14(suppl 1):44–49.
44 Okui M, Fukuda Y, Yamamoto I, Shintani S, Shimoyama T: Helicobacter pylori infection affects
gastric ulcer healing in Japanese monkeys. J Gastroenterol 1998;33(suppl 10):26–30.
45 Russo F, Messa C, Amati L, Caradonna L, Leoci C, Di Matteo G, Jirillo E, Di Leo A: The influ-
ence of Helicobacter pylori eradication on the gastric mucosal content of epidermal growth factor,
transforming growth factor-␣, and their common receptor. Scand J Gastroenterol 1998;33:
271–275.
46 Sato K, Watanabe S, Yoshizawa T, Hirose M, Murai T, Sato N: Ammonia, hydrogen peroxide and
monochloramine retard gastric epithelial restoration in rabbit cultured cell model. Dig Dis Sci
1999;44:2429–2434.
47 Takeuchi K, Suzuki K, Mizoguchi H, Araki H, Nishiwaka H: Monochloramine impairs mucosal
blood flow response and healing of gastric lesions in rats: Relation to capsaicin-sensitive sensory
neurons. J Gastroenterol Hepatol 2001;16:282–289.
48 Pai R, Sasaki E, Tarnawski AS: Helicobacter pylori vacuolating cytotoxin (VacA) alters cytoskeleton-
associated proteins and interferes with re-epithelialization of wounded gastric epithelial monolayers.
Cell Biol Int 2000;24:291–301.
49 Ricci V, Ciacci C, Zarrilli R, Sommi P, Tummuru MK, Del Vecchio Blanco C, Bruni CB, Cover
TL, Blaser MJ, Romano M: Effect of Helicobacter pylori on gastric epithelial cell migration and
proliferation in vitro: Role of VacA and CagA. Infect Immun 1996;64:2829–2833.
50 Xia HHX, Wong BCY, Talley NJ, Lam SK: Helicobacter pylori infection – Current treatment
practice. Exp Opin Pharmacother 2001;2:253–266.
51 Xia HHX, Phung N, Altiparmak E, Berry A, Matheson M, Talley NJ: Reduction of peptic ulcer
disease and Helicobacter pylori infection but increase of reflux esophagitis in western Sydney
between 1990 and 1998. Dig Dis Sci 2001;46:2716–2723.
52 Xia HHX, Wong BCY, Wong KW, Wong SY, Wong WM, Lai KC, Hu WHC, Chan CK, Lam SK:
Clinical and endoscopic characteristics of non-H. pylori, non-NSAID duodenal ulcers: A long-
term prospective study. Aliment Pharmacol Ther 2001;15:1875–1882.
53 Sung J, Russell RI, Nyeomans, Chan FK, Chen S, Fock K, Goh KL, Kullavanijaya P, Kimura K,
Lau C, Louw J, Sollano J, Triadiafalopulos G, Xiao S, Brooks P: Non-steroidal anti-inflammatory
drug toxicity in the upper gastrointestinal tract. J Gastroenterol Hepatol 2000;15(suppl):
G58–G68.
54 Barr M, Buckley M, O’Morain C: Non-steroidal anti-inflammatory drugs and Helicobacter pylori.
Aliment Pharmacol Ther 2000;14(suppl 3):43–47.
55 Chan FK, Sung JJ, Chung SC, et al: Randomised trial of eradication of Helicobacter pylori before
non-steroidal anti-inflammatory drug therapy to prevent peptic ulcers. Lancet 1997;350:975–979.
56 Cullen DJ, Hawkey CM, Greenwood DC, et al: Peptic ulcer bleeding in the elderly: The relative
roles of H. pylori and non-steroidal anti-inflammatory drugs. Gut 1997;41:459–462.
57 Porro Bianchi G, Parente F, Imbesi V, Fintrane F, Caruso I: Role of H. pylori in ulcer healing and
recurrence of gastric and duodenal ulcers in long-term NSAID users. Response to omeprazole
dual therapy. Gut 1996;39:22–26.
58 Hawkey CJ, Tulassay T, Szczepanski L, et al: Randomised controlled trials of H. pylori eradica-
tion in patients on non-steroidal anti-inflammatory drugs. HELP NSAIDS study. Lancet 1998;
352:1016–1021.
59 Ng TM, Fock KM, Khor JL, Teo EK, Sim CS, Tan AL, Machin D: Non-steroidal anti-inflamma-
tory drugs, Helicobacter pylori and bleeding gastric ulcer. Aliment Pharmacol Ther 2000;14:
203–209.
Xia/Wong/Lam 156
60 Konturek PC, Brzozowski T, Pierzchalski P, Kwiecien S, Pajdo R, Hahn EG, Konturek SJ:
Activation of genes for spasmolytic peptide, transforming growth factor-␣ and for cyclooxygenase
(COX)-1 and COX-2 during gastric adaptation to aspirin in rats. Aliment Pharmacol Ther 1998;
12:767–777.
61 Tarnawski AS, Jones MK: The role of epidermal growth factor (EGF) and its receptor in mucosal
protection, adaptation to injury and ulcer healing: Involvement of EGF-R signal transduction path-
ways. J Clin Gastroenterol 1998;27(suppl 1):S12–S20.
62 Konturek JW, Dembinski A, Konturek SJ, Stachura J, Domschke W: Infection of Helicobacter
pylori in gastric adaptation to continues administration of aspirin in humans. Gastroenterology
1998;114:245–255.
63 Lam SK, Talley NJ: Helicobacter pylori Consensus. Report of the 1997 Asia Pacific Consensus
Conference on the management of Helicobacter pylori infection. J Gastroenterol Hepatol
1998;13:1–12.
64 Wong BCY, Wong WM, Wang WH, Fung FMY, Lai KC, Chu KM, Yuen ST, Leung SY, Hu WHC,
Yuen MF, Lau GKK, Chan CK, Lam SK: One-week ranitidine bismuth citrate-based triple
therapy for the eradication of Helicobacter pylori in Hong Kong with high prevalence of metron-
idazole resistance. Aliment Pharmacol Ther 2001;15:403–409.
65 Wong BCY, Wang WH, Wong WM, Lau GKK, Fung FMY, Kung NNS, Chu KM, Lai KC, Hu
WHC, Hu FL, Liu XG, Chan CK, Hui WM, Lam SK: Three-day lansoprazole quadruple therapy
for Helicobacter pylori-positive duodenal ulcers: A randomized controlled study. Aliment
Pharmacol Ther 2001;15:843–849.
Dr. Benjamin Chun Yu Wong, Department of Medicine,

The University of Hong Kong, Queen Mary Hospital, Hong Kong SAR (China)
Tel. ⫹852 2855 4541, Fax ⫹852 2872 5828, E-Mail bcywong@hku.hk
Part 3: Experimental Therapeutics
Nitric Oxide-Releasing Agents –

A New Generation of Drugs for
Gastrointestinal Diseases
Tomasz Brzozowski a, Piotr C. Konturek b, Stanislaw J. Konturek a
a
Department of Physiology, Jagiellonian University School of Medicine,
Cracow, Poland and b Department of Medicine I, University of Medicine
Erlangen-Nuremberg, Erlangen, Germany
Non-steroidal anti-inflammatory drugs (NSAID), such as aspirin (ASA),

naproxen, sulindac or piroxicam, are widely used in a variety of diseases due to
their analgesic, anti-inflammatory, antipyretic and antithrombotic effects. The
major limitation of their clinical use is, however, serious adverse events such as
induction of acute hemorrhagic erosions, aggravation of stress ulceration and
interference with healing of pre-existing gastric ulcerations [1–5]. The gastric
mucosal damage by ASA was first reported by Douthwaite and Lintott in 1938
and subsequent observations have confirmed that ASA and other NSAID repre-
sent a major public health problem [7, 8].
Mechanisms of Gastric Damage by NSAID
NSAID cause gastroduodenal mucosal damage due to increased perme-

ability in the stomach and the rest of gastrointestinal canal. Their action on the
gastric mucosa is often described as ‘breaking the mucosal barrier’ through its
direct action on the mucosal cells and the inhibition of prostaglandin (PG)
biosynthesis recognized for the first time by Vane about 30 years ago.
Topical mucosal toxicity occurs following application of acidic NSAID
rather than those with a neutral pKa. Weak acids such as ASA are absorbed
easily into the gastric mucosal cells where they ionize or become trapped and
achieve high mucosal concentration that exerts direct effect on enzyme activity,
uncoupling of oxidative phosphorylation and inhibition of fatty acid metabo-
lism [10–12].
NSAID NO
}
Anti inflammatory
Analgesic COX-1 & COX-2
Antipyretic inhibition
Antithrombotic
Neutrophil adherence
Reduced mucosal blood flow
Reduced mucus secretion
TNF-␣ and oxyradical generation
} ⫺
Removal of ? ? Delay of
gastroprotection Gastric injury ulcer healing
Fig. 1. Schematic presentation of the action of NSAID on COX-1 and COX-2 activity
and the mucosal alterations caused by nonspecific NSAID resulting in the prevention of
gastroprotection and delay in gastric ulcerations. The addition of NO moiety to NSAID
counteracts the damaging effects of these substances.
Regarding the implication of PG inhibition in mucosal damage by ASA,

it is widely accepted that NSAID are nonspecific and potent inhibitors of both
COX-1 and COX-2, resulting in the deficiency of PG that are considered
as major mediators of mucosal defense mechanism. At present, numerous other
substances, especially nitric oxide (NO), and others such as sensory enteric neu-
rons with their sensory neuropeptide calcitonin-gene related peptide (CGRP) as
well as various growth factors have been involved in gastric mucosal defense,
which could attenuate the injurious action of NSAID on the stomach (fig. 1).
However, the influence of NSAID on these gastroprotectors has not been
clarified.
Nitric Oxide in Mucosal Homeostasis
Besides PG, NO appears to be a crucial mediator of gastrointestinal

mucosal defense, but, paradoxically, it can also contribute to the mucosal injury
in certain situations. Numerous NO donors are now available and some of
them such as S-nitroso-penicillamine (SNAP) are capable of prolonging release
of NO resulting in the reduction in the severity of the gastric lesions and in
attenuation of intestinal injury associated with the mucosal insults caused by
ethanol, stress or ischemia/reperfusion [18, 19]. We demonstrated that the
administration of L-arginine, that is a substrate for constitutive NO synthase
(NOS), enhances the healing of chronic gastric ulcers while the inhibition of
NOS delays this healing. Earlier studies revealed that endogenous NO released
NO-NSAID in Gastrointestinal Diseases 159

Vascular tone
Cellular adhesion
Vascular permeability
Inhibition of
platelet adhesion
Regulatory
NO
Protection and healing Deleterious
Antioxidant Inhibits enzyme function

Inhibition of Promotes DNA damage
leukocyte adhesion
Induces lipid peroxidation
Vasodilation and
angiogenesis Depletes antioxidant stores
Increased susceptibility to:
Protection of cells Radiation
against oxidant injury Alkylating agents
Toxic metals
Fig. 2. Regulatory, protective and deleterious effects of NO.
from vascular endothelium, sensory nerves or gastric epithelium cooperates

with PG in the maintenance of gastric mucosa integrity and microcirculation
[14, 15] (fig. 2).
Nitric Oxide-Releasing NSAID
Recently, a new class of NSAID has been developed by adding of NO moi-

ety to the native NSAID [22–25]. The rationale behind this strategy is that NO
released from these derivatives exerts beneficial influence on gastric mucosa by
enhancing the mucosal defense ability and preventing of pathogenic events
resulting from the suppression of prostanoid synthesis such as the reduction in
mucosal microcirculation and the leukocyte-endothelial adherence [26, 27].
Brzozowski/Konturek/Konturek 160
In contrast to native NSAID, their NO-releasing derivatives such as
NO-aspirin (NO-ASA) and NO-naproxen were found to exhibit lower gastric
toxicity despite inhibiting both COX-1 and COX-2 activity in the gastric
mucosa [25–27]. The major importance of NO included into the ASA structure
is its prolonged local release in the gastric mucosa and direct prevention of
mucosal damage and the preservation of ulcer healing ability. These beneficial
effects were expected based on previous studies showing that both endogenous
NO released by capsaicin or NO originating from L-arginine, a substrate for
NOS or that released from glyceryl trinitrate exerts gastroprotective activity
and accelerates ulcer healing. According to our experience, the major effect of
NO has been attributed to the increase in the gastric blood flow in the mucosa,
especially at the ulcer margin and enhancement of angiogenesis [20, 21] (fig. 2).
Fiorucci et al. demonstrated that ASA leads to the TNF-␣-dependent
activation of gastric caspases, a class of cysteine proteases, associated with the
enhanced apoptosis and cells damage. NO-ASA spares the gastric mucosa by inac-
tivation of caspase through its S-nitrosylation and reduction in the release and
activity of TNF-␣ (fig. 3). We confirmed recently that NO-ASA and NO-naproxen
reduced dose-dependently ethanol and stress-induced acute gastric lesions and
greatly enhanced the gastric blood flow [29, 30]. These gastroprotective and hyper-
emic effects of both NO-NSAID were completely abolished by 1H-[1,2,4]oxidia-
zolo[4,3,-a]quinoxalin-1-one (ODQ), a specific inhibitor of guanylyl cyclase,
supporting the involvement of the NO-guanylyl cyclase-cGMP system in the gas-
troprotective action of NO-NSAID. It was also confirmed that both NO-NSAID
exhibited potent inhibitory action on PGE2 generation indicating that these novel
agents inhibit COX activity with the extent similar to that exerted by their native
drugs. Unexpectedly, the gene expression of COX-2 was significantly upregulated
in gastric mucosa treated with NO-NSAID and this could account for the increased
gastric mucosal blood flow due to generation of various angiogenic substances by
COX-2. The major finding of our study was demonstration that NO-NSAID, in
contrast to their native agents, failed to delay the healing of chronic gastric ulcer-
ations and these effects were similar to those attained with SNAP that is known to
slowly release NO in the gastric mucosa. Somewhat stronger ulcer healing efficacy
of SNAP as compared to NO-NSAID could be attributed to significant increase of
PGE2 generation both in the intact and ulcerated mucosa confirming earlier find-
ings of Salvemini et al. [31] that NO can activate the COX-2 pathway. This could
serve as reasonable explanation why the treatment with NO-NSAID failed to influ-
ence significantly the spontaneous ulcer healing despite the fact that these NO
derivatives of NSAID suppressed COX activity in the nonulcerated and ulcerated
gastric mucosa as effectively as their parent NSAID. Thus, it is possible that NO
released from NO-releasing NSAID may counteract the potential mucosal impair-
ment resulting from the effect of the COX inhibition and subsequent deficiency of

Aspirin NO-aspirin
Bloodstream
M⌽ NO moiety
Aspirin
⫹/⫺ aspirin
1
M⌽
TNF-␣
2 TNF-␣
TNF-R1 TNF-R2
TRADD
3 TRADD
TADD 4
TADD 5
Caspase 8 Caspase 3
Caspase 8 Caspase 3
Activated
caspases NO cGMP Inhibited
caspases
Gastric damage Gastric protection

a b
Fig. 3. Diagram representing hypothetical pathways activated by aspirin (a) and

NO-ASA (b). ASA administration results in TNF-␣ release from gastric and macrophages
(M⌽) leading to activation of TNF-␣ receptors, caspase cascade and gastric damage (apo-
ptosis) [from 28, with permission].
PG. Another possibility of the action of NO released from NO-NSAID could be

the suppression of ROS generation caused by the native NSAID, that are known to
activate neutrophils and their interaction with endothelium with concomitant
enhancement of lipid peroxidation in the gastric mucosa (fig. 1). In our recent
studies on rats with ROS determined in the gastric mucosa by means of chemilu-
minescence and lipid peroxidation measured by malondialdehyde (MDA) in the
ulcerated gastric mucosa, the application of NO-ASA was as effective in the
suppression of ASA-provoked generation of ROS as the addition of vitamin C,
a known ROS scavenger to native ASA.
Comparison of NO-NSAID and COX-2 Inhibitors
The impact of NSAID on public health has driven a search for safer but
equally effective NSAID. These attempts originate from the recognition of
at least two isoforms of COX, one that is constitutive (COX-1) subserving
30
25 Rofecoxib
Ulcer area (mm2)
20 *
* *
15 * *
ASA *
10 Vehicle
NO-ASA
5 *p⬍0.05
0
0 3 7 10 14
Days upon ulcer induction
Fig. 4. Healing of chronic gastric ulcers in rats treated with aspirin (ASA), NO-aspirin
(NO-ASA) and rofecoxib.
physiological or housekeeping functions such as gastroprotection and mucosal

blood flow and another, COX-2, that mediates pathology, including joint
inflammation and pain [33–37]. COX-2 is very similar to COX-1 but differs in
amino acid substitution at point 523 that creates a defect in the ‘enzyme canal’,
leaving a side-pocket that can be accessed by COX-2-selective drugs named
‘coxibs’. Among these agents, two are currently available, rofecoxib and cele-
coxib [35, 36]. Numerous clinical trials proved the effectiveness of these agents
in arthritis and this was associated with lower levels of endoscopically visible
gastric lesions as compared to those observed with nonselective NSAID. These
agents do not appear to affect gastric mucosal PG synthesis or to interfere with
PG-dependent mucosal defense mechanisms. According to our experimental
data with rofecoxib, the generation of PG in the intact mucosa is not affected
by this selective COX-2 inhibitor but at the ulcer margin where this generation
is usually augmented, rofecoxib significantly suppressed this generation. This
suppression of PGE2 by rofecoxib at the ulcer margin was probably responsible
for the delayed ulcer healing indicating that COX-2 overexpressed in the ulcer-
ated mucosa and COX-2-induced excessive PG formation exert favorable
influence on the healing process. In this respect, the NO-NSAID seems to be
superior from coxibs because they do not influence the healing rate of chronic
or pre-existing gastric ulcerations (fig. 4).
Conclusion
In summary, NO-releasing NSAID, while inhibiting PGE2 generation in

the gastric mucosa, are not ulcerogenic and also do not impair the spontaneous

ulcer healing and, most important, exhibit gastroprotective activity against
acid-independent (ethanol) and acid-dependent (stress) induced gastric lesions
indicating that NO released from these compounds has a beneficial influence
on the gastric mucosa and may compensate, at least in part, for the deficiency
of endogenous PG induced by these NSAID.
References
1 Lanza FL: Endoscopic studies of gastric and duodenal injury after the use of ibuprofen, aspirin
and other nonsteroidal anti-inflammatory agents. Am J Med 1984;13:19–24.
2 Levi S, Goodlad RA, Lee CY et al: Inhibitory effect of non-steroidal anti-inflammatory drugs on
mucosal cell proliferation associated with gastric ulcer healing. Lancet 1990;336:840–843.
3 Konturek PK, Brzozowski T, Konturek SJ, Dembinski A: Role of epidermal growth factor,
prostaglandin and sulfhydryls in stress-induced gastric lesions. Gastroenterology 1990;99:
1607–1615.
4 Wang JY, Yamasaki S, Takeuchi K, Okabe S: Delayed healing of acetic acid-induced gastric ulcer
in rats with indomethacin. Gastroenterology 1989;96:393–402.
5 Wallace JL, Reuter B, Cicala C, McKnight W, Grishman M, Cirino G: A diclofenac derivative
without ulcerogenic properties. Eur J Pharmacol 1994;257:249–255.
6 Douthwaite AH, Lintott SAM: Gastroscopic observation of the effect of aspirin and certain other
substances on the stomach. Lancer 1938;ii1222–1225.
7 Fries JF, Williams CA, Bloch DA, Michel BA: Non-steroidal anti-inflammatory drug-associated
gastropathy: Incidence and risk factor models. Am J Med 1991;61:213–222.
8 Smalley WE, Rayt WA, Daughtery JR, Griffin MR: Nonsteroidal anti-inflammatory drugs and the
incidence of hospitalization for peptic ulcer disease in elderly persons. Am J Epidemiol 1995;
141:539–545.
9 Vane JR: Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs.
Nature 1971;231:232–235.
10 Smith MJH, Dawkins PD: Salicylates and enzymes. J Pharm Pharmacol 1971;23:729–744.
11 Jorgensen TG, Weiss-Fogh US, Nielsen HH, Olesen HP: Salicylate and aspirin-induced uncou-
pling of oxidative phosphorylation in mitochondria isolated from the mucosa membrane of the
stomach. Scand J Clin Lab Invest 1976;36:649–654.
12 Roediger WEW, Mollard S: Selective inhibition of fatty acid oxidation in colonocytes by
ibuprone: A cause of colitis? Gut 1995;36:55–59.
13 Konturek SJ: Mechanisms of gastroprotection. Scand J Gastroenterol 1990;25(suppl 174):15–28.
14 Brzozowski T, Drozdowicz D, Szlachcic A, Pytko-Polonczyk J, Majka J, Konturek SJ: Role of
nitric oxide and prostaglandins in gastroprotection induced by capsaicin and papaverine. Digestion
1993;54:24–31.
15 Whittle BJR, Lopez-Bolmonte J, Moncada S: Regulation of gastric mucosal integrity by endogenous
nitric oxide: Interactions with prostanoids and sensory neuropeptides in the rat. Br J Pharmacol
1990;99:607–611.
16 Brzozowski T, Majka J, Konturek SJ, Bielanski W, Slomiany BL, Garner A: Gastroprotective
activity and receptor expression of transforming growth factor-␣, epidermal growth factor
and basic fibroblast growth factor in the rat stomach. Eur J Gastroenterol Hepatol 1994;6:
337–343.
17 Muscara MN, Wallace JL: Nitric oxide therapeutical potential of nitric oxide donors and
inhibitors. Am J Physiol 1999;276:G1313–G1316.
18 MacNaughton WK, Cirino G, Wallace JL: Endothelium-derived relaxing factor (nitric oxide) has
protective actions in the stomach. Life Sci 1989;45:1869–1878.
19 Payne D, Kubes P: Nitric oxide donors reduce the rise in reperfusion-induced intestinal mucosal
permeability. Am J Physiol 1993;265:G189–G195.
20 Brzozowski T, Konturek SJ, Śliwowski Z, Drozdowicz D, Z·aczek M, Kedra D: Role of arginine,
a substrate for nitric oxide synthase in gastroprotection and ulcer healing. J Gastroenterol 1997;
32:442–452.
21 Konturek SJ, Brzozowski T, Majka J, Pytko-Polonczyk J, Stachura J: Inhibition of nitric oxide
synthase delays healing of chronic gastric ulcers. Eur J Pharmacol 1993;239:215–218.
22 Elliott SN, McKnight W, Cirino G, Wallace JL: A nitric oxide-releasing non-steroidal anti-inflam-
matory drug accelerates gastric ulcer healing in rats. Gastroenterology 1995;109:524–530.
23 Wallace JL, McKnigth W, Del Soldato P, Bavdoun AR, Cirino G: Anti-thrombotic effects of a
nitric oxide-releasing, gastric-sparing aspirin derivative. J Clin Invest 1995;96:2711–2718.
24 Wallace JL, Cirino G, McKnight G, Elliott SN: Reduction of gastrointestinal injury in acute endo-
toxic shock by flurbiprofen nitroxybutyl ester. Eur J Pharmacol 1995;280:63–68.
25 Takeuchi K, Suzuki K, Yamamoto H, Araki H, Mizoguchi H, Ukawa H: Cyclooxygenase-2
selective and nitric oxide-releasing nonsteroidal anti-inflammatory drugs and gastric mucosal
responses. J Physiol Pharmacol 1998;49:501–513.
26 Moncada S, Palmer RMJ, Higgs EA: Nitric oxide: Physiology, pathophysiology and pharma-
cology. Pharmacol Rev 1991;43:109–142.
27 Gaboury J, Woodman R, Granger DN, Reinhardt P, Kubes P: Nitric oxide prevents leukocyte
adherence: Role of superoxide. Am J Physiol 1993;265:H862–H867.
28 Fiorucci S, Antonelli E, Santucci L, Morelli O, Miglietti M, Federici B, Mannucci R, Del Soldato P,
Morelli A: Gastrointestinal safety of nitric oxide-derived aspirin is related to inhibition of
ICE-like cysteine proteases in rats. Gastroenterology 1999;116:1089–1096.
29 Brzozowski T, Konturek PC, Konturek SJ, Sliwowski Z, Drozdowicz D, Kwiecien S, Pajdo R,
Ptak A, Pawlik M, Hahn EG: Gastroprotective and ulcer healing effects of nitric oxide-releasing
non-steroidal anti-inflammatory drugs. Digest Liver Dis 2000;32:583–594.
30 Brzozowski T, Konturek PC, Konturek SJ, Sliwowski Z, Pajdo R, Drozdowicz D, Ptak A, Hahn
EG: Classic NSAID and selective cyclooxygenase (COX-1) and COX-2 inhibitors in healing of
chronic gastric ulcers. Microscopy Res Tech 2001;53:343–353.
31 Salvemini D, Currie MG, Mollace V: Nitric oxide-mediated cyclooxygenase activation. J Clin
Invest 1996;97:2562–2568.
32 Brzozowski T, Kwiecien S, Konturek P, Konturek SJ, Ptak A, Mitis-Musiol⁄ M, Duda A, Bielanski
W, Hahn EG: Comparison of nitric oxide-releasing NSAID and vitamin C with classic NSAID
in healing of chronic gastric ulcers: Involvement of reactive oxygen species. Med Sci Monit 2001;
7:529–599.
33 Masferrer JL, Zweifel BS, Manning PT, Hauser SD, Lleahy KM, Smith WG, Isakson PC, Seibert
K: Selective inhibition of inducible cyclooxygenase-2 in vivo is anti-inflammatory and nonul-
cerogenic. Proc Natl Acad Sci USA 1994;91:3228–3232.
34 Hawkey CJ: COX-2 inhibitors. Lancet 1999;353:307–314.
35 Scott LJ, Lamb HM: Rofecoxib. Drugs 1999;58:499–505.
36 Boyce EG, Breen GA: Celcoxib: A COX-2 inhibitor for treatment of osteoarthritis. Hosp Formul
1999;34:405–417.
37 Goldberg MM: Celecoxib, a selective cyclooxygenase-2 inhibitor for the treatment of rheumatoid
arthritis and osteoarthritis. Clin Ther 1999;21:1497–1513.
Prof. S.J. Konturek, Department of Physiology, Jagiellonian University School of Medicine,

16 Grzegorzecka Str., PL–31-531 Cracow (Poland)
Tel. +48 12 4211006, Fax +48 12 4211578, E-Mail mpkontur@cyf-kr.edu.pl

Cyclooxygenase Inhibitor, a Foe or

a Friend in the Mucosal Protection
and Repair
B.M. Peskar
Department of Experimental Clinical Medicine, Ruhr-University of Bochum,
Germany
The gastrointestinal mucosa is continuously exposed to potentially noxious

agents and has a remarkable ability to withstand insults that would cause damage
in other tissues. A complex system of functions regulated by interacting media-
tors is present in the gastrointestinal mucosa to strengthen resistance against dam-
age and to restore mucosal integrity after injury. Among the mediators involved
in the maintenance of mucosal integrity, prostaglandins (PGs) play a crucial role.
PGs are a family of polyunsaturated C20 fatty acids that are biosynthesized
from arachidonic acid after its liberation from cell membrane phospholipids. The
first step of PG biosynthesis is catalyzed by the enzyme cyclooxygenase (COX)
and results in the formation of the intermediate PGG2 which is then reduced to
PGH2 in a peroxidase reaction by the same enzyme. Metabolism of PGH2 by cell-
specific isomerases or reductases yields the different PGs and thromboxanes. Two
isoforms of COX have been identified and have been termed COX-1 and COX-2.
COX-1 is expressed constitutively and high mRNA and protein levels are found in
most tissues [1]. In contrast, under basal conditions, expression of COX-2 is
usually low [1] but rapid upregulation occurs in pathophysiological conditions
such as inflammation, tissue damage and malignant transformation [2]. Potent
inducers of COX-2 expression are cytokines, endotoxin and mitogens [3, 4].
The potent analgesic, antipyretic and anti-inflammatory activity of COX
inhibitors explains the extensive use of this type of drugs to treat inflamma-
tory disorders. Until recently, the available non-steroidal anti-inflammatory
drugs (NSAIDs) were nonselective and inhibited both COX-1 and COX-2 to a
comparable extent. In the mean time, compounds have been developed that
specifically inhibit either COX-1 or COX-2. COX-1 and COX-2 have a similar
molecular structure with a long narrow channel that contains a hairpin bend at
the end. The two isoforms differ, however, in the amino acid at position 120. At
this site, COX-1 contains isoleucine, whereas COX-2 contains the smaller
amino acid valine. This results in the formation of a small side pocket in the
inner lining of COX-2. The ability to bind within this side pocket is responsi-
ble for selective inhibition of the COX-2 isoform [for review, see 5].
With nonselective standard NSAIDs the therapeutic benefit is necessarily
associated with side effects since both are due to suppression of PG formation.
A major target organ of NSAID toxicity is the gastrointestinal tract. Dyspepsia
impairs the quality of life in over 30% of chronic NSAIDs users. With chronic
NSAID use the relative risk for development of gastric ulcer is 4.0, for upper
gastrointestinal bleeding 3.09, for perforation 5.93 and for death from ulcers or
their complications 7 [6, 7]. The identification of COX-2 which predominates
at sites of inflammation led to suggestions that inhibition of COX-2 accounts
for the therapeutic benefit of NSAIDs whereas inhibition of COX-1 underlies
the NSAID-induced toxicity, particularly in the gastrointestinal tract [2]. This
concept was supported by findings that selective COX-2 inhibitors did not
reduce PG formation in the gastric mucosa of either experimental animals
[8–14] or humans [15]. Furthermore, treatment with the COX-2 inhibitors cele-
coxib or rofecoxib caused significantly less erosions, ulcers and severe compli-
cations such as bleeds, perforation and obstruction than standard NSAIDs in
patients with rheumatoid arthritis or osteoarthritis [16–19]. Recent results of
experimental studies, however, indicate that both COX-1 and COX-2 are
involved in gastrointestinal mucosal defense and repair with specific functions
of the two isoenzymes depending on the pathophysiological setting.
Effect of COX-1 and COX-2 Inhibition in Normal Gastric Mucosa
In experimental animals, most NSAIDs damage the gastric mucosa. Thus,

in rats, indomethacin, given orally, induced dose-dependent gastric injury and
simultaneously near-maximally inhibited gastric mucosal PG formation. Oral
administration of the selective COX-1 inhibitor SC-560 inhibited mucosal PG
generation comparable to indomethacin but did not damage the gastric mucosa
[13, 14]. The selective COX-2 inhibitors rofecoxib and celecoxib caused neither
measurable inhibition of gastric PG formation nor damage. However, concur-
rent treatment with a COX-1 inhibitor and COX-2 inhibitor induced severe gas-
tric lesions (fig. 1a). Gastric mucosal PG formation was not more suppressed
after combined inhibition of COX-1 and COX-2 than after administration of the
COX-1 inhibitor alone [13, 14]. Obviously, inhibition of total gastric PG gen-
eration alone is not ulcerogenic and only simultaneous inhibition of both COX
Cyclooxygenase Inhibitor 167

a b
50 Normal mucosa Acid-challenged mucosa
*
40
Lesion index
30 *
* *
*
20
+Rof (20 mg/kg)

+Rof (20 mg/kg)
10
0
Co 20 20 20 20 5 20 20 20 mg/kg
Co
Indo Rof SC-560 Indo Rof SC-560
Fig. 1. a Lesion formation in normal rat gastric mucosa 5 h after oral administration of
the nonselective COX inhibitor indomethacin, the COX-1 inhibitor SC-560 and the COX-2
inhibitor rofecoxib. b Lesion formation after acid challenge. Rats were treated p.o. with
indomethacin, SC-560 or rofecoxib 60 min before instillation of 1 ml of 200 mM HCl and
gastric damage was assessed 60 min later. *p ⬍ 0.001 vs. vehicle-treated controls; 䊉 p ⬍
0.001, 䊏 p ⬍ 0.05 vs. SC-560 alone. Data are modified from Gretzer et al. [14].
isoenzymes interferes with mucosal integrity. This phenomenon could be due to

the specific effect of COX-1 inhibition (which reduces gastric mucosal blood
flow without affecting leukocyte adherence) and COX-2 inhibition (which
increases leukocyte adhesion but has no blood flow-reducing activity) [13].
Similar findings were observed in the rat small intestine where neither a selec-
tive COX-1 inhibitor nor a selective COX-2 inhibitor alone induced mucosal
damage whereas combined inhibition of both isoenzymes resulted in severe
lesions. Administration of the COX-1 inhibitor was associated with gastric and
small intestinal upregulation of COX-2 expression and this was suggested to
compensate for the suppression of COX-1 activity [20].
These findings are in keeping with phenomena found in COX-1- or COX-
2-deficient mice. Thus, no spontaneous gastrointestinal lesions occurred in
COX-1 knockout mice, although gastric PGE2 levels were ⬍1% than those in
wild-type animals [21]. Similarly, no gastric pathology developed in COX-2-
deficient mice [22].
Effect of COX-1 and COX-2 Inhibition on Mucosal Defense in the

Face of Pending Injury
Whereas instillation of acid (0.2 N HCl) alone did not damage the gastric
mucosa, this treatment caused severe injury in rats treated with the selective
Peskar 168
COX-1 inhibitor SC-560. Pretreatment with the COX-2 inhibitors DFU or rofe-
coxib did not cause injury when given alone but significantly enhanced the for-
mation of gastric lesions elicited by the COX-1 inhibitor in the acid-challenged
mucosa (fig. 1b). Instillation of acid substantially elevated levels of COX-2
mRNA but not of COX-1 mRNA. Pretreatment with dexamethasone prevented
the acid-induced increase in mucosal COX-2 expression and enhanced the inju-
rious effect of the COX-1 inhibitor on the acid-challenged mucosa similar to the
effect of COX-2 inhibitors [14].
These findings show that the effect of COX inhibitors differs in normal
gastric mucosa and in mucosa exposed to a potentially noxious agent. In nor-
mal mucosa, damage only develops when both COX-1 and COX-2 are inhib-
ited. When a potentially noxious agent is present in the gastric lumen, isolated
inhibition of COX-1 alone interferes with mucosal defense. Simultaneously,
pending injury induces the expression of COX-2 which then assists COX-1 in
the maintenance of mucosal integrity.
Various mediators including PGs, nitric oxide (NO) and afferent nerves
have been shown to act in concert to ascertain mucosal resistance against
injury [23]. Whereas inhibition of COX-2 did not induce gastric damage even
in the presence of intragastric acid, these compounds elicited severe and dose-
dependent injury in the acid-challenged mucosa when NO generation was sup-
pressed by L-NAME. Likewise, defunctionalization of afferent neurons
evoked by a high neurotoxic dose of capsaicin resulted in severe gastrotoxic-
ity of COX-2 inhibitors in the acid-challenged mucosa even without suppres-
sion of the NO system [24]. Comparable effects have been reported for
indomethacin [23]. Thus, in the face of pending injury, blockade of COX-2
activity results in breakdown of gastric mucosal resistance when in addition
one of the other factors involved in gastric mucosal defense is impaired. In
contrast, in the acid-challenged gastric mucosa, inhibition of COX-1 activity
alone elicits severe damage even in the presence of a functioning NO and
afferent nerval system.
Role of Cyclooxygenases in Minimizing Gastric Mucosal Damage

and Effect of Enzyme Inhibition
In rats, occlusion of the gastric artery followed by reperfusion increased

mRNA levels of COX-2 but not of COX-1 in a time-dependent manner [11, 25].
The up-regulation of COX-2 expression during ischemia-reperfusion was atten-
uated by pretreatment with dexamethasone [11].
Ischemia-reperfusion alone caused only negligible gastric injury. However,
dose-dependent up to fourfold aggravation of damage occurred after treatment with

60 Indo NS-398 DFU Dexa
50 *
Lesion index * * *
40
30
20
10
0
2⫻15 ng 2⫻4 ng 2⫻4 ng 2⫻4 ng
Co
16DM 16DM 16DM 16DM
PGE2 PGE2 PGE2 PGE2
Fig. 2. Aggravation of gastric damage during ischemia-reperfusion in rats treated with

indomethacin (10 mg/kg), NS-398 (4 mg/kg), DFU (2 mg/kg) or dexamethasone (1 mg/kg).
The injurious effects of COX inhibitors and dexamethasone were reversed by concurrent treat-
ment with 16,16-dimethyl-PGE2. *p ⬍ 0.001 vs. controls subjected to ischemia-reperfusion
alone; 䊉p ⬍ 0.001 vs. COX inhibitor or dexamethasone without PG treatment. Data are
derived from Maricic et al. [11].
indomethacin, COX-2 inhibitors or dexamethasone [11]. Ischemia-reperfusion-

induced gastric damage was also increased after administration of the COX-1
inhibitor SC-560, but considerably higher doses compared with those of COX-2
inhibitors were necessary for the effect [26]. Similarly, damage to the small intesti-
nal mucosa elicited by occlusion of the superior mesenteric artery followed by
reperfusion was substantially aggravated by indomethacin and the COX-2
inhibitor rofecoxib as well as dexamethasone whereas the COX-1 inhibitor SC-
560 only slightly increased injury [27]. In the stomach [11] (fig. 2) as well as in
the small intestine [unpubl. observations], the damage-aggravating effects of these
compounds were fully reversed by co-administration of 16,16-dimethyl-PGE2 at
very low doses that did not exert protection in the absence of COX inhibitors or
dexamethasone.
Short periods of gastric ischemia-reperfusion effectively prevented damage
induced by long periods of ischemia-reperfusion or intragastric instillation of
necrotizing agents. Various mediators including COX-1- and COX-2-derived PGs
were proposed to be involved in the protective effect of ischemia-reperfusion pre-
conditioning [28]. Furthermore, the healing of ischemia-reperfusion-induced
lesions in the rat stomach was impaired by treatment with resveratrol and COX-
2 inhibitors [29].
Obviously, in ischemia-reperfusion of the stomach and small intestine,
both COX-1 and COX-2 function to minimize damage but COX-2 appears to be
more essential in this pathophysiological setting. This may be related to the
Peskar 170
crucial role of reactive oxygen metabolites generated by activated neutrophils
that mediate the microvascular and parenchymal injury in ischemia-
reperfusion. As inhibition of COX-2 but not inhibition of COX-1 increases
leukocyte adherence to the vascular endothelium [13], mucosal integrity during
ischemia-reperfusion may be particularly sensitive to inhibition of COX-2.
COX-1 and COX-2 gene disruption in mice was used to characterize the
role of COX isoenzymes in the response to radiation injury in the proximal
jejunum. Radiation injury caused a large increase in the induction of apoptosis
of crypt epithelial cells and a decrease in crypt cell survival in COX-1-deficient
but not COX-2-deficient mice compared with wild-type littermates [30]. A sim-
ilar reduction in crypt cell survival was demonstrated in wild-type FVB/N mice
receiving indomethacin but not COX-2 inhibitors in the period after irradiation.
Administration of 16,16-dimethyl-PGE2 reversed the indomethacin-induced
decrease in crypt survival [31]. Mice with COX-1 gene disruption had dimin-
ished intestinal PGE2 levels compared with their wild-type littermates with or
without ␥-irradiation. Crypt cell survival after irradiation was inhibited by an
anti-PGE2 antibody, suggesting that the effects observed were specifically
caused by decreased formation of PGE2 via COX-1 [30].
These findings show that PGs derived from either COX-1 or COX-2 or
both COX isoenzymes can increase the resistance of the gastrointestinal
mucosa against injury. The relative importance of the COX isoforms differs
between various forms of mucosal damage and appears to depend on the patho-
mechanisms underlying the specific situation.
COX-2 as Modulator of Intestinal Immune Responses and

Inflammation
Feeding hen egg-white lysozyme to mice expressing a transgenic T-cell

receptor that recognizes hen egg-white lysozyme peptide 64–61 resulted in no
intestinal pathology. However, simultaneous administration of the COX-2
inhibitor NS-398 elicited increased proliferation of lamina propria mono-
nuclear cells and crypt epithelial cells, crypt expansion and villus blunting [32].
Lamina propria mononuclear cells constitutively produce high levels of COX-
2-derived arachidonic acid metabolites, which act as immunomodulators in the
immune response to dietary antigen [33]. These findings establish that COX-2-
dependent arachidonic acid metabolites are essential in the development and
maintenance of intestinal immune homeostasis.
The general concept is that COX-2-derived arachidonic acid metabolites
mediate inflammatory reactions and COX-2 inhibitors have anti-inflammatory
activity [2]. Induction of colitis with intrarectal trinitrobenzone sulfonic acid

markedly increased COX-2 mRNA and protein but not COX-1 expression.
Treatment with COX-2 inhibitors resulted in exacerbation of colitis, with per-
foration occurring when the compounds were administered for a week [34].
On the other hand, the selective COX-2 inhibitor SC-236 and the preferential
COX-2 inhibitor nimesulide were shown to significantly decrease the extent of
iodoacetamide-induced and acetic acid-induced colitis in rats [35]. Finally,
impaired mucosal defense to acute colonic injury induced by dextran sodium
sulfate in mice lacking either COX-1 or COX-2 could be demonstrated [36].
The final contribution of COX-1 and COX-2 to mucosal inflammation and
resistance in experimental colitis remains thus to be elucidated. In addition,
clinical trials are necessary to clarify whether COX-2 inhibitors share the
exacerbation-inducing effect of standard NSAIDs in patients with chronic
inflammatory bowel diseases.
Helicobacter pylori colonization is associated with an increase in the
expression of COX-2 mRNA which correlates positively with the degree of
gastritis [37]. Increased COX-2 expression is also found in metaplastic and
dysplastic epithelium and adenocarcinoma of Barrett’s esophagus [38] as well
as in colon cancer tissue [for review, see 39]. In these conditions, nonselective
COX inhibitors and COX-2 inhibitors could possibly exert beneficial effects by
antagonizing the development and progression of malignancy.
Effect of COX-1 and COX-2 Inhibition on Gastroprotective

Phenomena
Agents that stimulate endogenous PG formation such as mild irritants evoke

gastroprotective effects. This adaptive cytoprotection is inhibited by pretreatment
with indomethacin. Work from our group has shown that the adaptive protection
elicited by 20% ethanol is abolished by pretreatment with very low doses of
COX-2 inhibitors but not by pretreatment with dexamethasone suggesting that it
is mediated by a constitutive COX-2 [10] (fig. 3). Perfusion of the gastric lumen
with 8% peptone protected against damage induced by subsequent perfusion
with 50% ethanol. The protective effect of peptone was partially inhibited by pre-
treatment with indomethacin or COX-2 inhibitors. Dexamethasone did not
reverse the protective activity of peptone suggesting that upregulation of COX-2
is not involved [12]. The effect of COX-1 inhibitors on the protection conferred
by mild irritants or intragastric peptone has not been investigated so far.
Rebamipide, a 2-(1H)-quinolinone analog, protected the gastric mucosa
against acute injury caused by various ulcerogenic factors and improved the
speed and quality of ulcer healing in experimental animals. In Japan, rebamipide
is approved for therapeutic use in patients with gastric ulcers and acute gastritis.
Peskar 172
% Inhibition of protection
100 Indomethacin
L-745,337
NS-398
DFU
50 Dexa
0
0.002 0.02 0.2 2 20
COX inhibitor (mg/kg)
Fig. 3. Effect of COX inhibitors and dexamethasone on the gastroprotection conferred

by 20% ethanol. Pretreatment with indomethacin, NS-398, L-745,337, or DFU but not dexa-
methasone inhibited the protective effect of 20% ethanol against gastric damage caused by
70% ethanol in a dose-dependent manner. Data are derived from Gretzer et al. [10].
Recently, it was shown that rebamipide upregulates the expression of COX-2 in

rat gastric mucosa without modulating COX-1 expression. A COX-2 inhibitor
blocked both the rebamipide-induced gastroprotection and increase in mucosal
PGE2 [40].
In the rat stomach, repeated administration of endotoxin protected against
the damaging effect of ethanol and increased gastric mucosal mRNA levels for
both COX-1 and COX-2. The endotoxin-induced gastric resistance to injury
was abolished by pretreatment with indomethacin but not with the COX-2
inhibitor L-745,337 [41]. Thus, gastroprotective effects can be mediated by a
constitutive COX-2 (20% ethanol and intragastric peptone), by an inducible
COX-2 (rebamipide), or by an upregulated COX-1 without contribution of
COX-2 (repeated endotoxin).
Mucosal Repair and Ulcer Healing
To maintain the integrity of the gastric mucosa, the epithelial lining is

rapidly replaced by cells that migrate from the proliferative zone of the gastric
glands. Various growth factors have been shown to stimulate the mitogenic
response of gastric epithelial cells. Growth factors such as hepatocyte growth
factor increased the expression of COX-2. Simultaneously, the hepatocyte
growth factor-mediated restitution was delayed by a COX-2 inhibitor [42].
Furthermore, hepatocyte growth factor protected gastric epithelial cells against
ceramide-induced apoptosis through induction of COX-2 [43]. Finally, using

sponges implanted subcutaneously into the back of rats as angiogenesis model
it was demonstrated that basic fibroblast growth factor markedly elevated levels
of mRNA for COX-2 and vascular endothelial growth factor parallel to an
increase in angiogenesis. The COX-2 inhibitor NS-398 significantly suppressed
the basic fibroblast growth factor-induced stimulation of angiogenesis and
upregulation of vascular endothelium growth factor [44]. Indomethacin and
diclofenac as well as the COX-2 inhibitors L-745,337 and NS-398 inhibited
angiogenesis in the ulcer base of chronic gastric cryoulcers and acetic acid-
induced ulcers in rats [9, 45].
In contrast to normal gastric mucosa where expression of COX-2 is low,
abundant expression of COX-2 occurs in ulcerated gastric mucosa. Thus, in
mice, COX-2 mRNA and protein were upregulated in gastric ulcers elicited by
subserosal injection of acetic acid. Expression of COX-1 was not different in
ulcerated and nonulcerated mucosa. In the ulcerated tissue, PG formation was
threefold higher than in normal gastric tissue and was inhibited in vitro by
the COX-2 inhibitor NS-398 [46]. In rats with chronic ulcers, COX-2
immunoreactivity was negligible in the normal gastric wall but after ulceration
occurred in abundance in the cytoplasm of monocytes, macrophages, fibro-
blasts and endothelial cells in regions of maximal repair activity below and in
late stages between the regenerative glands. The time course of COX-2 induc-
tion closely paralleled the increase in epithelial cell proliferation as assessed by
the uptake of bromodeoxyuridine showing a maximum 5 days after ulcer induc-
tion. COX-1 immunoreactivity was located mainly in the mucus neck cells of
the nonulcerated mucosa and decreased after gastric ulceration in the mucosa
adjacent to the ulcer crater. After day 5 the COX-1 immunoreactivity reap-
peared in the apical cytoplasm of the regenerative epithelial cells [9]. These
findings show that in chronic ulcers COX-1 and COX-2 have a different location
and a different time sequence of expression.
In acetic acid-induced ulcers in rats, in addition to elevated PGE2 produc-
tion and COX-2 mRNA levels, increased expression of interleukin-1␤, tumor
necrosis factor-␣ and transforming growth factor-␤1 mRNAs occurred in the
ulcerated but not normal tissue. In a culture of isolated ulcer base, blockade of
interleukin-1 and tumor necrosis factor-␣ reduced the expression of COX-2
mRNA and production of PGE2. In contrast, COX-2 mRNA expression and
PGE2 production were promoted by preventing the action of transforming
growth factor-␤1 [47]. These findings confirm the upregulation of COX-2 in
gastric ulcers in rats and suggest that COX-2 expression is under the regulation
of cytokines and growth factors.
Standard NSAIDs which inhibited both COX-1 and COX-2 impair gastric
ulcer healing in experimental animals and man. In mice with acetic acid-
induced gastric ulcers, the COX-2 inhibitor NS-398 significantly delayed ulcer
Peskar 174
healing [46]. In rats with gastric ulcers, treatment with indomethacin or
diclofenac for 8 or 15 days resulted in a dose-dependent significant delay in
ulcer healing which was evident in the second week after ulcer induction. Ulcer
healing was also impaired by treatment with the COX-2 inhibitor L-745,337.
Epithelial cell proliferation in the ulcer margin and microvessel density in the
ulcer bed was decreased and the thickness of the granulation tissue below the
ulcer crater and the gap between both edges of the muscularis mucosae was
increased to the same extent by indomethacin, diclofenac and L-745,337 [9]. In
chronic ulcers 15 days after induction when the initially increased expression of
COX-2 protein had returned to normal values, L-745,337 did not inhibit PG
levels in the intact gastric mucosa or in the mucosal ulcer margin and did not
inhibit platelet thromboxane release indicating that inhibition of COX-1 was
not involved in the L-745,337-induced impairment of ulcer healing [9].
Similarly, indomethacin and the COX-2 inhibitor NS-398 impaired the
healing of acetic acid-induced ulcers in rats by preventing regeneration of the
mucosa, maturation of the ulcer base and angiogenesis in the base [45]. Injection
of hepatocyte growth factor or gastrin locally around acetic acid-induced ulcers
accelerated the rate of healing, raised mucosal blood flow at the ulcer margin
and caused further upregulation of COX-2 mRNA and protein in the ulcerated
mucosa. Treatment with indomethacin or the COX-2 inhibitors NS-398 and
rofecoxib inhibited generation of PGE2, reduced mucosal blood flow at the ulcer
margin and attenuated the acceleration of ulcer healing by hepatocyte growth
factor and gastrin [48]. The role of COX-1 was studied using resveratrol which
also delayed ulcer healing [49]. However, resveratrol in addition to inhibiting
COX-1 enzyme activity is a potent inhibitor of COX-2 mRNA and protein
induction [50] and could thus delay healing via suppression of COX-2-derived
PGs. Studies with selective COX-1 inhibitors are necessary to clearly define the
role of COX-1 in gastric ulcer healing. As healing of gastric ulcers is associated
with an increase in gastric blood flow [48, 49, 51] and inhibition of COX-1 was
found to decrease gastric mucosal blood flow [13], selective inhibitors of COX-1
could possibly impair the healing process.
Conclusions
PGs play an important role in the regulation of physiological functions and

in the modulation of pathophysiological reactions of the gastrointestinal tract.
Whereas initially it has been proposed that only COX-1 is involved in the main-
tenance of gastrointestinal mucosal integrity, recent findings clearly estab-
lish important contributions of COX-2 to mucosal defense. Thus, in normal
gastrointestinal mucosa only simultaneous inhibition of COX-1 and COX-2

elicits damage. This is in keeping with the results of clinical trials showing that
treatment with selective COX-2 inhibitors causes significantly less erosions,
ulcers and severe complications in patients with rheumatoid arthritis and
osteoarthritis than standard NSAIDs which inhibit both COX isoenzymes. In
contrast, in the acid-challenged stomach, inhibition of COX-1 alone is ulcero-
genic whereas specific inhibition of COX-2 has no effect. However, when the
function of other factors enrolled in mucosal resistance such as NO, afferent
nerves or COX-1 is impaired, specific inhibition of COX-2 is detrimental to
mucosal integrity. Furthermore, whereas during gastric and intestinal ischemia-
reperfusion inhibition of COX-2 markedly augments injury, COX-1 limits radi-
ation-induced intestinal damage. Similarly, gastroprotective effects can be
mediated either by COX-1 or COX-2. COX-2 appears to be essential for the
development of intestinal immune homeostasis and for the limitation of
mucosal inflammation in certain types of experimental colitis. COX-2, upregu-
lated under the influence of growth factors and cytokines, plays a crucial role
in restitution and healing of the gastrointestinal mucosa by stimulating cell
migration, angiogenesis and possibly blood flow. Nonselective COX inhibitors
and COX-2 inhibitors delay ulcer healing. Thus, with the exception of H. pylori-
induced gastritis, Barrett’s esophagus and colon cancer where COX inhibition
may antagonize the development and progression of malignancies, COX
inhibitors have a negative impact on gastrointestinal mucosal defense and repair
with specific effects of COX-1 inhibitors and COX-2 inhibitors depending on
the physiological and pathophysiological setting.
References
1 Kargman S, Charleson S, Cartwright M, Frank J, Riendeau D, Mancini J, Evans J, O’Neill G:

Characterization of prostaglandin G/H synthase-1 and -2 in rat, dog, monkey, and human
gastrointestinal tracts. Gastroenterology 1996;111:445–454.
2 Vane JR, Botting RM: New insights into the mode of action of anti-inflammatory drugs. Inflamm
Res 1995;44:1–10.
3 Raz A, Wyche A, Needleman P: Temporal and pharmacological division of fibroblast cyclooxy-
genase expression into transcriptional and translational phases. Proc Natl Acad Sci USA 1989;86:
1657–1661.
4 Kujubu DA, Fletcher BS, Varnum BC, Lim RW, Herschman HR: TIS10, a phorbol ester tumor
promoter-inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthase/
cyclooxygenase homologue. J Biol Chem 1991;266:12866–12872.
5 Jackson LM, Hawkey CJ: COX-2 selective nonsteroidal anti-inflammatory drugs: Do they really
offer any advantages? Drugs 2000;59:1207–1216.
6 Hawkey CJ: Non-steroidal anti-inflammatory drugs and peptic ulcers: Facts and figures multiply,
but do they add up? BMJ 1990;300:278–284.
7 Hawkey CJ, Hudson N: Mucosal injury caused by drugs, chemicals and stress; in Haubrich WS,
Schaffner F, Berck JE (ed): Bockus Gastroenterology. Philadelphia, Saunders, 1994, vol 5,
pp 656–699.
Peskar 176
8 Masferrer JL, Zweifel BS, Manning PT, Hauser SD, Leahy KM, Smith WG, Isakson PC,
Seibert K: Selective inhibition of inducible cyclooxygenase-2 in vivo is anti-inflammatory and
nonulcerogenic. Proc Natl Acad Sci USA 1994;91:3228–3232.
9 Schmassmann A, Peskar BM, Stettler C, Netzer P, Stroff T, Flogerzi B, Halter F: Effects of
inhibition of prostaglandin endoperoxide synthase-2 in chronic gastrointestinal ulcer models in
rats. Br J Pharmacol 1998;123:795–804.
10 Gretzer B, Ehrlich K, Maricic N, Lambrecht N, Respondek M, Peskar BM: Selective cyclo-
oxygenase-2 inhibitors and their influence on the protective effect of a mild irritant in the rat stom-
ach. Br J Pharmacol 1998;123:927–935.
11 Maricic N, Ehrlich K, Gretzer B, Schuligoi R, Respondek M, Peskar BM: Selective cyclo-
oxygenase-2 inhibitors aggravate ischaemia-reperfusion injury in the rat stomach. Br J Pharmacol
1999;128:1659–1666.
12 Ehrlich K, Plate S, Stroff T, Gretzer B, Respondek M, Peskar BM: Peptidergic and cholinergic
neurons and mediators in peptone-induced gastroprotection: Role of cyclooxygenase-2. Am
J Physiol 1998;274:G955–G964.
13 Wallace JL, McKnight W, Reuter BK, Vergnolle N: NSAID-induced gastric damage in rats:
Requirement for inhibition of both cyclooxygenase-1 and -2. Gastroenterology 2000;119:706–714.
14 Gretzer B, Maricic N, Respondek M, Schuligoi R, Peskar BM: Effects of specific inhibition of
cyclo-oxygenase-1 and cyclo-oxygenase-2 in the rat stomach with normal mucosa and after acid
challenge. Br J Pharmacol 2001;132:1565–1573.
15 Wight NJ, Gottesdiener K, Garlick NM, Atherton CT, Novak S, Gertz BJ, Calder NA, Cote J,
Wong P, Dallob A, Hawkey CJ: Rofecoxib, a COX-2 inhibitor, does not inhibit human gastric
mucosal prostaglandin production. Gastroenterology 2001;120:867–873.
16 Silverstein FE, Faich G, Goldstein JL, Simon LS, Pincus T, Whelton A, Makuch R, Eisen G,
Agrawal NM, Stenson WF, Burr AM, Zhao WW, Kent JD, Lefkowith JB, Verburg KM, Geis GS:
Gastrointestinal toxicity with celecoxib vs. nonsteroidal anti-inflammatory drugs for osteoarthri-
tis and rheumatoid arthritis: the CLASS study: A randomized controlled trial. Celecoxib Long-
Term Arthritis Safety Study. JAMA 2000;284:1247–1255.
17 Laine L, Harper S, Simon T, Bath R, Johanson J, Schwartz H, Stern S, Quan H, Bolognese J:
A randomized trial comparing the effect of rofecoxib, a cyclooxygenase-2-specific inhibitor, with
that of ibuprofen on the gastroduodenal mucosa of patients with osteoarthritis. Rofecoxib
Osteoarthritis Endoscopy Study Group. Gastroenterology 1999;117:776–783.
18 Bombardier C, Laine L, Reicin A, Shapiro D, Burgos-Vargas R, Davis B, Day R, Ferraz MB,
Hawkey CJ, Hochberg MC, Kvien TK, Schnitzer TJ: Comparison of upper gastrointestinal toxic-
ity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J
Med 2000;343:1520–1528.
19 Langman MJ, Jensen DM, Watson DJ, Harper SE, Zhao PL, Quan H, Bolognese JA, Simon TJ:
Adverse upper gastrointestinal effects of rofecoxib compared with NSAIDs. JAMA 1999;282:
1929–1933.
20 Tanaka A, Araki H, Komoike Y, Hase S, Takeuchi K: Inhibition of both COX-1 and COX-2 is
required for development of gastric damage in response to nonsteroidal antiinflammatory drugs.
J Physiol Paris 2001;95:21–27.
21 Langenbach R, Morham SG, Tiano HF, Loftin CD, Ghanayem BI, Chulada PC, Mahler JF,
Lee CA, Goulding EH, Kluckman KD, Kim SH, Smithies O: Prostaglandin synthase-1 gene
disruption in mice reduces arachidonic acid-induced inflammation and indomethacin-induced
gastric ulceration. Cell 1995;83:483–492.
22 Morham SG, Langenbach R, Loftin CD, Tiano HF, Vouloumanos N, Jennette JC, Mahler JF,
Kluckman KD, Ledford A, Lee CA, Smithies O: Prostaglandin synthase-2 gene disruption causes
severe renal pathology in the mouse. Cell 1995;83:473–482.
23 Whittle BJ, Lopez-Belmonte J, Moncada S: Regulation of gastric mucosal integrity by endoge-
nous nitric oxide: Interactions with prostanoids and sensory neuropeptides in the rat. Br J
Pharmacol 1990;99:607–611.
24 Maricic N, Ehrlich KH, Schuligoi R, Peskar BM: Cyclooxygenase-2-derived prostaglandins
interact with nitric oxide and afferent neurons in the maintenance of rat gastric mucosal integrity.
Gastroenterology 1999;116:A280–A281.

25 Kishimoto Y, Wada K, Nakamoto K, Kawasaki H, Hasegawa J: Levels of cyclooxygenase-1 and -2
mRNA expression at various stages of acute gastric injury induced by ischemia-reperfusion in rats.
Arch Biochem Biophys 1998;352:153–157.
26 Maricic N, Ehrlich KH, Peskar BM: Effect of cyclooxygenase-1 inhibitor SC-560 on mucosal
damage during ischemia-reperfusion in the rat stomach. Gastroenterology 2001;120:A152.
27 Maricic N, Ehrlich KH, Peskar BM: Role of cyclooxygenase-1 and cyclooxygense-2 in ischemia-
reperfusion injury in the rat small intestine. Gastroenterology 2001;120:A194.
28 Konturek SJ, Brzozowski T, Pajdo R, Konturek PC, Kwiecien; S, Sliwowski Z, Pawlik M, Ptak A,
Drozdowicz D, Hahn EG: Gastric preconditioning induced by short ischemia: The role of
prostaglandins, nitric oxide and adenosine. Med Sci Monit 2001;7:610–621.
29 Brzozowski T, Konturek PC, Konturek SJ, Sliwowski Z, Drozdowicz D, Stachura J, Pajdo R,
Hahn EG: Role of prostaglandins generated by cyclooxygenase-1 and cyclooxygenase-2 in
healing of ischemia-reperfusion-induced gastric lesions. Eur J Pharmacol 1999;385:47–61.
30 Houchen CW, Stenson WF, Cohn SM: Disruption of cyclooxygenase-1 gene results in an impaired
response to radiation injury. Am J Physiol 2000;279:G858–G865.
31 Cohn SM, Schloemann S, Tessner T, Seibert K, Stenson WF: Crypt stem cell survival in the mouse
intestinal epithelium is regulated by prostaglandins synthesized through cyclooxygenase-1. J Clin
Invest 1997;99:1367–1379.
32 Newberry RD, Stenson WF, Lorenz RG: Cyclooxygenase-2-dependent arachidonic acid metabo-
lites are essential modulators of the intestinal immune response to dietary antigen. Nat Med
1999;5:900–906.
33 Newberry RD, McDonough JS, Stenson WF, Lorenz RG: Spontaneous and continuous cyclooxy-
genase-2-dependent prostaglandin E2 production by stromal cells in the murine small intestine
lamina propria: directing the tone of the intestinal immune response. J Immunol 2001;166:
4465–4472.
34 Reuter BK, Asfaha S, Buret A, Sharkey KA, Wallace JL: Exacerbation of inflammation-associated
colonic injury in rat through inhibition cyclooxygenase-2. J Clin Invest 1996;98:2076–2085.
35 Karmeli F, Cohen P, Rachmilewitz D: Cyclo-oxygenase-2 inhibitors ameliorate the severity of
experimental colitis in rats. Eur J Gastroenterol Hepatol 2000;12:223–231.
36 Morteau O, Morham SG, Sellon R, Dieleman LA, Langenbach R, Smithies O, Sartor RB:
Impaired mucosal defense to acute colonic injury in mice lacking cyclooxygenase-1 or cyclooxy-
genase-2. J Clin Invest 2000;105:469–478.
37 Fu S, Ramanujam KS, Wong A, Fantry GT, Drachenberg CB, James SP, Meltzer SJ, Wilson KT:
Increased expression and cellular localization of inducible nitric oxide synthase and cyclooxyge-
nase-2 in Helicobacter pylori gastritis. Gastroenterology 1999;116:1319–1329.
38 Shirvani VN, Ouatu-Lascar R, Kaur BS, Omary MB, Triadafilopoulos G: Cyclooxygenase-2
expression in Barrett’s esophagus and adenocarcinoma: Ex vivo induction by bile salts and acid.
39 Williams C, Shattuck-Brandt RL, DuBois RN: The role of COX-2 in intestinal cancer. Ann NY
Acad Sci 1999;889:72–83.
40 Sun WH, Tsuji S, Tsujii M, Gunawan ES, Kawai N, Kimura A, Kakiuchi Y, Yasumaru M, Iijima H,
Okuda Y, Sasaki Y, Hori M, Kawano S: Induction of cyclooxygenase-2 in rat gastric mucosa by
rebamipide, a mucoprotective agent. J Pharmacol Exp Ther 2000;295:447–452.
41 Ferraz JG, Sharkey KA, Reuter BK, Asfaha S, Tigley AW, Brown ML, McKnight W, Wallace JL:
Induction of cyclooxygenase-1 and -2 in the rat stomach during endotoxemia: Role in resistance
to damage. Gastroenterology 1997;113:195–204.
42 Horie-Sakata K, Shimada T, Hiraishi H, Terano A: Role of cyclooxygenase-2 in hepatocyte growth
factor-mediated gastric epithelial restitution. J Clin Gastroenterol 1998;27:S40–S46.
43 Shimada T, Hiraishi H, Terano A: Hepatocyte growth factor protects gastric epithelial cells against
ceramide-induced apoptosis through induction of cyclooxygenase-2. Life Sci 2000;68:539–546.
44 Majima M, Hayashi I, Muramatsu M, Katada J, Yamashina S, Katori M: Cyclo-oxygenase-2
enhances basic fibroblast growth factor-induced angiogenesis through induction of vascular
endothelial growth factor in rat sponge implants. Br J Pharmacol 2000;130:641–649.
45 Shigeta J, Takahashi S, Okabe S: Role of cyclooxygenase-2 in the healing of gastric ulcers in rats.
J Pharmacol Exp Ther 1998;286:1383–1390.
Peskar 178
46 Mizuno H, Sakamoto C, Matsuda K, Wada K, Uchida T, Noguchi H, Akamatsu T, Kasuga M:
Induction of cyclooxygenase-2 in gastric mucosal lesions and its inhibition by the specific antag-
onist delays healing in mice. Gastroenterology 1997;112:387–397.
47 Takahashi S, Shigeta J, Inoue H, Tanabe T, Okabe S: Localization of cyclooxygenase-2 and regu-
lation of its mRNA expression in gastric ulcers in rats. Am J Physiol 1998;275:G1137–G1145.
48 Brzozowski T, Konturek PC, Konturek SJ, Pajdo R, Schuppan D, Drozdowicz D, Ptak A,
Pawlik M, Nakamura T, Hahn EG: Involvement of cyclooxygenase (COX)-2 products in acceler-
ation of ulcer healing by gastrin and hepatocyte growth factor. J Physiol Pharmacol 2000;51:
751–773.
49 Brzozowski T, Konturek PC, Konturek SJ, Sliwowski Z, Pajdo R, Drozdowicz D, Ptak A,
Hahn EG: Classic NSAID and selective cyclooxygenase (COX)-1 and COX-2 inhibitors in
healing of chronic gastric ulcers. Microsc Res Tech 2001;53:343–353.
50 Subbaramaiah K, Chung WJ, Michaluart P, Telang N, Tanabe T, Inoue H, Jang M, Pezzuto JM,
Dannenberg AJ: Resveratrol inhibits cyclooxygenase-2 transcription and activity in phorbol
ester-treated human mammary epithelial cells. J Biol Chem 1998;273:21875–21882.
51 Hirose H, Takeuchi K, Okabe S: Effect of indomethacin on gastric mucosal blood flow around
acetic acid-induced gastric ulcers in rats. Gastroenterology 1991;100:1259–1265.
Brigitta M. Peskar, MD, Department of Experimental Clinical Medicine,

Ruhr-University of Bochum, Universitätsstrasse 150, D–44801 Bochum (Germany)
Tel. ⫹49 234 32 24994, Fax ⫹49 234 32 14206, E-Mail dorothea.seier@ruhr-uni-bochum.de

Polysaccharides: A New Role in

Gastrointestinal Protection
Carmen Chi Min Cho, Edgar Shiu Lam Liu, Vivian Yvonne Shin,
Chi Hin Cho
Faculty of Medicine, The University of Hong Kong, SAR, China
Wound repair in the gastrointestinal tract involves a close interplay

between cell migration and proliferation. Stimulation of angiogenesis would
promote the progression of tissue repair at the later stage of wound recovery.
The amelioration of inflammation during ulceration may also help reduce the
severity of wounds and thereby enhance ulcer healing. Recent studies found
that unfractionated heparin, a polysaccharide from animal sources, not only
increased ulcer healing but also possessed anti-inflammatory action in the gas-
trointestinal mucosa. Moreover, polysaccharides derived from vegetal origins
were shown to produce similar antiulcer actions in the same system. These
intriguing findings prompt us to believe that studying the pharmacological
actions of polysaccharides represents a new facet of drug development for
wound repair in the gastrointestinal mucosa. This article reviews these actions
and also some basic mechanisms to show how polysaccharides promote tissue
repair through various growth factors and produce anti-inflammatory action by
suppressing the neutrophil/cytokine cascade in the gastrointestinal tract. These
actions could have significant implications for wound repair in other parts of
the human body.
Sources and Compositions of Polysaccharides
There are various sources of polysaccharides with different composi-

tions. Table 1 summarizes the sources of polysaccharides and their known
pharmacological actions on the gastrointestinal tract. These sources range
from plant to animal origins with different molecular structures. They are
mostly sulfated molecules with uronic acid as part of the residue. It was
Table 1. Origins and pharmacological actions of polysaccharides
Sources Pharmacological actions Ref.
Angelica sinensis Antiulcer on the stomach 53

Promote gastric cell migration 54, 55
and proliferation
Antihepatic damage 56
Bupleurum falcatum Antiulcer on the stomach 50, 51
Basidiomycetes mushroom Antitumor on the stomach and colon 46, 47
Lentinus edodes Anticancer on the stomach and colon 48, 49
Panax ginseng Antiulcer on the stomach 52
Porcine intestinal mucosa Stimulate gastric ulcer healing 17, 22, 29
Antihepatic damage 45
Anticolitis 38–42
Antipancreatitis 44
Ulva lactuca Inhibit colon cancer cell proliferation 1
(green seaweed)
shown that the sulfated polysaccharides, which contribute to the active com-
ponents in the extract, produce various pharmacological actions [1–3].
Desulfation of the polysaccharides abolished the inhibitory action on the pro-
liferation of arterial smooth muscle cells [4]. Also, some of the actions of
these compounds are partially dependent on the molecular size of the poly-
saccharides. Bigger molecules in the unfractionated heparin are necessary to
achieve a better action on the mitogenic activity of endothelial cells initiated
by acidic fibroblast growth factor [5]. However, a sulfated polysaccharide
extracted from brown seaweeds produced antiproliferative activity on vascu-
lar smooth muscle with a molecular weight threshold of about 30 saccharidic
units, which were necessary to give observable pharmacological effects [6].
In contrast, molecular sizes of heparin smaller than 10 saccharides, determine
heparin’s ability to block the biological activity of basic fibroblast growth fac-
tor (bFGF) on Caco-2 colon cancer cells [7]. All these studies demonstrate
that the length of polysaccharides determines the ability of the drugs for
different biological systems. Experimental evidence also shows that the ␤-
1,3- or -1,6-linked polysaccharides are responsible for different pharmaco-
logical actions on vascular and tumor cells, as the ␣-1,4 linkage can be
digested by the saccharidases to monomers in the gastrointestinal tract [8, 9].
Indeed the ␤-glucan receptors were first identified on human monocytes as
phagocytic receptors, which initiated phagocytosis [10].
Polysaccharides: A New Role in Gastrointestinal Protection 181

Effects of Heparin Beyond Its Anticoagulation Action
Heparin, generally referred to as unfractionated heparin, is a mixture of

highly sulfated glycosaminogylcans (GAG) found in mast cell secretory granules.
It is composed of modified D-glucosamine alternating with either modified glu-
curonic acid or uronic acid residues. Heparin consists of polysaccharides of dif-
ferent chain sizes, ranging from 3 to 30 kD, with an average molecular weight of
15 kD. The antithrombin III activity depends on its unique pentasaccharide
sequence that potentiates the inhibitory action on thrombin and activated factor X
by antithrombin III [11]. Besides the anticoagulant function mentioned above, new
heparin actions have been reported recently, and are now widely investigated for
their underlying modulating mechanisms and potentials in clinical application.
Modulation of Angiogenesis
Angiogenesis is the formation of new blood vessels from parent microves-
sels. Persistent capillary blood vessel growth is often associated with disease,
such as diabetic reinopathy, neovascular glaucoma, rheumatoid arthritis and
hemangioma [12]. Progressive tumor growth and metastases also appear to
depend on angiogenesis [13, 14]. However, stimulation of angiogenesis acceler-
ates wound healing [15]. Angiogenesis is a complicated process modulated by a
number of different growth factors. Some of these have been characterized to be
heparin binding. Among them, the relationship between the bFGF and heparin is
well defined [16, 17]. Heparin and heparan sulfate have a high affinity for bFGF,
and heparin and its polysaccharides stabilize bFGF. Fragments of heparin or
heparan sulfate may act as natural chaperones to shuttle bFGF to different cellu-
lar compartments. Heparin-like low-affinity receptors on the surface of endothe-
lial cells prepare FGFs for binding to their specific high-affinity receptors [18].
Furthermore, heparin itself is also demonstrated to promote angiogenesis [19].
However, it is observed that heparin has antiangiogenic actions under certain
conditions, such as in the presence of cortisone or hydrocortisone [20]. Moreover,
fragments of heparin that lack anticoagulant activity, e.g., hexasaccharides pro-
duced by enzymatic cleavage of heparin [21] or a synthetic pentasaccharide [20],
also inhibit angiogenesis when administered with steroids.
Preliminary study has shown that heparin given in doses of 500 and
1,000 U/kg can increase ulcer healing in rat stomachs [17, 22]. This effect is
thought to be a result of the enhancement of defensive mechanisms due to
increasing mucosal prostaglandin E2 levels as well as blood flow and, at the
same time, reducing inflammation in the gastric mucosa [22]. The drug also
promotes angiogenesis at the ulcer margin. Li et al. [17] showed that this
action was likely to be stimulated by bFGF, epidermal growth factor (EGF) and
constitutive nitric oxide synthase (cNOS) activity in the gastric mucosa.
Cho/Liu/Shin/Cho 182
Recently, studies have been focused on the role of these growth factors in the
process of ulcer healing, including the acceleration of healing of acute and
chronic lesions in the gastrointestinal tract [23, 24], which is in part through
the regeneration of the microvascular system (angiogenesis) in the mucosal
and submucosal layers [25].
Modulation of Cell Proliferation

Besides its angiogenic action, heparin is reported to have effects on cell pro-
liferation. D’Amore [26] found that heparin influences endothelial cell prolifera-
tion in a way that it is unrelated to its anticoagulant properties. Moreover, Flint et
al. [27] reported that heparin was a trophic factor in primary cultures of rat small
intestinal epithelium. This effect is dependent upon extensive molecular sulfation.
In contrast, the inhibitory effect of heparin on cell proliferation was also observed
in smooth muscle cells, hepatoma cells and promyelocytes [28]. All these results
indicate the complexity of heparin’s action on cell proliferation, which depends on
different tissues and situations. In the stomach, heparin given either by the intra-
venous or intragastric route accelerates ulcer healing. This effect was reported to
be nitric oxide-dependent as heparin increased the mucosal NOS activity and the
inhibitor of NOS blocked such action [17, 29].
Inhibition of Inflammation
Recent studies have indicated that heparin and related GAGs can modulate
the activities of a number of inflammatory cells, including T cells [30] and neu-
trophils [31]. Low-molecular-weight heparin (2 kD) also blocks neutrophil
enzyme release from azurophilic granules as well as the homotypic aggregation
of neutrophils [31]. Besides, heparin derivatives with low anticoagulant activities
can block the superoxide anion generation. Superoxide radicals produced by acti-
vated leukocytes can be neutralized indirectly by heparin through its association
with superoxide dismutase [32]. These studies indicate that heparin has the abil-
ity to attenuate damage of endothelium and tissue parenchyma by neutrophils
through a number of mechanisms. In addition, mast cell activation induced by
nonimmunological and immunological stimuli can be attenuated by heparin [33].
It has been postulated that heparin blocks inositol 1,4,5-triphosphate receptors of
the endoplasmic reticulum and thus prevents the release of intracellular Ca2⫹ con-
comitant with the downstream signals necessary for mast cell degranulation [33].
A considerable body of evidence indicates that heparin and its related molecules
are able to inhibit leukocyte adherence to the vascular endothelium and the
subsequent trafficking of cells into tissues [34, 35]. Heparin can bind to both
the leukocyte (L-type) and platelet (P-type) selectins but not the endothelium
(E-type) selectin [36]. It might reduce the number of rolling leukocytes and, as a
consequence, the number of cells recruited into the inflamed tissues. Heparin and

low-molecular-weight heparin have also been shown to reduce neutrophil chemo-
taxis stimulated by zymosan-activated serum [37].
The potential therapeutic applications and mechanisms of action of
heparin in inflammatory bowel disease have been extensively reviewed [38].
Clinical investigation showed that heparin produced remission of ulcerative col-
itis; thus heparin might have a role in the treatment of this inflammatory con-
dition [39]. This beneficial effect was further evidenced in 16 cases of
corticosteroid-resistant ulcerative colitis, which were attenuated after treatment
with a standard heparin [40]. It has been concluded that heparin exhibits a
broad spectrum of immunomodulating and anti-inflammatory properties by
inhibiting the recruitment of neutrophils and reducing proinflammatory
cytokines. Moreover, it can restore the high-affinity receptor binding of bFGF
and this could enhance ulcer healing in the inflamed bowel [38].
Experimental studies revealed that heparin treatment improved microangio-
graphic features and reduced inflammation in trinitrobenzene sulfonic acid [41]
and acetic acid-induced colitis [42] in rats. However, heparin did not ameliorate the
murine colitis induced by dextran sulfate [43]. Heparin was also found to have
beneficial effect on increasing microcirculatory values, decreasing serum IL-6
concentration, and improving morphological changes in experimental acute pan-
creatitis in rats [44]. Low-molecular-weight heparin can maximally attenuate liver
damage induced by concanavalin A. It can also reduce the levels of TNF-␣ and
IL-6, while significantly increasing the level of IL-10 [45].
Other Sources of Polysaccharides and Their Potential

Applications in Gastrointestinal Disorders
Polysaccharides from Mushroom and Seaweed

Higher Basidiomycetes mushrooms have been used in folk medicine
throughout the world since ancient times. It has been known for many years that
certain mushrooms from a higher Basidiomycetes origin are effective against can-
cers of the stomach and esophagus [46, 47]. However, the active components
responsible for these activities are largely unknown. Several antitumor polysac-
charides such as hetero-␤-glucans and their protein complexes, as well as dietary
fibers, lectins and terpenoids, have been isolated from medicinal mushrooms.
Using standard methods of fractionation and purification of polysaccharides,
Chihara [48] isolated from fruiting bodies of Lentinus edodes, a water-soluble
antitumor polysaccharide, which was found to be a ␤-1,3-D-glucan with ␤-1,6-
D-glucopyranoside branches. Results of clinical data suggested that this lentinan
polysaccharide was responsible for prolonging the life span of patients with
advanced and recurrent stomach and colorectal cancer with little toxicity [49].
Table 2. Effect of polysaccharide extract from A. sinensis (AS) on gas-
tric ulcer healing in rats (3 and 7 days of treatment after ulcer induction)
Dose Ulcer area, mm2
3 days 7 days
Control 90.14 ⫾ 9.47 38.25 ⫾ 2.62

AS 25 mg/kg 85.84 ⫾ 11.74** 22.29 ⫾ 2.55
AS 50 mg/kg 56.00 ⫾ 5.40** 22.00 ⫾ 2.78*
Ulcer was induced by 60% acetic acid applied locally in the gastric lumen
with the aid of a mold. See Li et al. [29] for details. Drug treatment given
intragastrically once daily started 1 day after ulcer induction. Values shown
are means ⫾ SEM of 6–8 rats. *p ⬍ 0.01, **p ⬍ 0.001 when compared to
the control.
In seaweed, another source of polysaccharides, ulvans, the sulfated polysaccha-

rides, inhibited the colonic epithelial cancer cell line, but were inactive on normal
colonocytes [1].
Polysaccharides from Other Plant Origins

There have been a number of reports showing that polysaccharides isolated
from different plant origins possess pharmacological actions on various kinds
of cells in the gastrointestinal tract. Acidic polysaccharide fraction from the
roots of Bupleurum falcatum irrespective to the route of administration inhib-
ited the formation of gastric lesions induced by necrotizing agents such as HCl-
ethanol and ethanol [50]. The protection was due to the antisecretory activity
in the stomach, the stimulation of the defensive mechanism through mucus
secretion, and the radical scavenging effect in the gastric mucosa. However,
such protective effect was prostaglandin-independent [50, 51]. Similarly, acidic
polysaccharides from the leaves of Panax ginseng also inhibited the formation
of gastric lesions through the same mechanisms [52].
Recent studies showed that a crude extract from the roots of Angelica sinen-
sis (Dong Quai) mainly consisting of polysaccharides prevented neutrophil-
dependent ulcers, such as indomethacin- and ethanol-induced ulceration in rat
stomachs [53]. It also enhanced ulcer healing in the gastric mucosa (table 2),
perhaps through its anti-inflammatory action on the stomach. In the normal gas-
tric epithelial cells, the same extract stimulated cell proliferation and migration
[54, 55]. As both actions are the major mechanisms accounting for wound repair,
they, together with the anti-inflammatory action of the polysaccharides, could

Ulcer induction
Chemokines Cytokines
Inflammation PS
Ulcer formation
Proliferation
Apoptosis PS
Migration
PS Angiogenesis
⫽ Stimulation,
Ulcer healing ⫽ Inhibition,
PS ⫽ polysaccharides
Fig. 1. The pathogenesis of ulcer formation and healing; the mechanisms of how poly-
saccharides prevent inflammation through free radical scavenging action and reduction of
neutrophil infiltration and promote ulcer healing by stimulation of cell migration, prolifera-
tion and angiogenesis at the ulcer site.
explain the antiulcer action of Dong Quai in the stomach. Furthermore, it was
observed that polysaccharides from Dong Quai enhanced the stimulatory action
induced by ulceration on EGF and c-myc expression. This was followed by the
enhancement of ornithine decarboxylase (ODC) activity, the crucial factor
responsible for cell proliferation and migration and finally wound repair in the
stomach [54, 55].
EGF
Dong Quai ODC Wound repair
c-myc
Polysaccharides isolated from Dong Quai also had a protective effect

against liver damage provoked by paracetamol and carbon tetrachloride. This
effect was partially mediated through the inhibition of NOS in hepatocytes
[56].
Conclusions
Polysaccharides derived from various sources produce different actions on

the gastrointestinal tract with low systemic toxicity. These actions range from
anti-inflammatory actions (free radical scavenging and antineutrophil actions to
the reduction of margination) to ulcer healing (stimulation of cell proliferation,
migration and angiogenesis) (fig. 1). These polysaccharides represent a new type
of drug for ulcer diseases throughout the gastrointestinal tract and perhaps also
in other organs. Large-scale and detailed clinical studies are needed to affirm
such a revolutionary therapeutic application of polysaccharides in the treatment
of ulcer disease in humans. Interestingly, the polysaccharides isolated from
mushroom and seaweed could have a great potential for cancer therapy in the
gastrointestinal tract. A thorough mechanistic study of this anticancer action
would be a key issue for future investigations in polysaccharide medicine.
Acknowledgements
The project was supported in part by the University of Hong Kong and the Hong Kong
Research Grant Council. We also thank Dr. Joanne Zhong for her comments on the manuscript.
References
1 Kaeffer B, Benard C, Lahaye M, Blottiere HM, Cherbut C: Biological properties of ulvan, a new
source of green seaweed sulfated polysaccharides, on cultured normal and cancerous colonic
epithelial cells. Planta Med 1999;65:527–531.
2 Benitz WE, Lessler DS, Coulson JD, Bernfield M: Heparin inhibits proliferation of fetal vascular
smooth muscle cells in the absence of plate-derived growth factor. J Cell Physiol 1986;127:1–7.
3 Klein-Soyer C, Beretz A, Cazenave JP, Wittendrop-Rechenmann E, Vonesch JL, Rechenmann RV,
Driot F, Maffrand JP: Sulfated polysaccharides modulate effects of acidic and basic fibroblast growth
factors on repair of injured confluent human vascular endothelium. Arterosclerosis 1989; 9:147–153.
4 Vischer P, Buddecke E: Different action of heparin and fucoidan on arterial smooth muscle cell
proliferation and thrombospondin and fibronectin metabolism. Eur J Cell Biol 1991;56:407–414.
5 Sudhalter J, Folkman J, Svahn CM, Bergendal K, D’Amore PA: Importance of size, sulfation, and
anticoagulant activity in the potentiation of acidic fibroblast growth factor by heparin. J Biol
Chem 1989;264:6892–6897.
6 Logeart D, Prigent-Richard S, Boisson-Vidal C, Chaubet F, Durand P, Jozefonvicz J, Letourneur D:
Fucans, sulfated polysaccharides extracted from brown seaweeds, inhibit vascular smooth muscle
cell proliferation. II. Degradation and molecular weight effect. Eur J Cell Biol 1997;74:385–390.
7 Jayson GC, Gallagher JT: Heparin oligosaccharides: Inhibitors of the biological activity of bFGF
on Caco-2 cells. Br J Cancer 1997;75:9–16.
8 Miao HQ, Ishai-Michaeli R, Peretz T, Vlodavsky I: Laminarin sulfate mimics the effects of
heparin on smooth muscle cell proliferation and basic fibroblast growth factor-receptor binding
and mitogenic activity. J Cell Physiol 1995;164:482–490.
9 Chihara G: Recent progress in immunopharmacology and therapeutic effects of polysaccharides.
Dev Biol Stand 1992;77:191–197.
10 Czop JK, Kay J: Isolation and characterization of ␤-glucan receptors on human mononuclear
phagocytes. J Exp Med 1991;173:1511–1520.

11 Hirsh J, Raschke R, Warkentin TE, Dalen JE, Deykin D, Poller L: Heparin: Mechanism of action,
pharmacokinetics, dosing considerations, monitoring, efficacy and safety. Chest 1995;108:
258S–275S.
12 Folkman J: XIth Congress of Thrombosis and Haemostasis. Leuven, Leuven University Press,
1987, pp 583–596.
13 Folkman J, Klagsburn M: Angiogenic factors. Science 1987;235:442–447.
14 Folkman J: Advances in Cancer Research. New York, Academic Press, 1985, pp 175–203.
15 Robson MC, Philips LG, Lawrence WT, Bishop JB, Youngerman JS, Hayward PG, Broemeling
LD, Heggers JP: The safety and effect of topically applied recombinant basic fibroblast growth
factor on the healing of chronic pressure sores. Ann Surg 1992;216:401–406.
16 Montesano R, Vassalli JD, Baird A, Guilemin R, Orci L: Basic fibroblast growth factor induces
angiogenesis in vitro. Proc Natl Acad Sci USA 1986;83:7297–7301.
17 Li Y, Wang HY, Cho CH: Association of heparin with basic fibroblast growth factor, epidermal
growth factor, and constitutive nitric oxide synthase on healing of gastric ulcer in rats. J Pharmacol
Exp Ther 1999;290:789–796.
18 Folkman J, Shing Y: Control of angiogenesis by heparin and other sulfated polysaccharides. Adv
Exp Med Biol 1992;313:355–364.
19 Azizkhan RG, Azizhan JC, Zetter BR, Folkman J: Mast cell heparin stimulates migration of
capillary endothelial cells in vitro. J Exp Med 1980;152:931–944.
20 Crum R, Szabo S, Folkman J: A new class of steroid inhibits angiogenesis in the presence of
heparin or a heparin fragment. Science 1985;230:1375–1378.
21 Folkman J, Langer R, Linhardt RJ, Haudenschild C, Taylor S: Angiogenesis inhibition and tumor
regression caused by heparin or a heparin fragment in the presence of cortisone. Science 1983;
221:719–725.
22 Li Y, Mei QB, Cho CH: Healing effects of heparin on acetic acid-induced gastric ulcers in rats.
Chin Med J 1998;111:12–16.
23 Tarnawski A, Stachura J, Krause WJ, Douglass TG, Gergely H: Quality of gastric ulcer healing:
A new, emerging concept. J Clin Gastroenterol 1991;13(suppl 1):S42–S47.
24 Konturek PC, Konturek SJ, Brzozowski T, Ernst H: Epidermal growth factor and transforming
growth factor-␣: Role in the protection and healing of gastric mucosal lesions. Eur J Gastroenterol
Hepatol 1995;7:933–937.
25 Folkman J, Szabo S, Stovroff M, McNeil P, Li W, Shing Y: Duodenal ulcer: Discovery of a new
mechanism and development of angiogenic therapy that accelerates healing. Ann Surg 1991;
214:414–427.
26 D’Amore PA: Heparin-endothelial cell interactions. Haemostasis 1900;20(suppl 1):159–165.
27 Flint N, Cove FL, Evans GC: Heparin stimulates the proliferation of intestinal epithelial cells in
primary culture. J Cell Sci 1994;107:401–411.
28 Tiozzo R, Reggiano D, Cingi MR, Bianchini P, Osima B, Calandra S: Effect of heparin-derived frac-
tions on the proliferation and protein synthesis of cells in culture. Thromb Res 1991;62: 177–188.
29 Li Y, Wang WP, Wang HY, Cho CH: Intragastric administration of heparin enhances gastric ulcer
healing through a nitric oxide-dependent mechanism in rats. Eur J Pharmacol 2000;399:205–214.
30 Ihrcke NS, Wrenshall LE, Lindman BJ, Platt JL: Role of heparan sulfate in immune system-blood
vessel interactions. Immunol Today 1993;14:500–505.
31 Bazzoni G, Beltran-Nunez A, Mascellani G, Bianchini P, Dejana E, Del-Maschio A: Effect of
heparin, dermatan sulfate and related oligo-derivatives on human polymorphonuclear leukocyte
functions. J Lab Clin Med 1993;212:268–275.
32 Oyanagui Y, Sato S: Heparin, a potent releasing agent of extracellular superoxide dismutase, sup-
presses ischemic paw oedema in mice. Free Radic Res Commun 1990;9:87–99.
33 Lucio J, D’Brot J, Guo CB, Abraham WM, Lichtenstein LM, Kagey-Sobotka A, Ahmed T:
Immunologic mast cell-mediated responses and histamine release are attenuated by heparin.
J Appl Physiol 1992;73:1093–1101.
34 Salas A, Sans M, Soriano A, Reverter JC, Anderson DC, Pique JM, Panes J: Heparin attenuates
TNF-␣ induced inflammatory response through CD11b-dependent mechanism. Gut 2000;47:88–96.
35 Tangelder GJ, Arfors KE: Inhibition of leukocyte rolling in venules by protamine and sulfated
polysaccharides. Blood 1991;77:1565–1571.
36 Nelson RM, Cecconi O, Roberts WG, Aruffo A, Linhardt RJ, Bevilacqua MP: Heparin oligosac-
charides bind L- and P-selectin and inhibit acute inflammation. Blood 1993;82:3253–3258.
37 Matzner Y, Marx G, Drexler R, Eldor A: The inhibitory effect of heparin and related glycosamino-
glycans on neutrophil chemotaxis. Thromb Haemost 1984;52:134–137.
38 Papa A, Danese S, Gasbarrini A, Gasbarrini G: Potential applications and mechanisms of action
of heparin in inflammatory bowel disease. Aliment Pharmacol Ther 2000;14:1403–1409.
39 Gaffney PR, Doyle CT, Gaffney A, Hogan J, Hayes DP, Annis P: Paradoxical response to heparin
in 10 patients with ulcerative colitis. Am J Gastroenterol 1995;90:220–223.
40 Evans RC, Wong VS, Morris AI, Rhodes JM: Treatment of corticosteroid-resistant ulcerative
colitis with heparin – A report of 16 cases. Aliment Pharmacol Ther 1997;11:1037–1040.
41 Fries W, Pagiaro E, Canova E, Carraro P, Gasparini G, Pomerri F, Martin A, Carlotto C, Mazzon E,
Sturniolo GC, Longo G: The effect of heparin on trinitrobenzene sulphonic acid-induced colitis in
the rat. Aliment Pharmacol Ther 1998;12:229–239.
42 Dobosz M, Mionskowska L, Dobrowolski S, Dymecki D, Makarewicz W, Hrabowska M, Wajda Z:
Is nitric oxide and heparin treatment justified in inflammatory bowel disease? An experimental
study. Scand J Clin Lab Invest 1996;56:657–663.
43 Korzenik JR, Hsu A, Robert ME: Effect of heparin on dextran sulfate sodium-induced colitis. Dig
Dis Sci 1998;43:1800–1805.
44 Dobosz M, Wajda Z, Hac S, Mysliwska J, Mionskowska L, Bryl E, Roszkiewicz A, Mysliwski A:
Heparin and nitric oxide treatment in experimental acute pancreatitis in rats. Forum Genova
1998;8:303–310.
45 Hershkoviz R, Bruck R, Aeed H, Shirin H, Halpern Z: Treatment of concanavalin A-induced
hepatitis in mice with low molecular weight heparin. J Hepatol 1999;31:834–840.
46 Yang QY, Jong SC: Medicinal mushroom in China. Mushr Sci 1989;12:631–643.
47 Wasser SP, Weis AL: Therapeutic effects of substances occurring in higher Basidiomycetes mush-
room: A modern perspective. Crit Rev Immunol 1999;19:65–96.
48 Chihara G: Antitumor and immunological properties of polysaccharides from fungal origin.
Mushr Sci 1978;9:797–814.
49 Chihara G: Recent progress in immunopharmacology and therapeutic effects of polysaccharides.
Dev Biol Stand 1992;77:191–197.
50 Sun XB, Matsumoto T, Yamada H: Effects of a polysaccharide fraction from the roots of
Bupleurum falcatum L. on experimental gastric ulcer models in rats and mice. J Pharm Pharmacol
1991;43:699–704.
51 Yamada H: Structure and pharmacological activity of pectic polysaccharides from the roots of
Bupleurum falcatum L. Folia Pharmacol Jpn 1995;106:229–237.
52 Sun XB, Matsumoto T, Yamada H: Anti-ulcer activity and mode of action of the polysaccharide
fraction from the leaves of Panax ginseng. Planta Med 1992;58:432–435.
53 Cho CH, Mei QB, Shang P, Lee SS, So HL, Guo X, Li Y: Study of the gastrointestinal protective
effects of polysaccharides from Angelica sinensis in rats. Planta Med 2000;66:348–351.
54 Ye YN, Koo MWL, Li Y, Matsui H, Cho CH: Angelica sinensis modulates migration and prolif-
eration of gastric epithelial cells. Life Sci 2001;68:961–968.
55 Ye YN, Liu ESL, Shin VY, Koo MWL, Li N, Wei EQ, Matsui H, Cho CH: A mechanistic study of
proliferation induced by Angelica sinensis in a normal gastric epithelial cell line. Biochem
Pharmacol 2001;61:1439–1448.
56 Ye YN, Li Y, So HL, Cho CCM, Sheng HP, Lee SS, Cho CH: Protective effects of polysaccharide-
enriched fraction from Angelica sinensis on hepatic injury. Life Sci 2001;69:637–646.
Prof. Chi Hin Cho, Department of Pharmacology, Faculty of Medicine,

The University of Hong Kong, 5 Sassoon Road, Hong Kong, SAR (China)
Tel. ⫹852 2819 9252, Fax ⫹852 2817 08959, E-Mail chcho@hkusua.hku.hk

Modulators of Inducible Nitric Oxide

Synthase: Potential Drugs for the
Therapy of Gut Inflammation?
Brendan J.R. Whittle, Maryan Cavicchi, Dominique Lamarque
William Harvey Research Institute, St. Bartholomew’s & The Royal London
School of Medicine, London, UK
The inducible isoform of nitric oxide synthase, variously termed iNOS,

NOS II or NOS-2 (EC 1.14.13.39), is found in many different cell types in asso-
ciation with the gut, including the epithelium and inflammatory cells [1].
Unlike the two constitutively expressed isoforms, eNOS (NOS 3) and nNOS
(NOS 1), this isoform is functionally calcium-independent and can be induced
by cytokines and bacterial lipopolysaccharides (fig. 1). Once expressed, iNOS
has the ability to produce sustained and substantial amounts of nitric oxide
(NO), which can be cytotoxic under appropriate environmental conditions and
provoke tissue injury and inflammation [2]. Moreover, the production of other
cytotoxic moieties by the combination of these levels of NO with reactive oxy-
gen radicals [3] appears to be of major relevance to role of NO in the pathology
of tissue damage and in the inflammatory response in the gut [4].
Macrophages and polymorphonuclear cells were initially thought to be the
main source of NO produced during inflammatory processes [2]. It is now clear
that in the gut, intestinal and colonic epithelial cells are a major site of iNOS
expression, both in models of gut inflammation and in human inflammatory
bowel disease (IBD) [2, 4–7]. Early studies reported the expression of iNOS in
human intestinal epithelial cell lines [8, 9]. In the extensively studied DLD-1
intestinal human epithelial cell line for example, in which the cloned iNOS gene
shows cDNA and amino acid sequences very similar to cloned human iNOS
from other sources, iNOS induction by cytokines with or without bacterial
products, is transcriptionally controlled involving both NF-␬B and a tyrosine
protein kinase [9, 10].
NF-␬B I␬B
LPS
iNOS
gene
Cytokines
mRNA
L-Arginine Citrulline
iNOS
NO
Fig. 1. Expression of the NOS-inducible isoform, iNOS involving the nuclear factor,
NF-␬B, with the production of NO following exposure of epithelial and endothelial cells to
lipopolysaccharide (LPS) or cytokines.
The immunologically evoked induction of NO synthesis in macrophages

appears to be part of a host defence mechanism [2]. In addition, NO can modu-
late the production of cytokines, for example by down-regulating the production
of interleukin-6, while up-regulating the production of tumour necrosis factor-␣
in a macrophage cell line and provoking cytotoxicity to these cells [1, 2]. NO is
produced by neutrophils via iNOS activity [11], being induced by endotoxin and
interferon-␥, an effect potentiated by tumour necrosis factor-␣. The inhibition of
NO induction by glucocorticoids, one of the main and most effective treatments
in IBD, which can be demonstrated in macrophages and neutrophils, may thus
contribute to the therapeutic mechanisms of such agents [2].
Detection of iNOS in IBDs
Animal Models
The potential role of iNOS in gastrointestinal disease obviously requires
that iNOS is expressed and can be detected in the affected tissues. The expres-
sion of the iNOS enzyme activity in colonic tissue has been observed in a range
of models of IBD, such as that provoked by trinitrobenzene sulphonic acid
(TNBS) in the rat [1, 4, 12–15]. In addition to tissue injury, expression of iNOS
has been associated with functional changes such as the dilatation of the colon
observed in the TNBS colitis [16]. In other studies on the inflamed guinea-pig
iNOS, Superoxide and Peroxynitrite in Gut Inflammation 191

ileum produced by local application of TNBS, the levels of nitrite in the lumi-
nal lavage were elevated when measured 7 days following challenge, along with
iNOS gene expression [13, 17].
Elevated plasma of nitrate and nitrite are also observed in the chronic, but
not the acute phases of colonic inflammation induced by bacterial wall poly-
mers in the rat [18]. An increased production of nitrite and nitrate, as well as
iNOS activity, in the HLA-B27 transgenic rat that exhibits spontaneous colitis
has also been observed [19, 20]. In rhesus macaques displaying idiopathic col-
itis, increased nitrogen intermediates, iNOS gene expression and iNOS activity
have been detected [21].
Human Studies
The involvement of NO in IBD in patients also has support from a number
of different studies on the production of NO and metabolites and the presence
of iNOS activity in colonic tissue. It was known from early work that nitrite lev-
els in rectal dialysates are elevated in patients with active ulcerative colitis [22].
Augmented levels of citrulline, the co-product of NOS activity, were also found
in biopsies of inflamed human colon while substantially increased luminal
levels of NO gas was detected directly in the colon of colitic patients [23, 24].
In the first direct study on colonic iNOS enzyme activity in human IBD,
a sixfold increase in calcium-independent iNOS activity was found in colonic
mucosal biopsies from patients with ulcerative colitis [25]. Subsequent studies
with colonic mucosal explants demonstrated elevated nitrite production and
iNOS activity in patients with ulcerative colitis or Crohn’s disease [26].
Increased mRNA for interleukin-8 and for iNOS, along with iNOS protein
expression has also been detected in colonic biopsies from such patients [7, 27,
28]. In other studies, toxic megacolon in patients with IBD was associated with
the appearance of iNOS in the colonic muscularis propria [29]. More recently,
iNOS expression has been detected in colonic tissue, primarily the epithelium,
from patients with collagenous colitis, along with elevated nitrite and nitrate
efflux into colonic perfusates [30].
iNOS and Endothelial and Epithelial Injury
Excessive NO synthesis following the induction of iNOS has been impli-

cated in the splanchnic dysfunction and hypotension following endotoxaemia
[1, 2, 31]. In the gut, administration of endotoxin leads after a 2-hour lag period
to the expression of iNOS activity in the rat jejunal, ileal and colonic tissue
[32]. This iNOS induction is also associated with time-dependent micro-
vascular permeability changes in both small and large intestine, indicative of
Whittle/Cavicchi/Lamarque 192
Albumin
LPS
iNOS
1400W NO Peroxynitrite
SOD
Fig. 2. Expression of iNOS following systemic challenge with LPS leads to microvas-
cular injury throughout the gastrointestinal mucosa. NO in combination with the superoxide
radical forms the peroxynitrite species. The endothelial cell injury that leads to the leakage of
plasma proteins, such as albumin into the interstitium, can be attenuated by either selective
iNOS inhibitors such as 1400W or by SOD.
endothelial cell injury and is a cardinal sign of the inflammatory response

(fig. 2). Pretreatment with dexamethasone prevents both the induction of iNOS
and the vascular permeability changes in intestinal tissue. Furthermore, admin-
istration of isoform non-selective inhibitors such as NG-monomethyl-L-arginine
(L-NMMA) or NG-nitroarginine-methyl ester (L-NAME), at a time of iNOS
expression, reduces the microvascular injury in jejunum ileum and colon
[1, 32].
These findings thus indicate that iNOS expression is associated with
gastrointestinal microvascular injury, reflecting the production of NO. One indi-
rect process by which excessive NO may contribute to microvascular leakage
could involve hyperaemia, augmenting the actions of other pro-inflammatory
mediators which have a direct injurious action on the microvascular endothelium.
This effect of endotoxin is not limited to the microvasculature and indeed,
expression of iNOS is also been observed in epithelial cells isolated from rat
small intestine and colon following endotoxin challenge in vivo [5, 6]. Such
expression of iNOS is associated with a reduction in epithelial cell viability.
Both the induction of iNOS and the cell injury is prevented by in vivo pretreat-
ment with dexamethasone, while administration of L-NAME in vivo can also
prevent the epithelial damage, again clearly demonstrating the dependency of
the cytotoxic effects on NO production [5, 6].

The expression of iNOS in intestinal epithelial cells resulting in injury
contrasts however with the role of NO formed by the constitutive NOS enzymes
in maintaining epithelial integrity [2, 13, 14]. The induction of iNOS in epithe-
lial cells may also contribute to the accumulation of fluid within the intestinal
lumen in response to exposure to bacterial toxins by disrupting the physiologi-
cal processes regulating electrolyte transport in the villus cells, as well as by the
stimulation of crypt cell secretion [1].
NO and Peroxynitrite as Cytotoxic Mediators
NO from iNOS
The suggestion of the potential role of iNOS in gut inflammation has
sparked some controversy, with evidence being produced from animal models
both for and against the cytotoxicity of NO in the pathological process [4, 14].
However, such conflicting data may well be reconciled by consideration of the
prevailing experimental conditions [4, 14, 15]. There is little doubt that the
physiological release of NO formed by eNOS and possibly nNOS, has an
important function in regulating gastrointestinal microvascular tone and vascu-
lar integrity [1]. There is good evidence that NO prevents the activation of
platelets, and NO derived from eNOS is considered to be a key modulator in the
adhesion of neutrophils to the microvasculature [2, 33]. NO has been shown to
modulate mast cell activity, and hence could modulate the release of mediators
in those inflammatory events involving these cells [13]. Experimental studies
also suggest that eNOS is involved in the control of intestinal epithelial barrier
function [34], which would thus be important in the control of toxin and bacte-
rial ingress into the mucosal tissue.
NO donors in low doses can prevent gut damage in a range of models
[1, 13, 35], and although this cannot be construed as necessarily reflecting the
action of enzymatically formed endogenous NO from eNOS, it is most probable
that it is the NO liberated from such agents that exerts the significant protective
actions. Indeed, such findings on the beneficial actions of NO form the phar-
macological basis for the development of the NO-containing non-steroidal anti-
inflammatory agents, which have considerably less gastrointestinal damage
than their parent classical anti-inflammatory drug [35]. It is unlikely that these
latter agents would release sufficient free NO to produce local cytotoxic
actions in therapeutic doses, while the pharmacokinetic disposition of such
NO-containing compounds in terms of NO release may well prevent the for-
mation of subsequent injurious metabolites.
The involvement of NO, formed by the constitutive NOS isoforms, in all
these protective and physiological anti-inflammatory events, does not necessarily
Nitric oxide Superoxide Peroxynitrite Hydroxyl
•
NO ⫹ O2•⫺ ONOO⫺ •
OH
Cytotoxicity
Fig. 3. Interaction of NO and the reactive oxygen radical, superoxide, to form the cyto-
toxic moiety, peroxynitrite, and with subsequent decomposition, production of the hydroxyl
radical.
preclude an involvement of NO, known to have cytotoxic potential, at distinct

stages in the inflammatory process. It is now well established that other media-
tors such as the prostanoids also exhibit both potent protective actions and pro-
inflammatory properties in the gut, a concept exploited in the development of the
anti-inflammatory COX-2 inhibitors. Indeed, the enzymatic source of the NO
may not be the deciding factor in the profile of the subsequent pharmacological
responses, since the site of its release and the stage of the inflammatory response
may be critical, while the local milieu may also be all important in defining the
ultimate actions of NO.
Peroxynitrite
The possible temporal distinction in the beneficial or detrimental role of
NO, especially that produced by iNOS, in the inflammatory process may be
dependent on the cellular environment as interaction of NO with the superoxide
anion can give rise to the peroxynitrite species [3], as depicted in figure 3.
Reactive oxygen radicals have long been implicated in the pathogenesis of
inflammatory diseases of the gut. Peroxynitrite is also cytotoxic, oxidizing a
number of key molecular species including ascorbate, sulphydryls and thiols as
well as producing membrane lipid peroxidation, causing DNA injury and acti-
vating poly(ADP)-ribose synthase [2, 3, 15]. Decomposition of peroxynitrite
can also give rise to the highly reactive hydroxyl radical [3], well known to pro-
duce cell damage and injury, and has been implicated in the vascular disruption
in the gut that follows ischaemia-reperfusion. Thus, a number of cytotoxic moi-
eties can be potentially generated from NO in the inflammatory environment,
where reactive oxygen species are produced by both inflammatory cells and the
involved tissue. Using immunohistochemical detection of nitrotyrosine as an
index of peroxynitrite formation and activity, co-localization with iNOS has

been observed in the colonic tissue from colitic patients, suggesting that indeed
both radicals are formed at these sites and may be involved in the epithelial
inflammation [7, 28]. Interestingly, despite pronounced expression of iNOS
protein in the epithelium from patients with collagenous colitis, nitrotyrosine
co-staining could not be detected, while no overt injury to the colonic tissue was
seen, suggesting that NO production from iNOS per se is not distinctly injuri-
ous to the human colonic epithelium, although it may involved with diarrhoea
observed in this condition [30].
Direct evidence for the damaging actions of peroxynitrite in vivo on the gut
comes from a study in which the colonic instillation of a peroxynitrite-
producing mixture provoked distinct mucosal inflammation in the rat [36]. In
addition, it has been shown that local intra-arterial infusion of high doses of NO
donors that can liberate significant local levels of NO rapidly cause cellular
injury and lipid peroxidation in the rat stomach. Since these effects are abol-
ished by concurrent administration of superoxide dismutase (SOD), the involve-
ment of superoxide, and hence peroxynitrite, is implicated in this mucosal
damage [37, 38]. The source of superoxide in those in vivo studies on acute
injury the stomach is not the circulating neutrophil, as depletion of these cells
does not ameliorate this injury provoked by NO donors. The xanthine oxidase
inhibitor, allopurinol, does however reduce the gastric injury. This could suggest
that this latter enzyme, which is located in many cell types including the
endothelial cell, may liberate superoxide under these conditions, to interact with
NO [37, 38]. Other sources of oxygen metabolites including the inflammatory
cell are, however, likely to be involved in the more chronic inflammatory
conditions such as colitis.
Actions of iNOS Inhibitors
The presence of iNOS in gastrointestinal inflammation provides only

circumstantial evidence of the involvement of NO in the process of inflamma-
tion and tissue injury, its presence perhaps only reflecting a bystander role in
some circumstances. The rigorous appraisal of the pathological role of NO
requires careful pharmacological and gene-manipulative analysis. There is, how-
ever, a significant potential risk for confusion if the experimental conditions are
not critically controlled [31, 32], as demonstrated in early work in a model of
microvascular inflammation provoked by endotoxin [32]. Pretreatment or con-
current administration of a number of isoform non-selective NOS inhibitors with
the endotoxin challenge results in a profound acute microvascular damage and
haemorrhage throughout the gut. This reflects the inhibition of the protective
eNOS, allowing the cytotoxic actions of the pro-inflammatory mediators,
released early following endotoxin challenge. By contrast, administration of
these NOS inhibitors at a time of iNOS expression, caused a marked reduction
in the subsequent microvascular injury [32], demonstrating the cytotoxic actions
resulting from iNOS expression in vivo, and the protective actions of iNOS inhi-
bition. Such time-dependent dual actions were seen not only with the now clas-
sical non-selective NOS inhibitors such as L-NAME or L-NMMA, but also with
aminoguanidine, an agent widely thought of as a selective inhibitor of iNOS, and
L-N-iminoethyl ornithine (L-NIO), an agent with some in vitro inhibitory selec-
tivity, but not with a selective iNOS inhibitor [39]. Thus the dose and timing of
the administration of such non-selective NOS inhibitors may greatly influence
the eventual outcome. Moreover, it has been argued that the route of administra-
tion also may have importance in gastrointestinal inflammation [14], the oral
route perhaps offering a more pronounced or selective effect on the iNOS-
expressing superficial epithelium in the gut.
In a model of ileitis in rats following challenge with TNBS, administration
of L-NAME in the drinking water reduced the inflammatory response, as deter-
mined by myeloperoxidase (MPO) levels, and lowered both protein and nitrite
levels in the lavage fluid [40]. Systemic administration of high doses of L-NAME
via an implanted osmotic minipump tended to augment colonic injury in TNBS
colitis [41]. However, in a time-course study in this model, where pretreatment
with L-NAME in the drinking water augmented the inflammation seen after 72 h,
delay of the administration of L-NAME until the time of expression of iNOS in
this model significantly attenuated the inflammatory response [12]. Treatment of
rats orally with L-NAME, every 24 h after challenge with TNBS, abolished both
the macrophage infiltration into the colonic muscle, and the muscle hyperplasia
associated with the colitis [42].
The chronic colitis produced by local application of sulphydryl-blocking
agents, and the associated NOS activity, was also prevented by the administra-
tion of L-NAME in the drinking water [43]. Furthermore, in a model of
chronic granulomatous colitis, both L-NAME and aminoguanidine adminis-
tered in the drinking water reduced the degree of colonic colitis [44], while
both agents in the drinking water also reduced colonic MPO levels and
mucosal thickening and crypt depth in the colitis that develops in HLA-B27
transgenic rats [19].
In other studies however, treatment of monkeys with spontaneous colitis
for 10 days with L-N-iminoethyl lysine or aminoguanidine, used as selective
iNOS inhibitors, failed to demonstrate any reduction in histological inflamma-
tory score or attenuated the associated diarrhoea [45]. In that study, aminoguani-
dine failed to reduce the index of NOS activity, again suggesting that this is a
poor iNOS inhibitor for such in vivo studies. In recent studies in the colitic
HLA-B27 transgenic rat, subcutaneous administration of L-N-iminoethyl lysine

also failed to prevent the tissue injury, but as the doses used were not fully
selective for the iNOS isoform, the concurrent effects on the other NOS iso-
forms may have influenced the outcome [20].
As the degree of selectivity against the NOS isoforms may dictate the
response, the development of novel highly potent and highly selective iNOS
inhibitors such as 1400W [46], will have importance in defining the role of
iNOS at different stages of the inflammatory response. This agent potently
inhibits the microvascular injury in the small and large intestine associated with
iNOS induction following challenge with endotoxin [46], and its effect is inde-
pendent of the stage at which it is administered. Studies in a model of colitis
has shown that subcutaneous administration of 1400W reduces both the acute
tissue injury and the raised MPO levels seen 24 h after colonic challenge with
TNBS [47]. These findings thus suggest that highly selective inhibitors of iNOS
may have therapeutic benefit, but these actions will need to be explored further
at different phases in more chronic colitis.
Consequences of iNOS Gene Deletion
Other experimental studies in iNOS gene-deleted animals have also cre-

ated some confusion on the role of iNOS in gut inflammation. In a transgenic
mouse with the iNOS gene deleted, the chemically induced acute colonic dam-
age following instillation of acetic acid was not attenuated when compared to
the wild-type mouse, and indeed, appeared to be enhanced [48]. Resolution
of this tissue injury over the following 7 days was also retarded in the iNOS-
deficient animals. Moreover, in the more chronic colitis induced by TNBS, the
acute inflammatory response was exacerbated in iNOS gene-deleted animals,
but with no difference to wild-type animals after 7 days [49]. Such findings
thus could point to a potential beneficial role of iNOS, rather than an injurious
one, but this may reflect the extent of initial injury and the degree of expression
of iNOS after such challenge. However, in marked contrast, in other studies
with iNOS gene-deleted animals, the lethality of the colitis was substantially
reduced, as was the degree of colonic injury and MPO levels in the more
chronic phase of inflammation from 4 to 7 days as compared to the wild-type
animals [50]. It will therefore be essential to characterize the phenotype of
these different groups of animals, since it is perhaps unlikely that gene deletion
will not evoke other developmental changes that may obscure the interpreta-
tion. Furthermore, as the basis for these conflicting findings between the gene-
deleted animals is still enigmatic [14], conclusions either for or against an
involvement of iNOS in this model of colitis in these studies must be drawn
cautiously.
iNOS
LPS
NO Peroxynitrite
SOD
1400W
Fig. 4. The cytotoxic actions of NO, produced by iNOS following its induction by LPS,
on gastrointestinal epithelial cells. The cellular injury is brought about by the production of
NO, which in combination with the superoxide radical forms the cytotoxic moiety, peroxyni-
trite, and can be attenuated by either selective iNOS inhibitors such as 1400W or by SOD.
Bacterial Products and iNOS in Gut Inflammation
It is likely that NO produced from iNOS or its subsequent products, are not
involved in all aspects of the inflammatory response in the gut, and because of
their profile on leukocyte function, would not, for example, be expected directly
to promote cellular infiltration. However, there is a wealth of convincing litera-
ture supporting the cytotoxic actions of NO in many different inflammatory sit-
uations both in and beyond the gastrointestinal tract. This is particularly true
under conditions where bacterial products are formed, such endogenously released
bacterial products including endotoxin and lipopolysaccharide (LPS) being
potent stimuli for iNOS expression.
It has been demonstrated that the organism implicated in the pathogenesis of
peptic ulceration, Helicobacter pylori, generates a factor that can induce iNOS in
macrophage cell lines, although this effect was weak [51], while a water extract
could induce iNOS in duodenal epithelial cells after in vivo challenge [52].
Increased expression of iNOS protein was observed in epithelium, endothelium
and inflammatory cells in gastritis in human subjects resulting from H. pylori
infection [53]. More recently, the purified LPS derived from H. pylori has been
shown to be highly active in vivo following oral or parental administration in
stimulating the expression of iNOS in rat duodenal epithelial cells [54] and is
associated with epithelial cellular injury (fig. 4). Likewise, challenge with the
H. pylori LPS induced microvascular leakage in gastric and duodenal tissue [55].
Both the injurious effects of the LPS on the epithelium and the microvasculature

could be attenuated by inhibition of iNOS using the isoform-selective agent,
1400W [54, 55].
The involvement of NO in the tissue injury associated with infectious
diseases of the gut have also been explored in a range of experimental and clin-
ical settings. In a model of jejunal infection with Trichinella spiralis, oral treat-
ment with L-NAME lowered bacterial count and MPO levels, although effects
on iNOS activity were complex [56]. In patients in the acute phase of shigellosis,
inflammation in rectal biopsies was associated with expression of iNOS in the
epithelium [57]. Enhanced nitrate synthesis as a consequence of increased NO
production has also been observed in patients with infective gastroenteritis [58].
Failure of the intestinal barrier, with the subsequent facilitation of intestinal
bacterial ingress, has been implicated in necrotizing enterocolitis and has been
shown to involve the expression of iNOS [59]. The localized NO synthesis by
iNOS under these conditions may also have a beneficial bacteriostatic function
and reflect initiation of a host-defence mechanism. The importance of such a
general bacterial defence may explain why iNOS knockout mice have shown
susceptibility to infection, in comparison with their normal counterparts [2]. In
contrast however, iNOS gene-deleted mice exhibit an increased resistance to
intestinal injury and bacterial translocation following ischaemia-reperfusion
[60], suggesting that the products of iNOS expression could be involved in the
aggravation of epithelial barrier dysfunction.
There is also involvement of iNOS in the chronic microvascular leakage
and inflammatory injury in the small intestine that follows the administration
of non-steroid anti-inflammatory agents such as indomethacin [61–65]. This
slowly developing enteropathy involves indigenous bacteria, as shown by the
protective actions of antimicrobials, treatment, which reduces the site-specific
iNOS expression in the jejunum [61, 63]. Moreover, polymyxin B, which binds
and inactivates LPS, prevents the induction of iNOS activity and the entero-
pathy following indomethacin administration. This agent, as with the anti-
bacterials, did not directly affect the activity of the iNOS enzyme following
in vitro incubation [61, 63]. Such findings not only confirm that the induction
of iNOS and the subsequent enteropathy is due to indigenous bacteria from
the gut lumen, but also suggests that the LPS fraction of these bacteria is the
stimulus which initiates the expression of iNOS.
It is possible the local release of the LPS from the translocated bacteria in
the intestinal mucosa is also a stimulus for the observed local production of the
cytokines, including tumour necrosis factor-␣, in the tissue following administra-
tion of non-steroid anti-inflammatory agents [65], these also being potent stimuli
for iNOS induction. However, the initiating process that leads to acute epithelial
barrier disruption and bacterial translocation from the gut lumen following
administration of these agents is not clear, but could involve pro-inflammatory
mediators such as platelet-activating factor. The release of such mediators may
follow the initial fast-onset inhibition of prostanoid production by the cyclo-
oxygenase enzymes, the cyclo-oxygenase-1 isoform being found extensively in
intestinal tissue.
Role of Superoxide and Peroxynitrite
As discussed above, the prime mechanism by which NO can mediate this

microvascular injury and mucosal damage may involve its interaction with
superoxide radicals, to generate the more reactive peroxynitrite radical [1, 3, 14,
15, 66]. Studies with the superoxide scavenger SOD, conjugated with polyeth-
ylene glycol (PEG) to reduce clearance from the circulation, has demonstrated
that it can prevent tissue injury from local infusion of NO donors [37, 38].
Moreover, SOD-PEG can reduce the iNOS-associated injury to duodenal
epithelial cells following challenge with the LPS from H. pylori [54].
Other studies have also shown that the damage induced by intravenous
challenge with Escherichia coli LPS in rat small intestinal epithelial cells can
be attenuated by a SOD mimetic [67]. Expression of iNOS in rat intestinal
epithelial cells is associated with the induction of apoptosis in these cells and is
reduced by SOD-PEG treatment [54]. Studies have demonstrated that the
iNOS-associated microvascular leakage induced by indomethacin in the
jejunum was significantly reduced by pretreatment with SOD-PEG [63]. In
addition, recent work has reported that administration of a SOD mimetic can
reduce the colonic injury, MPO elevation and nitrotyrosine staining over 4 days
in the TNBS model of colitis [68], implicating both NO and superoxide, along
with peroxynitrite as pathogenic mediators in this gut inflammation.
Apart from inhibiting iNOS or scavenging superoxide, prevention of per-
oxynitrite formation or acceleration of its decomposition provides a novel
approach to reducing NO-dependant cytotoxicity [15, 66]. Thus, agents that act
as peroxynitrite decomposition catalysts reduce the intestinal microvascular
injury, lipid peroxidation and epithelial cytotoxicity following LPS challenge
[67]. Studies utilizing human intestinal epithelial cells indicate that peroxynitrite-
induced apoptosis and cell death is reduced in the presence of mesalamine, an
antioxidant salicylate preparation used in colitis, which appears to accelerate
the decomposition of peroxynitrite [69]. In other work, the agent mercapto-
ethylguanidine, that can both inhibit iNOS activity and act as a peroxynitrite
scavenger, reduced the inflammatory indices of TNBS-induced colitis [70].
Thus, the present literature suggests that NO alone, produced from iNOS,
does not itself provoke the tissue injury in models of gut injury, such as that
induced in the intestinal microvasculature and epithelial cells by LPS challenge,

the jejunal injury produced by indomethacin or the colonic injury provoked by
TNBS, in the absence of the superoxide moiety. Whereas it is feasible that NO
and superoxide act synergistically to bring about cellular injury, the experi-
mental findings obtained so far make it is more likely that they combine to form
the injurious species, peroxynitrite.
Interactions with Endogenous Antioxidant Systems:

Heme Oxygenase-1
HO-1 in Inflammation
The NO pathway may also interact with other systems that may be involved
in the modulation of the inflammatory response. One such system is heme oxy-
genase-1 (HO-1; EC 1.14.99.3), a microsomal-inducible enzyme, which con-
verts heme into biliverdin, carbon monoxide (CO) and free ferrous iron, the
biliverdin being subsequently reduced to bilirubin [71]. HO-1 is considered to
provide a potent antioxidant system leading to removal of heme, a promoter of
lipid peroxidation and reactive oxygen intermediates formation. In addition, bile
pigments resulting from HO-1 activity possess antioxidant properties and are
anticomplement agents. The associated induction of ferritin also provides
antioxidant activity and, because of its ability to sequestrate free iron, limits the
subsequent production of reactive oxygen intermediates via the Fenton reaction.
This 32-kDa heat-shock protein (HSP 32) can be expressed in numerous cell
types following a number of different stimuli including endotoxin or cytokine
stimulation, heavy metals, NO donors or heme.
HO-1 has been found to modulate inflammation in vivo in a number of
models, with an early demonstration of it role in a murine model of pleurisy [72].
Induction of HO-1 can also attenuate venular leukocyte adhesion provoked by
pro-oxidant stimuli or inhibition of constitutive NO synthesis [73]. Moreover, a
recent report indicates that HO-1 can modulate experimental colitis in rats [74].
In that study, administration of tin mesoporphyrin, which inhibits HO-1, potenti-
ated the colonic injury induced by TNBS over 3 days, while increasing the
production of reactive oxidant species and iNOS activity [74].
Interactions between HO-1 and iNOS

Interactions between the two inducible enzyme systems HO-1 and iNOS
may be of importance in understanding the pathophysiology of inflammatory
gut diseases [75, 76]. In a recent study, three different classes of compounds
that are known to induce HO-1, namely heme, NO donors and heavy metals
such as bismuth, have been shown to decrease cytokine-induced iNOS activity
or expression in the human intestinal epithelial cell line DLD-1. The effect of
endogenous HO-1 inducers on iNOS expression or activity provides another
mechanism in addition to antioxidant activity by which the HO-1 system could
be beneficial in inflammatory conditions. The NO donor’s sodium nitroprusside
or S-nitroso-acetylpenicillamine inhibited iNOS transcription but were also
potent inhibitors of iNOS enzyme activity in these human epithelial cells [76],
and such feedback inhibition may itself reflect a self-regulating process limit-
ing the actions of iNOS.
Conclusions
It is as yet premature to dismiss NO produced by iNOS, as having any role

in the complex inflammatory processes that can occur in the gut. The involve-
ment of NO may well differ with the nature of the insult, the tissue and the envi-
ronment involved, as well as the local response and interaction with other
mediators present at the inflammatory site. As with many other inflammatory
mediators, it is entirely possible that a low level of expression of iNOS will
reflect a positive host-defence response to challenge in the gut, both at a cellu-
lar level and towards hostile challenge such as invading bacteria, but that exag-
gerated or uncontrolled expression of iNOS itself becomes detrimental. The
apparent paradox of conflicting beneficial and damaging roles of NO, may also
be reconciled by consideration of the time-course of the inflammatory response
and the apparent requirement for superoxide to be present for the injurious
actions of NO to be elicited. Under conditions and environments where super-
oxide or other oxygen moieties are not present in sufficient quantities to allow
cytotoxic products to be formed, it is possible that NO, even produced in high
rates by iNOS, will not be detrimental, and could, in contrast, be helpful to the
resolution of the tissue damage.
The expression of endogenous antioxidant systems may modulate the
activity of NO, experimental findings pointing to the potential for interplay
between the iNOS- and HO-1-inducible enzyme systems, for example. A pri-
mary response of intestinal epithelial cells in IBD may involve iNOS induction
following pro-inflammatory cytokine release, with epithelial injury due to NO
in combination with reactive oxygen intermediates to form peroxynitrite.
Subsequent feedback HO-1 induction by high levels of NO would permit the
deactivation and clearance of reactive oxygen intermediates, as well as a
decrease in NO production through iNOS inhibition, thus limiting the potential
cytotoxicity of NO.
Selective inhibitors of iNOS that have appropriate pharmacokinetic char-
acteristics may be of clinical benefit in gut inflammation, although any positive
benefit may be attenuated at certain stages of the inflammatory process if NO

produced from iNOS has a beneficial action in promoting the resolution of
tissue injury at that site. Thus, the consequences of careful posology and tim-
ing of the administration of iNOS inhibitors, perhaps early or in the acute
inflammatory phase such as in the relapse of colitis, where iNOS is expressed
and NO is released with other mediators, will have to be explored.
Agents that prevent the formation or remove peroxynitrite are also likely
to exhibit a positive profile in such conditions. Other approaches such as scav-
enging oxygen metabolites either by antioxidant agents or SOD mimics could
also be of clinical value, either alone or in combination with iNOS inhibitors,
especially in the early stages of inflammation. Therapeutic intervention aimed
at inducing endogenous antioxidant systems such as HO-1 are also worthy of
pharmaceutical consideration. Thus, knowledge of these interacting pathways
of NO, superoxide and HO-1 that regulate the production of peroxynitrite and
other cytoxic nitrogen moieties may thus provide new therapeutic approaches to
the control of inflammatory conditions in the gut.
References
1 Whittle BJR: Nitric oxide in gastrointestinal physiology and pathology; in Johnson LR (ed):
Physiology of the Gastrointestinal Tract, ed 3. New York, Raven Press, 1994, pp 267–294.
2 Moilanen E, Whittle BJR, Moncada S: Nitric oxide as a factor in inflammation; in Gain JI,
Snyderman R (eds): Inflammation: Basic Principles and Clinical Correlates. Philadelphia,
Lippincott/Williams & Wilkins, 1999, pp 787–800.
3 Beckman JS, Beckman TW, Chen J, Marshall PA, Freeman BA: Apparent hydroxyl radical pro-
duction by peroxynitrite: Implications for endothelial injury from nitric oxide and superoxide.
4 Whittle BJR: Nitric Oxide – A mediator of inflammation or mucosal defence. Eur J Gastroenterol
Hepatol 1997;9:1026–1032.
5 Tepperman BL, Brown JF, Whittle BJR: Nitric oxide synthase induction and intestinal epithelial
cell viability in rats. Am J Physiol 1993;265:G214–G218.
6 Tepperman BL, Brown JF, Korolkiewicz R, Whittle BJR: Nitric oxide synthase activity, viability
and cyclic GMP levels in rat colonic epithelial cells: Effect of endotoxin challenge. J Pharmacol
Exp Ther 1994;271:1477–1482.
7 Singer II, Kawka DW, Scott S, Weidner JR, Mumford RA, Riehl TE, Stenson WF: Expression of
inducible nitric oxide synthase and nitrotyrosine in colonic epithelium in inflammatory bowel
disease. Gastroenterology 1996;111:871–885.
8 Kolios G, Rooney N, Murphy CT, Robertson DAF, Westwick J: Expression of inducible nitric
oxide synthase activity in human colon epithelial cells: Modulation by T lymphocyte-derived
cytokines. Gut 1998;43:56–63.
9 Salzman AL, Eaves-Pyles T, Linn SC, Denenberg AG, Szabo, C: Bacterial induction of inducible
nitric oxide synthase in cultured human intestinal epithelial cells. Gastroenterology 1998;114:
93–102.
10 Caviccchi M, Whittle BJR: Potentiation of cytokine-induced iNOS expression in the human
intestinal epithelial cell line, DLD-1, by cyclic AMP. Gut 1999;45:367–374.
11 McCall TB, Boughton-Smith NK, Palmer RMJ, Whittle BJR, Moncada S: Synthesis of nitric
oxide from L-arginine by neutrophils. Release and interaction with superoxide anion. Biochem J
1989;261:293–296.
12 Kiss J, Lamarque D, Delchier JC, Whittle BJR: Time-dependent actions of nitric oxide synthase
inhibition on colonic inflammation induced by trinitrobenzene sulphonic acid in rats. Eur J
Pharmacol 1997;336:219–234.
13 Wallace JL, Miller MJ: Nitric oxide in mucosal defense: A little goes a long way. Gastroenterology
2000;119:512–20.
14 Kubes P: Inducible nitric oxide synthase: A little bit of good in all of us. Gut 2000;47:6–9.
15 McCafferty DM: Peroxynitrite and inflammatory bowel disease. Gut 2000;46:436–439.
16 Mourelle M, Vilaseca J, Guarner F, Salas A, Malagelada JR: Toxic dilation of colon in a rat model
of colitis is linked to an inducible form of nitric oxide synthase. Am J Physiol 1996;270:
G425–G430.
17 Miller MJ, Thompson JH, Zhang XJ, Sadowska-Krowicka H, Kakkis JL, Munshi UK, Sandoval
M, Rossi JL, Eloby-Childress S, Beckman JS: Role of inducible nitric oxide synthase expression
and peroxynitrite formation in guinea pig ileitis. Gastroenterology 1995;109:1475–1483.
18 Grisham MB, Ware K, Gilelland HE Jr, Gilelland LB, Abell CL, Yamada T: Neutrophil-mediated
nitrosamine formation: Role of nitric oxide in rats. Gastroenterology 1992;103:1260–1266.
19 Aiko S, Fuseler J, Grisham MB: Effects of nitric oxide synthase inhibition or sulfasalazine on the
spontaneous colitis observed in HLA-B27 transgenic rats. J Pharmacol Exp Ther 1998;28:722–727.
20 Blanchard HS, Dernis-Labous E, Lamarque D, Nhieu JT, Szepes Z, Flejou JF, Wollman E, Whittle
BJR, Breban M: Inducible nitric oxide synthase attenuates chronic colitis in human histocompat-
ibility antigen HLA-B27/human ␤2-microglobulin transgenic rats. Eur Cytokine Netw 2001;12:
111–118.
21 Ribbons KA, Zhang XJ, Thompson JH, Greenberg SS, Moore WM, Kornmeier CM, Currie MG,
Lerche N, Blanchard J, Clark DA: Potential role of nitric oxide in a model of chronic colitis in rhe-
sus macaques. Gastroenterology 1995;108:705–711.
22 Roediger WE, Lawson MJ, Nance SH, Radcliffe BC: Detectable colonic nitrite levels in inflam-
matory bowel disease – Mucosal or bacterial malfunction? Digestion 1986;35:199–204.
23 Middleton SJ, Shorthouse M, Hunter JO: Increased nitric oxide synthesis in ulcerative colitis.
Lancet 1993;341:465–466.
24 Lundberg JO, Hellstrom PM, Lundberg JM, Alving K: Greatly increased luminal nitric oxide in
ulcerative colitis. Lancet 1994;344:1673–1674.
25 Boughton-Smith NK, Evans SM, Hawkey CJ, Cole AT, Balsitis M, Whittle BJR, Moncada S:
Differential changes in nitric oxide synthase activity in ulcerative colitis and Crohn’s disease.
Lancet 1993;342:338–340.
26 Rachmilewitz D, Stamler JS, Bachwich D, Karmeli F, Ackerman Z, Podolsky DK: Enhanced
colonic nitric oxide generation and nitric oxide synthase activity in ulcerative colitis and Crohn’s
disease. Gut 1995;36:718–723.
27 McLaughlan JM, Seth R, Vautier G, Robins RA, Scott BB, Hawkey CJ, Jenkins D: Interleukin-8
and inducible nitric oxide synthase mRNA levels in inflammatory bowel disease at first presenta-
tion. J Pathol 1997;181:87–92.
28 Kimura H, Hokari R, Miura S, Shigematsu T, Hirokawa M, Akiba Y, Kurose I, Higuchi H,
Fujimori H, Tsuzuki Y, Serizawa H, Ishii H: Increased expression of an inducible isoform of nitric
oxide synthase and the formation of peroxynitrite in colonic mucosa of patients with active ulcer-
ative colitis. Gut 1998;42:180–187.
29 Mourelle M, Casellas F, Guarner F, Salas A, Riveros-Moreno V, Moncada S, Malagelada JR:
Induction of nitric oxide synthase in colonic smooth muscle from patients with toxic megacolon.
30 Perner A, Andresen L, Normark M, Fischer-Hansen B, Sorensen S, Eugen-Olsen J, Rask-Madsen J:
Expression of nitric oxide synthases and effects of L-arginine and L-NMMA on nitric oxide pro-
duction and fluid transport in collagenous colitis. Gut 2001;49:387–394.
31 Kilbourn RG, Szabo C, Traber DL: Beneficial versus detrimental effects of nitric oxide synthase
inhibitors in circulatory shock: Lessons learnt from experimental and clinical studies. Shock
1997;7:235–246.
32 Laszlo F, Whittle BJR, Moncada S: Time-dependent enhancement or inhibition of endotoxin-
induced vascular injury in rat intestine by nitric oxide synthase inhibitors. Br J Pharmacol
1994;111:1309–1315.

33 Kubes P, Suzuki M, Granger DN: Nitric oxide: An endogenous modulator of leukocyte adhesion.
34 Alican I, Kubes P: A critical role for nitric oxide in intestinal barrier function and dysfunction.
Am J Physiol 1996;270:G225–G237.
35 Muscara MN, Wallace JL: Nitric oxide. V. Therapeutic potential of nitric oxide donors and
inhibitors. Am J Physiol 1999;276:G1313–G1316.
36 Rachmilewitz D, Stamler JS, Karmeli F, Mullins ME, Singel DJ, Loscalzo J, Xavier RJ, Podolsky DK:
Peroxynitrite-induced rat colitis – A new model of colonic inflammation. Gastroenterology 1993;
105:1681–1688.
37 Lamarque D, Whittle BJR: Role of oxygen-derived metabolites in the rat gastric mucosal injury
induced by nitric oxide donors. Eur J Pharmacol 1995;277:187–194.
38 Lamarque D, Whittle BJR: Involvement of peroxynitrite in the lipid peroxidation induced by nitric
oxide in rat gastric mucosa. Eur J Pharmacol 1996;313:R5–R7
39 Whittle BJR, Laszlo F: Actions of isoform-selective and non-selective nitric oxide synthase
inhibitors on endotoxin-induced vascular leakage in rat colon. Eur J Pharmacol 1997;334:
99–102.
40 Miller MJS, Sadowska-Krowicka H, Chotinaruemol S, Kakkis JL, Clark DA: Amelioration of
chronic ileitis by nitric oxide synthase inhibition. J Pharmacol Exp Ther 1993;264:11–16.
41 Pfeiffer CJ, Qiu BS: Effects of chronic nitric oxide synthase inhibition on TNB-induced colitis in
rats. J Pharm Pharmacol 1995;47:827–832.
42 Hogaboam CM, Jacobson K, Collins SM, Blennerhassett MG: The selective beneficial effects of
nitric oxide inhibition in experimental colitis. Am J Physiol 1995;268:G673–G684.
43 Rachmilewitz D, Karmeli F, Okon E: Sulfhydryl blocker-induced rat colonic inflammation is ame-
liorated by inhibition of nitric oxide synthase. Gastroenterology 1995;109:98–106.
44 Grisham MB, Specian RD, Zimmerman TE: Effects of nitric oxide synthase inhibition on the
pathophysiology observed in a model of chronic granulomatous colitis. J Pharmacol Exp Ther
1994;271:1114–1121.
45 Ribbons KA, Currie MG, Connor JR, Manning PT, Allen PC, Didier P, Ratterree MS, Clark DA,
Miller MJ: The effect of inhibitors of inducible nitric oxide synthase on chronic colitis in the rhe-
sus monkey. J Pharmacol Exp Ther 1997;280:1008–1015.
46 Garvey EP, Oplinger JA, Furfine ES, Kiff RJ, Laszlo F, Whittle BJR, Knowles RG: 1400W is a
slow, tight binding, and highly selective inhibitor of inducible nitric-oxide synthase in vitro and
in vivo. J Biol Chem 1997;272:4959–4963.
47 Kankuri E, Vaali K, Knowles RG, Lahde M, Korpela R, Vapaatalo H, Moilanen E: Suppression
of acute experimental colitis by a highly selective inducible nitric-oxide synthase inhibitor,
N-[3-(aminomethyl)benzyl]acetamidine. J Pharmacol Exp Ther 2001;298:1128–1132.
48 McCafferty DM, Mudgett JS, Swain MG, Kubes P: Inducible nitric oxide synthase plays a criti-
cal role in resolving intestinal inflammation. Gastroenterology 1997;112:1022–1027.
49 McCafferty DM, Miampamba M, Sihota E, Sharkey KA, Kubes P: Role of inducible nitric oxide
synthase in trinitrobenzene sulphonic acid induced colitis in mice. Gut 1999;45:864–873.
50 Zingarelli B, Szabo C, Salzman AL: Reduced oxidative and nitrosative damage in murine experi-
mental colitis in the absence of inducible nitric oxide synthase. Gut 1999;45:199–209.
51 Wilson KT, Ramanujam KS, Mobley HL, Musselman RF, James SP, Meltzer SJ: Helicobacter
pylori stimulates inducible nitric oxide synthase expression and activity in a murine macrophage
cell line. Gastroenterology 1996;111:1524–1533.
52 Lamarque D, Kiss J, Tankovic J, Flejou JF, Delchier JC, Whittle BJR: Induction of nitric oxide syn-
thase in vivo and cell rat injury duodenal epithelium by a water-soluble extract Helicobacter
pylori. Br J Pharmacol 1998;123:1073–1078.
53 Fu S, Ramanujam KS, Wong A, Fantry GT, Drachenberg CB, James SP, Meltzer SJ, Wilson KT:
Increased expression and cellular localization of inducible nitric oxide synthase and cyclooxyge-
nase 2 in Helicobacter pylori gastritis. Gastroenterology 1999;116:1319–1329.
54 Lamarque D, Moran AP, Szepes Z, Delchier JC, Whittle BJR: Cytotoxicity associated with
induction of nitric oxide synthase in rat duodenal epithelial cells in vivo by lipopolysaccharide
of Helicobacter pylori: Inhibition by superoxide dismutase. Br J Pharmacol 2000;130:
1531–1538.
55 Kiss J, Lamarque D, Moran AP, Pozsar J, Morschl E, Laszlo F, Whittle BJR: Helicobacter pylori
lipopolysaccharide-provoked injury to rat gastroduodenal microvasculature involves inducible
nitric oxide synthase. Eur J Pharmacol 2001;420:175–179.
56 Hogaboam CM, Collins SM, Blennerhassett MG: Effects of oral L-NAME during Trichinella
spiralis infection in rats. Am J Physiol 1996;271:G338–G346.
57 Islam D, Veress B, Bardham PK, Lindberg AA, Christensson B: In situ characterization of inflam-
matory responses in the rectal mucosae of patients with shigellosis. Infect Immun 1997;65:
739–749.
58 Forte P, Dykhuizen RS, Milne E, McKenzie A, Smith CC, Benjamin N: Nitric oxide synthesis in
patients with infective gastroenteritis. Gut 1999;45:355–361.
59 Ford H, Watkins S, Reblock K, Rowe M: The role of inflammatory cytokines and nitric oxide in
the pathogenesis of necrotizing enterocolitis. J Pediatr Surg 1997;32:275–282.
60 Suzuki Y, Deitch EA, Mishima S, Lu Q, Xu D: Inducible nitric oxide synthase gene knockout mice
have increased resistance to gut injury and bacterial translocation after an intestinal ischemia-
reperfusion injury. Crit Care Med 2000;28:3692–3696.
61 Whittle BJR, Laszlo F, Evans SM, Moncada S: Induction of nitric oxide synthase and microvas-
cular injury in the rat jejunum provoked by indomethacin. Br J Pharmacol 1995;116:2286–2290.
62 Evans SM, Laszlo F, Whittle BJR: Site-specific lesion formation, inflammation and inducible
nitric oxide synthase expression by indomethacin in the rat intestine. Eur J Pharmacol 2000;388:
281–285.
63 Evans SM, Whittle BJR: Interactive roles of superoxide and inducible nitric oxide synthase in rat
intestinal injury provoked by non-steroidal anti-inflammatory drugs. Eur J Pharmacol 2001;429:
287–296.
64 Konaka A, Nishijima M, Tanaka A, Kunikata T, Kato S, Takeuchi K: Nitric oxide, superoxide
radicals and mast cells in pathogenesis of indomethacin-induced small intestinal lesions in rats.
J Physiol Pharmacol 1999;50:25–38.
65 Bertrand V, Guimbaud R, Sogni P, Lamrani A, Mauprivez C, Giroud JP, Couturier D, Chauvelot-
Moachon L, Chaussad S: Role of tumour necrosis factor-␣ and inducible nitric oxide synthase in
the prevention of nitroflurbiprofen small intestine toxicity. Eur J Pharmacol 1998;356:245–253.
66 Cuzzocrea S, Riley DP, Caputi AP, Salvemini D: Antioxidant therapy: A new pharmacological
approach in shock, inflammation and ischemia/reperfusion injury. Pharmacol Rev 2001;53:
135–159.
67 Salvemini D, Riley DP, Lennon PJ, Wang ZQ, Currie MG, Macarthur H, Misko TP: Protective
effects of a superoxide dismutase mimetic and peroxynitrite decomposition catalysts in endotoxin-
induced intestinal damage. Br J Pharmacol 1999;127:685–692.
68 Cuzzocrea S, Mazzon E, Dugo L, Caputi AP, Riley DP, Salvemini D: Protective effects of
M40403, a superoxide dismutase mimetic, in a rodent model of colitis. Eur J Pharmacol 2001;
432:79–89.
69 Sandoval M, Liu X, Mannick EE, Clark DA, Miller MJ: Peroxynitrite-induced apoptosis in
human intestinal epithelial cells is attenuated by mesalamine. Gastroenterology 1997;113:
1480–1488.
70 Zingarelli B, Cuzzocrea S, Szabo C, Salzman AL: Mercaptoethylguanidine, a combined inhibitor
of nitric oxide synthase and peroxynitrite scavenger, reduces trinitrobenzene sulfonic acid-induced
colonic damage in rats. J Pharmacol Exp Ther 1998;287:1048–1055.
71 Maines MD: The heme oxygenase system: A regulator of second messenger gases. Annu Rev
Pharmacol Toxicol 1997;37:517–554.
72 Willis D, Moore AR, Frederick R, Willoughby DA: Heme oxygenase: A novel target for the mod-
ulation of the inflammatory response. Nat Med 1996;2:87–90.
73 Hayashi S, Takamiya R, Yamaguchi T, Matsumoto K, Tojo SJ, Tamatani T, Kitajima M, Makino N,
Ishimura Y, Suematsu M: Induction of heme oxygenase-1 suppresses venular leukocyte adhesion
elicited by oxidative stress: Role of bilirubin generated by the enzyme. Circ Res 1999;85:
663–671.
74 Wang WP, Guo X, Koo MW, Wong BC, Lam SK, Ye YN, Cho CH: Protective role of heme oxy-
genase-1 on trinitrobenzene sulfonic acid-induced colitis in rats. Am J Physiol 2001;281:
G586–G594.

75 Guo X, Shin VY, Cho CH: Modulation of heme oxygenase in tissue injury and its implication in
protection against gastrointestinal diseases. Life Sci 2001;69:3113–3119.
76 Cavicchi M, Gibbs L, Whittle BJR: Inhibition of inducible nitric oxide synthase in the human
intestinal epithelial cell line, DLD-1, by the inducers of heme oxygenase-1, bismuth salts, heme
and nitric oxide donors. Gut 2000;47:771–778.
Prof. B.J.R. Whittle, William Harvey Research Institute,

St. Bartholomew’s & The Royal London School of Medicine,
Charterhouse Square, London EC1M 6BQ (UK)
Tel. ⫹44 207 882 6176, Fax ⫹44 207 882 6177, E-Mail b.j.whittle@qmul.ac.uk
Peptide and Gene Therapy with

Angiogenic Growth Factors bFGF,
PDGF or VEGF in Gastrointestinal
Ulcers in Rats
Tetyana Khomenko, Xiaoming Deng, Hideki Ishikawa,
Zsuzsa Sandor, Sandor Szabo
Pathology & Laboratory Medicine Service, Diagnostic & Molecular Medicine
Health Care Group, VA Medical Center, Long Beach, Calif. and Departments of
Pathology & Pharmacology, University of California, Irvine, Calif., USA
Recent advances in basic sciences such as molecular physiology, pharma-

cology and pathology, as well as cell and molecular biology, provide new
methods and yield new results in the study complex and multifactorial disorders
such as ulcer disease of the gastrointestinal tract. The etiologic prevention and
treatment are the primary goal in medical practice, but it is rarely possible with
gastrointestinal ulceration. Ulcerative and inflammatory disease in the upper
and lower gastrointestinal tract were actually treated empirically and indirectly.
Until recently the most widely used approach to gastroduodenal ulceration
has been the suppression of gastric acid secretion or the stimulation of mucus
and bicarbonate secretion. The recent emphasis on Helicobacter pylori also
involves only one of the many etiological agents in gastroduodenal ulceration.
H. pylori might be one of the main reasons for poor healing and high recurrence
of gastroduodenal ulcers, but its primary etiological role has been proven only
in the initiation of acute and chronic gastritis.
The treatment of inflammatory bowel disease (IBD) has also been indirect,
that is, suppression of the inflammation that may only be an aggravating response
to the initial cell and tissue damage. Compounds with direct healing effect, e.g.,
growth factors, may offer a common strategy for ulcer healing throughout the
gastrointestinal tract. These agents act locally and reduce rapid and good quality
healing. Growth factors stimulate virtually all the cellular responses of ulcer
healing, e.g., angiogenesis, granulation tissue, and re-epithelialization.
The new developments in the molecular underpinning of the angiogenic
process in reproduction, development and tissue repair have implications for
several diseases. Angiogenesis is the generation of new blood vessels from an
existing vascular bed. The process involves extensive interaction between
several cells and molecules. Endothelial cells potentially express all the biologic
signals to instill a vascular network by basement membrane dislocation, migra-
tion, proliferation and generation of capillary tubes. The switch to angiogenesis
involves a change in the local equilibrium between positive and negative regu-
lators of microvessels. The proliferation of endothelial cells and tube formation
are crucial elements in granulation tissue production. The formation of granu-
lation tissue, i.e., angiogenesis followed by proliferation of fibroblasts and
deposition of collagen, on the other hand, is a rate-limiting step in the repair of
major tissue injury (e.g., after the loss of cardiac or stomach muscle). Only in
certain organs such as liver, adrenal and renal cortex, does the regeneration
involve the proliferation of original parenchymal cells that rapidly replace the
lost tissue [1]. The healing process needs the granulation tissue, which forms
the basis for proliferation and migration of epithelial cells. The migrating cells
cannot grow over necrotic tissue, unless it is gradually replaced by angiogenesis-
dependent granulation tissue. Thus, stimulation of only epithelial cell pro-
liferation is counterproductive in the healing of internal and external wounds,
unless it is accompanied by the expression of a loose or solid granulation
tissue, which provides the basis and physical framework for the migration of
epithelial cells to complete the healing process [2]. The healing of deep ulcers
that reach or penetrate the muscularis propria almost always requires initial
angiogenesis and proliferation of other elements of granulation tissue [3].
The main endogenous regulators of cell regeneration and proliferation are
polyamines, cytokines or growth factors and certain hormones (e.g., steroids).
Various angiogenic peptides have been discovered with different activities.
Vascular endothelial cells can be stimulated by peptides to proliferate, migrate
or form capillary tubes or a combination of these. Furthermore, angiogenic
peptides may stimulate vascular endothelial cells indirectly to release endothe-
lial cell growth factors. Synergism between angiogenic factors has also been
demonstrated, e.g., a potent synergism between vascular endothelial growth
factor (VEGF) and basic fibroblast growth factor (bFGF) in the induction of
angiogenesis [4]. Because cell proliferation and granulation production are
stimulated by growth factors, these peptides represent a new molecular tool to
gain insights into the mechanisms of ulcer healing. The angiogenesis-targeted
tissue repair is thus a novel cellular and molecular approach to ulcer healing,
which until recently has been stimulated only indirectly [5, 6]. Namely, by
reduction or elimination of endogenous and exogenous aggressive factors such
as acid, pepsin, H. pylori, non-steroidal anti-inflammatory drugs (NSAID)
Khomenko/Deng/Ishikawa/Sandor/Szabo 210
Table 1. Comparative effect of growth factors on gastroduodenal secretion, cell prolif-
eration and angiogenesis
Peptides Gs-acid Du-bicarb. Epith.↑ Fibrobl.↑ Angiogenesis
EGF ↓ ↑ / / /

bFGF /↑
PDGF
VEGF NT
Mild, strong, no effect, NT not tested.
ulcers have been left alone to heal themselves. Our approach, on the other hand,
has always been to stimulate directly ulcer healing by either growth factor pep-
tides or related genes, which preferentially affect angiogenesis and granulation
tissue production [6–8].
We thus review here our initial and recent pharmacologic experiments with
administration of peptides or genetic vectors of VEGF, PDGF (platelet-derived
growth factor) or bFGF in the healing of gastrointestinal ulcers.
Ulcer Healing with Administration of Angiogenic Peptides
Growth factors, such as epidermal growth (EGF), bFGF, PDGF and, more
recently, VEGF, have been used extensively to heal experimental gastric, duo-
denal and colonic ulcers in animal models [7, 9–14]. Among these, only EGF
has an effect on gastroduodenal secretion while the other peptides stimulate
virtually all the cellular elements of ulcer healing (table 1).
Table 1 demonstrates that the most consistent common effect of these
growth factors in ulcer healing is the stimulation of cell proliferation, especially
angiogenesis. The molar potency of bFGF, PDGF and VEGF is 2–7 million
times higher than that of cimetidine in the healing of cysteamine-induced
chronic duodenal ulcer in rats [7, 12]. Since VEGF apparently has no relevant
effect other than stimulation of angiogenesis, it is probably safe to conclude
that enhancement of angiogenesis which is accompanied by granulation tissue
production is the most important cellular event leading to ulcer healing.
Apparently spontaneous proliferation and migration of epithelial cells over
dense granulation tissue then complete the healing process.
In most countries of the world, duodenal ulcer is the most prevalent form
of “peptic ulcers”. For that reason, we first used animal models of duodenal
Angiogenic Factors in Gastrointestinal Ulcer Healing 211

ulcers to test the hypothesis that angiogenic peptides like bFGF might acceler-
ate the healing of chronic duodenal ulcers. Our laboratory was the first to
investigate the effects of bFGF, PDGF and most recently VEGF and their
derivatives on ulcer healing. In these studies we focused on investigating the
role of angiogenesis and its modulation in the natural healing of gastrointestinal
ulceration.
bFGF
Chronic Duodenal Ulcer. We used the modified model of chronic duode-
nal ulcer induced by cysteamine [7, 9, 15]. These studies demonstrated that per os
treatment with either naturally occurring bFGF or its acid-resistant mutein, in
which the second and third cysteines are replaced by serine residues [16],
bFGF-CS23 (100 ng/100 g twice daily) accelerated the healing of cysteamine-
induced chronic duodenal ulcer [7, 12]. This effect was about 7 million times
more potent on a molar basis than the oral administration of cimetidine (10 mg/
100 g, twice a day for 3 weeks), however only the treatment with acid-resistant
mutein bFGF-CS23 (100 ng/100 g twice a day for 3 weeks) significantly
decreased both size (p 0.01 versus control) and prevalence (p 0.05 versus
control) of the remaining chronic ulcers.
Secretory studies showed that a single dose of bFGF had no effect on
gastric output of acid and pepsin, whereas daily treatment for 2 or 3 weeks
resulted in enhanced outputs of both products [7, 12, 17].
To test both the direct and indirect mode of action of antiulcer drugs, we
compared cimetidine (10 mg/100 g) and bFGF (50 ng/100 g) alone and in com-
bination in the healing of cysteamine-induced chronic duodenal ulcers. We
found that the ulcer size in the combination group was significantly lower
(2.0 0.6 mm2, p 0.05) than in the group that received cimetidine alone
(7.5 2.1 mm2). Histologic evaluation showed that the combination treatment
reduced the extent of necrosis and inflammation in the ulcer crater [18, 19].
Novel derivatives of human recombinant (hr) bFGF were also tested in
treatment of experimental chronic duodenal ulcer in rats [20], e.g., Ser 78,96-
hrbFGF, which is bioequivalent to rbFGF-CS23, CMC-hrbFGF, a car-
boxymethyl cysteine derivative of hrbFGF and PEG-hrbFGF, a polyethylene
glycol derivative of hrbFGF. Oral administration of these novel derivatives for
3 weeks accelerated the healing of cysteamine-induced chronic duodenal ulcer,
stimulated angiogenesis and PEG-hrbFGF was more active than the other
analogs.
The antiulcerogenic effect of bFGF was confirmed in preliminary human
studies which demonstrated that previously therapy-resistant duodenal and
antral ulcers healed within 4 weeks after oral treatment with bFGF-CS23, with-
out any adverse effect or systemic absorption [21]. Subsequently, the potent
ulcer healing effect of bFGF-CS23 was confirmed only in NSAID-induced
gastric and duodenal ulcers [22, 23].
Chronic Gastric Ulcer. In the studies performed by H. Satoh, a former
co-worker of ours, per os administration of a mutein of rhbFGF twice a day for
2 weeks significantly accelerated the healing of acetic acid-induced gastric
ulcer in rats: size of the ulcers decreased while the regeneration of the mucosa
was enhanced [24]. The same mutein of rhbFGF had no effect on the develop-
ment of ethanol-induced acute gastric erosions when given to the rats prior to
ethanol administration [24]. These results suggest that a mucosal protective
effect may not be involved in the healing effect of bFGF. Furthermore, bFGF
the above-mentioned mutein of hrbFGF was shown to be able to prevent the
indomethacin-induced relapse of acetic acid-induced gastric ulcer in rats that
were given either prior to or with indomethacin [25, 26].
In the cryoprobe-induced experimental gastric ulcer the interaction of
indomethacin with bFGF and omeprazole was investigated [27]. Contrary to
omeprazole, bFGF accelerated healing only in the late phase (days 10–15).
Omeprazole reversed all indomethacin-induced effects, e.g., angiogenesis, cell
proliferation, maturation of granulation tissue and ulcer healing rate, while
bFGF, despite stimulation of angiogenesis, did not reverse indomethacin-
induced delay in ulcer healing.
The involvement of bFGF in gastric ulcer was also studied using other
experimental gastric ulcer model of mice with acetic anhydride applied to the
serosal surface of the stomach [28]. Electron microscopic immunohistochemi-
cal studies of stomach were performed at 5 and 21 days after the treatment with
bFGF. bFGF was localized in fibroblasts in the ulcer bed with distribution
throughout the cytoplasm, excluding organelles involved in the usual secretory
system, such as rough endoplasmic reticulum, Golgi apparatus and secretory
vesicles, but it was present also in the nucleus.
Besides the ulcer healing and angiogenic actions, bFGF was found to
promote the reinnervation of the newly formed microvessels, regeneration of
autonomic nerves in the granulation tissue in the experimental gastric ulcer
induced by acetic acid in rats [29]. One human study tested the efficacy and
safety of bFGF-CS23 in healing of NSAID-associated gastric ulcers, which
were resistant to or relapsed after conventional treatment [30]. Five patients
with 9 ulcers were treated per os and after 4 weeks, 4 ulcers had healed and
there was significant reduction in the area of others.
Chronic Gastritis. Inflammatory gastrointestinal diseases such as chronic
gastritis are more frequent than duodenal ulcer in certain countries (e.g., Japan).
Unfortunately, contrary to the plethora of models of acute gastric erosions
and ulcers, there has been no appropriate animal model of diffuse acute and
chronic gastritis. Based on previous investigations on the role of sulfhydryl

(SH) compounds as gastroprotective agents [31, 32], our laboratory developed
a new model for gastritis: ingestion of low concentrations of SH alkylators
(e.g., 0.1% (w/v) iodoacetamide) in drinking water induced severe, diffuse
acute and chronic gastritis in rats [33, 34].
We investigated the effect of bFGF and sucralfate in this animal model.
The results demonstrate that only acid-resistant bFGF was effective in acceler-
ating the healing of chronic chemically induced gastritis. Neither native bFGF-w
nor sucralfate had any effect at low doses on chronic gastritis. However, both
native bFGF and the acid-resistant mutein in combination with sucralfate at low
dose were more efficient in the repair of mucosal injury and significantly
(p 0.001) more effective than either of these agents alone. Thus, bFGF and
sucralfate may act synergistically in healing chronic erosive gastritis.
Inflammatory Bowel Disease. The effect of bFGF was also studied in ani-
mal models of IBD. As in experimental gastritis (see above), these experiments
were also mainly performed with new animal models of IBD induced by SH
alkylators [35–37]. Rats were given the SH alkylator iodoacetamide (6%) per
rectum once to initiate IBD-like lesions. From the 2nd day, animals were treated
daily with bFGF per rectum. bFGF was found to accelerate the healing of
iodoacetamide-induced ulcerative colitis and reduced the extent of necrosis and
inflammation [35, 37].
Immunohistochemically bFGF was shown to be most prominent in the
extracellular matrix of inflammatory lesions. The expression of fibrogenic
cytokines increased in rat small intestine after irradiation [38], most probably
due to inflammatory reaction caused by radiation.
The development of colitis and rectosigmoiditis is a well-known side effect
of abdominal irradiation for the treatment of certain malignant neoplasms. This
so-called radiation colitis or enterocolitis or rectosigmoiditis is a major clinical
problem and it is relatively unresponsive to the usual therapy. Hence, new animal
models of radiation have been developed to study the pathogenesis, prevention
and treatment of these lesions [39–42]. Most recently, we adapted and sim-
plified an animal model of radiation-induced enterocolitis in our laboratory
[42, 43]. Groups of rats after receiving 20 Gy -radiation on abdominal parts
were given sucralfate (10, 20 and 50 mg/100 g) or bFGF-CS23 (100 or 500 ng/
100 g) by intragastric gavage twice daily from the 2nd day of experiment.
Autopsy was performed on the 10th day when the area of the ileal and colonic
lesions as well as the wet weight of the ileum and colon were measured. All
clinical signs and parameters were dose-dependently decreased by sucralfate.
bFGF-CS23, in the small doses used, exerted a beneficial effect only on a
few parameters of enterocolitis. Light microscopic examination of inflamed
bowel sections revealed that the lesions were reduced in irradiated and sucral-
fate or bFGF-treated rats, e.g., the submucosal edema was absent and the
necrotic-mucosa was replaced mainly by proliferating granulation tissue which
was covered by flat epithelium.
Thus, these results suggest a beneficial effect of sucralfate and bFGF in
radiation-induced enterocolitis, but extended animal studies and clinical con-
firmation are essential to explore the full therapeutic benefit of these agents in
this application.
PDGF
Chronic Duodenal Ulcer. After bFGF, PDGF was the second growth factor
tested in our laboratories to accelerate the healing of experimental duodenal
ulcers. Duodenal ulcer was induced by cysteamine as in the previous experi-
ments with bFGF, and after 3 weeks per os treatment with PDGF, rats were
killed, and the ulcer size was evaluated macroscopically and with stereomicro-
scopic planimetry. Administration of PDGF-BB (100 and 500 ng/100 g, twice
daily for 3 weeks) significantly accelerated the healing of chronic cysteamine-
induced ulcer: ulcer sizes were 2.5 1.1 mm2 (p < 0.05) and 2.0 1.4 mm2
(p 0.05), respectively versus 16.9 6.8 mm2 in the control group. It is
important to stress that gastric acid secretion was not influenced by any of the
doses of PDGF [13, 44].
As a follow-up to our studies, the effect of bFGF and PDGF was investi-
gated on the migration and proliferation of cultured rabbit gastric epithelial
cells [45]. While any dose of bFGF had virtually no effect on the epithelial
cells, PDGF significantly increased the migration of cultured epithelial cells in
a dose-dependent manner. The action of PDGF-BB was investigated in addition
to the confluent rabbit gastric epithelial cells after wounding [46]. PDGF-BB
dose-dependently accelerated the migration and proliferation of cultured cells.
Therefore, PDGF-BB may have a role in gastric epithelial cell restoration
during the healing of gastric ulcers.
Chronic Gastric Ulcer. In our gastroprotective studies, PDGF was tested
first for the prevention of acute ethanol-induced gastric erosions, and sub-
sequently for the acceleration of healing of indomethacin-induced gastric
ulcers [47]. In our studies, groups of fasted rats were given PDGF at doses of
500 ng/100 g, 1 or 2.5 g/100 g subcutaneously or by intragastric gavage,
30 min prior to the per os administration of 1 ml 75% ethanol. As a positive
control, an additional group of rats received SH-containing taurine (50 mg/
100 g). All of the animals were killed 1 h after receiving ethanol and the area
of hemorrhagic mucosal lesions in the glandular stomach was measured by
computerized stereomicroscopic planimetry. The results indicated that only
2.5 g/100 g of PDGF administered intragastrically reduced the area of acute
mucosal lesions at the borderline of statistical significance (p 0.095), while
pretreatment with taurine resulted in about 50% reduction of gastric damage

(p 0.027). It was also reported that PDGF-BB accelerated the repair of
gastric mucosa after indomethacin-induced damage, without influencing gastric
acid secretion [47].
Chronic Gastritis. We found as a follow-up to our investigation on the
gastroprotective role of SH [31, 32] that the ingestion of low concentrations of SH
alkylators (e.g., 0.1% iodoacetamide) in drinking water induced severe diffuse
acute and chronic erosive gastritis in rats [33]. We used this new animal model
of gastritis to investigate the possible beneficial effect of bFGF and PDGF.
After the induction of gastritis using 0.1% iodoacetamide in drinking water
for 1 week, rats were treated with PDGF in doses of 500 ng/100 g or 2.5 g/
100 g per os twice daily. At autopsy on the 14th day, macroscopic and histologic
involvement of gastric glandular mucosa was quantified and wet weights of the
stomach were obtained. Oral treatment with PDGF dose-dependently decreased
the severity of gastritis induced by iodoacetamide as exemplified by the dose-
dependent decrease in the wet weight of the stomach [48].
Inflammatory Bowel Disease. We used our previously developed new
ulcerative colitis model to test the effect of PDGF on the healing of colonic
ulcers and inflammation. Colitis-like lesions were induced in rats by giving 6%
iodoacetamide solution per rectum once on the 1st day [36, 37]. From the 2nd
day, PDGF-BB was given to the rats per rectum twice daily at doses of 100 and
500 ng/100 g [49, 50]. PDGF dose-dependently decreased the wet weight of
the colon, the lesion area and the severity of colonic lesions as well as the extent
of adhesions after 10 days of treatment. All parameters were significantly
improved in the high PDGF dose group. Histologically, the ulcer size was
smaller, the signs of inflammation were reduced, and in the majority of cases
extensive re-epithelization was seen.
VEGF/VPF
Chronic Duodenal Ulcer. The list of growth factors which stimulate
angiogenesis (e.g., bFGF, acidic FGF, PDGF) has been extended by the recent
availability of VEGF for pharmacologic studies. Also, there is a potent syner-
gism between VEGF and bFGF in the induction of angiogenesis in vitro and
probably also in vivo [4].
Our previous studies demonstrated that bFGF and PDGF which stimulate
angiogenesis and granulation tissue production accelerated experimental duo-
denal ulcer healing that was 2–7 million times more potent on a molar basis
than the similar effect of cimetidine. Since VEGF is highly specific for
endothelial cells, we have recently tested the hypothesis that stimulation of
angiogenesis alone is sufficient for chronic ulcer healing [51]. To induce
chronic duodenal ulcers, groups of rats were given cysteamine-HCl on the 1st
day, as in previous experiments with bFGF and PDGF. To randomize rats with
equally severe penetrating or perforated duodenal ulcers, the animals were
anesthetized and laparotomized on the 3rd day when treatment started with
vehicle saline or rhVEGF-165, 1 g/100 g once daily by gavage for 21 days. At
autopsy, the size of duodenal ulcers was measured, evaluated by stereomicro-
scopic planimetry and histologic sections taken. The results revealed that the
size of duodenal ulcers in controls was 7.4 1.6 mm2 while in the VEGF group
1.9 0.9 mm2 (p 0.05), histologically accompanied by complete healing or
prominent angiogenesis and granulation tissue production. The density of blood
vessels per 400 magnification field in the granulation tissue at the ulcer edge
was 8.0 0.5 in controls, and 16.9 1.9 (p 0.001) in the VEGF-treated rats.
More recently, we also demonstrated a potent ulcer healing effect of VEGF,
administered rectally in the above described rat model of IBD induced by
iodoacetamide [52]. Conceptually, the potent ulcer healing effect of VEGF in
the upper and lower gastrointestinal tract is very important because it demon-
strates that stimulation of vascular factors alone, i.e., angiogenesis, is sufficient
for ulcer healing, probably because epithelial cells proliferate and migrate
spontaneously over dense granulation tissue to complete the healing process.
Thus, the molecular and cellular basis of ulcer healing remains a produc-
tive research area [5], and previous pharmacologic studies with bFGF and
PDGF are now greatly expanded with VEGF.
Ulcer Healing after Gene Therapy Related to Angiogenic

Growth Factors
Although our previous studies demonstrate that administration of angio-

genic growth factor peptides such as bFGF, PDGF or VEGF accelerates ulcer
healing in the rat duodenum, stomach and colon [12–14, 51, 52], intraluminal
administration of peptide growth factors is usually limited by acid-proteolytic
degradation and possible immunologic reactions, and large-scale production of
human recombinant proteins is still an expensive and technologically limited
process. Some of these problems may be overcome by gene transfer of the
cDNA for angiogenic growth factors [53].
Gene therapy refers to administration of exogenous DNA to initiate the
cascade of DNA to RNA to protein to the enhanced synthesis of peptide(s) and
changed biologic function. Molecular biology in general has greatly advanced
our understanding of the pathogenesis of many human diseases, and gene
therapy is poised to implement that new knowledge [54].
Gene therapy with growth factors has been used successfully for the treat-
ment of ischemic diseases in limbs and myocardium. Direct injection of naked
DNA and using viral vector for gene transfer with growth factors (e.g., VEGF)

has been investigated for the treatment of ischemic diseases in both animal
models and clinical settings such as limb ischemia [55–57] and cardiovascular
diseases [58, 59].
Thus, after our pharmacologic experiments on the potent ulcer healing
effects of bFGF, PDGF and VEGF, we tested the effects of naked DNA and ade-
noviral plasmids of PDGF and VEGF to accelerate the healing of experimental
chronic duodenal ulcer [8] and ulcerative colitis.
PDGF
Chronic Duodenal Ulcer. In gene therapy studies, we first compared the
effects of naked DNA (ND) and adenoviral vector (AV) of VEGF and PDGF to
treat chronic duodenal ulcer induced by the ulcerogen cysteamine in rats.
Groups of Sprague-Dawley female rats (180–210 g) were given cysteamine-
HCl (25 mg/100 g by gavage 3 times with 4-hour intervals) to cause duodenal
ulcers. Laparotomy on the 3rd day was performed to evaluate ulcer formation
and rats with equal severity of duodenal ulcer were randomly divided into the
control groups which received either intraduodenal injection of 0.1 ml/rat of
Tris-EDTA buffer or 5 108 pfu/rat of AV-LacZ, and the treatment groups
which received 100 g/rat of ND-PDGF once, intraduodenally on the 3rd day
or 200 g/rat twice, intravenously on the 3rd and 5th days, and 5 108 pfu/rat
of AV-PDGF once, intraduodenally or intravenously on the 3rd day. Gross and
histologic evaluation of ulcer healing was performed on the 7th and 14th days:
rats with superficial nonperforated ulcers were killed on the 7th day, while
rats with perforated or penetrated ulcers were euthanized on the 14th day after
cysteamine. Duodenal ulcer crater was measured in millimeters and ulcer area
was calculated by the ellipsoid formula. Mucosal scrapings of 3 cm of proximal
duodenum were homogenized and tested for VEGF, PDGF and bFGF by
Western blot analysis and ELISA. The results demonstrated that the ulcer
area in controls was about 8–10 mm2. The ulcer areas in the groups with 100 g
of ND-PDGF intraduodenally were significantly smaller than in control
(p 0.0291) in 7 days and accelerated healing was seen in all treatment groups
with AV-PDGF in both 7 and 14 days (fig. 1). Histologic evaluation indicated
that in the control, the normal duodenal mucosa was interrupted with a sharply
demarcated ulcer crater which consists of necrotic and inflamed tissue and
mucus-secreting epithelium that does not go over the necrotic ulcer crater, while
in the treatment with AV of PDGF the granulation tissue was re-epithelized and
extremely dense collagen replaced the ulcer crater. Western blot analysis
showed both 23 kDa VEGF (fig. 2) and 30 kDa PDGF (fig. 3) increased by
50–70% in most treatment groups after 7 days, while only some elevation of
both 23 kDa VEGF and 30 kDa PDGF was seen 14 days after cysteamine.
ELISA showed a similar change of the levels of VEGF or PDGF as in Western
20
Superficial nonperforated ulcers (7 days)
15
10
*
*
Size of duodenal ulcers (mm2)
* ** **
5
* **
20
Perforated or penetrating ulcers (14 days)
15
10
**
**
5 ** ** **
**
0
Control 100g 200 g 100 g 200g 5 108 5 108 5 108 5 108 5 108
i.d. i.d. i.v. i.d. i.v. i.d. i.d. i.v. i.d. i.v.
TE buffer ND of VEGF ND of PDGF Control AV of VEGF AV of PDGF
Fig. 1. The effects of naked DNA (ND) or adenoviral vectors (AV) of VEGF or PDGF
in 7 and 14 days on the healing of duodenal ulcers induced by cysteamine in rats. *p 0.05;
**p 0.01.
blot, while surprisingly the concentration of bFGF in duodenal mucosa was

also increased by 50–70% after ND or AV of PDGF treatment in both 7 and
14 days.
Inflammatory Bowel Disease. After our successful gene therapy with ND
or AV of PDGF in chronic duodenal ulcer healing, we thus tested the hypothe-
sis that gene therapy with ND- or AV-PDGF might also accelerate the healing
of experimental ulcerative colitis. Groups of unfasted Sprague-Dawley female
rats (180–210 g) were randomly divided and given 0.1 ml/rat of 6% iodo-
acetamide once intracolonically to induce ulcerative colitis. On the 2nd day rats
received either intracolonical or intravenous injection of 100 g or 200 g/rat
of ND-PDGF or 5 108 pfu/rat of AV-PDGF. Additional groups received
5 108 pfu/rat of AV-PDGF intracolonically or intravenously on the 3rd day.
Controls were given 5 108 pfu/rat of AV-LacZ. Lethargy and diarrhea were
monitored daily until rats were euthanized on the 10th day after iodoacetamide.
The area of colonic lesion, the severity of colitis, colon wet weight, colon dilation,

60
50
7 days
40
Density of 23 kDa VEGF (arbitrary units)
30
20
10
60
50
14 days
40
30
20
10
0
Control 100g 200 g 100 g 200 g 5 108 5 108 5 108 5 108 5 108
Fig. 2. VEGF (23 kDa) Western blot in duodenal mucosa 7 and 14 days after induc-
tion of duodenal ulcer by cysteamine and treatment with naked DNA (ND) or adenoviral
vector (AV) of VEGF or PDGF in rats.
colon thickness and pericolonic adhesions were measured and samples were
fixed in 10% formalin for histologic evaluation. The levels of endogenous
bFGF, PDGF and VEGF were detected by Western blot.
The results with ND-PDGF treatment showed no significant changes in
comparison with control. However, the gene therapy with AV-PDGF demon-
strated that the lethargy was significantly decreased in all groups in compari-
son with controls (p 0.05). The area of colonic lesions and colon wet weight
were at least 50% lower than controls in all groups with gene therapy although
only the groups with single or double doses of 5 108 pfu of AV-PDGF intra-
venously demonstrated significant decreases (p 0.0496 or p 0.032, respec-
tively). The body weight loss and severity of colitis in these groups also showed
significant improvements (p 0.007 or p 0.037, respectively). Colon
dilation, thickness and pericolonic adhesions were significantly decreased
with double doses of AV-PDGF intravenously (p 0.026, p 0.037, and
p 0.044, respectively). Western blot showed that 19 kDa of bFGF was
70
60 7 days
50
Density of 30 kDa PDGF (arbitrary units)
40
30
20
10
0
70
60
14 days
50
40
30
20
10
0
Control 100g 200g 100 g 200g 5 108 5 108 5 108 5 108 5 108
Fig. 3. PDGF (30 kDa) Western blot in duodenal mucosa 7 and 14 days after induction
of duodenal ulcer by cysteamine and treatment with naked DNA (ND) or adenoviral vector
(AV) of VEGF or PDGF in rats.
markedly increased in the groups with intravenous injection of AV-PDGF and

intracolonical administration of larger dose of AV-PDGF in comparison with
controls, while almost no expression of VEGF was seen in all groups.
VEGF
Chronic Duodenal Ulcer. We also administered ND or AV of VEGF to
treat chronic duodenal ulcer induced by the ulcerogen cysteamine in rats.
Controls received either intraduodenal injection of 0.1 ml/rat of Tris-EDTA
buffer or 5 108 pfu/rat of AV-LacZ and the treatment groups were given
100 g/rat of ND-VEGF once, intraduodenally on the 3rd day after cysteamine
or 200 g/rat of ND-VEGF twice, intravenously on the 3rd and 5th days, and
5 108 pfu/rat of AV-VEGF once, intraduodenally or intravenously on the 3rd
day. Gross and histologic evaluation of ulcer healing was performed on the 7th
and 14th days: rats with superficial nonperforated ulcers were killed on the
7th day, while rats with perforated or penetrated ulcers were euthanized on the
14th day after cysteamine. Duodenal ulcer crater was measured in millimeters

and ulcer area was calculated by the ellipsoid formula. Mucosal scrapings of
3 cm of proximal duodenum were homogenized and tested for VEGF, PDGF
and bFGF by Western blot analysis and ELISA.
The results demonstrated that the ulcer area in controls was about
8–10 mm2. The ulcer areas in the groups with ND of VEGF in both 7 and 14
days were significantly smaller than in control, e.g. 100 g of ND of VEGF,
intraduodenally (p 0.0455) in 7 days as well as 100 g of ND of VEGF,
intraduodenally (p 0.0498) and 200 g of ND of VEGF, intravenously
(p 0.0392) in 14 days, and significant healing was seen in all treatment
groups with AV of VEGF in both 7 and 14 days. Histologic evaluation
also indicated that in the treatment with AV of VEGF the granulation tissue
was re-epithelized and extremely dense collagen replaced the ulcer crater.
Western blot analysis also showed both 23 kDa VEGF and 30 kDa PDGF
increased with both ND or AV of VEGF treatment in 7 days, while a slight
elevation of both 23 kDa VEGF and 30 kDa PDGF was seen 14 days after
cysteamine. ELISA showed a similar change of the levels of VEGF or PDGF as
in Western blot.
Chronic Gastric Ulcers. More recently, gene therapy for gastric ulcer
with ND-VEGF was also shown to be successful. Nonviral DNA encoding
VEGF accelerated ulcer healing through enhanced angiogenesis [60].
Gastric ulcers were induced in rats by acetic acid applied to the serosal sur-
face of the stomach, and the site around the ulcer was injected with nonviral
plasmid-encoding full-length complementary DNA (cDNA) of human
recombinant (rh) VEGF165. For some animals, neutralizing anti-VEGF anti-
body was administered. The results showed single local injection of plasmids
encoding VEGF165 significantly increased neovascularization and acceler-
ated ulcer healing. A neutralizing anti-VEGF antibody significantly reduced
the acceleration of ulcer healing resulting from the treatment. Thus, VEGF
gene therapy significantly accelerates not only duodenal but gastric ulcer
healing as well. Inhibition of accelerated healing by a neutralizing anti-
VEGF antibody indicates an essential role for VEGF and enhanced angio-
genesis in ulcer healing.
Inflammatory Bowel Disease. Previously our studies demonstrated an
accelerated ulcer healing by VEGF in a rat model of IBD induced by iodo-
acetamide [52]. We therefore tested if gene therapy with ND or AV of VEGF
might be also effective on the healing of experimental ulcerative colitis.
The ulcerative colitis was induced by 0.1 ml/rat of 6% iodoacetamide once,
intracolonically. On the 2nd day rats received either intracolonical injection
of 100 g ND-VEGF/rat or intravenous injection of 200 g ND-VEGF/rat or
5 108 pfu/rat of AV-VEGF. A combination of 5 108 pfu/rat of AV-PDGF
and AV-VEGF as well as two doses of 5 108 pfu/rat of AV-VEGF were
given intracolonically on the 2nd and 3rd days, respectively, after iodoac-
etamide. Controls were given 0.1 ml of Tris-EDTA buffer or 5 108 pfu/rat
of AV-LacZ. Lethargy and diarrhea were monitored daily until rats were
euthanized on the 10th day after iodoacetamide. The area of colonic lesion, the
severity of colitis, colon wet weight, colon dilation, colon thickness and
pericolonic adhesions were measured and samples were fixed for histologic
evaluation.
The lethargy, colonic lesions and colon dilation showed close to significant
decrease in the combination group with gene therapy in comparison with con-
trols (p 0.061, p 0.072 or p 0.098, respectively). The colon wet weight,
colon thickness and pericolonic adhesions in the combination with PDGF and
VEGF gene therapy were 2-fold lower than controls. The body weight loss in
this group also showed 4-fold lower than controls. Gene therapy with ND-VEGF
or two doses of AV-VEGF did not demonstrate significant changes although in
the group with 200 g ND-VEGF the area of colonic lesions and colonic wet
weight were 50% lower than controls.
Advances in basic sciences such as cell and molecular biology and phar-
macology of growth factors provide new methods and new results in the study
complex and multifactorial disorders such as ulcer and inflammatory diseases
of the upper and lower gastrointestinal tract. Until recently, the pathogenesis of
gastroduodenal ulceration has been investigated mostly from the point of view
of aggressive factors and the therapeutic interventions affected the healing
process only indirectly. In this review we summarized mostly our data on the
ulcer healing with either peptides or genes of growth factors, such as bFGF,
PDGF and VEGF in direct ulcer treatment which is now possible without
affecting HCl and pepsin secretion. Studies performed in animal models and
humans also demonstrate a key role of these endogenous angiogenic peptides
in ulcer healing. We thus conclude that stimulation of cell proliferation is the
most consistent mechanism of ulcer healing by growth factors either with
peptides or gene transfer. Furthermore, enhancement by VEGF of angiogenesis
and granulation tissue production is sufficient for ulcer healing. Hence, growth
factors are potent, endogenously derived antiulcer agents which directly stimu-
late ulcer healing in which angiogenesis seems to be the most important
process. In comparison with peptide growth factors, gene therapy with single or
double doses is more efficient for ulcer healing. Thus, VEGF and PDGF gene
therapy seems to be a new option to achieve a rapid ulcer healing in the upper
and lower gastrointestinal tract.

References
1 Wynendaele W, van Oosterom AT, Pawinski A, de Bruijn EA, Maes RA: Angiogenesis:
Possibilities for therapeutic interventions. Pharm World Sci 1998;20:225–235.
2 Szabo S, Shing Y, Folkman MJ, Vincze A, Gombos Z, Deng XM, Khomenko T, Yoshida M:
Angiogenesis and growth factors in ulcer healing; in Fan Tai-Ping D, Kohn EC (eds): New
Angiogenesis. Totowa, Humana Press, 2001, pp 119–211.
3 Szabo S, Vincze A: Growth factors in ulcer healing: Lessons from recent studies. J Physiol (Paris)
2000;94:77–81.
4 Papper MS, Ferrara N, Orci L, Montesano R: Potent synergism between vascular endothelial
growth factor and basic fibroblast growth factor in the induction of angiogenesis in vitro. Biochem
5 Szabo S, Kusstatscher S, Sandor Z et al: Molecular and cellular basis of ulcer healing. Scand
J Gastroenterol 1995;30(suppl 208):3–8.
6 Szabo S, Kusstatcher S, Sakoulas G, Sandor Z, Vincze A, Jadus M: Growth factors: ‘New endo-
genous drugs’ for ulcer healing. Scand J Gastroenterol 1995;30(suppl 210):15–18.
7 Szabo S, Folkman J, Vattay P et al: Duodenal ulcerogens: The effect of FGF on cysteamine-
induced duodenal ulcer; in Halter F, Garner A, Tytgat GNJ (eds): Mechanism of Peptic Ulcer
Healing. London, Kluwer Academic, 1991, pp 139–150.
8 Deng XM, Szabo S, Jadus MR, Khomenko T, Yoshida M, Herlyn M, Nesbit M, Matsumoto H,
Florsheim WH: Gene therapy with naked DNA or adenoviral plasmids of VEGF and PDGF
increases endogenous VEGF, PDGF and bFGF expression and accelerates chronic duodenal ulcer
healing in rats. Gastroenterology 2000;118:A660–A661.
9 Poulsen SS, Olsen KS, Kierkegaard P: Healing of cysteamine-induced duodenal ulcers in rats.
Dig Dis Sci 1985;30:161–167.
10 Garner A: Strategies for the development of novel anti-ulcer drugs; in Garner A, Whittle BJR
(eds): Advances in Drug Therapy of Gastrointestinal Ulceration. New York, Wiley-Interscience,
1989, pp 275–288.
11 Konturek SJ, Brzozowski T, Dembinski A, Warzecha Z, Yamazaki J: Gastric protective and
ulcer-healing action of epidermal growth factor; in Garner A, Whittle BJR (eds): Advances in
Drug Therapy of Gastrointestinal Ulceration. New York, Wiley-Interscience, 1989, pp 261–273.
12 Szabo S, Folkman J, Vattay P et al: Accelerated healing of duodenal ulcers by oral administration
of mutein of fibroblast growth factor in rats. Gastroenterology 1994;106:1106–1111.
13 Szabo S, Sandor Z: Basic fibroblast growth factor and PDGF in GI diseases. Baillière’s Clin
Gastroenterol 1996;10:97–112.
14 Szabo S, Vincze A, Sandor Z, Jadus M, Gombos Z, Pedram A, Levin E, Hagar J, Iaquinto G:
Vascular approach to gastroduodenal ulceration. New studies with endothelins and VEGF. Dig Dis
Sci 1998;43:40S–45S.
15 Szabo S: Animal model of human disease: duodenal ulcer disease. Animal model: cysteamine-
induced acute and chronic duodenal ulcer in the rat. Am J Pathol 1978;93:273–276.
16 Seno K, Sasada R, Iwane K et al: Stabilizing basic fibroblast growth factor using protein engi-
neering. Biochem Biophys Res Commun 1988;151:701–708.
17 Konturek SJ, Brzozowski T, Majka J et al: Fibroblast growth factor in gastroprotection and ulcer
healing: Interaction with sucralfate. Gut 1993;34:881–887.
18 Kusstatscher S, Nagy L, Morales RE, Szabo S: Additive effect of basic fibroblast growth factor
and cimetidine on chronic duodenal ulcer healing in rats. Gastroenterology 1993;102:A104.
19 Szabo S, Sakoulas G, Kusstatscher S et al: Effects of endogenous and exogenous basic fibroblast
growth factor in ulcer healing. Eur J Gastroenterol 1993;5(suppl 3):S53–S57.
20 Kusstatscher S, Sandor Z, Szabo S et al: Different molecular forms of basic fibroblast growth
factor accelerate duodenal ulcer healing in rats. J Pharmacol Exp Ther 1995;275:456–461.
21 Wolfe MM, Bynum TE, Parson WG et al: Safety and efficacy of an angiogenic peptide, basic
fibroblast growth factor, in treatment of gastroduodenal ulcers: A preliminary report.
22 Syngal S, Bynum TE, Folkman J et al: Randomized, double-blind placebo-controlled trial of basic
fibroblast growth factor in the healing of gastroduodenal ulcers. Gastroenterology 1996;100:A267.
23 Hull MA, Knifton A, Filipowicz B et al: Basic fibroblast growth factor reduces indomethacin-
induced relapse in a human model of gastric ulceration. Gastroenterology 1996;110:A138.
24 Satoh H, Shino A, Inatomi N et al: Effect of rhbFGF mutein CS23 (TGP-580) on the healing of
gastric ulcers induced by acetic acid in rats. Gastroenterology 1991;100:A155.
25 Satoh H, Shino A, Sato F et al: Role of endogenous and exogenous bFGF in the healing of gastric
ulcers in rats. Gastroenterology 1992;102:A159.
26 Satoh H, Shino A, Sato F, Asano S, Murakami I, Inatomi N, Nagaya H, Kato K, Szabo S, Folkman J:
Role of endogenous basic fibroblast growth factor in the healing of gastric ulcers in rats. Jpn J
Pharmacol 1997;73:59–71.
27 Schmassman A, Tarnawski A, Peskar BM et al: Influence of acid and angiogenesis on kinetics of
gastric ulcer healing in rats: Interaction with indomethacin. Am J Physiol 1995;286:G267–G285.
28 Yabu M, Shinomura Y, Minami T et al. Immunohistochemical localization of basic fibroblast
growth factor in the healing stage of mouse gastric ulcer. Histochemistry 1993;100:409–413.
29 Nakamura M, Oda M, Inoue J et al: Effect of basic fibroblast growth factor on reinnervation of
gastric microvessels. Possible relevance to ulcer recurrence. Dig Dis Sci 1995;40:1451–1458.
30 Hull MA, Cullen DJ, Hudson N et al: Basic fibroblast growth factor treatment for non-steroidal
anti-inflammatory drug-associated gastric ulceration. Gut 1995;37:610–612.
31 Szabo S, Trier JS, Frankel PW: Sulphydryl compounds may mediate gastric cytoprotection.
Science 1981;214:200–202.
32 Dupuy D, Raza A, Szabo S: The role of endogenous nonprotein and protein sulfhydryls in gastric
mucosal injury and protection; in Szabo S, Pfeiffer CJ (eds): Ulcer Disease: New Aspects of
Pathogenesis and Pharmacology. Boca Raton, CRC Press, 1989, pp 421–434.
33 Szabo S, Trier JS, Brown A et al: Sulfhydryl blockers induce severe inflammatory gastritis in the
rat. Gastroenterology 1984;86:A271.
34 Stovroff M, Vattay P, Marino B et al: Healing of experimental gastritis by oral fibroblast growth
factor. Surg Forum 1991;42:174–175.
35 Satoh H, Szabo S: New animal model of ulcerative colitis induced by sulfhydryl blockers in the
rat. Gastroenterology 1990;98:A202.
36 Sandor Z, Nagata M, Kusstatscher S, Szabo S: New animal model of ulcerative colitis associated
with depletion of glutathione and protein SH alkylation. Gastroenterology 1994;106:A766.
37 Satoh H, Sato F, Takami K, Szabo S: New ulcerative colitis model induced by sulfhydryl blockers
in rats and the effects of anti-inflammatory drugs on the colitis. Jpn J Pharmacol 1997;73:299–309.
38 Langberg CW, Martin HJ, Sung CC et al: Expression of fibrogenic cytokines in rat small intestine
after fractionated irradiation. Radiother Oncol 1994;32:29–36.
39 Saclarides TJ, King DG, Franklin JL, Doolas A: Formalin instillation for refractory radiation-
induced hemorrhagic proctitis. Dis Colon Rect 1996;39:196–199.
40 Henriksson R, Frazen L, Littbrand B: Prevention and therapy of radiation-induced bowel discom-
fort. Scand J Gastroenterol 1992;191:7–11.
41 Dubray BM, Thames HD: Chronic radiation damage in the rat rectum: An analysis of the influences
of fractionation, time and volume. Radiother Oncol 1994;33:41–47.
42 Szabo S, Sandor Z, Vincze A, Gombos Z, Mohiuddin A, Viravathana T: Radiation-induced
enterocolitis: Basic and applied science. Eur J Surg 1998;suppl 582:85–89.
43 Sandor Z, Vincze A, Gombos Z et al: The effect of sucralfate and basic fibroblast growth factor
in a model of radiation-induced enterocolitis in rats. J Gastroenterol 2002 (in press).
44 Vattay P, Gambier E, Morales RE et al: Effect of orally administered platelet-derived growth
factor on healing of chronic duodenal ulcers and gastric secretion in rats. Gastroenterology 1991;
100:A180.
45 Watanabe S, Maehiro K, Hirose M et al: Modulation of gastric mucosal restoration by growth
factors in a culture cell model. Gastroenterology 1994;106:A209.
46 Watanabe S, Wang XE, Hirose M et al: Platelet-derived growth factor accelerates gastric epithe-
lial restoration in a rabbit cultured cell model. Gastroenterology 1996;110:775–779.
47 Guglietta A, Hervada T, Nardi RV et al: Effect of platelet-derived growth factor-BB on
indomethacin-induced gastric lesions in rats. Scand J Gastroenterol 1992;27:673–676.
48 Kusstatscher S, Szabo S: Effect of platelet-derived growth factor on the healing of chronic
gastritis in rats. Gastroenterology 1993;104:A125.

49 Sandor Z, Szeli D, Charette M, Szabo S: Platelet-derived growth factor accelerates the healing of
experimental ulcerative colitis in rats. Gastroenterology 1995;108:A887.
50 Sandor Z, Kusstatscher S, Szeli D, Szabo S: Effect of platelet-derived growth factor in experi-
mental colitis in rats. Orv Hetil 1995;136:1059–1061.
51 Szabo S, Folkman J, Vincze A, Sandor Z, Gombos A: Modulation of vascular factors by vascular
endothelial cell growth factor/vascular permeability factor is sufficient for chronic ulcer healing
and acute gastroprotection. Gastroenterology 1997;112:A303.
52 Sandor Z, Singh G, Szabo S: The effect of vascular endothelial growth factor on experimental
ulcerative colitis in rats. Gastroenterology 1998;114:G4403.
53 Safi J Jr, Gloe TR, Riccioni T, Kovesdi I, Capogrossi MC: Gene therapy with angiogenic factors:
A new potential approach to the treatment of ischemic diseases. J Mol Cell Cardiol 1997;29:
2311–2325.
54 Moldawer LL, Edwards PD, Josephs M, Minter RM, Copeland EM III, Mackay SLD: Application
of gene therapy to acute inflammatory diseases. Shock 1999;12:83–101.
55 Tsurumi Y, Takeshita S, Chen D, Kearney M, Rossow ST, Passeri J, Horowitz JR, Symes JF, Isner
JM: Direct intramuscular gene transfer of naked DNA encoding vascular endothelial growth
factor augments collateral development and tissue perfusion. Circulation 1996;94:3281–3290.
56 Isner JM, Pieczek A, Schainfeld R, Blair R, Haley L, Asahara T, Rosenfield K, Razvi S, Walsh K,
Symes JF: Clinical evidence of angiogenesis after arterial gene transfer of phVEGF165 in patient
with ischemic limb. Lancet 1996;348:370–374.
57 Baumgartner I, Pieczek A, Manor O, Blair R, Kearney M, Walsh K, Isner JM: Constitutive expres-
sion of phVEGF165 after intramuscular gene transfer promotes collateral vessel development in
patients with critical limb ischemia. Circulation 1998;97:1114–1123.
58 Banai S, Jaklitsch MT, Shou M, Lazarous DF, Scheinowits M, Biro S, Epstein SE, Unger EF:
Angiogenic-induced enhancement of collateral blood flow to ischemic myocardium by vascular
endothelial growth factor in dogs. Circulation 1994;89:2183–2189.
59 Losordo DW, Vale PR, Symes JF, Dunnington CH, Esakof DD, Maysky M, Ashare AB, Lathi K,
Isner JM: Gene therapy for myocardial angiogenesis initial clinical results with direct myocardial
injection of phVEGF165 as sole therapy for myocardial ischemia. Circulation 1998;98: 2800–2804.
60 Jones MK, Kawanaka H, Baatar D, Szabo IL, Tsugawa K, Pai R, Koh GY, Kim I, Sarfeh IJ,
Tarnawski AS: Gene therapy for gastric ulcers with single local injection of naked DNA encoding
VEGF and angiopoietin-1. Gastroenterology 2001;121:1040–1047.
Sandor Szabo, MD, PhD, MPH, Pathology & Laboratory Medicine Service,
VA Medical Center, 5901 East 7th Street, Long Beach, CA 90822–5201 (USA)
Tel. 1 562 494 5921, Fax 1 562 494 5623
Gastrointestinal Protective Action of

Prostaglandin E2 and EP Receptor
Subtypes
Koji Takeuchi, Shinichi Kato, Akiko Tanaka
Department of Pharmacology and Experimental Therapeutics,
Kyoto Pharmaceutical University, Kyoto, Japan
Prostaglandins (PGs), produced from arachidonic acid by two isoforms of

cyclooxygenase (COX), are present throughout the gastrointestinal tract and are
known to bring about a wide variety of actions in the gut, including control of
acid secretion, bicarbonate secretion, mucus production, mucosal blood flow
and maintenance of mucosal integrity [1]. Indeed, PGs administered exoge-
nously protect the gastrointestinal mucosa against various ulcerogenic stimuli
such as stress, necrotizing agents and non-steroidal anti-inflammatory drugs
(NSAIDs). The pioneering study by Robert et al. [2] in 1979 was the first to
demonstrate that PGs protect the stomach against a variety of necrotizing
agents, a phenomenon called ‘gastric cytoprotection’. PGE2 is one of the major
PGs most effective in this action.
Recent pharmacological studies have classified PGE2 receptors into four
specific G protein-coupled subtypes, EP1 to EP4 [3]. The distribution of these
receptors is considered to explain the multiple effects of PGE2 in various
tissues including the gastrointestinal tract. In addition, mice lacking various
receptors for prostanoids have been established [4–6], and by using these
‘knockout mice’ the roles of specific PG receptors in various biological
actions of PGs have been demonstrated [6–8]. We have recently performed
a series of experiments to determine the EP receptor subtypes mediating the
gastrointestinal protection afforded by PGE2, using various models in both
rats and EP receptor knockout mice [9–13]. In these studies, we also used
prostanoids, subtype-specific EP receptor agonists and an antagonist, as a tool
to characterize the EP receptor subtypes involved in gastrointestinal protection
(table 1).
Table 1. Various prostanoids, subtype-specific EP
receptor agonists and an antagonist used
Prostanoids EP subtype selectivity
17-phenyl-PGE2 EP1 agonist

Sulprostone EP1/EP3 agonist
Enprostil EP1/EP3 agonist
Butaprost EP2 agonist
ONO-NT-012 EP3 agonist
11-deoxy-PGE1 EP3/EP4 agonist
ONO-AE1-329 EP4 agonist
ONO-AE-829 EP1 antagonist
In this article, we review our recent publications on the relation between

EP receptor subtypes and the gastrointestinal protections afforded by endoge-
nous or exogenous PGE2 and discuss possible functional alterations responsible
for the protective action of PGE2 in the stomach, the duodenum and the small
intestine.
Gastric Cytoprotection by PGE2
A variety of models have been used for assessing antiulcer drugs, and
PGE2 is shown to be effective in most [1, 2]. Among them, gastric lesions
induced by necrotizing agents such as ethanol or NSAIDs are considered
the most suitable for examining the protective action of PGE2 in the stomach
[9, 10, 12].
HCl/Ethanol-Induced Gastric Damage

Oral administration of HCl/ethanol (1 ml; 60% ethanol in 150 mM HCl)
produced multiple band-like lesions in the glandular mucosa, along the long
axis of the stomach. PGE2 given prior to HCl/ethanol prevented the develop-
ment of these lesions, in a dose-dependent manner. This action of PGE2 was
mimicked by prostanoids, such as 17-phenyl-PGE2 or sulprostone specific to
the EP1 receptor, and was significantly attenuated by ONO-AE-829 the selec-
tive EP1 antagonist [9]. Neither butaprost, ONO-NT-012, nor 11-deoxy-PGE1
had any effect on the gastric ulcerogenic response to HCl/ethanol. These results
suggest that the protective action of PGE2 against HCl/ethanol is mediated
by activation of the EP1 receptors. It is known that when the stomach is pre-
exposed to a mild irritant such as taurocholate (TC), the resistance of the
Takeuchi/Kato/Tanaka 228
(1) Exogenous PGE2
EP1-receptor Gastric cytoprotection
(2) Mild irritant Endogenous

PGs
ONO-AE-829
Indomethacin
Sensory deafferentation
(3) Capsaicin Sensory neurons Gastric cytoprotection

EP2- and/or IP-receptor
Endogenous
Indomethacin
PGs ⫾
: Inhibition : Increase
Fig. 1. EP receptor subtypes responsible for gastric cytoprotection induced by exoge-

nous PGE2, a mild irritant and capsaicin. (1) Exogenous PGE2 provides direct gastric cyto-
protection mediated by activation of EP1 receptors, and this effect is totally blocked by the
EP1 antagonist, ONO-AE-829. (2) A mild irritant increases endogenous PGE2 production
in the stomach and offers adaptive gastric cytoprotection. This action is prevented by
indomethacin as well as by the EP1 antagonist. (3) Capsaicin provides gastric cytoprotection,
essentially mediated by capsaicin-sensitive afferent neurons. Although capsaicin does not
increase PGE2 production in the gastric mucosa, this protective action is facilitated by
endogenous PGs through EP2 and IP receptors.
mucosa to subsequently applied necrotizing agents increases, a phenomenon

called ‘adaptive cytoprotection’ [14]. Since this effect disappears in the presence
of indomethacin, a COX inhibitor, it is assumed that it is mediated through
an enhanced production of endogenous PGs. Indeed, 20 mM TC given p.o.
increased the PGE2 content in the stomach and prevented the formation of
gastric lesions induced by a subsequent challenge with HCl/ethanol [12]. Of
interest, this effect of TC was antagonized by ONO-AE-829, the EP1 antago-
nist, suggesting that the adaptive gastric cytoprotection is mediated mainly by
endogenous PGE2 through EP1 receptors (fig. 1).
The results obtained in rats were confirmed using EP receptor knockout
mice. Oral administration of HCl/ethanol produced similar band-like lesions in
the stomachs of wild-type mice and mice lacking EP1 or EP3 receptors. The
development of these lesions was prevented by prior administration of PGE2 in
both wild-type and EP3 receptor knockout mice but not in the animals lacking
EP1 receptors [9]. Likewise, TC acted as a mild irritant in the mouse stomach
to increase production of PGE2, which resulted in prevention of HCl/ethanol-
induced damage. This effect of TC was significantly mitigated by pretreatment
with indomethacin as well as ONO-AE-829, the EP1 antagonist. In addition,
the protective action of TC was observed in EP3 receptor knockout mice but
EP Receptors and PGE2-Induced Gastrointestinal Protection 229

Table 2. EP receptor subtype responsible for the cyto-
protective action of PGE2 in various gastrointestinal lesion
models
Lesion model EP receptor subtype Ref.
Stomach
HCl/ethanol EP1 receptor 9, 12, 13
Indomethacin EP1 receptor 10, 37
Duodenum
Acid injury EP3/EP4 receptors 8, 9
Small intestine
Indomethacin EP3/EP4 receptors 11, 37
totally disappeared in EP1 receptor knockout animals [12]. These results

strongly suggest that EP1 receptors are essential for the cytoprotective action of
PGE2, either generated endogenously or administered exogenously, in the stom-
ach against necrotizing agents (table 2).
Endogenous PGs play a role in the gastric cytoprotection induced by cap-
saicin and some antiulcer drugs. Capsaicin in particular is unique in that it
causes a selective stimulation of capsaicin-sensitive afferent neurons through
interaction with vaniloid type 1 receptors [15]. The protective action of cap-
saicin was totally blocked by chemical ablation of these afferent neurons and
significantly attenuated by the antagonist of calcitonin gene-related peptide
(CGRP) as well as nitric oxide (NO) synthase inhibitors. Thus, it is considered
that capsaicin exhibits gastroprotective action through capsaicin-sensitive
afferent neurons mediated by both CGRP and NO. Interestingly, the protec-
tive action of capsaicin was also significantly mitigated in the presence of
indomethacin, suggesting an involvement of endogenous PGs, similar to the
case of adaptive cytoprotection induced by a mild irritant [16, 17]. However,
this effect of capsaicin was not affected by the selective EP1 antagonist ONO-
AE-829, in contrast to that of TC as a mild irritant [13]. It should also be noted
that neither stimulation of sensory neurons by capsaicin nor sensory deaf-
ferentation affected mucosal PGE2 levels in the stomach. These results suggest
that although endogenous PGs are involved in the gastric protection induced by
both mild irritants and capsaicin, the mode of action seems to be different in
these two cases [12, 13]. It is assumed that the stimulation of afferent neurons
by capsaicin does not increase the generation of PG in the stomach, yet it exerts
a gastroprotective action partly dependent on endogenous PGs. We found that
the protective action of capsaicin was significantly restored even in the pres-
ence of indomethacin by prior administration of butaprost, the EP2 agonist, but
not EP3 or EP4 agonist. Since the capsaicin-induced gastric protection was not
affected by the EP1 antagonist, it is unlikely that EP1 receptors are involved in
the facilitation by endogenous PGs of this action. Indeed, significant protection
by capsaicin was observed even in the knockout mice lacking EP1 and EP3
receptors, confirming that the capsaicin-induced gastric protection has nothing
to do with the EP1 and EP3 receptors. However, we found that capsaicin did
not provide gastric cytoprotection against HCl/ethanol in IP receptor knockout
animals [13]. These findings in knockout mice suggest that IP receptors are also
involved in the protective action of capsaicin in the stomach, in addition to EP2
receptors. At present, the exact mechanism by which endogenous PGs con-
tribute to the protective action of capsaicin remains unknown. Boku et al. [18]
recently reported a lack of release of CGRP in response to mild injury in the
stomach of IP receptor knockout mice. Thus, it is assumed that endogenous
PGI2 plays a supportive role in the mechanism of capsaicin-induced gastric
cytoprotection, probably by sensitizing capsaicin-sensitive afferent neurons.
Further study is needed to clarify this point.
Indomethacin-Induced Gastric Damage

NSAIDs such as indomethacin damage the stomach of experimental animals
and humans as through adverse reactions. Since these drugs induce a depletion
of endogenous PGs by inhibiting COX activity, it is considered that a deficiency
of PG is a major pathogenic factor in this model. Indeed, gastric lesions induced
by indomethacin were effectively and dose-dependently prevented by supple-
mentation of exogenous PGE2 [10, 19]. This effect of PGE2 was mimicked by
sulprostone and 17-phenyl-PGE2, both having a potent affinity to EP1 recep-
tors, and significantly attenuated by the EP1 antagonist ONO-AE-829, the
result being similar to the protective action against HCl/ethanol [10]. Neither
butaprost, ONO-NT-012 nor 11-deoxy-PGE1 afforded significant protection
against indomethacin-induced gastric lesions. In addition, indomethacin caused
gastric damage similarly in both wild-type and knockout mice lacking EP1 or
EP3 receptors, yet the protective action of PGE2 was observed in wild-type and
EP3 receptor knockout mice but not in mice lacking EP1 receptors (fig. 2).
Given the above findings, it is assumed that PGE2 prevents indomethacin-
induced gastric lesions through the activation of EP1 receptors (table 2).
Functional Alterations Related to Gastric Cytoprotection

Endogenous PGs play a role in the regulation of various gastric functions,
such as acid secretion, mucus/bicarbonate secretion, mucosal blood flow and
motility, that may contribute to gastric cytoprotection. According to previous
studies including our own [7, 8, 20–22], PGE2 inhibits acid secretion through
EP3 receptors and increases mucus and bicarbonate secretion through EP4 and
EP1 receptors, respectively (table 3). In a preliminary study, we observed that

15
n ⫽4~5
*p⬍0.05
Gastric damage (mm2)
10
*
Saline
Saline
5
Saline
PGE2
PGE2
PGE2
0
Wild-type EP1(⫺/⫺) EP3(⫺/⫺)
Fig. 2. Effect of PGE2 on gastric lesions induced by indomethacin in wild-type mice

and those lacking EP1 or EP3 receptors. The animals were given indomethacin (35 mg/kg)
s.c. and killed 4 h later. PGE2 (0.3 mg/kg) was administered s.c. 30 min before indomethacin
treatment. All values are presented as the means ⫾ SE from 4 to 5 rats per group.
*Significantly different from saline in the corresponding group, at p ⬍ 0.05.
Table 3. EP receptor subtype responsible for the various actions of PGE2

in the gastrointestinal tract
Function Action EP receptor subtype Ref.
Acid secretion Decrease EP3 receptor 9

Bicarbonate secretion
Stomach Increase EP1 receptor 7
Duodenum Increase EP3/EP4 receptors 7, 8
Mucus secretion
Stomach Increase EP4 receptor 21
Small intestine Increase EP3/EP4 receptors 11
Gastric mucosal Increase EP2/EP3/EP4 9, 22
blood flow receptors
Motility (circular smooth muscle)
Stomach Decrease EP1 receptor 9, 10
Small intestine Decrease EP4 receptor 11
gastric mucosal blood flow was increased by EP2, EP3 and EP4 agonists but
not by EP1 agonists [9]. Of interest, prostanoids exhibiting a preference for
only EP1 receptors affected gastric motility and provided mucosal protection
against gastric lesions induced by HCl/ethanol or indomethacin (fig. 3) [9, 10].
a PGE2 d
Butaprost
(cm H2O) (0.3 mg/kg, i.v.) (cm H2O) (3 mg/kg, i.v.)
Indomethacin Indomethacin
80
Gastric motility
Gastric motility
(35 mg/kg, s.c.) 80
(35 mg/kg, s.c.)
60 60
40 40
20 20
0 0
⫺30 0 30 60 90 120 150 180 ⫺30 0 30 60 90 120 150 180
Time (min) Time (min)
b e
Sulprostone ONO-NT-012
(cm H2O) (cm H2O) (3 mg/kg, i.v.)
(1 mg/kg, i.v.)
80 Indomethacin 80 Indomethacin
Gastric motility
Gastric motility
(35 mg/kg, s.c.) 60 (35 mg/kg, s.c.)
60
40 40
20 20
0 0
⫺30 0 30 60 90 120 150 180 ⫺30 0 30 60 90 120 150 180
c f
11-deoxy PGE1
17-phenyl PGE2 (cm H2O) Indomethacin (1 mg/kg, i.v.)
(cm H2O) (1 mg/kg, i.v.)
80 Indomethacin 80 (35 mg/kg, s.c.)
Gastric motility
Gastric motility
60 (35 mg/kg, s.c.) 60

40 40
20 20
0 0
⫺30 0 30 60 90 120 150 180 ⫺30 0 30 60 90 120 150 180
Fig. 3. Representative figures showing the effects of PGE2 (0.3 mg/kg), sulprostone
(1 mg/kg), 17-phenyl-PGE2 (1 mg/kg), butaprost (3 mg/kg), ONO-NT-012 (3 mg/kg) and
11-deoxy-PGE1 (1 mg/kg) on gastric hypermotility caused by indomethacin (35 mg/kg) in
rats. Each prostanoid was given i.v. as a single injection 2 h after s.c. administration of
indomethacin.
Certainly, these effects were both antagonized by ONO-AE-892, the EP1 antag-
onist, suggesting that the motility effect of PGE2 is paralleled by a reduction
in gastric mucosal damage. We have reported that a variety of compounds
afforded gastric cytoprotection at doses that inhibit gastric motility [19, 23–25].
The inhibition of gastric motility may lead to a flattening of the mucosal fold-
ings and a decrease in mucosal vulnerability to irritants, resulting in prevention
of the fold-related band-like lesions, as observed following administration
of HCl/ethanol. A role for muscle elements in the pathogenic mechanism of
indomethacin-induced gastric lesions has also been demonstrated [19, 23, 24].
Mersereau and Hinchey [23] were the first to show the importance of stomach

hypermotility and mucosal foldings in the genesis of gastric lesions in response
to NSAIDs. We also reported that indomethacin at an ulcerogenic dose
enhances gastric motility and induces microcirculatory disturbances due to
abnormal mucosal compression of the gastric wall [24]. Since neither
butaprost, ONO-NT-012 nor 11-deoxy-PGE1 provided any gastric protection
against HCl/ethanol or indomethacin, despite causing an increase in gastric
mucosal blood flow, it is unlikely that the gastric cytoprotection afforded by
PGE2 is functionally associated with an increase of gastric mucosal blood flow
[9]. Certainly, because inhibition of gastric motility may lead to attenuation of
microvascular disturbances due to stomach contraction, it is possible that
prostanoids through EP1 receptors help to maintain mucosal blood flow during
exposure to noxious agents.
On the other hand, neutrophils have been implicated in the damage associ-
ated with NSAIDs [26]. It is known that PGE2 has an inhibitory effect on
neutrophil functions, including chemotaxis [27]. We confirmed that PGE2
exhibited an inhibitory effect on the migration of neutrophils caused by formyl-
methionyl-leucyl-phenylalanine in vitro [10]. The same inhibitory action was
shown by both butaprost and 11-deoxy-PGE1, but not by either 17-phenyl-
PGE2, sulprostone or ONO-NT-012, clearly indicating that the antineutrophil
chemotaxis action of PGE2 is mediated by activation of EP2 and EP4 receptors.
Thus, it is assumed that the inhibition of neutrophil migration by itself is not
sufficient to reduce the overall expression of gastric lesions in response to
indomethacin. Since the increase in myeloperoxidase activity as well as lesion
formation induced by indomethacin was prevented when the enhanced gastric
motility response was inhibited by atropine [24], it is likely that the neutrophil
infiltration is secondary to the event associated with gastric hypermotility
following indomethacin. Indeed, Melange et al. [28] even showed that NSAID-
induced gastric injury is neutrophil-independent in the neutropenic rat.
The mechanism by which PGE2 inhibits gastric motility through EP1
receptors remains unknown. Milenov and Golenhofen [29] reported that PGE2
relaxed the circular muscle but contracted the longitudinal muscle of the canine
stomach. Narumiya’s group [30, 31] reported the localization of mRNAs of the
EP receptors along the gastrointestinal tract. They showed that strong signals
for EP1 transcripts occurred in the smooth muscle cells in the muscularis mucosa
throughout the tract. Since EP1 receptors are coupled to phosphatidylinositol
turnover [6], it is assumed that contraction of longitudinal smooth muscle by
PGE2 is associated with an increase of cytosolic calcium. Contraction of circular
smooth muscle leads to the appearance of mucosal folds, which have been
implicated in the pathogenesis of several ulcer models including indomethacin-
induced gastric lesions [23]. At present, the mechanism by which PGE2 relaxes
circular smooth muscle through activation of EP1 receptors is unknown.
Duodenal Cytoprotection and HCOⴚ3 Stimulation by PGE2
Duodenal mucosal HCO⫺ 3 secretion is a key process that aids in preventing

acid-peptic injury. This is most exemplified by the finding that the tissues
respond to acid by secreting more HCO⫺3 [32]. Although this process has been
shown to involve both humoral and neural factors as well as PGs [33], it is
thought that endogenous PGs are particularly important in the local control of
this secretion. Indeed, PGE2 and its analogues, whether applied luminally or
vascularly, stimulate duodenal HCO⫺ 3 secretion in vivo and in vitro, in a variety
of species, and in this way may contribute to protection of the mucosal epithe-
lium against acid-induced injury [33]. We have recently shown that COX-1 but
not COX-2 is a key enzyme in regulating this process and maintaining the
mucosal integrity against acid in the duodenum [34].
PGE2 increased HCO⫺ 3 secretion by the rat duodenal mucosa; this action
was verapamil-sensitive and potentiated by isobutylmethylxanthine, an inhibitor
of phosphodiesterase [7]. This effect was mimicked by enprostil, sulprostone,
ONO-NT-012 and ONO-AE1-329 but not by butaprost or 17-phenyl-PGE2 [7].
These results strongly suggest that PGE2 stimulates duodenal HCO⫺ 3 secretion
via both EP3 and EP4 receptors, and this action is coupled with both Ca2⫹ and
adenosine 3⬘,5⬘-cyclic monophosphate (cAMP). Although EP3 receptors are
coupled with Gi, leading to inhibition of adenylate cyclase, a recent study
showed the existence of four splicing variants of EP3 receptors, coupled to
different signaling pathways [6]. The EP3A receptor is linked to activation of
Gi protein, while EP3B and EP3C are coupled with activation of Gs protein,
resulting in stimulation of adenylate cyclase activity. Thus, it is possible that
EP3B and EP3C receptors are involved in stimulation of HCO⫺ 3 secretion in
the duodenum. Morimoto et al. [31] demonstrated by Northern blot analysis
significant expression of EP3 and EP4 receptors in the gastroduodenal mucosal
layer containing epithelial cells. These results are compatible with our observa-
tion that HCO⫺ 3 secretion, one of the epithelial functions, is mediated by both
EP3 and EP4 receptors in the duodenum (tables 2, 3).
In the duodenum of wild-type mice, secretion of HCO⫺ 3 increased in
response to luminal perfusion of PGE2 and forskolin as well as mucosal acidi-
fication [8]. The latter effect was significantly inhibited by prior administration
of indomethacin. The HCO⫺ 3 response to acid was observed in EP1 receptor
knockout mice but disappeared in the animals lacking EP3 receptors, although
the acidification increased mucosal PGE2 levels to a similar degree in all
groups. Consistent with the results obtained with rats, the HCO⫺ 3 stimulatory
action of PGE2 was markedly reduced in EP3 receptor knockout but not EP1
receptor knockout mice, but the forskolin effect was observed in both groups of
animals, similar to wild-type mice. It is believed that the acid-induced HCO⫺ 3

secretion is mediated via an axonal reflex pathway, in addition to endogenous
PGs, and the mediator on the efferent side of this reflex pathway may be
vasoactive intestinal peptide [33]. Since this response is substantially inhibited
by indomethacin, it is also speculated that the afferent side of this reflex path-
way is influenced by PGs, probably by facilitating neuronal excitation in
response to H⫹. We have previously reported that acid-induced HCO⫺ 3 secretion
was significantly attenuated by chemical ablation of capsaicin-sensitive afferent
neurons and that the HCO⫺ 3 stimulatory action of capsaicin is also suppressed
by indomethacin [35]. The EP3 receptors, which are prerequisite for the acid-
induced duodenal HCO⫺ 3 secretion, might be on cells on the afferent side of the
reflex pathway. Thus, it is assumed that local PGE2 release would stimulate the
reflex pathway on the afferent side and may also directly stimulate the epithe-
lial cells, both resulting in an increase in HCO⫺ 3 secretion.
As mentioned above, the HCO⫺ 3 secretion plays an important role in pro-
tection of the duodenal mucosa against luminal acid [33]. Indeed, perfusion of
the proximal duodenum with 20 mM HCl for 4 h caused only a few hemorrhagic
lesions in wild-type mice. Gene disruption of EP1 receptors did not affect the
duodenal ulcerogenic response to acid perfusion, and the lesion score was not
significantly different compared to that of wild-type mice. In EP3 receptor
knockout mice, however, the acid perfusion for 4 h provoked severe lesions over
almost the entire proximal duodenum; the lesion score was about 6 times
greater than that observed in wild-type littermates [8]. Certainly, increased duo-
denal ulcerogenicity to acid perfusion was also observed in wild-type mice after
indomethacin pretreatment. It is assumed that a decrease of HCO⫺ 3 secretion in
EP3 receptor knockout mice leads to a progressive breakdown of the mucosal
defensive response to acid and increases the mucosal susceptibility to acid
injury. Thus, the presence of EP3 receptors is essential for maintaining duode-
nal HCO⫺ 3 secretion and mucosal integrity against luminal acid. The involve-
ment of EP4 receptors in mucosal protection should be verified in a future
study.
Intestinal Cytoprotection by PGE2
NSAIDs such as indomethacin are known to cause intestinal damage,

including ulcers complicated by bleeding and perforation, in experimental ani-
mals and in humans. Although several factors have been postulated as patho-
genic elements of intestinal ulceration induced by indomethacin, including a
deficiency of PGs, bile acid, bacterial flora and NO [36], the exact mechanisms
remain unexplored. It is, however, certain that PG deficiency plays a critical
role in the pathogenesis of these lesions. Indeed, all these events caused by
35
n ⫽4~5
*p⬍ 0.05
30
Intestinal damage (mm2)
25
20
15
* dmPGE2
* dmPGE2
Saline
Saline
Saline
10
dmPGE2
5
0
Wild-type EP1(⫺/⫺) EP3(⫺/⫺)
Fig. 4. Effect of dmPGE2 on intestinal lesions induced by indomethacin in wild-type

mice and those lacking EP1 or EP3 receptors. The animals were given indomethacin
(30mg/kg) s.c., and killed 24 h later. dmPGE2 (30 ␮g/kg) was administered s.c. twice, 30 min
before and 9 h after indomethacin treatment. *Significantly different from saline in the
corresponding group, at p ⬍ 0.05.
indomethacin are effectively prevented by supplementation with exogenous

PGE2 [11].
Intestinal Cytoprotection against Indomethacin

Indomethacin caused hemorrhagic lesions in the rat small intestine, mainly
in the jejunum and ileum, accompanied by an increase in enterobacterial translo-
cation. The development of these lesions was prevented by pretreatment of the
animals with 16,16-dimethyl-PGE2 (dmPGE2) in a dose-dependent manner [11].
Other prostanoids such as ONO-NT-012 and ONO-AE1-329 also provided dose-
dependent protection against indomethacin-induced intestinal damage, while nei-
ther 17-phenyl-PGE2 nor butaprost had any effect on these lesions. These results
strongly suggest that the intestinal protection by dmPGE2 against indomethacin
is brought about by activation of EP3 and EP4 receptors, similar to the protective
action in the duodenum (table 2). We confirmed this using EP receptor knockout
mice and showed that dmPGE2 provided less protection against indomethacin-
induced intestinal damage in the animals lacking EP3 receptors, although the
agent exhibited potent inhibition in both wild-type and EP1 receptor knockout
mice (fig. 4) [37]. The fact that even in EP3 receptor knockout mice dmPGE2
provided partial protection against these lesions, supports the involvement of
another EP receptor subtype, EP4, in the protective action of dmPGE2.

Functional Alterations Associated with Intestinal Protection
Although multiple factors are implicated in the pathogenesis of
indomethacin-induced intestinal damage, enterobacteria and NO play a key
pathogenic role in this model; the release of bacterial products such as endo-
toxin contributes to the development of intestinal damage through overproduc-
tion of NO by upregulation of inducible NO synthase (iNOS) in the mucosa
[38]. Indeed, the prevention of these lesions was observed on blockade of NO
production through the inhibition of the iNOS activity by the NO synthase
inhibitor or the iNOS expression by dexamethasone [39]. It is also suggested
that NO interacts with the superoxide radicals to produce a cytotoxic peroxy-
nitrite, which has a deleterious influence on the intestinal mucosal integrity.
Certainly, the development of intestinal lesions as well as the bacterial translo-
cation and the upregulation of iNOS activity following treatment with
indomethacin were both potently prevented by supplementation with dmPGE2,
suggesting a pathogenic role for PG deficiency in this lesion model [11]. These
effects of dmPGE2 were reproduced by ONO-NT-012 and ONO-AEI-329 but
not by 17-phenyl-PGE2 or butaprost, confirming a close relationship between
intestinal protection and prevention of bacterial translocation as well as iNOS
activity.
It is known that mucin plays an important part in the innate host defense
against intestinal pathogens and irritants. We found that both dmPGE2, ONO-
NT-012 and ONO-AEI-329 increased mucus secretion in the small intestine,
suggesting involvement of EP3/EP4 receptors in the mucus stimulatory action
of PGE2 [11]. Belly and Chadee [40] demonstrated that PGE2 coupled to the
EP4 receptor stimulates cAMP-dependent mucin exocytosis in the rat colon.
Although the reason for these different results remains unknown, it might be
due to experimental conditions such as the tissues used. In any case, it is possi-
ble that PGE2 by stimulating the mucus secretion and by increasing the mucus
gel thickness may hamper the bacterial invasion in the mucosa, which is respon-
sible for excessive NOx production through induction of iNOS. In addition,
secretion of intestinal fluid may prevent the process of bacterial translocation,
by washing out these microorganisms. The enteropooling was increased by
both dmPGE2, ONO-NT-012 and ONO-AEI-329, suggesting stimulation of this
process by EP3 and EP4 receptors [11]. Since the amount of fluid accumulated
in the intestine can be affected by changes in secretion, absorption, transit and
the volume of upper gastrointestinal secretions, the interpretation of these
results should be limited. Yet, this event is largely influenced by intestinal fluid
(Cl⫺) secretion. Several studies reported the PGE2 effect on Cl⫺ secretion in the
gastrointestinal tract. Bunce and Spraggsl [41] reported that gastrointestinal
Cl⫺ secretion was stimulated by PGE2 through activation of both EP3 and EP4
receptors. Since prostanoids exhibiting a preference for EP3 and EP4 receptors
showed mucus and fluid stimulatory action and intestinal protection against
indomethacin, it is likely that these processes contribute to intestinal protec-
tion by PGE2, through suppression of bacterial translocation. Interestingly,
indomethacin caused a marked enhancement of intestinal motility, resulting in
an increase of both the amplitude and frequency of the contraction [11, 37, 42].
Because the spasmodic nature of the intestinal motility results in a disruption
of unstirred mucus layer over the epithelium, leading to an increase in mucosal
susceptibility to pathogens and irritants, the enhanced intestinal contractions
may also be part of the pathogenic mechanism for indomethacin-induced small
intestinal lesions. The enhanced intestinal motility caused by indomethacin was
antagonized by both dmPGE2 and another prostanoid specific to EP4 receptors.
Since EP4 receptors are coupled to adenylate cyclase, it is speculated that the
relaxation of circular smooth muscle by PGE2 is associated with an increase of
intracellular cAMP.
Thus, intestinal protection by PGE2 may be functionally associated with
stimulation of mucus and fluid secretions as well as inhibition of intestinal
hypermotility, the former two processes being mediated by both EP3 and EP4
receptors, the latter mediated by EP4 receptors (table 3). These functional
changes strengthen the barrier against intestinal pathogens and irritants, result-
ing in prevention of bacterial translocation and inhibition of the iNOS upregu-
lation, and by so doing prevent the development of small intestinal lesions.
Summary and Future Prospect
Endogenous PGs play a central role in the mucosal defensive mechanism

of the gastrointestinal tract, and among them PGE2 is most important in their
actions. This paradigm is largely based on the finding of ‘gastric cytoprotec-
tion’ by Robert et al. [2] in 1979. Since then, a number of studies have been
conducted to elucidate the factors involved in this phenomenon, yet the true
mechanism underlying this action remains still unexplored. As reviewed in this
chapter, exogenous PGE2 confers protection of the stomach against ulcerogenic
stimuli, irrespective of whether it is necrotizing agent (HCl/ethanol) or NSAID
(indomethacin), mainly through activation of EP1 receptors. As observed in
the adaptive cytoprotection induced by a mild irritant, endogenous PGE2 also
exhibits gastroprotection mediated by EP1 receptors. On the other hand, PGE2
affords protection of the intestinal mucosa, including the duodenum and the
small intestine, through activation of both EP3 and EP4 receptors. The under-
lying mechanism related to these actions of PGE2 in the stomach, duodenum or
small intestine may be related to inhibition of stomach contraction (EP1), stim-
ulation of duodenal alkaline secretion (EP3/EP4), or suppression of bacterial

translocation due to inhibition of intestinal contraction (EP4) as well as stimu-
lation of mucus secretion (EP3/EP4), respectively. Since the results introduced
in this chapter were obtained in rats using subtype-specific EP agonists and
were further confirmed in EP receptor knockout mice, they would be reliable
and have a high reproducibility when compared to those obtained in either rats
or knockout mice alone. Anyway, it is interesting to note that the EP receptor
subtypes responsible for cytoprotection are different depending upon the tissues
and that the functional alterations responsible for the protective action also
differ depending on the tissues. These approaches should contribute to further
understanding of the mechanism of ‘cytoprotection’ in the gastrointestinal tract
and also to future development of new strategies for treatment of peptic ulcer
diseases.
References
1 Miller TA: Protective effects of prostaglandins against gastric mucosal damage: Current knowl-
edge and proposed mechanisms. Am J Physiol 1983;245:G601–G623.
2 Robert A, Nezamis JE, Lancaster C, Hanchar AJ: Cytoprotection by prostaglandins in rats.
Prevention of gastric necrosis produced by alcohol, HCl, NaOH, hypertonic NaCl and thermal
injury. Gastroenterology 1979;77:433– 443.
3 Coleman RA, Kennedy I, Humphrey PPA, Bunce K, Lumley P: Prostanoids and their receptors;
in Emmett JC (ed): Comprehensive Medicinal Chemistry, Membranes and Receptors. Oxford,
Pergamon Press, 1990, vol 3, pp 643–714.
4 Sugimoto Y, Namba T, Honda A, Hayashi Y, Negishi M, Ichikawa A, Narumiya S: Cloning and
expression of a cDNA for mouse prostaglandin E receptor EP3 subtype. J Biol Chem 1992;
267:6463– 6466.
5 Watabe A, Sugimoto Y, Honda A, Irie A, Namba T, Negishi M, Ito S, Narumiya S, Ichikawa A:
Cloning and expression of cDNA for a mouse EP1 subtype of prostaglandin E receptor. J Biol
Chem 1993;27:20175–20178.
6 Narumiya S, Sugimoto Y, Ushikubi F: Prostanoid receptors: Structure, properties and functions.
Physiol Rev 1999;79:1193–1226.
7 Takeuchi K, Yagi K, Kato S, Ukawa H: Roles of prostaglandin E-receptor subtypes in gastric and
duodenal bicarbonate secretion in rats. Gastroenterology 1997;113:1553–1559.
8 Takeuchi K, Ukawa H, Kato S, Furukawa O, Araki H, Sugimoto Y, Ichikawa A, Ushikubi F,
Narumiya S: Impaired duodenal bicarbonate secretion and mucosal integrity in mice lacking
prostaglandin E receptor subtype EP3. Gastroenterology 1999;117:1128–1135.
9 Araki H, Yagi K, Suzuki K, Furukawa O, Takeuchi K: Roles of prostaglandin E receptor subtypes
in cytoprotective action of prostaglandin E2 in rat stomachs. Aliment Pharmacol Ther 2000;
14(suppl 1):18–25.
10 Suzuki K, Araki H, Mizoguchi H, Takeuchi K: E-type prostaglandin inhibits indomethacin-
induced gastric lesions in rats through EP1 receptors: Importance of antigastric motility action.
Digestion 2001;63:92–101.
11 Kunikata T, Umeda M, Tanaka A, Kato S, Takeuchi K: E type prostaglandin inhibits
indomethacin-induced small intestinal lesions through EP3 and EP4 receptors: A study using rats
and knockout mice (abstract). Gastroenterology 2000;118:A692.
12 Takeuchi K, Araki H, Komoike Y, Umeda M, Suzuki K: Adaptive gastric cytoprotection is
mediated by prostaglandin EP1 receptors: A study using rats and knockout mice. J Pharmacol Exp
Ther 2001;297:1160–1165.
13 Takeuchi K, Kato S, Ogawa Y, Kanatsu K, Umeda M: Role of endogenous prostacyclin in gastric
ulcerogenic and healing response: Study using IP receptor knockout mice. J Physiol Paris
2001;95:75–80.
14 Robert A, Nezamis JE, Lancaster C, Davis JP, Field SO, Hanchar AJ: Mild irritants prevent
gastric necrosis through ‘adaptive cytoprotection’ mediated by prostaglandins. Am J Physiol
1983;245:G113–G121.
15 Holzer P: Neural emergency system in the stomach. Gastroenterology 1998;114:828–839.
16 Takeuchi K, Niida H, Matsumoto J, Ueshima K, Okabe S: Gastric motility changes in capsaicin-
induced cytoprotection in the rat stomach. Jpn J Pharmacol 1991;55:147–155.
17 Uchida M, Yano S, Watanabe K: The role of capsaicin-sensitive afferent nerves in protective effect
of capsaicin against absolute ethanol-induced gastric lesions in rats. Jpn J Pharmacol 1991;
55:279–282.
18 Boku K, Ohno T, Saeki T, Hayashi H, Hayashi I, Katori M, Murata T, Narumiya S, Saigenji K,
Majima M: Adaptive cytoprotection mediated by prostaglandin I2 is attributable to sensitization of
CRGP-containing sensory nerves. Gastroenterology 2001;120:134–143.
19 Takeuchi K, Ueki S, Okabe S: Importance of gastric motility in the pathogenesis of indomethacin-
induced gastric lesions in rats. Dig Dis Sci 1986;31:1114–1121.
20 Yokotani K, Okuma Y, Osumi Y: Inhibition of vagally mediated gastric acid secretion by activa-
tion of central prostanoid EP3 receptors in urethane-anesthetized rats. Br J Pharmacol 1996;
117:653–656.
21 Takahashi S, Takeuchi K, Okabe S: EP4 receptor mediation of prostaglandin E2-stimulated mucus
secretion by rabbit gastric epithelial cells. Biochem Pharmacol 1999;58:1997–2002.
22 Ohno T, Katori M, Majima M, Saeki T, Boku K, Nishiyama K, Hayashi H, Saigenji K: Dilatation
and constriction of rat gastric mucosal microvessels through prostaglandin EP2 and EP3 receptors.
Aliment Pharmacol Ther 1999;13:1243–1250.
23 Mersereau WA, Hinchey EJ: Role of gastric mucosal folds in formation of focal ulcers in the rat.
Surgery 1982;91:150–155.
24 Takeuchi K, Ueshima K, Hironaka Y, Fujioka Y, Matsumoto J, Okabe S: Oxygen free radicals and
lipid peroxidation in the pathogenesis of gastric mucosal lesions induced by indomethacin in rats.
Digestion 1991;49:175–184.
25 Takeuchi K, Matsumoto J, Ueshima K, Okabe S: Gastric motility changes in the cytoprotective
action of N-ethyl-maleimide and capsaicin in the rat stomach; in Tsuchiya M (ed): Frontier of
Mucosal Immunology. Amsterdam, Elsevier Science, 1991, vol 2, pp 581–584.
26 Wallace JL, Granger DN: Pathogenesis of NSAID gastropathy: Are neutrophils the culprits? Trend
Pharmacol Sci 1992;3:129–131.
27 Roma A, Armstrong I: Investigation of inhibitory effects of PGE2 and selective EP agonists on
chemotaxis of human neutrophils. Br J Pharmacol 1995;116:2903–2908.
28 Melange R, Gentry C, Toseland N, Smith PH, Fuller J: Neutropenia does not prevent etodolac- or
indomethacin-induced gastrointestinal damage in the rat. Dig Dis Sci 1995;40:2694–2703.
29 Milenov K, Golenhofen K: Contractile responses of longitudinal and circular smooth
muscle of the canine stomach to prostaglandin E2 and F2␣. Prostaglandins Leukot Med 1982;
8:287–300.
30 Ding M, Kinoshita Y, Kishi K, Nakata H, Hassan S, Kawanami C, Sugimoto Y, Katsuyama M,
Negishi M, Narumiya S, Ichikawa A, Chiba T: Distribution of prostaglandin E receptors in the rat
gastrointestinal tract. Prostaglandins 1997;53:199–216.
31 Morimoto K, Sugimoto Y, Katsuyama M, Oida H, Tsuboi K, Kishi K, Kinoshita Y, Negishi M,
Chiba T, Narumiya S, Ichikawa A: Cellular localization of mRNAs for prostaglandin E receptor
subtypes in mouse gastrointestinal tract. Am J Physiol 1997;272:G681–G687.
32 Heylings JR, Garner A, Flemstrom G: Regulation of gastroduodenal HCO⫺ 3 transport by luminal
acid in the frog in vitro. Am J Physiol 1984;246:G235–G246.
33 Flemstrom G, Garner A: Gastroduodenal HCO⫺ 3 transport: Characteristics and proposed role in
acidity regulation and mucosal protection. Am J Physiol 1982;242:G183–G193.
34 Hirata T, Ukawa H, Kato S, Takeuchi K: Effects of selective cyclooxygenase-2 inhibitors on
acid-induced alkaline secretory and mucosal ulcerogenic responses in the rat duodenum. Life Sci
1997;61:1603–1611.

35 Takeuchi K, Matsumoto J, Ueshima K, Okabe S: Role of capsaicin-sensitive afferent neurons in
alkaline secretory response to luminal acid in the rat duodenum. Gastroenterology 1991;101:
954–961.
36 Reuter BK, Davies NM, Wallace JL: Nonsteroidal anti-inflammatory drug enteropathy in rats:
Role of permeability, bacteria and enterohepatic circulation. Gastroenterology 1997;112:109–117.
37 Kunikata T, Araki H, Takeeda M, Kato S, Takeuchi K: Prostaglandin E prevents indomethacin-
induced gastric and intestinal damage through different EP receptor subtypes. J Physiol Paris
2001;95:157–163.
38 Boughton-Smith N, Evans SM, Laszlo F, Whittle BJR, Moncada S: The induction of nitric oxide
synthase and intestinal vascular permeability by endotoxin in the rat. Br J Pharmacol 1993;110:
1189–1195.
39 Konaka A, Tanaka A, Kato S, Nishijima M, Takeuchi K: Nitric oxide, superoxide radicals and mast
cells in pathogenesis of indomethacin-induced intestinal lesions in rats. J Pharmacol Physiol
1999;50:25–38.
40 Belly A, Chadee K: Prostaglandin E2 stimulates rat and human colonic mucin exocytosis via the
EP4 receptor. Gastroenterology 1999;117:1352–1362.
41 Bunce KT, Spraggs CF: Prostanoids stimulation of anion secretion in guinea-pig gastric and ileum
mucosa is mediated by different receptors. Br J Pharmacol 1990;101:889–895.
42 Takeuchi K, Hase S, Mizoguchi H, Miyazawa T, Tanaka A: Protection by aspirin of indomethacin-
induced small intestinal damage in rats: Mediation by salicylic acid. J Physiol Paris 2001;95:
51–57.
Koji Takeuchi, PhD, Department of Pharmacology and Experimental Therapeutics,

Kyoto Pharmaceutical University, Misasagi, Yamashina, Kyoto 607-8414 (Japan)
Tel. ⫹81 75 5954679, Fax ⫹81 75 5954774, E-Mail takeuchi@mb.kyoto-phu.ac.jp
Author Index
Baatar, D. 101 Lam, S.K. 143 Shin, V.Y. 180

Basson, M.D. 1 Lamarque, D. 190 Szabo, S. 209
Brzozowski, T. 158 Liu, E.S.L. 180
Lotz, M.M. 14 Taher, M.M. 69
Cavicchi, M. 190 Takeuchi, K. 227
Cho, C.C.M. 180 Ma, L. 117 Tanaka, A. 227
Cho, C.H. 180 Matthews, J.B. 14 Tarnawski, A.S. 101
Coy, L.M. 69 McCormack, S.A. 43 Tso, P. 129
Miller, T.A. 69 Turner, J.R. 1
Deng, X. 209
Osada, T. 57 Wallace, J.L. 117
Wang, J.-Y. 29, 82
Ishikawa, H. 209 Watanabe, S. 57
Pai, R. 101
Peskar, B.M. 166 Whittle, B.J.R. 190
Johnson, L.R. 43 Wollin, A. 129
Jones, M.K. 101 Rao, J.N. 29 Wong, B.C.Y. 143
Ray, R.M. 43
Kato, S. 227 Redlak, M.J. 69 Xia, H.H.X. 143
Khomenko, T. 209
Konturek, P.C. 158 Sandor, Z. 209 Yoo, J. 14
Konturek, S.J. 158 Sato, N. 57
243
Subject Index
Actin nonsteroidal anti-inflammatory drug

cellular forms 48, 49 inhibition 112, 113
epidermal growth factor, cytoskeletal platelet factor modulation
change induction in mucosal repair endostatin 121–123
22, 23 thrombospondin 123, 124
mechanical strain effects in gastric signal transduction 111, 112
wound repair 64, 65 submucosal vessels 101, 210
polyamine effects on cytoskeleton 49–51 vascular damage in mucosal injury 103,
S-Adenosylmethionine decarboxylase 104, 210
inhibitors 45, 46, 48, 49 vascular endothelial growth factor
polyamine biosynthesis 44, 45 response 105–107, 109–112
Angiogenesis AP-1, protooncogene product
alcohol injury response and growth interactions 84
factors 104, 105 Apoptosis
angiogenic factors Helicobacter pylori induction in
angiopoietins 110–112 epithelial cells 147, 148
basic fibroblast growth factor 110 ischemia/reperfusion injury response in
gene therapy 217–223 intestinal mucosa 137, 138
peptide therapy Aspirin, see also Nonsteroidal
basic fibroblast growth factor anti-inflammatory drugs
212–215 challenge protection by
overview 210–212 diacylglycerol/protein kinase C
platelet-derived growth factor 215, pathway 71, 72
216 gastric damage mechanisms 158, 159
vascular endothelial growth factor
216, 217 Basic fibroblast growth factor
platelet-derived growth factor 110 angiogenesis promotion 110
prostaglandins 112 peptide therapy
vascular endothelial growth factor 110 chronic duodenal ulcer 212, 213
granulation tissue and ulcer healing 109 chronic gastric ulcer 213
heparin modulation 182, 183 chronic gastritis 213, 214
hyperemic response to injury 101, 102 inflammatory bowel disease 214, 215
244
ulcer healing role 118 gastric mucosal damage minimization
Bicarbonate, prostaglandins in duodenal and inhibition effects 169–171,
cytoprotection 235, 236 175, 176
Helicobacter pylori induction of
COX-2 172
Calcium signaling, epithelial restitution
inhibitor specificity 166, 167
cell mobility modulation 30, 32–34
intestinal immune response and
diacylglycerol/protein kinase C pathway,
inflammation role of COX-2
see Diacylglycerol/protein kinase C
171, 172
pathway
isoforms 162, 163, 166
homeostasis regulation
Cytokines, see also specific cytokines
membrane potential and calcium
cell-cell interaction alterations in
driving force 34–36, 39
mucosal repair 23, 24
transient receptor potential
cell-matrix interaction alterations in
channels 36
mucosal repair 24–26
voltage-gated calcium channels 34, 35
cell migration stimulation in mucosal
in vitro studies 31–34
restitution 17–21
in vivo studies 30, 31
cytoskeletal change induction in mucosal
intracellular sources 29
repair 22, 23
polyamine signaling interactions 49, 50
mucosal repair regulation 21, 22
prospects for study 40
receptor signaling 17
targets
secretion at mucosal injury site 15, 17
overview 36
Cytoskeleton, see Actin, Myosin
Rho 37, 38
effectors 39
Capsaicin, gastroprotection 230, 231 Diacylglycerol/protein kinase C pathway
␤-Catenin, cell-cell interaction alterations aspirin challenge gastric protection
in mucosal repair 24 studies 71, 72
Cornea, protooncogene role in healing intestinal injury protection
88, 89 Caco-2 cell studies 72–74
Cyclooxygenase colonic mucosal sample studies 75
COX-1 inhibition experimental colitis model 75
gastric mucosa effects 167, 168 IEC-18 studies 75, 76
gastroprotective agent interactions modeling of gastrointestinal injury and
172, 173 protection 78, 79
mucosal defense effects 168, 169 OAG protection against cellular injury
mucosal repair and healing effects 70, 74
174, 175 prostaglandin cytoprotection
COX-2 inhibition modulation 70
comparison with NO-NSAIDs protein kinase C isoforms and families
162, 163 gastric cytoprotection 72
gastric mucosa effects 168 injury vs protection roles 76–78
gastroprotective agent interactions intestinal cytoprotection 74–76
172, 173 overview 70
mucosal defense effects 168, 169
mucosal repair and healing effects Endostatin, angiogenesis modulation
173–175 121–123
rationale for specific inhibition 167 EP receptors, see Prostaglandins
Subject Index 245

Epidermal growth factor vascular endothelial growth factor
actin cytoskeletal change induction in chronic duodenal ulcer 221, 222
mucosal repair 22, 23 chronic gastric ulcer 222
cell-cell interaction alterations in inflammatory bowel disease 222, 223
mucosal repair 24 vectors 217, 218
Helicobacter pylori
enterocyte migration role 5, 6
COX-2 induction 172
receptor signaling 23
eradication regimens 153, 154
recombinant protein therapy 149
gastroduodenal mucosa injury
ulcer healing role 118, 119
acid secretion induction 146, 147
Epithelium
apoptosis of epithelial cells 147, 148
barrier function, restoration phases 14
inflammation and immune response
intestinal, see Intestinal epithelium
146, 209
restitution, see Wound healing
virulence factors
ERM proteins, cytoskeletal regulation 39
alcohol dehydrogenase 145
Ethanol, cytoskeletal effects in gastric
CagA 145
wound repair 6
lipopolysaccharide 145
Extracellular matrix
phospholipases 145
components 24, 25
proteases 145
cytokines, cell-matrix interaction
urease 144
alterations in mucosal repair 24–26
VacA 144, 145
intestinal epithelium healing
healing inhibition
effects 4, 5
cell proliferation effects 150
mechanical strain effects in gastric
growth factor interference 148–150
wound repair 65–67
mucosal healing 150, 151
protease functions 26
inducible nitric oxide synthase gene
Extracellular-regulated kinase
induction 199, 200
integrin receptor signaling 6, 7
nonsteroidal anti-inflammatory drug
myosin light chain kinase as target 7
interactions
mucosal adaptation to aspirin 152
Fibroblast, cytokine secretion 15, 17 mucosal healing 152
Fish oil, ischemia/reperfusion injury ulceration and bleeding incidence
protection of intestinal mucosa 138, 139 151, 152
Focal adhesion kinase use effects on infection incidence 151
integrin receptor signaling 6, 7 prevalence of infection 143
mechanical strain effects in gastric Heme oxygenase-1
wound repair 65, 66 inducible nitric oxide interactions 202,
polyamine modulation 51, 52 203
c-Fos, see Protooncogenes inflammation role 202
Heparin
Gene therapy angiogenesis modulation 182, 183
definition 217 anticoagulation 182
platelet-derived growth factor cell proliferation effects 183
chronic duodenal ulcer 218, 219 Helicobacter pylori binding 149
inflammatory bowel disease 219–221 inflammation inhibition 183, 184
prospects 223 structure 182
Subject Index 246

Hepatocyte growth factor, ulcer healing epithelial cell phenotypic shift 2
role 118, 120, 121 extracellular matrix, direct and
Histamine, ischemia/reperfusion injury independent effects in healing 4, 5
response in intestinal mucosa 134–136 fence assay of healing 4
Hypoxia-inducible factor-1, vascular integrins
endothelial growth factor gene activation cell migration role 5, 6
107, 108 receptor signal transduction 6, 7
mucosal sheet model 3
prospects for study 10, 11
Inflammatory bowel disease
wound healing overview 1, 2
gene therapy
mechanical strain effects 63
platelet-derived growth factor 219–221
polyamines in restitution,
vascular endothelial growth factor
see Polyamines, intestinal epithelium
222, 223
restitution
growth factor therapy, see specific
Ischemia/reperfusion injury, intestinal
factors
mucosa
inducible nitric oxide synthase detection
apoptosis of enterocytes 137, 138
animal models 191, 192
assessment 129
human studies 192
fish oil protection 138, 139
Insulin-like growth factor, ulcer healing
lipid absorption
role 118
mechanisms of impairment 132, 133
Integrins
monitoring of injury 130, 139, 140
cell migration and wound closure role in
rat studies 130–132
intestine 5, 6
phases 129
repair and restoration 134–137
mucosal repair 25
ITF
mechanical strain effects in gastric
cell-cell interaction alterations in
wound repair 65, 66
polyamine role in signaling 51
epithelial cell restitution role 21
receptor signal transduction 6, 7
Interferon-␥, epithelial cell restitution
role 18 c-Jun, see Protooncogenes
Interleukin-1, epithelial cell restitution role
interleukin-1␣ 19, 20 Keratinocyte growth factor, epithelial cell
interleukin-1␤ 18 restitution role 20
Interleukin-2
epithelial cell restitution role 17, 18 Macrophage, cytokine secretion 15
sources 17 Mechanical strain
Interleukin-4, endothelial activation 89 airway epithelium effects 62, 63
Interleukin-8, epithelial cell restitution gastric forces 57
role 20 gastric wound repair
Intestinal epithelium clinical implications 67
cell lines 3 Flexercell Strain Unit stretching of
functions 1 cells 59, 61, 62
injury, see also Ischemia/reperfusion mechanisms of inhibition
injury, intestinal mucosa cytoskeletal system 64, 65
actomyosin contraction in cell extracellular matrix adhesion
movement 7–10 65–67
Subject Index 247

Mechanical strain (continued) Nitric oxide synthase
gastric wound repair (continued) cell distribution 190
RGM-1 cell model 58, 59, 61, 62 inducible enzyme
smooth muscle effects 62 cytoprotective vs cytotoxic effects
growth factor induction 57 194, 195
intestinal epithelium effects 63 detection in inflammatory bowel
Mesenchymal cell, cytokine diseases
secretion 15 animal models 191, 192
Mitogen-activated protein kinase, human studies 192
see also Extracellular-regulated kinase endothelial and epithelial injury role
mechanical strain effects in gastric 192–194
wound repair 65, 66 heme oxygenase-1 interactions 202,
Ras pathway in vascular endothelial 203
growth factor activation 106, 107 induction of gene expression
vascular endothelial growth factor Helicobacter pylori 199, 200
signaling 110, 111 lipopolysaccharide 200, 201
Mucin, prostaglandin regulation of inhibition rationale 203
secretion 238, 239 inhibitor studies 196–198
Mucosal restitution, see Wound healing knockout consequences 198
c-Myc, see Protooncogenes peroxynitrite cytotoxic effects 195, 196
Myosin isoforms 102, 103, 190
intestinal epithelial cell Nonsteroidal anti-inflammatory drugs,
contraction and migration 7, 8 see also Aspirin
myosin light chain kinase role in angiogenesis inhibition 112, 113
movement 8–10 gastric damage mechanisms 158, 159
phosphorylation 7, 8 Helicobacter pylori interactions
polyamine effects on cytoskeleton 49, mucosal adaptation to aspirin 152
50, 52 mucosal healing 152
ulceration and bleeding incidence
Nitric oxide 151, 152
cytoprotective vs cytotoxic effects 194, use effects on infection incidence 151
195, 238 nitric oxide conjugates
donors 159, 160 aspirin conjugates 161
immune response 191 cyclooxygenase-2 inhibitor
mucosal homeostasis role 159, 160 comparisons 162, 163
mucosal injury response 102 mechanisms of action 161, 162
nonsteroidal anti-inflammatory drug rationale for use 160, 163, 164
conjugates prostaglandin protection against damage,
aspirin conjugates 161 see Prostaglandins
cyclooxygenase-2 inhibitor Nuclear factor-␬B, gastric ulcer healing
comparisons 162, 163 role 87
mechanisms of action 161, 162
rationale for use 160, 163, 164 Ornithine decarboxylase
peroxynitrite cytotoxic effects 195, Dong Quai induction 185, 186
196, 201 inhibitors 45, 90, 91
superoxide interactions 201, 202 ischemia/reperfusion injury response in
synthases, see Nitric oxide synthase intestinal mucosa 134, 137
Subject Index 248

polyamine biosynthesis 44, 45 RhoA interactions 50
stress effects on expression 91 toxicity 45, 46
transport 46
Peroxynitrite, cytotoxic effects 195, 196, Polysaccharides, gastrointestinal protection
201 Dong Quai polysaccharides 185, 186
Phosphatidylinositol 3⬘-kinase, ginseng polysaccharides 185, 186
angiogenesis signaling 111, 112 heparin, see Heparin
Platelet mushroom polysaccharides 184, 185
angiogenesis modulation seaweed polysaccharides 184, 185
endostatin 121–123 sources and pharmacological actions,
thrombospondin 123, 124 overview 180, 181
granules 118 ulcer formation and healing,
ulcer healing role polysaccharide mechanisms 187
accumulation at injury site 125, 126 Prostaglandins
clinical manipulation 126, 127 cyclooxygenase inhibition effects,
growth factors 118–121 see Cyclooxygenase diacylglycerol/
overview 124, 125 protein kinase C signaling,
Platelet-derived growth factor see Diacylglycerol/protein kinase C
angiogenesis promotion 110 pathway
gene therapy inhibition in mucosal damage 158, 159
chronic duodenal ulcer 218, 219 metabolism 166
inflammatory bowel disease 219–221 PGE2
peptide therapy angiogenesis promotion 112
chronic duodenal ulcer 215 calcium modulation of effects 32, 33
chronic gastric ulcer 215, 216 duodenal cytoprotection 235, 236
chronic gastritis 216 EP receptor subtypes and agonists
inflammatory bowel disease 216 227, 228, 231, 232, 239, 240
ulcer healing role 118, 120 gastric cytoprotection
Polyamines, intestinal epithelium restitution acid secretion inhibition 231
biosynthesis 44, 45 EP receptor knockout effects
calcium signaling interactions 49, 50 229–231, 240
cell migration and proliferation hydrochloric acid/ethanol-induced
effects 43 gastric damage 228–231
cytoskeletal modulation 49–52 indomethacin-induced gastric
deficiency 46, 47 damage 231
discovery 43 motility effects 232–234
in vitro models 47, 48 neutrophil inhibition 234
inhibitor studies 46, 48, 49 overview 69, 70
integrin signaling role 51 intestinal cytoprotection
model 52 functional alterations in protection
nucleic acid interactions 90, 95 238, 239
protooncogene expression regulation indomethacin-induced damage 236,
inhibitor studies 91, 93, 94 237
posttranscriptional regulation 95–97 synthesis 167, 227
spermidine response 93, 94 therapeutic prospects 240
transcriptional regulation mechanisms Protein kinase C, see Diacylglycerol/
95 protein kinase C pathway
Subject Index 249

Protooncogenes Thrombospondin, angiogenesis modulation
cell cycle regulation 82, 83 123, 124
corneal healing role 88, 89 Transforming growth factor-␣, ulcer
endothelial cell expression in response to healing role 118–120
injury 89 Transforming growth factor-␤
heterodimerization of proteins 83, 84 cell-matrix interaction alterations in
polyamine regulation of expression mucosal repair 25, 26
posttranscriptional regulation 95–97 epithelial cell restitution role 18, 19, 22
regulation inhibitor studies 91, 93, 94 ulcer healing role 118
spermidine response 93, 94 Tumor necrosis factor-␣, epithelial cell
transcriptional regulation mechanisms restitution role 19, 20
95
prospects for wound healing studies 97, Ulcer healing, see Wound healing
98
ulcer healing role Vascular endothelial growth factor
cell renewal and ulceration 84, 85 angiogenesis promotion and signaling
cellular sources 87, 88 110–112
expression antibody inhibition studies of
c-fos 85, 86 angiogenesis 105, 106
c-jun 87 ethanol triggering of overexpression
c-myc 85–87 105
wounding response and activation of gene activation
expression 83, 84 hypoxia 107, 108
Putrescine, see Polyamines, intestinal Ras pathway 106, 107
epithelium restitution gene therapy
chronic duodenal ulcer 221, 222
Rebamipide, cyclooxygenase interactions chronic gastric ulcer 222
172, 173 inflammatory bowel disease 222,
Rho 223
activation 37 mechanical strain induction 57
calcium regulation 37, 38 peptide therapy
effectors 39 chronic duodenal ulcer 216, 217
mechanical strain effects in gastric inflammatory bowel disease 217
wound repair 65, 66 structure 118
RhoA ulcer healing role 118, 119
cell migration role during wound
healing 37, 40 Wound healing
polyamine interactions 50 angiogenesis, see Angiogenesis
calcium signaling, see Calcium
Spermidine, see Polyamines, intestinal signaling, epithelial restitution
epithelium restitution cyclooxygenase role,
Spermine, see Polyamines, intestinal see Cyclooxygenase
epithelium restitution cytokine role, see Cytokines
Stress, see Mechanical strain diacylglycerol/protein kinase C pathway,
Superoxide dismutase, nitric oxide see Diacylglycerol/protein kinase C
cytotoxicity protection 201, 204 pathway
Subject Index 250

early primary response genes, intestinal epithelium, see Intestinal
see Protooncogenes epithelium
epithelial barrier function, restoration polyamines in restitution,
phases 14 see Polyamines, intestinal epithelium
gene therapy, see Gene therapy restitution
growth factor therapy, see specific stress effects, see Mechanical strain
factors
Helicobacter pylori effects,
see Helicobacter pylori
Subject Index 251

Gastrointestinal - Mucosal.repair - And.experimental - Therapeutics Ublog - TK

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gastrointestinal - Mucosal.repair - And.experimental - Therapeutics Ublog - TK

Uploaded by

Copyright:

Available Formats

Gastrointestinal Mucosal Repair and Experimental Therapeutics

I.M. Modlin New Haven, Conn.

C.-H. Cho Hong Kong

57 figures, and 9 tables, 2002

Basel · Freiburg · Paris · London · New York ·

Library of Congress Cataloging-in-Publication Data

© Copyright 2002 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland)

Part 1: Epithelial Restitution

1 Roles of Extracellular Matrix and Cytoskeleton in Intestinal Epithelial

Part 2: Mucosal Repair and Ulcer Healing

82 Expression of Early Primary Response Genes in Healing of

Part 3: Experimental Therapeutics

158 Nitric Oxide-Releasing Agents – A New Generation of Drugs for

243 Author Index

Chi-Hin Cho, PhD

Roles of Extracellular Matrix and

an intact cytoskeleton. While each of these events is critical to ultimate healing

In vivo Processes and in vitro Models

As noted above, migrating epithelial cells abandon their columnar shape

Extracellular Matrix and the Cytoskeleton 3

The earliest phase of intestinal epithelial restitution involves rapid migra-

Differences in migration rates over different matrixes are, in part, deter-

Extracellular Matrix and the Cytoskeleton 5

Signal Transduction Processes

When bound to extracellular ligands, integrin receptors cluster and activate

Although intracellular signaling processes and membrane receptors that

Extracellular Matrix and the Cytoskeleton 7

migration. In the absence of tMLCK expression, lamellipodia were easily iden-

Extracellular Matrix and the Cytoskeleton 9

Intestinal epithelial biology is profoundly affected by complex interactions

Extracellular Matrix and the Cytoskeleton 11

Marc D. Basson, MD, PhD, Chief, Surgical Service (11S),

Extracellular Matrix and the Cytoskeleton 13

Cytokines play a critical role in the response to mucosal injury, regulat-

Cytokines Are Secreted at Sites of Mucosal Injury

Various lines of evidence suggest that restitution occurs on a routine basis

[Cytoskeleton] [Zonula adherens]

Fig. 2. Cytokines regulate restitution by leading to changes in cell structure, cell-cell

Cytokines Stimulate Migration

The contributory role of cytokines in restitution has been demonstrated in

IL-2 is a key cytokine involved in the initiation of the immune response,

IL-1␤ and IFN-␥

IL-1␤ and IFN-␥ are found at sites of mucosal injury or inflammation.

IL-1␣ and TNF-␣

KGF-2 (FGF-10) is a member of the FGF family. It is secreted by fibro-

In response to mucosal injury, IL-8, like many other cytokines, appears to

Cytokines Promote Other Pro-Migratory Factors

Cytokines Regulate Early and Late Phases of Mucosal Repair

Cytokines may regulate the process of mucosal healing by coordinating the

Cytokines Lead to Changes in Cell Structure

Following mucosal injury, epithelial cells abutting the damage rapidly

Cytokines Alter Cell-Cell Interactions

In an intact monolayer, epithelial cells form a selectively permeable barrier

Cytokines Alter Cell-Matrix Interactions

Jeffrey B. Matthews, MD, FACS, 231 Albert B. Sabin Way,

Ca2ⴙ Signaling in Epithelial Restitution

Ca2ⴙ in Restitution in vivo

Restitution of gastrointestinal mucosa was originally studied in vivo and

Ca2ⴙ in Restitution in vitro

Gastrointestinal epithelial restitution is a complex process, for it necessi-

Ca2 Signaling in Epithelial Restitution 31

Fig. 1. Effects of removal of extracellular Ca2 (a) and 4-aminopyridine (4-AP; b) on

ileum [31]. Lysophosphatidic acid (LPA) stimulates intestinal epithelial cell