Acta Biomaterialia 7 (2011) 3237–3247

Contents lists available at ScienceDirect

Acta Biomaterialia
journal homepage: www.elsevier.com/locate/actabiomat

Review

Regeneration and repair of tendon and ligament tissue using collagen

fibre biomaterials
S.J. Kew a,⇑, J.H. Gwynne b, D. Enea c, M. Abu-Rub d, A. Pandit d, D. Zeugolis d, R.A. Brooks c,
N. Rushton c, S.M. Best b, R.E. Cameron b
a
Tigenix Ltd., Byron House, Cambridge Business Park, Milton Road, Cambridge CB4 0WZ, UK
b
Department of Materials Science and Metallurgy, University of Cambridge, Pembroke Street, Cambridge CB2 3QZ, UK
c
Orthopaedic Research Unit, University of Cambridge, Box 180, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 2QQ, UK
d
Network of Excellence for Functional Biomaterials (NFB), National University of Ireland Galway (NUI Galway), Galway, Ireland

a r t i c l e i n f o a b s t r a c t

Article history: Collagen fibres are ubiquitous macromolecular assemblies in nature, providing the structures that sup-
Received 22 December 2010 port tensile mechanical loads within the human body. Aligned type I collagen fibres are the primary
Received in revised form 25 May 2011 structural motif for tendon and ligament, and therefore biomaterials based on these structures are con-
Accepted 1 June 2011
sidered promising candidates for mediating regeneration of these tissues. However, despite considerable
Available online 13 June 2011
investigation, there remains no collagen-fibre-based biomaterial that has undergone clinical evaluation
for this application. Recent research in this area has significantly enhanced our understanding of these
Keywords:
complex and challenging biomaterials, and is reinvigorating interest in the development of such struc-
Collagen fibre
Tendon
tures to recapitulate mechanical function. In this review we describe the progress to date towards a lig-
Ligament ament or tendon regeneration template based on collagen fibre scaffolds. We highlight reports of
Regenerative medicine particular relevance to the development of the underlying biomaterials science in this area. In addition,
Review the potential for tailoring and manipulating the interactions between collagen fibres and biological sys-
tems, as hybrid biomaterial–biological ensembles, is discussed in the context of developing novel tissue
engineering strategies for tendon and ligament.
Ó 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction However, these approaches inevitably result in damage and conse-

Tendon and ligament regeneration has proved an elusive goal for
tissue engineering owing to the specialised nature of these tissues
and the high mechanical demands placed on the extracellular matrix
(ECM) of these structures in the human body. There is a significant
clinical need for augmentative and substitutional approaches that
enhance the structural performance of damaged and degenerated
tissues. The clinical problem may be chronic or acute, and may occur
in a range of extra- and intra-articular environments. Furthermore,
the mechanical demands supported by these structures vary within
a wide range, making a generic reconstructive approach particularly
challenging. Hence, several different surgical modalities and fixation
strategies have been developed to partially replicate the structure
and function of these critical tissues.
Surgical autografting procedures have been developed that can
be used in a number of anatomical locations with considerable
success; for example, the anterior cruciate ligament (ACL) can be
replaced with bone–patellar–bone [1] or hamstring tissue [2].
⇑ Corresponding author. Tel.: +44 0 1223 438254.
E-mail address: sjkew@hotmail.com (S.J. Kew).
quent morbidity at the donor site, necessitating a second invasive
procedure [3,4]. Furthermore, sacrifice of these tissues may pre-
clude a secondary revision operation and, perhaps most signifi-
cantly, can cause pain and impair the native biomechanics of the
harvest site [5]. Despite these drawbacks, autografts remain the
‘‘gold-standard’’ option for reconstructive tendon and ligament
surgeries owing to the high mechanical strength of the tissues,
excellent compatibility with the host tissues and their propensity
to revascularise and remodel when transplanted [6]. However,
the process of biological remodelling remains incompletely under-
stood [7], with in vivo studies indicating a multifactorial nature of
the repair, but clearly implicating mechanical loading [8] and
hence the rehabilitation protocol as significant factors for the
success of the procedure in humans [9].
Allograft tissues are increasingly being used clinically for the
replacement of ruptured ACL, with some success reported for
short-term outcomes [10]. However, these tissues are expensive,
limited in availability and carry a limited risk of disease transmis-
sion from the donor. Furthermore, the sterilisation procedures
employed to prepare these tissues are not standardised and can
have a negative effect on biomechanical properties, leading to
increased graft failure compared with autograft [11]. Xenograft

1742-7061/$ - see front matter Ó 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.actbio.2011.06.002
3238 S.J. Kew et al. / Acta Biomaterialia 7 (2011) 3237–3247
materials have also been investigated for ACL reconstruction, with
the overall aim being to provide full mechanical substitution. How-
ever, the ability to recapitulate the mechanical properties of the
ACL was limited and rupture of the graft was observed in a number
of cases [12]. Replacement of ligaments such as the ACL with syn-
thetic polymers has been attempted and, whilst offering adequate
initial mechanical strength, they have been largely unsuccessful
owing to long-term deterioration of mechanical properties and
inflammatory response due to release of degradation products
[13–16].
Biological ECM materials have been investigated as degradable
scaffolds for augmentative tendon repair, utilising decellularised
xenograft or allograft tissues to promote native tendon repair pro-
cesses [17]. There has been some in vivo evidence indicating that
tendon remodelling can be mediated by these matrices [18]. How-
ever, despite their prevalence in the marketplace, ECM-derived
materials have yet to find widespread acceptance for this applica-
tion. The results from clinical usage of ECM materials have to date
been mixed; the various commercially available products have
been reviewed by Chen et al. who described the clinical issues with
these materials in detail [19]. Significant drawbacks associated
with the clinical use of xenograft/allograft ECM materials include
poor mechanical properties compared to tendon (i.e. not mechan-
ically supporting the repair), an antigenic effect of the residual
non-collagenous ECM components, and potential for pathogenic
transmission of virus or bacterial species [20]. In particular, the
use of decellularised tissue can be associated with retention of
antigenic components [21], which may cause an inflammatory re-
sponse and potentially lead to serious clinical complications, such
as post-operative oedema [22].
It is widely recognised that the natural three-dimensional ma-
trix of connective tissues, which surrounds the musculoskeletal
cells, plays a key biological and mechanical role [23]. Collagen is
the principle load-bearing component of the extracellular matrix
and confers high mechanical strength to tissues such as tendon.
It is therefore theoretically capable of providing the structural
properties required for load bearing in musculoskeletal tissues
[24]. Unfortunately, the mechanical properties of collagen gels or
porous collagen fabricated via either biological or physical means
are vastly inferior to those of biological tissues. Furthermore, the
ability of cells to rapidly synthesise load-bearing collagen matrix
is limited [25]. Therefore, fabrication of aligned, mechanically
strong collagen fibres constitutes an attractive first step in
achieving a ‘‘bottom-up’’ fabrication method for biomaterials that
mimic both the mechanical and biological ligament or tendon
environment.
Aligned collagen fibre scaffolds based on self-assembled colla-
gen were first reported by Kato et al. who found that a homoge-
nised insoluble collagen gel could be extruded as a fibre by
utilising fibrillogenesis in vitro to fabricate pseudo-native struc-
tures [26]. The alignment of fibrous constructs, on both the nano-
and micro-scales, has been shown in numerous studies to direct
the axial alignment of tissue synthesis and is a useful scaffolding
motif for aligned tissues [27]. Furthermore, misalignment of colla-
gen fibrils in repair tissue is believed to be associated with the
mechanical inferiority of these tissues [28,29]. However, it has
proved challenging to replicate the mechanical strength of native
collagen in vitro and, importantly, to achieve adequate long-term
mechanical properties in vivo, which has led to the exploration
of synthetic [30] and collagen/synthetic composites [31]. It has
remained an elusive goal for the field to replicate the structural
features of the native collagen fibre, in particular the nanoscale
features of the collagen fibril recently characterised in detail and
summarised in Fig. 1B [37,110]. In this review we describe the
recent research activity focusing on novel biomaterials fabrication
processes using ‘‘bottom-up’’ aligned collagen scaffold fabrication.
Particular emphasis is placed on recent developments in the field
relating to the development of the collagen fibres and their under-
lying biomaterials science, crosslinking strategies and biological
interactions. In addition, we discuss the interactions between col-
lagen fibre materials and biological systems both in vivo and
in vitro. Finally, we review the prospects for the development of
collagen fibre/biologic combinations, which may enable novel
modes of ligament and tendon regeneration and provide routes
by which biofabricated scaffold materials can be developed.
2. Collagen fibre structures in tendon and ligament
Collagen is the most abundant protein in vertebrates, account-
ing for approximately 30% of all body proteins; it is a key and ubiq-
uitous component of the ECM, providing the tensile strength
required to fulfil the demanding biomechanical requirements of
human tissues. Despite extensive investigations of synthetic mate-
rials for biomedical applications, purified and crosslinked collagen
remains a preferred base material for ligament/tendon tissue engi-
neering owing to its low antigenicity [32], chemotactic surface
structure for fibroblasts [33], biocompatibility and proteolytic deg-
radation pathways [34,35]. The crosslinking process may, however,
reduce the number of cellular attachment sites (binding via cell
surface receptors such as integrins) if the amino acid residues are
utilised or occluded in the reaction; often making an inverse rela-
tion between mechanical properties and biocompatibility.
Tendons and ligaments connect bone to muscle and bone to
bone, respectively, and are predominantly composed of type I col-
lagen fibrils arranged such that they are axially load bearing. Their
structural hierarchy, which is illustrated in Fig. 1A–C, is exceed-
ingly complex [36]. Recent detailed fibre X-ray analysis of the col-
lagen structure has characterised further complexity at the
microfibrillar level, which indicates a super-twisted discontinuous
right-handed helical structure that forms inter-unit packing via an
interdigitated motif [37]. The fibril diameters vary from 10 to
500 nm, depending on a variety of biological factors. Tendon and
ligament additionally exhibit a characteristic nanoscale axial band-
ing pattern of light and dark regions, observed using transmission
electron microscopy or atomic force microscopy with a 64–67 nm
periodicity (Fig. 2C). This is due to the axial alignment of collagen
fibrils that result from alternating overlap and gap zones, produced
by the specific packing arrangement of the 300 nm long, 1.5 nm
diameter collagen molecules. The ability of type I collagen to form
striated fibrils is complex and involves specific charge–charge and
hydrophobic interactions [38–40]. Microscopically, collagen is
birefringent, and appears bright when viewed between crossed
polarising filters due to its oriented structure. In addition, a further
microscopic axial periodic zigzag pattern arises due to a crimped
structure, in which straight fibrils are kinked with respect to the
axis of orientation [41]. Under tensile stress, the crimps gradually
disappear until they are no longer microscopically visible [28,42],
suggesting that extension of tendon and ligament initially involves
straightening of the crimps. As shown in Fig. 1D, this gives rise to a
toe region in the force–time curve for human ligaments, which is
followed by a linear region in which the tissue is reversibly extend-
able and then finally by yielding and failure.
Collagen type I can be extracted and purified from connective
tissues and reconstituted. It is extracted as an insoluble crosslinked
material or as soluble collagen, which can be either acid solubilised
or enzyme solubilised (customarily pepsin extracted via cleavage
of telopeptides). The extraction of collagen typically involves the
acidification of a protein gel or suspension to form a fraction that
can be reprecipitated. It has been known for over 50 years that
collagen fibres assemble at physiological pH and temperature to
form fibril bundles that exhibit the periodic nanoscale D-banding
 S.J. Kew et al. / Acta Biomaterialia 7 (2011) 3237–3247 3239

Fig. 1. Properties and structure of human tendon/ligament tissues. (A) Hierarchical assembly. Structure of tendon collagen showing the organisation of different length scales
in the material. Adapted from Ref. [36]. (B) Structure of the type I collagen fibril. The left image shows the type I collagen unit cell determined using X-ray diffraction. The cell
is compressed 5 in the c-direction for viewing. The unit cell contains five triple helices, assembled into a supertwisted discontinuous collagen microfibre [37]. The right
image shows an AFM image of a disassembling D-banded type I collagen exhibiting a ‘‘rope-like’’ substructure [110]. (C) Collagen fascicle. Scanning electron micrograph of a
rat tail tendon in transverse section showing the fascicle units (asterisked) that make up the tendon. Adapted from Ref. [111]. (D) Ligament mechanical properties. Typical
mechanical properties of ACL for younger human with mean age 22 showing the lower modulus ‘‘toe’’ region (0–20% strain), the linear region (20–50% strain) and the yield
and failure of the ligament (>50% strain). The ACL examples shown here fail at 1–2.5 kN, demonstrating the significant challenge associated with matching these mechanical
properties using reconstituted type I collagen fibre prosthesis. Strain rate = one ligament length per second. Adapted from Ref. [112]. All figures are used with permission.

patterns of native tendon/ligament tissue [43,44]. The oriented
structure of ligament and tendon can be readily observed using
atomic force microscopy (AFM), shown in Fig. 1B, or transmission
electron microscopy (TEM), shown in Fig. 2C, which is character-
ised by a repeating pattern of light and dark regions. The collagen
fibril itself is an assembly of fibres, which is observed with a vari-
ety of characteristic diameters, which often depend on the biolog-
ical process by which the fibril was formed; for example, collagen
type V plays a role in regulation of fibril dimensions [45]. The
microscale crimp patterns found in tendon/ligament tissue are
not observed in the reconstituted type I collagen systems.
3. Collagen fibre fabrication techniques
Collagen-based fibres have been in contemporary clinical use
for over 80 years [46] and may have been in medical use for over
1000 years as the ‘‘cattle gut’’ suture known as ‘‘catgut’’, which is
a split sheep intestine submucosa. This material remains commer-
cially available for a limited number of applications; however, it is
not desirable for tissue engineering applications due to inflamma-
tory tissue reactions [47]. As the inflammatory response is
primarily associated with non-collagen epitopes, an alternative
methodology is to purify the collagen and reconstitute the fibre
from a collagen gel or solution. The majority of collagen fibre
fabrication methods start with acidified collagen solutions or sus-
pension because acidification disassembles the fibrillar structure
to form a flowable gel. Collagen fibrillogenesis is commonly trig-
gered via a combination of the following processes: pH shift to
neutral; dewatering of the dilute acidified gel; increasing temper-
ature to approximately 30–37 °C. This type of approach was first
utilised on an industrial scale in the 1960s using acetone and
ammonia to cause a dehydrative pH shift to produce reconstituted
collagen sutures from bovine dermis [48].
3.1. Fabrication of assembled microfibres from acidified collagen gels
using in vitro fibrillogenesis
Extrusion of collagen on a laboratory scale under aqueous con-
ditions from acidified collagen gels was first reported by Kato and
Silver in 1989 [26]. Their process relied on the extrusion of acidi-
fied insoluble collagen into a pH neutral buffer and incubation
for 1 h at 37 °C, before dewatering with isopropyl alcohol, rehydra-
tion and air-drying under tension. Whilst usable on a laboratory
scale, this process was not particularly amenable for large-scale
manufacture, and the process was developed on a continuous basis
with a limited production rate of approximately 5 m h 1 [49].
These fibres can typically be fabricated between 10 and 2000 lm
in diameter (Fig. 2B shows an example scanning electron micros-
copy (SEM) image of a collagen microfibre) and, although consider-
ably larger in diameter than the collagen fibrils (<500 nm), they
exhibit axial alignment that mimics that of the native tissue. Solu-
ble collagen has also been used in this in vitro self-assembly pro-
cess, conferring enhanced purity and improved mechanical
properties – perhaps owing to improved assembly kinetics [50].
3240 S.J. Kew et al. / Acta Biomaterialia 7 (2011) 3237–3247

Fig. 2. Diagram of a collagen fibre extrusion set-up for production of microscopic collagen fibres. (A) A typical microscopic fibre production sequence showing extrusion of an
acidified collagen solution into a fibre formation buffer at physiological pH and temperature. The nascent fibre is transferred to a fibre incubation buffer to continue the
process of collagen fibrillogenesis and finally transferred to a crosslinking or functionalisation solution to confer the structure with desired properties. (B) SEM image of a
fractured collagen microfiber. (C) TEM images of the banding patterns observed on a nanoscale which represent the organised 64 nm periodicity staggered packing found for
collagen fibres which undergo fibrillogenesis. Adapted from Ref. [57] and used with permission.

The uniaxial alignment of fibres extruded from acid-soluble colla-
gen solutions was also found to be superior to that of the insoluble
collagen used for the initial studies of this method [51]. In addition,
it was found that the strain imparted to the fibres prior to drying
has a significant effect on the final mechanical properties of ex-
truded collagen fibres, with prestretched fibres (50% strain) exhib-
iting an ultimate tensile strength of five times that of unstretched
ones [52].
Cavallaro et al. also reported the extrusion of collagen fibres
from soluble collagen into a circulating solution of a neutral buf-
fered solution of 20 wt.% poly(ethylene glycol) (PEG) [53]. The os-
motic pressure gradient between the collagen and the PEG solution
increased the speed of dehydration and coagulation, such that the
residence time in the bath could be reduced to approximately
4 min compared to 60. More recently, Zeugolis et al. have extended
this methodology, shown schematically in Fig. 2A, and have com-
pared the properties of fibres extruded into fibre formation buffers
consisting of a variety of co-agents, co-agent concentrations and
collagen sources to produce microscale collagen fibres (such as
that shown in Fig. 2B) [54–56].
The challenge of continuously spinning reconstituted collagen
fibres at a rate suitable for commercial manufacture has provided
a consistent limitation to the usage of collagen fibres owing to
the extended timescales required for dewatering and assembly of
nascent collagen fibres. The reports of Kato and Silver [49] and Cav-
allaro et al. [53] indicate that production rates of 5 and 100 m h 1,
respectively are possible using an in-line extrusion, fibrillogenesis,
washing and drying system. However, these systems are con-
strained by the requirement to achieve fibrillogenesis within the
temporal domain defined by the dimensions of the equipment
(see the schematic in Fig. 2A). Furthermore, it is not clear that
the collagen structures formed via these methods are consistently
a native-type assembly including the characteristic D-banded
structure. A recent report of Caves et al. has addressed this limita-
tion via the use of an off-line fibrillogenesis process to reduce the
time and space required for a continuous fibre extrusion process
and reporting a production rate of 60 m h 1 to form D-banded
collagen fibres, confirmed via TEM and shown in Fig. 2C [57].
However, in all cases, the quantification of gelatine within the final
collagen fibre remains an area of experimental difficulty for all the
systems discussed. Whilst observation of characteristic D-banding
may indicate collagen assembly on a nanoscale, the potential for
gelatinised regions within a macroscopic collagen system remain
challenging to determine.
3.2. Fabrication of physically aligned type I collagen scaffolds
The use of electrochemical control to direct the pH of the colla-
gen fibre assembly microenvironment has been utilised to produce
a collagen fibre structure with a particularly aligned structure. Re-
ported by Cheng et al. in 2008 and shown in Fig. 3A, a pH gradient
created between parallel anode and cathode based on the redox
reactions of water enabled isoelectric focusing of collagen mole-
cules [113]. Interestingly, this study demonstrated that it was pos-
sible to align reconstituted collagen without the mechanical action
of extrusion and achieve crosslinked fibre properties that were
comparable with extruded fibres. Furthermore, the aligned colla-
gen construct was microscopically comparable with native tendon
tissue (Fig. 3B) and was able to be extensively colonised with ten-
don-derived fibroblast cells after 7 days in culture (Fig. 3C).
Reduction of fibre diameter is also anticipated to result in highly
aligned structures, with the ultimate limit of a single chain fibre
offering the possibility of increased axial alignment and a biomi-
metic fibril structure. Since the absolute mechanical properties of
the fibre scale with dimensions, it becomes increasingly difficult
to handle fibres of <100 lm, particularly when hydrated and not
crosslinked. Electrospinning is considered an attractive and rela-
tively straightforward method to fabricate and handle submicron
diameter fibres, potentially providing the high porosity and
tensile strength considered useful for many tissue engineering
 S.J. Kew et al. / Acta Biomaterialia 7 (2011) 3237–3247 3241

Fig. 3. Examples of novel methods to form collagen fibre structures: electrospinning and electrochemical alignment and assembly. (A) Diagrammatic representation of an
electrochemical cell capable of producing an electrochemically driven pH gradient. The pH gradient enabled collagen fibre assembly and orientation, as shown by polarised
optical microscopy in (B), indicating a comparable orientation with native tendon. (C) Tendon-derived fibroblast cells were seeded onto twisted collagen fibre bundles and
imaged after 7 days in culture; cells imaged using Alexa Fluor 488 Phalloidin staining and fluorescent microscopy. Adapted from Ref. [113] and used with permission. (D)
Diagrammatic representation of a typical experimental set-up for electrospinning type I collagen fibres. Collagen nanofibres are typically collected on an aluminium drum in
an unoriented mesh. (E) Characteristic D-banding observed using TEM (scale bar is 100 nm) for electrospun type I collagen. Adapted from Ref. [58] and used with permission.

applications (Fig. 3D/E). In 2002, Matthews et al. reported the pro-
duction of a three-dimensional matrix of electrospun collagen fi-
bres with an average diameter of 100 nm and exhibiting the
characteristic 67 nm axial D-banding pattern of quarter staggered
collagen fibrils (Fig. 3E) [58]. Furthermore, Rho et al. have also de-
scribed the fabrication of a similar scaffold consisting of fibres with
a larger average diameter of 460 nm, which indicated favourable
results in cell culture experiments [59]. In addition, the mechanical
properties of glutaraldehyde-crosslinked electrospun type I colla-
gen fibres (no D-banding observed) of 225–425 nm diameter were
found to have a bending modulus of 0.26–0.07 GPa, which also
indicated the expected inverse relation between fibre diameter
modulus [60]. However, electrospinning of collagen using solvents
such as hexafluoropropanol may have a denaturing effect on the
triple-helix structure and supramolecular fibril. Furthermore, if re-
tained by the implant, these solvents may have a deleterious effect
on the biological environment. In a detailed study of the collagen
structures resulting from electrospinning acid-soluble collagen
from fluoroalcohols, near complete denaturation was determined
by Zeugolis et al. from second harmonic generation/circular
dichroism spectra and low denaturation temperatures, casting
doubt over whether gelatinisation can be avoided using this pro-
cessing route [61].
Enhanced strength and alignment can be achieved via a number
of processes without the use of fibre extrusion or electrospinning.
Plastic compression (mechanical densification) has been used to
strengthen non-aligned collagen gels and, although mechanically
weaker than tendon or ligament tissues, these materials have been
shown to retain the viability of fibroblasts seeded into the
structure [62]. Electrostatically and magnetically aligned collagen
structures have also recently been reported. In 2003, Gigante
et al. described the production of aligned multilayer collagen
membranes by pouring successive layers of collagen gel onto
electrostatically charged plates and allowing each layer to partially
dry before pouring on the next [63]. A similar method, reported in
2007 by Torbet et al. involved the use of a superconducting magnet
to align collagen films, allowing a scaffold to be built up consisting
of multiple orthogonal layers [64].
4. Crosslinking collagen fibres assembled in vitro
4.1. Mechanical properties
Reconstituted collagen structures are generally rather mechan-
ically weak when formed and degrade rapidly in vivo due to the
action of proteolytic enzymes such as the collagenases. Crosslink-
ing is commonly employed in order to improve their mechanical
properties and increase the material’s resistance to enzymatic
lysis. The native crosslinking pathway does not occur in vitro and
hence there is a requirement for a synthetic process that can confer
collagen with suitable mechanical characteristics whilst retaining
its biocompatibility and the chemotactic behaviour necessary for
cellular infiltration and proliferation. Despite the extensive
3242 S.J. Kew et al. / Acta Biomaterialia 7 (2011) 3237–3247
investigation, there is no accepted standard crosslinking method
for the production of strong and biocompatible materials. The
chemical approaches investigated include aldehydes such as glu-
taraldehyde (GA) and formaldehyde, epoxides, carbodimiides and
azides, alongside several physical methods such as extreme dehy-
dration and dye-mediated photo-oxidation [65]. GA, a crosslinker
with widespread application as a highly effective bifunctional
crosslinker, is a well-known example, which is considered poten-
tially problematic due to its toxicity [66], ill-defined structure
and in vivo degradation profile, which can lead to fibrous encapsu-
lation [76].
The most attractive crosslinking methodologies are those which
do not involve the addition of chemical functionality to the
structure. Kato et al. [26] reported on reconstituted collagen fibres
that were crosslinked with dehydrothermal treatment (DHT) and
cyanamide vapour to achieve break strengths of approximately
30 MPa and a low strain modulus of 175 MPa. A further investiga-
tion of the mechanical properties of collagen fibres crosslinked
using DHT only, which utilises elevated temperature and reduced
pressure, was performed by Wang et al. who found that collagen
fibres could be fabricated with a particularly high hydrated
breaking strength of 91.8 ± 31.4 MPa [50]. Chemical crosslinking
methods utilising carbodiimide activation to produce zero-length
crosslinks are also attractive owing to the lack of additional
functionality and ability to approach the mechanical break stress
of human ligaments at 24.7–50.0 MPa [67]. However, in the case
of the human ACL, the ultimate load to failure of 2160 ± 157 N
[68] correlates with a breaking stress of approximately 50 MPa
(estimated based on an ACL cross-sectional area of approximately
40 mm2), which is achieved by only a few of the optimised cross-
linking strategies developed to date (listed in Table 1).
Glucose has been studied extensively in the context of diabetes
research for its ability to crosslink collagens via glycosylation and
subsequent Maillard reactions. It was recently demonstrated that
ultraviolet (UV) irradiation in combination with glucose could be
used to confer collagen films with both increased mechanical
strength and resistance to collagenase and trypsin [69]. Nord-
ihydroguaiaretic acid (NDGA) has been used in a crosslinking
methodology to impart exceptionally high tensile strength to colla-
gen fibres, with a break strength of 90 MPa and an elastic modulus
of 580 MPa reported [70]. The details of the mechanism for this
reaction remain to be elucidated; however, it is possible that a
multicomponent composite is produced with collagen fibres
embedded in a polymerised NDGA matrix. A recent example of
crosslinking using Myrica rubra presents the possibility that an
enzymatic mechanism could be developed (mechanical properties
shown in Table 1) [71]. Steps towards the utilisation of the native
mechanism of transglutaminases to crosslink collagen fibres have
Table 1
Mechanical properties of selected crosslinked type I collagen fibres.
Type I collagen source Cross-linking Mechanical properties (hydrated in phosphate-buffered saline)
methodology
Maximum tensile stress
at failure (MPa)
Insoluble bovine dermis Glutaraldehyde (2 days) 66.2 ± 17.2
(i) DHT 31.3 ± 4.70
(ii) Carbodiimide
Soluble acid extracted DHT 91.8 ± 31.4
bovine dermis
Soluble fetal bovine tendon NGDA 90.7 ± 15.3
Insoluble bovine achilles tendon EDC 50 ± 13.4
Soluble bovine dermis (i) pH gradient 22–88
(ii) Genipin
Soluble rat tail collagen Glutaraldehyde vapour 94 ± 19
Soluble bovine tendon Myrica rubra (enzymatic) 28.13 ± 8.51
Data reported for 1989–2010 using the collagen microfibre extrusion process as shown in Fig. 3, except for Cheng et al. [113], where the electrochemical assembly process
was used.
been made, although the mechanical properties remain inferior
to those of more conventional chemical crosslinking strategies
[72].
A comprehensive experimental survey of the main crosslinking
strategies reported in the literature to date has recently been per-
formed by Zeugolis et al. [56]. It was found that the mechanical
properties reported for other fibre systems were not satisfactorily
replicated and that some crosslinking conditions tested exhibited
an order of magnitude difference in mechanical properties com-
pared to previous literature reports. This perhaps indicates the
intrinsic variability in the collagen sources and structures that have
been evaluated and highlights the inability of investigators to iden-
tify and standardise a particular crosslinking methodology for col-
lagen fibres.
5. Biological interactions with reconstituted collagen fibres
5.1. In vitro studies using collagen fibres
The in vitro biocompatibility and mechanical properties of col-
lagen fibres crosslinked with a variety of chemistries have been
extensively investigated. A study of long-term fibroblast cell cul-
ture to 8 weeks has shown that the cells continue to proliferate
during this period. However, in the case of the UV crosslinked fi-
bres tested, the mechanical properties were reduced despite en-
hanced cellularity compared with an EDC crosslinked fibre [73].
This may be understood by the possibility of UV-induced denatur-
ation, which may reveal otherwise cryptic cellular recognition mo-
tifs and thus cellular activity remains high. The in vitro rate of
fibroblast migration on collagen fibres has been measured by Corn-
well et al. who found that rates ranged from 0.75 to 1.25 mm day 1
[74], indicating that population of a typical tendon or ligament
scaffold would occur over several months. Both DHT- and UV-
crosslinked fibres have been found to be more susceptible to tryp-
sin digestion compared with non-crosslinked controls [75], which
again implies that these physical methods cause denaturation of
the collagen to some extent. The DHT samples in particular were
found to be particularly susceptible to enzymatic lysis.
5.2. In vivo studies of collagen fibres: evidence for tendon and ligament
regeneration
5.2.1. Achilles tendon model
Reconstituted collagen fibre structures have been evaluated in
a number of in vivo ligament and tendon models, both intra- and
extra-articular. Goldstein et al. evaluated a rabbit Achilles tendon
model, comparing GA with cyanamide and dehydrothermal (CDI/
References
Tensile strain Tangent modulus (MPa)
at failure (%)
16.1 ± 2.70 407 ± 96.6 (low strain toe region) Kato et al. [26]
17.7 ± 2.20 170.1 ± 32.9 (low strain toe region)
11.87 ± 1.92 895.8 ± 205.9 (High strain Wang et al. [50]
linear region)
10 ± 1 582 ± 58 Koob et al. [70]
Not reported 484.7 ± 76.3 Gentleman et al. [67]
7–24 277–671 Cheng et al. [113]
14.2 ± 1.9 775 ± 173 Caves et al. [57]
15 ± 4 23.10 ± 9.45 (2% strain) Zeugolis et al. [71]
 S.J. Kew et al. / Acta Biomaterialia 7 (2011) 3237–3247
DHT) crosslinking methods and evaluating explants at 3, 10 and
20 weeks. A marked difference in the histology of collagen fibres
crosslinked with CDI/DHT compared with GA was reported [76].
In the case of GA fibres, histology indicated encapsulation and,
in addition, minimal implant degradation was observed. In con-
trast, a substantial repair of native structure was obtained for
the CDI/DHT treated fibres after 10 weeks, with only remnants
of the prosthesis remaining. Interestingly, it was reported that
the mechanical properties of both the implanted collagen fibre
prosthesis types improved from 3 to 20 weeks in vivo. This indi-
cated that a load-bearing matrix was being deposited into the
scaffold and/or that the microenvironment was having a cross-
linking or strengthening effect on the residual implant. The effect
was particularly pronounced in the CDI/DHT-crosslinked prosthe-
sis, with an ultimate tensile strength after 20 weeks superior to
that of the autograft control. One year data for the rabbit Achilles
model was reported by Kato et al. [77], who confirmed the fibrous
encapsulation of the GA fibres, with minimal degradation of the
prosthesis. In contrast, the CDI/DHT implant had completely de-
graded, leaving a neotendon structure with many, though not
all, microstructural features reminiscent of native tissue, includ-
ing the characteristic crimp pattern observed under polarised
microscopy.
5.2.2. Anterior cruciate ligament models
The reconstituted collagen fibre structures of Kato et al. and
Goldstein et al. were further evaluated in a rabbit ACL model by
Dunn et al. [78]. The intra-articular environment may be consid-
ered more hostile to collagen-based biomaterials due to nearly
avascular conditions and the presence of synovial fluid. As in the
previous study, glutaraldehyde crosslinking was compared with
CDI/DHT in two implant groups, which were evaluated mechani-
cally and histologically at 4 and 20 weeks post-implantation. It
was found that the GA-crosslinked prosthesis did not degrade as
rapidly, and at 20 weeks it was encapsulated by fibrous tissue
and inflammatory cells. In comparison, the CDI/DHT-crosslinked
prostheses appeared to have been replaced by fibrous repair tissue
exhibiting neoligamentous crimp patterns at 20 weeks. The load to
failure of both prosthesis types increased after 20 weeks implanta-
tion such that they were not significantly different. However,
approximately 50% of the prosthesis ruptured in this study due
to the mechanical loads present in this site. To increase the
mechanical strength, the collagen fibres were further investigated
in a solvent-cast poly(lactic acid) (PLA) matrix [31]. This study
demonstrated that improved mechanical properties could be
achieved using PLA instead of collagen with retention of similar
subcutaneous in vivo biodegradation. In addition, in the rabbit
ACL model, neoligament repair tissue was identified in 7/8 implan-
tations after 4 weeks, with 3/7 exhibiting continuous tibia–femur
fibres. However, again approximately 50% of the prostheses had
ruptured, thus highlighting the significant challenge of regenerat-
ing the ACL site.
Regeneration of tissue at the soft–hard tissue interface also
constitutes a major challenge for tissue engineering owing to the
dramatic changes to both structure and properties at these
interfaces [79]. The tendon–bone interface consists of four distinct
tissue zones: bone, calcified fibrocartilage, fibrocartilage and
tendon [80]. The incorporation of collagen fibre prostheses into
non-loaded osseous bone tunnels has been investigated in a rabbit
model and it was found that the CDI/DHT-crosslinked collagen fi-
bres actually degraded within the bone tunnel within 8 weeks,
with new bone and fibrous tissue infiltrating the scaffold [81].
The production of braided collagen fibres was reported by
Cavallaro et al. where a collagen fibre prosthesis was fabricated
for ACL reconstruction [53]. A pilot canine ACL study (n = 2) was
reported, demonstrating cellular infiltration and remodelling after
3243
12 weeks, with concomitant neoligament repair tissue formation
(Fig. 4A–C). However, no longer-term follow-up was reported.
The reconstruction of goat ACL was investigated by Chvapil
et al. using a collagen fibre-based scaffold with two different
collagen fibre purification routes [82]. Unlike the studies using
reconstituted collagen, the collagen fibres used were mechanically
recovered from tendon. The fibres were purified via multiple wash
steps and crosslinked using hexamethylenediisocyanate in an
organic solvent. Unlike the studies of Kato et al. and Dunn et al.
the collagen fibres exhibited a dramatic reduction in mechanical
strength following implantation. Interestingly, partially purified
collagen fibres, which exhibited a lower extent of crosslinking, also
exhibited more rapid healing and remodelling, with more
extensive cellular infiltration and neovascularisation at 3 months
compared with the more highly purified type. The histological
analysis revealed that almost 50% of the fibres were degraded
and replaced by neoligamentous structures after 6 months.
5.2.3. Remodelling of allograft ligament/tendon tissues
The in vivo remodelling of ligament and tendon allograft tissues
used in the reconstruction of the ligament and tendon can be con-
sidered an ideal template for a type I collagen fibre bioscaffold be-
cause allografts almost replicate the native structure; however,
these tissues are non-self. Instructive examples of the effect of
their non-self nature arise in the comparison of allograft vs. auto-
graft tissues for anterior cruciate reconstruction.
Although many studies in humans do not report different clin-
ical outcomes for the two groups [83,84], several publications have
shown an increased knee laxity and rerupture rate after allograft
implantation compared with autograft [85]. In vivo studies have
shown the remodelling activity of allograft tissue to be delayed
with respect to that of autograft, although for both implants, com-
mon phases of remodelling were graft necrosis, repopulation and
complete substitution (even in autografts) by host cells [86,87].
However, the autograft has been shown to undergo the so-called
‘‘proliferation phase’’ as early as 6 weeks after implantation [88]
with increased cellularity and vascularity, whereas the allograft
has shown lower cellularity and no revascularisation, apart from
in the synovial sheets, at that early time point [89,90]. A progres-
sive increase in vascularity was observed in allograft from 6 to
52 weeks whilst, in contrast, autograft vascularity was decreasing;
in both cases the central part of the implant was relatively avascu-
lar with respect to the subsynovial and intermediate parts [91].
The crimp length tended to decrease for both allograft and auto-
graft from 6 to 52 weeks, becoming not significantly different from
the crimp pattern of native ACL at the latest time point. Conversely,
myofibroblasts tended to increase from 6 to 52 weeks in both auto-
grafts and allografts [92]. Overall, any gross morphological differ-
ences between autografts and allografts tended to disappear after
1 year from implantation, even if a native ACL appearance was
never obtained [91].
Ultrastructural remodelling of the allograft collagen composi-
tion has been shown to take place beyond 6 months from the
implantation in humans, with the progressive loss of large-diame-
ter fibrils and of the typical bimodal fibril distribution. These
changes persisted for many years after implantation; large fibrils
were never reformed and were replaced by newly synthesised
small-diameter fibrils [93]. Moreover, the loss of large-diameter
fibrils occurred quicker in autografts with respect to allografts in
a goat model [8].
The mechanical properties of soft tissue grafts were shown to be
significantly superior for autografts than for allografts at 1 year
follow-up in terms of knee laxity and stress to failure, whilst no
differences were observed at earlier time points [91,92]. In
contrast, for bone–tendon–bone grafts, allografts already showed
significantly decreased stress to failure and increased knee laxity
3244 S.J. Kew et al. / Acta Biomaterialia 7 (2011) 3237–3247

Fig. 4. Reconstituted collagen microfibres in an in vivo ACL site. Histological examples of collagen fibre scaffolds explanted from a canine ACL model after 12 weeks are
shown. The three images, A, B and C, show the morphology of explanted collagen threads in a canine ACL site. Scale bar = 300 lm. (A) Histology adjacent to the fixation staple.
(B) Longitudinal section in the substance of the explant taken between the femoral and tibial insertion points. (C) New collagenous tissue interdigitation with articular
cartilage at the bony interface. Tissue stained with Masson’s trichrome to show tissue alignment with embedded collagen fibres. Scale bar = 20 lm. Adapted from Ref. [53]
and used with permission.

with respect to controls at 6 months after implantation [90]. In
both cases the mechanical properties of native ACL were never re-
stored [90,91].
In summary, the current knowledge on graft remodelling sug-
gests that, even if a self autograft is provided, avascular necrosis,
revascularisation and collagen fibril reorganisation have to take
place, leading to a weaker implant with respect to the native struc-
ture. The whole process was shown to be even slower and more
disorganised when non-self allografts were used, leading to a fur-
ther decrease in mechanical properties of the implant. It seems of
paramount importance for a biosynthetic scaffold to allow a fast
vascular and cellular ingrowth, in order to support collagen fibril
synthesis and matrix reorganisation and to accelerate the strength-
ening of the tissue.
6. Collagen fibre scaffolds and biologics
Biologics such as growth factors (GFs) are potentially useful sig-
nalling strategies to enhance scaffold-mediated tissue repair. GFs
are fundamental signals in both the natural growth and the healing
process of soft tissues, and are able to stimulate cell chemotaxis,
proliferation, differentiation and matrix protein production. The
combination of collagen scaffolds with active biological species,
such as cell (and growth factor) concentrates, including platelet-
rich plasma [94], and autologous or allogenic cell types [95], pres-
ent a significant opportunity to achieve tendon or ligament regen-
eration (vs. repair) via trophic and/or cellular scaffold colonisation
mechanisms. A large variety of GFs have been tested to augment
the natural tendon and ligament healing response, both alone
[96] and in association with scaffolds [97]. For example, transform-
ing growth factor-beta, which enhances cellular proliferation,
demonstrated a statistically significant increase in regenerated
tendon mechanical properties in vivo [96]. Furthermore, insulin-
like growth factor, platelet-derived growth factor and basic fibro-
blast growth factor were also shown to enhance in vitro tendon
fibroblast proliferation and to have a synergistic effect when used
together [98]. Cartilage-derived morphogenetic proteins and vas-
cular endothelial growth factor have also demonstrated some
enhancement to tendon repair [99,100].
The process of preseeding cells on a scaffold prior to implanta-
tion has become a favoured approach for circumventing the time

Fig. 5. Collagen fibre scaffold utilising cellular preseeding: an in vitro study. (A) Stress–strain curves for collagen fibre–collagen gel composites preseeded with dermal
fibroblasts and maintained in culture for 5 days. (B) Fluorescence micrographs of same scaffolds after 5 days in culture, the cells being labelled with a ‘‘Live/Dead’’ staining kit.
(C and D) Histological transverse (C) and longitudinal (D) sections of the cell-seeded scaffolds after 20 days of culture and stained with Masson’s trichrome to show tissue
alignment with embedded collagen fibres. Scale bar = 20 lm. Adapted from Ref. [109] and used with permission.
 S.J. Kew et al. / Acta Biomaterialia 7 (2011) 3237–3247
required to achieve cellular population of the scaffold and to
potentially select the appropriate cell types. The overall aim is to
enhance tissue regeneration by increasing the implanted scaffold
cellularity and by secretion of relevant growth factors, thus accel-
erate ECM production. It may be particularly important in avascu-
lar (e.g. rotator cuff) or intrasynovial (e.g. ACL) environments, in
which cell recruitment from the surrounding tissues can be lim-
ited. Mesenchymal stem cells (MSC), which are multipotent cells
capable of differentiation into mesenchymal tissue types, have
been proposed as biologics for tendon/ligament tissue engineering.
The main advantage of using these cells is that they can be ob-
tained from bone marrow aspirate or used as an allogenic thera-
peutic, owing to their immunoprivileged status [101]. MSCs
seeded onto collagen gels and poly(lactic-co-glycolide) scaffolds
have been shown to enhance the mechanical properties and the
histological appearance of hybridised scaffolds in tendon sites
[102,103]. Alternatively, autologous tendon and ligament fibro-
blasts, which are the main residing cells in tendons and ligaments
respectively, could be used. However, such hybridisation may be
clinically impractical because it would require a two-step surgery:
a tissue harvest (with the related donor site morbidity; cf. auto-
graft procedures), cell expansion and cultivation on the scaffold;
and implantation. Nevertheless, they have been shown to improve
the mechanical strength of tendon repair in vivo when associated
with poly(glycolic acid) scaffolds [104]. They also enhanced the
in vitro tensile strength of chitosan/hyaluronan and chitosan/algi-
nate mixed scaffolds [105,106].
A further possible autologous cell source is dermal fibroblasts
[107], which are perhaps more attractive than tendon and ligament
fibroblasts because they are accessible and easier to harvest, with a
low donor site morbidity. A study of this approach utilised both lig-
ament and dermal fibroblasts seeded onto UV-crosslinked collagen
fibres and implanted into non-load-bearing subcutaneous and in-
tra-articular sites [108] where it was found that the seeded cells
could be detected up to 6 weeks after implantation. Pre-seeding
autologous dermal fibroblasts into a collagen fibre/collagen gel
composite in vitro has indicated that the presence of cells can sig-
nificantly enhance the breaking stress of the hybrid biomaterial
[109]. The peak stress at break for the cell-seeded constructs was
significantly higher (p < 0.05) than the cell-free controls, although
the peak stresses at ca. 9 MPa (as shown in Fig. 5A) are still consid-
erably lower than human ligament tissue at 30–50 MPa. The scaf-
folds consisted of a parallel array of type I collagen fibres in a
matrix of type I collagen gel and dermal fibroblasts, hence the
non-load-bearing component will have reduced the stress at break.
The cell seeding efficiency reported was nearly 100% as a result of
using a gel-based encapsulation of the cells and, as shown in
Fig. 5B, this environment was favourable for the growth and prolif-
eration of these cells. As shown in Fig. 5C and D, the tissue struc-
ture found after 20 days in culture was ligament-like, with a
remarkably aligned cellular organisation; thus these preliminary
results indicated that this may be a viable method to form a
‘‘farmed’’ autograft ex vivo using harvested dermal fibroblasts in
combination with a collagen fibre biomaterial.
7. Conclusions
Collagen-based biomaterials have a significant history and
successful clinical track record as functional supports for repair
of soft tissues. Their utilisation as fibres for structural supports in
load-bearing sites has been less widely exploited, owing to the
multifaceted challenges associated with achieving both mechani-
cal properties which match and maintain in vivo those of native
load-bearing tissues and provision of a microenvironment that is
inducive of tissue regeneration. In vivo studies of collagen fibre
biomaterials have so far failed to consistently demonstrate tendon
3245
and ligament regeneration, although the early indications of cellu-
lar remodelling and repair of these structures are encouraging. The
recent developments in our understanding of native collagen fibrils
and of the synthetic collagen-processing routes used to fabricate
these structures indicate that there is now considerable potential
to tailor the properties of collagen fibre biomaterials to achieve
the goal of regenerating ligament and tendon tissue. Furthermore,
the more detailed understanding of autograft remodelling in cases
such as the ACL bone–patellar–bone reconstruction may provide a
useful basis from which to design these new biomaterial strategies.
Our ability to tailor not only the physical structure but also the sur-
face chemistry and nanoscale topography of these materials now
provides a highly flexible and promising biomaterials approach
for the development of novel and rational strategies. Furthermore,
there is a largely unexplored area involving the combination of col-
lagen fibres biomaterials with biological species which may afford
an additional level of control over the microenvironment required
to achieve the complete regeneration of ligament and tendon tis-
sue. Hence we believe that the approach of using collagen fibre
scaffolds alone or in combination with biological species may
now be well placed to provide the solution to a longstanding and
widespread clinical problem of repairing or regenerating damaged
ligament and tendon tissues in the human body.
Acknowledgements
S.J.K., J.H.G., D.E., R.B., N.R., S.M.B. and R.E.C. would like to thank
the Technology Strategy Board for support under Grant No. TP/8/
BIO/6/I/Q0052K. A.P., D.Z. and M.A. would like to thank Science
Foundation Ireland for support under Grant Nos. 07/SRC/B1163
and 09/RFP/ENM2483; the Engineering and Physical Sciences Re-
search Council, UK; the Faculty Research Committee of the Faculty
of Engineering, NUS, Singapore; and the Irish Research Council for
Science Engineering and Technology, Ireland.
Appendix A. Figures with essential colour discrimination
Certain figures in this article, particularly Figures 1, 3, and 5, are
difficult to interpret in black and white. The full colour images can
be found in the on-line version, at doi:10.1016/j.actbio.2011.
06.002.
References
[1] Jones KG. Reconstruction of the anterior cruciate ligament using the central
one-third of the patellar ligament. J Bone Joint Surg Am 1970;52:838–9.
[2] Pinczewski LA, Lyman J, Salmon LJ, Russell VJ, Roe J, Linklater J. A 10-year
comparison of anterior cruciate ligament reconstructions with hamstring
tendon and patellar tendon autograft: a controlled, prospective trial. Am J
Sports Med 2007;35(4):564–74.
[3] Barrett GR, Noojin FK, Hartzog CW, Nash CR. Reconstruction of the anterior
cruciate ligament in females: a comparison of hamstring versus patellar
tendon autograft. Arthroscopy 2002;18:46–54.
[4] Beynnon BD, Johnson RJ, Fleming BC, Kannus P, Kaplan M, Samani J, et al.
Anterior cruciate ligament replacement: comparison of bone–patellar
tendon–bone grafts with two-strand hamstring grafts: a prospective,
randomized study. J Bone Joint Surg Am 2002;84:1503–13.
[5] Sachs RA, Daniel DM, Stone ML, et al. Patellofemoral problems after anterior
cruciate ligament reconstruction. Am J Sports Med 1989;17(6):760–5.
[6] Janssen R, van der Wijk J, Fiedler A, Schmidt T, Sala H, Scheffler S. Remodelling
of human hamstring autografts after anterior cruciate ligament
reconstruction. Knee Surg Sports Traumatol Arthrosc 2011;1–8.
[7] Ménétrey J, Duthon V, Laumonier T, Fritschy D. Biological failure of the
anterior cruciate ligament graft. Knee Surg Sports Traumatol Arthrosc
2008;16(3):224–31.
[8] Muneta T, Lewis JL, Stewart NJ. Load affects remodeling of transplanted,
autogenous bone–patellar tendon–bone segments in a rabbit model. J Orthop
Res 1994;12(1):138–43.
[9] Scheffler S, Unterhauser F, Weiler A. Graft remodeling and ligamentization
after cruciate ligament reconstruction. Knee Surg Sports Traumatol Arthrosc
2008;16(9):834–42.
3246 S.J. Kew et al. / Acta Biomaterialia 7 (2011) 3237–3247
[10] Carey JL, Dunn WR, Dahm DL, Zeger SL, Spindler KP. A systematic review of
anterior cruciate ligament reconstruction with autograft compared with
allograft. J Bone Joint Surg Am 2009;91(9):2242–50.
[11] Krych AJ, Jackson JD, Hoskin TL, Dahm DL. A meta-analysis of patellar tendon
autograft versus patellar tendon allograft in anterior cruciate ligament
reconstruction. Arthroscopy 2008;24(3):292–8.
[12] van Steensel CJ, Schreuder O, van den Bosch BF, van Paassen HC, Menke HE,
Voorhorst G, et al. Failure of anterior cruciate-ligament reconstruction using
tendon xenograft. J Bone Joint Surg Am 1987;69:860–4.
[13] McPherson GK, Mendenhall HV, Gibbons DF, et al. Experimental mechanical
and histologic evaluation of the Kennedy ligament augmentation device. Clin
Orthop 1985;196:186–95.
[14] Yahia L. Ligaments and ligamentoplasties. Berlin: Springer; 1997.
[15] Jenkins DHR, Forster IW, McKibbon B, et al. Induction of tendon and
ligament formation by carbon implants. J Bone Joint Surg 1977;59B:
53–7.
[16] Mody BS, Howard L, Harding ML, Parmar HV, Learmonth DJ. The ABC carbon
and polyester prosthetic ligament for ACL-deficient knees. J Bone joint Surg
1993;75B:818–21.
[17] Badylak SF. The extracellular matrix as a biologic scaffold material.
Biomaterials 2007;28:3587–93.
[18] Gilbert TW, Stewart-Akers AM, Simmons-Byrd A, Badylak SF. Degradation
and remodeling of small intestinal submucosa in canine achilles tendon
repair. J Bone Joint Surg Am 2007;89:621–30.
[19] Chen J, Xu J, Zheng M. Scaffolds for tendon and ligament repair: review
of the efficacy of commercial products. Expert Rev Med Devices
2009;6(1):61–73.
[20] Aurora A, McCarron J, Ianotti JP, Derwin K. Commercially available
extracellular matrix materials for rotator cuff repairs: state of the art and
future trends. J Shoulder Elbow Surg 2007;16(Suppl.):S171–8.
[21] Zheng MH, Chen J, Kirilak Y, Willers C, Xu J, Wood D. Porcine small intestine
submucosa (SIS) is not an acellular collagenous matrix and contains porcine
DNA: possible implications in human implantation. J Biomed Mater Res B
Appl Biomater 2005;73:61–7.
[22] Walton JR, Bowman NK, Khatib Y, Linklater J, Murrell GAC. Restore
orthobiologic implant: not recommended for augmentation of rotator cuff
repairs. J Bone Joint Surg Am 2007;89:786–91.
[23] Ghosh K, Ingber DE. Micromechanical control of cell and tissue development:
implications for tissue engineering. Adv Drug Deliv Rev 2007;59:1309–18
[and references therein].
[24] Buehler MJ. Nature designs tough collagen: explaining the nanostructure of
collagen fibrils. PNAS 2006;103(33):12285–90.
[25] Cheema U, Nazhat NN, Alp A, Foroughi F, Anandagoda N, Mudera V, et al.
Fabricating tissues: analysis of farming versus engineering strategies.
Biotechnol Bioprocess Eng 2007;12(1):9–14.
[26] Kato YP, Christiansen DL, Hahn RA, Shieh SJ, Goldstein JD, Silver FH.
Mechanical properties of collagen fibres: a comparison of reconstituted and
rat tail tendon fibres. Biomaterials 1989;10(1):38–42.
[27] Baker BM, Mauck RL. The effect of nanofiber alignment on the maturation of
engineered meniscus constructs. Biomaterials 2007;28(11):1967–77.
[28] Wang JHC. Mechanobiology of tendon. J Biomech 2006;39(9):, 1563–1582.
[29] Reddy GK, Stehno-Bittel L, Enwemeka CS. Matrix remodeling in healing rabbit
Achilles tendon. Wound Repair Regen 1999;7(6):518–27.
[30] Liu Y, Ramanath HS, Wang DA. Tendon tissue engineering using scaffold
enhancing strategies. Trends Biotechnol 2008;26(4):201–9.
[31] Dunn MG, Bellincampi LD, Tria AJ, Zawadsky JP. Preliminary development of a
collagen–PLA composite for ACL reconstruction. J Appl Polym Sci
1997;63(11):1423–8.
[32] Lynn AK, Yannas IV, Bonfield W. Antigenicity and immunogenicity of
collagen. J Biomed Mater Res B Appl Biomater 2004;71B(2):343–54.
[33] Zhang ZG, Bothe I, Hirche F, Zweers M, Gullberg D, Pfitzer G, et al. Interactions
of primary fibroblasts and keratinocytes with extracellular matrix proteins:
contribution of a2b1 integrin. J Cell Sci 2006;119(9):1886–95.
[34] Yahyouche A, Zhidao X, Czernuszka JT, Clover AJP. Macrophage-mediated
degradation of cross-linked collagen scaffolds. Acta Biomater 2011;7(1):
278–86.
[35] Weadock KS, Miller EJ, Keuffel EL, Dunn MG. Effect of physical cross-linking
methods on collagen-fiber durability in proteolytic solutions. J Biomed Mater
Res 1996;32:221–6.
[36] Kastelic J, Galeski A, Baer E. The multicomposite structure of tendon. Connect
Tissue Res 1978;6:11–23.
[37] Orgel JPRO, Irving TC, Miller A, Wess TJ. Microfibrillar structure of type I
collagen in situ. PNAS 2006;103(24):9001–5.
[38] Nicholls SP, Gathercole LJ, Shah JS. Morphology of human palmaris longus
tendon. Ann Rheum Dis 1984;43(3):477–82.
[39] Wolman M, Kasten FH. Polarized light microscopy in the study of the
molecular structure of collagen and reticulin. Histochem Cell Biol
1986;85(1):41–9.
[40] Vidal BdC. Image analysis of tendon helical superstructure using interference
and polarized light microscopy. Micron 2003;34(8):423–32.
[41] Raspanti M, Manelli A, Franchi M, Ruggeri A. The 3D structure of crimps in the
rat Achilles tendon. Matrix Biol 2005;24(7):503–7.
[42] Screen HRCDL, Bader DL, Lee DA, Shelton JC. Local strain measurement within
tendon. Strain 2004;40:157–63.
[43] Gross J, Kirk D. The heat precipitation of collagen from neutral salt solutions:
some rate-regulating factors. J Biol Chem 1958;233(2):355–60.
[44] Silver FH, Trelstad RL. Type I collagen in solution. Structure and properties of
fibril fragments. J Biol Chem 1980;255(19):9427–33.
[45] Birk DE. J Cell Sci 1990;95(part 4):649–57.
[46] Marshall GD. Prevention of adhesions to tendons in the hand and wrist: with
report of two cases. J Bone Joint Surg Am 1928;10:816–8.
[47] Yaltirik M, Dedeoglu K, Bilgic B, Koray M, Ersev H, Issever H, et al. Comparison
of four different suture materials in soft tissues of rats. Oral Dis
2003;9(6):284–6.
[48] Griset EJ, Reissmann TL, Nichols J, Method of producing a collagen strand. US
Patent 3114, 593;1964.
[49] Kato YP, Silver FH. Formation of continuous collagen fibres: evaluation of
biocompatibility and mechanical properties. Biomaterials 1990;11(3):
169–75.
[50] Wang M-C, Pins GD, Silver FH. Collagen fibres with improved strength for the
repair of soft tissue injuries. Biomaterials 1994;15(7):507–12.
[51] Pins GD, Silver FH. A self-assembled collagen scaffold suitable for use in soft
and hard tissue replacement. Mater Sci Eng C 1995;3(2):101–7.
[52] Pins GD, Christiansen DL, Patel R, Silver FH. Self-assembly of collagen fibers.
Influence of fibrillar alignment and decorin on mechanical properties.
Biophys J 1997;73(4):2164–72.
[53] Cavallaro JF, Kemp PD, Kraus KH. Collagen fabrics as biomaterials. Biotechnol
Bioeng 1994;43(8):781–91.
[54] Zeugolis DI, Paul RG, Attenburrow G. Extruded collagen-polyethylene glycol
fibers for tissue engineering applications. J Biomed Mater Res A
2008;85B(2):343–52.
[55] Zeugolis DI, Paul RG, Attenburrow G. Engineering extruded collagen fibers for
biomedical applications. J Appl Polym Sci 2008;108(5):2886–94.
[56] Zeugolis DI, Paul GR, Attenburrow G. Cross-linking of extruded collagen
fibers: a biomimetic three-dimensional scaffold for tissue engineering
applications. J Biomed Mater Res A 2009;89A(4):895–908.
[57] Caves JM, Kumar VA, Wen J, Cui W, Martinez A, Apkarian R, et al.
Fibrillogenesis in continuously spun synthetic collagen fiber. J Biomed
Mater Res 2010;93B(1):24–38.
[58] Matthews JA, Wnek GE, Simpson DG, Bowlin GL. Electrospinning of collagen
nanofibers. Biomacromolecules 2002;3(2):232–8.
[59] Rho KS, Jeong L, Lee G, Seo BM, Park YJ, Hong SD, et al. Electrospinning of
collagen nanofibers: effects on the behavior of normal human keratinocytes
and early-stage wound healing. Biomaterials 2006;27(8):1452–61.
[60] Yang L, Fitie CFC, van der Werf OK, Bennink ML, Dijkstra PJ, Feijen J.
Mechanical properties of single electrospun type I collagen fibres.
Biomaterials 2008;29(8):955–62.
[61] Zeugolis DI, Khew ST, Yew ESY, Ekaputra AK, Tong YW, Yung LY, et al. Electro-
spinning of pure collagen nano-fibres–just an expensive way to make
gelatin? Biomaterials 2008;29(15):2293–305.
[62] Brown RA, Wiseman M, Chuo C-B, Cheema U, Nazhat SN. Ultra-rapid
engineering of biomimetic materials and tissues: fabrication of nano- and
micro-structures by plastic compression. Adv Funct Mater 2005;15(11):
1762–70.
[63] Gigante A, Cesari E, Busilacchi A, Manzotti S, Kyriakidou K, Greco F, et al.
Collagen I membranes for tendon repair: effect of collagen fiber orientation
on cell behavior. J Orthop Res 2009;27(6):826–32.
[64] Torbet J, Malbouyres M, Builles N, Justin V, Roulet M, Damour O, et al.
Orthogonal scaffold of magnetically aligned collagen lamellae for corneal
stroma reconstruction. Biomaterials 2007;28(29):4268–76.
[65] Khor E. Methods for the treatment of collagenous tissues for bioprostheses.
Biomaterials 1997;18(2):95–105.
[66] Sung H-W, Huang R-N, Huang LLH, Tsai C-C. In vitro evaluation of cytotoxicity
of a naturally occurring cross-linking reagent for biological tissue fixation. J
Biomater Sci Polymer Ed 1999;10:63–78.
[67] Gentleman E, Lay AN, Dickerson DA, Nauman EA, Livesay GA, Dee KC.
Mechanical characterization of collagen fibers and scaffolds for tissue
engineering. Biomaterials 2003;24(21):3805–13.
[68] Woo SLY, Hollis JM, Adams DJ, Lyon RM, Takai S. Tensile properties of the
human femur–anterior cruciate ligament–tibia complex. Am J Sports Med
1991;19(3):217–25.
[69] Ohan MP, Weadock KS, Dunn MG. Synergistic effects of glucose and
ultraviolet irradiation on the physical properties of collagen. J Biomed
Mater Res 2002;60(3):384–91.
[70] Koob TJ, Hernandez DJ. Material properties of polymerized NDGA–collagen
composite fibers: development of biologically based tendon constructs.
Biomaterials 2002;23(1):203–12.
[71] Zeugolis DI, Paul RG, Attenburrow G. The influence of a natural cross-linking
agent (Myrica rubra) on the properties of extruded collagen fibres for tissue
engineering applications. Mater Sci Eng C 2010;30(1):190–5.
[72] Zeugolis DI, Panengad PP, Yew ESY, Sheppard C, Phan TT, Raghunath M. An
in situ and in vitro investigation for the transglutaminase potential in tissue
engineering. J Biomed Mater Res Part A 2010;92A(4):1310–20.
[73] Caruso AB, Dunn MG. Changes in mechanical properties and cellularity
during long-term culture of collagen fiber ACL reconstruction scaffolds. J
Biomed Mater Res Part A 2005;73A(4):388–97.
[74] Cornwell KG, Downing BR, Pins GD. Characterizing fibroblast migration on
discrete collagen threads for applications in tissue regeneration. J Biomed
Mater Res Part A 2004;71A(1):55–62.
[75] Weadock KS, Miller EJ, Keuffel EL, Dunn MG. Effect of physical cross-linking
methods on collagen-fiber durability in proteolytic solutions. J Biomed Mater
Res 1996;32(2):221–6.
 S.J. Kew et al. / Acta Biomaterialia 7 (2011) 3237–3247
[76] Goldstein JD, Tria AJ, Zawadsky JP, Kato YP, Christiansen D, Silver FH.
Development of a reconstituted collagen tendon prosthesis. A preliminary
implantation study. J Bone Joint Surg Am 1989;71(8):1183–91.
[77] Kato YP, Dunn MG, Zawadsky JP, Tria AJ, Silver FH. Regeneration of Achilles
tendon with a collagen tendon prosthesis. Results of a one-year implantation
study. J Bone Joint Surg Am 1991;73(4):561–74.
[78] Dunn MG, Tria AJ, Kato YP, Bechler JR, Ochner RS, Zawadsky JP, et al. Anterior
cruciate ligament reconstruction using a composite collagenous prosthesis.
Am J Sports Med 1992;20(5):507–15.
[79] Moffat KL, Wang IE, Rodeo SA, Lu HH. Orthopaedic interface tissue
engineering for the biological fixation of soft tissue grafts. Clin Sports Med
2009;28(1):15–76.
[80] Thomopoulos S, Williams GR, Gimbel JA, Favata M, Soslowsky LJ. Variation of
biomechanical, structural, and compositional properties along the tendon to
bone insertion site. J Orthop Res 2003;21(3):413–9.
[81] Dunn MG, Maxian SH, Zawadsky JP. Intraosseous incorporation of composite
collagen prostheses designed for ligament reconstruction. J Orthop Res
1994;12(1):128–37.
[82] Chvapil M, Speer DP, Holubec H, Chvapil TA, King DH. Collagen fibers as a
temporary scaffold for replacement of ACL in goats. J Biomed Mater Res
1993;27(3):313–25.
[83] Bach Jr BR, Aadalen KJ, Dennis MG, Carreira DS, Bojchuk J, Hayden JK, et al.
Primary anterior cruciate ligament reconstruction using fresh-frozen,
nonirradiated patellar tendon allograft: minimum 2-year follow-up. Am J
Sports Med 2005;33:284–92.
[84] Kustos T, Balint L, Than P, Bardos T. Comparative study of autograft or
allograft in primary anterior cruciate ligament reconstruction. Int Orthop
2004;28:290–3.
[85] Prodromos C, Joyce B, Shi KA. Meta-analysis of stability of autografts
compared to allografts after anterior cruciate ligament reconstruction. Knee
Surg Sports Traumatol Arthrosc 2007;15:851–6.
[86] Bosch U, Kasperczyk WJ. Healing of the patellar tendon autograft after
posterior cruciate ligament reconstruction–a process of ligamentization? An
experimental study in a sheep model. Am J Sports Med 1992;20:558–66.
[87] Kleiner JB, Amiel D, Roux RD, Akeson WH. Origin of replacement cells for the
anterior cruciate ligament autograft. J Orthop Res 1986;4:466–74.
[88] Amiel D, Kleiner JB, Roux RD, Harwood FL, Akeson WH. The phenomenon of
‘‘ligamentization’’: anterior cruciate ligament reconstruction with
autogenous patellar tendon. J Orthop Res 1986;4:162–72.
[89] Arnoczky SP, Tarvin GB, Marshall JL. Anterior cruciate ligament replacement
using patellar tendon. An evaluation of graft revascularization in the dog. J
Bone Joint Surg Am 1986;64:217–24.
[90] Jackson DW, Grood ES, Goldstein JD, Rosen MA, Kurzweil PR, Cummings JF,
et al. A comparison of patellar tendon autograft and allograft used for anterior
cruciate ligament reconstruction in the goat model. Am J Sports Med
1993;21:176–85.
[91] Scheffler SU, Schmidt T, Gangey I, Dustmann M, Unterhauser F, Weiler A.
Fresh-frozen free-tendon allografts versus autografts in anterior cruciate
ligament reconstruction: delayed remodeling and inferior mechanical
function during long-term healing in sheep. Arthroscopy 2008;24:448–58.
[92] Dustmann M, Schmidt T, Gangey I, Unterhauser FN, Weiler A, Scheffler SU.
The extracellular remodeling of free-soft-tissue autografts and allografts for
reconstruction of the anterior cruciate ligament: a comparison study in a
sheep model. Knee Surg Sports Traumatol Arthrosc 2008;16:360–9.
[93] Shino K, Oakes BW, Horibe S, Nakata K, Nakamura N. Collagen fibril
populations in human anterior cruciate ligament allografts. Electron
microscopic analysis. Am J Sports Med 1995;23:203–8.
[94] Murray MM, Spindler KP, Ballard P, Welch TP, Zurakowski D, Nanney LB.
Enhanced histologic repair in a central wound in the anterior cruciate
ligament with a collagen platelet-rich plasma scaffold. J Orthop Res
2007;25(8):1007–17.
3247
[95] Butler DL, Juncosa-Melvin N, Boivin GP, Galloway MT, Shearn JT, Gooch C,
et al. Functional tissue engineering for tendon repair: a multidisciplinary
strategy using mesenchymal stem cells, bioscaffolds, and mechanical
stimulation. J Orthop Res 2008;26(1):1–9.
[96] Yuko A, Kazunori Y, Tokifumi M, Harukazu T, Akio M, Kozaburo H. The effect
of transforming growth factor-beta on mechanical properties of the fibrous
tissue regenerated in the patellar tendon after resecting the central portion.
Clin Biomech (Bristol, Avon) 2005;20(9):959–65.
[97] Sahoo S, Toh SL, Goh JC. A bFGF-releasing silk/PLGA-based biohybrid scaffold
for ligament/tendon tissue engineering using mesenchymal progenitor cells.
Biomaterials 2010;31(11):2990–8.
[98] Costa MA, Wu C, Pham BV, Chong AKS, Pham HM, Chang J. Tissue engineering
of flexor tendons: optimization of tenocyte proliferation using growth factor
supplementation. Tissue Eng 2006;12(7):1937–43.
[99] Virchenko O, Fahlgren A, Skoglund B, Aspenberg P. CDMP-2 injection
improves early tendon healing in a rabbit model for surgical repair. Scand J
Med Sci Sports 2005;15(4):260–4.
[100] Wang XT, Liu PY, Tang JB. Tendon healing in vitro: modification of tenocytes
with exogenous vascular endothelial growth factor gene increases expression
of transforming growth factor beta but minimally affects expression of
collagen genes. J Hand Surg Am 2005;30(2):222–9.
[101] Chamberlain G, Fox J, Ashton B, Middleton J. Concise review: mesenchymal
stem cells: their phenotype, differentiation capacity, immunological features,
and potential for homing. Stem cells 2007;25(11):2739–49.
[102] Juncosa-Melvin N, Boivin GP, Gooch C, Galloway MT, West JR, Dunn MG, et al.
The effect of autologous mesenchymal stem cells on the biomechanics and
histology of gel–collagen sponge constructs used for rabbit patellar tendon
repair. Tissue Eng 2006;12(2):369–79.
[103] Ouyang HW, Goh JCH, Thambyah A, Teoh SH, Lee EH. Knitted
poly-lactide-co-glycolide scaffold loaded with bone marrow stromal
cells in repair and regeneration of rabbit Achilles tendon. Tissue Eng
2003;9(3):431–9.
[104] Cao Y, Liu Y, Liu W, Shan Q, Buonocore SD, Cui L. Bridging tendon defects
using autologous tenocyte engineered tendon in a hen model. Plast Reconstr
Surg 2002;110 (5).
[105] Majima T, Funakosi T, Iwasaki N, Yamane ST, Harada K, Nonaka S, et al.
Alginate and chitosan polyion complex hybrid fibers for scaffolds in ligament
and tendon tissue engineering. J Orthop Sci 2005;10(3):302–7.
[106] Funakoshi T, Majima T, Iwasaki N, Yamane S, Masuko T, Minami A, et al.
Novel chitosan-based hyaluronan hybrid polymer fibers as a scaffold in
ligament tissue engineering. J Biomed Mater Res Part A 2005;74A(3):338–46.
[107] Liu W, Chen B, Deng D, Xu F, Cui L, Cao Y. Repair of tendon defect with dermal
fibroblast engineered tendon in a porcine model. Tissue Eng
2006;12(4):775–8.
[108] Bellincampi LD, Closkey RF, Prasad R, Zawadsky JP, Dunn MG. Viability of
fibroblast-seeded ligament analogs after autogenous implantation. J Orthop
Res 1998;16(4):414–20.
[109] Gentleman E, Livesay G, Dee K, Nauman E. Development of ligament-like
structural organization and properties in cell-seeded collagen scaffolds
in vitro. Ann Biomed Eng 2006;34:726–36.
[110] Bozec L, van der Heijden G, Horton M. Collagen fibrils: nanoscale ropes.
Biophy J 2007;92(1):, 70–75.
[111] Silver FH, Freeman JW, Seehra GP. Collagen self-assembly and the
development of tendon mechanical properties. J Biomech
2003;36(10):1529–53.
[112] Noyes FR, Grood ES. The strength of the anterior cruciate ligament in humans
and Rhesus monkeys. J Bone Joint Surg Am 1976;58:1074–82.
[113] Cheng X, Gurkan UA, Dehen CJ, Tate MP, Hillhouse HW, Simpson GJ, et al. An
electrochemical fabrication process for the assembly of anisotropically
oriented collagen bundles. Biomaterials 2008;29(22):3278–88.