Embed Size (px)
Citation preview
Development of a novel mesenchymal
stromal cell (MSC) therapy
for repairing the cornea
Elham Nili Ahmadabadi
B.Econ., M.Biotech.
A thesis submitted in fulfilment of the requirements for
the degree of Doctor of Philosophy
2018
School of Biomedical Sciences
Institute of Health and Biomedical Innovation
Queensland University of Technology
i
Abstract
The tissue residing at the edge of the cornea, or corneal limbus, contains at least two
types of progenitor/stem cells. Epithelial progenitor cells are concentrated within the
basal layer of the limbal epithelium and stromal progenitor cells can be isolated from
the adjacent connective tissue. Injuries and diseases affecting the corneal limbus can
therefore lead to significant alterations in corneal tissue structure and function. It is
therefore fortunate that studies of the limbal epithelium have led to significant
advances in the treatment of limbal dysfunction. For example, biopsies of limbal tissue
can be used to generate sheets of epithelial cells for repairing the corneal surface. In
contrast, however, the clinical value of limbal stromal progenitor cells remains at
present largely theoretical.
The goal of the present project, therefore, has been to further our understanding of the
biology and clinical potential of limbal stromal progenitor cells. This goal has been
pursed under three experimental aims; (1) evaluation of techniques for optimal
isolation and cultivation of limbal stromal progenitor cells, (2) exploration of their
effects when applied to the wounded ocular surface, and (3) development of a novel
strategy for the co-implantation of epithelial/stromal progenitor cells using membranes
constructed from the silk structural protein, fibroin.
In Chapter 3 of this thesis, I demonstrate efficient initiation of stromal cell cultures
from tissue explants, when pieces of stromal tissue are seeded into collagen gels.
Moreover, optimal cell outgrowth is observed when serum-supplemented growth
medium is used (DMEM supplemented with 10% FBS, L-glutamine and anti-biotics;
SSM). Under these conditions the stromal cells adopt a mesenchymal stromal cell
(MSC) phenotype characterized by expression of the cell surface markers
CD73/CD90/CD105 and absence of CD34/CD45. Cultures can also be initiated when
using a serum-free medium that has been recently shown to encourage retention of a
progenitor cell phenotype (DMEM supplemented with 10% Knockout Serum
Replacement and growth factors/Stem Cell Medium or SCM). Nevertheless, the
degree of cellular outgrowth is less than that seen in SSM and the cultures failed to
thrive when removed from the collagen gel. On this basis, SSM was chosen for
ii
subsequent studies of stromal cell biology in later chapters. Nevertheless, SCM-type
media used in conjunction with either collagen or other extracellular matrix (ECM)
factors may yet prove to be more appropriate depending upon the intended clinical use
of limbal stromal progenitor cells.
In Chapter 4 of this thesis, the clinical potential of limbal stromal progenitor cells is
examined using a rabbit model of ocular surface trauma. The rabbit stromal cells are
grown under conditions that promote development of a mesenchymal stromal cell
(MSC) phenotype and are applied to the surface of corneas that have been depleted of
epithelial cells using the Algerbrush II rotating burr tool. Human amniotic membrane
(HAM) is used as a vehicle implanting the rabbit limbal MSC (RL-MSC) directly onto
the wounded ocular surface. The effects of RL-MSC are examined when applied alone
(n = 3) and in conjunction with a stratified culture of human limbal epithelial cells
(HLE) grown on the opposing side of the HAM (n = 3). The effects of each treatment
are monitored over a period of 3 months in comparison with animals receiving no
treatment (n = 3) or treatment with HLE alone on HAM (n = 3). All animals receiving
RL-MSC displayed a faster rate of re-epithelization with best results being observed
for cultures grown in the presence of HLE. While all animals displayed an abnormal
ocular surface, characterized by the presence of conjunctival epithelial cells (via
positive staining for cytokeratin 13) and vascularization of the stroma, the clearest
evidence for corneal epithelial cells (via positive staining for keratin 3; K3) was
observed in animals that received RL-MSC in the presence of HLE. Based upon the
absence of staining for human nuclear antigen (HNA), however, the keratin 3-positive
cells would appear to be of rabbit origin. Further studies are therefore required to
confirm whether the cultured RL-MSC are the origin of the K3+ cells. Nevertheless, a
very different outcome is observed when RL-MSC are applied to the ocular surface in
the absence of HLE. All three animals treated with RL-MSC on HAM alone displayed
significantly greater vascularization of the corneal stroma and so this would not be
encouraged as a future therapy.
In Chapter 5 of this thesis, a potential alternative to HAM was tested as a vehicle for
delivering co-cultures of L-MSC and limbal epithelial cells to the ocular surface. The
HAM substitute was manufactured from the silk structural protein fibroin. Membranes
prepared from fibroin were optimized for cell attachment by using a recombinant
formulation containing the RGD cell binding motif. Moreover, the membranes were
iii
fabricated in the presence of a porogen (low molecular weight poly(ethylene glycol))
and a cross-linking agent (horseradish peroxidase; HRP) to facilitate permeability and
strength respectively. Membranes prepared from RGD-fibroin supported optimal
growth of the L-MSC. Moreover, L-MSC enhanced stratification of the HLE cultures
grown on the opposing surface of the RGD-fibroin membrane. Attachment of the flat
RGD-fibroin membrane to the domed ocular surface was facilitated by using a petal
wrapping strategy. The membrane could be successfully sutured to the ocular surface,
but dislodged within 1-week following surgery. A pilot treatment of RL-MSC/HLE
implantation using RGD-fibroin membrane (n =1) initially produced promising results
(nearly complete re-epithelialization in 14 days), but subsequently led to a prominent
epithelial defect and conjunctivalization of the ocular surface. An alternative strategy
for cell transfer using fibroin is therefore most likely warranted.
In summary, these studies have produced advances in our understanding of the biology
and potential clinical application of limbal stromal progenitor cells. In particular, these
studies indicate that stromal cells in the form of L-MSC will provide a useful adjunct
to treatment with limbal epithelial progenitor cells. Further studies are nonetheless
necessary to better understand the mechanism of action when L-MSC are applied in
conjunction with limbal epithelial cells and in particular whether the stromal cells are
somehow conditioned/licensed by being grown in the presence of epithelial cells prior
to clinical use.
iv
v
Keywords
Cornea, limbus, stem cells, limbal stem cell deficiency, tissue engineering, limbal
mesenchymal stromal cells, MSC therapy, stem cell therapy, corneal epithelial cells,
silk fibroin, RGD.
vi
vii
List of Publications
Refereed journal publications relevant to this thesis:
Fiona J. Li, Elham Nili, Cora Lau, Neil A. Richardson, Jennifer Walshe,
Nigel L. Barnett, Brendan G. Cronin, Lawrence W. Hirst, Ivan R. Schwab,
Traian V. Chirila, Damien G. Harkin, (2016), “Evaluation of the AlgerBrush
II rotating burr as a tool for inducing ocular surface failure in the New
Zealand White rabbit”, Experimental Eye Research, Vol.147, pp. 1-11
Refereed conference publications and presentations relevant to this thesis:
Elham Nili, Neil Richardson, Rebecca Dawson, Damien G. Harkin,
“Optimisation of an explant technique for initiation and propagation of
limbal mesenchymal stromal cell (L-MSC) cultures from human cadaveric
tissue”, Asia- Association for research in Vision and Ophthalmology
(Asia-ARVO) congress, February 2017, Brisbane, Australia
Elham Nili, Neil Richardson, Rebecca Dawson, Shuko Suzuki, Damien G.
Harkin, “Enhancement of limbal mesenchymal stromal cell adhesion and
proliferation on RGD fibroin”, International Society for Eye Research
(ISER) congress, September 2016, Tokyo, Japan
Elham Nili, Neil Richardson, Rebecca Dawson, Shuko Suzuki, Damien G.
Harkin, “Recombinant fibroin containing the RGD motif enhances limbal
mesenchymal stromal cell adhesion and proliferation”, Royan
International Tween Congress, Reproductive Biomedicine & Stem Cell,
August 2016, Tehran, Iran
Fiona J. Li, Elham Nili, Cora Lau, Neil A. Richardson, Jennifer Walshe,
Nigel L. Barnett, Brendan G. Cronin, Lawrence W. Hirst, Ivan R. Schwab,
Traian V. Chirila, Damien G. Harkin, “Evaluation of the AlgerBrush II
rotating burr as a tool for inducing ocular surface failure in the New Zealand
White rabbit”, Asia Pacific Academic of Ophthalmology (APAO)
Congress, March 2016, Taipei, Taiwan
viii
ix
Statement of Original Authorship
The work contained in this thesis has not been previously submitted to meet
requirements for an award at this or any other higher education institution. To the
best of my knowledge and belief, the thesis contains no material previously
published or written by another person except where due reference is made.
Signature:
Date: 24th of October 2018
QUT Verified Signature
x
xi
Acknowledgements
Coming to the end of my PhD I would like to review this entire experience. In
fact, my PhD journey has been far beyond solely an academic milestone, but was
a bundle of life lessons and a chance to know and develop myself as an individual.
All the emotions, joys, fears, confidence, and disappointment became part of me.
I appreciated every single moment of it, in particular the most difficult ones
which were definitely my best teachers. None of this would have been
experienced and achieved without the support of many people along the way.
To my supervisors - Professor Damien Harkin, Dr Neil Richardson, and Mrs
Rebecca Dawson, I thank you so much for your guidance and support. I thank
you for being my teacher, generously sharing your knowledge, often with big
doses of patience. I am deeply thankful to you.
Damien - anyone in my life who knows about my PhD, knows also how much I
am grateful and feeling blessed to be your student. You have been a tremendous
mentor for me. I immensely thank you for trusting me and accepting me as your
student, giving me the opportunity, and supporting me throughout this incredible
journey. I am indebted to you. I dream to become a teacher and pass all the things
that I have learnt from you, and not to be the last ring in this chain of love,
generosity and kindness.
I am sincerely grateful to Queensland University of Technology. Thank you for
awarding me a Research Training Program scholarship and supporting my
candidature. A big thank to the Queensland Eye Bank for supplying human donor
tissue and the University of Queensland staff for looking after our bunnies and
assisting us in performing our animal trial.
I would like to thank the Queensland Eye Institute, specially Professor Mark
Radford and Professor Triaian Chrilia, for supporting me to conduct my research.
A special thanks to the QEI research team. I truly enjoyed working amongst you
amazing people these years. A heartfelt thank you to my QEI friends/family
Audra, Najla, Natalie and Nadine. Our lab and bright glass door office felt like
home having you present in there.
xii
A heartfelt thank you to my Iranian friends in Australia, Paria, Elnaz,
Mahboubeh, Arezoo and Hamideh, thank you for all your genuine caring and
support.
To Dr Alireza Shiri (personal development mentor) - no amount of words can do
justice to the extent of my appreciation to you and all the life lessons that you
thought me. I wouldn’t be the person I am now without the life skills that I learnt
and still learning from you. I thank you so much!
To my Mum and Dad - thank you for loving me, believing in me, and
continuously encouraging me to achieve my goals. I thank you so much for
overcoming your concerns for me being miles and miles away from you and
traveling all the way to Australia every year to support me in perusing my dreams.
To my beloved brother Ehsan, thank you for your never-ending love and support.
Words are not enough for me to express how much I feel blessed to be the girl of
this amazing family!
And finally, to my beloved husband Hamed. I’m sure you can hear my voice
reading this as I’m repeating it almost every day: “how many people in the entire
world are as lucky as I am for having such a wonderful husband?” You’ve been
always with me and supporting me even though we were geographically apart for
the second half of my PhD journey. This thesis is dedicated to you and my
parents. Thank you for loving me, understanding me, for your patience and
sharing your precious PhD experience. You are my strength, my friend and my
love. Thank you for all those moments that I should have spent with you, but stayed
in Australia so that I could follow my dreams. I am more excited now than I’ve
ever been for the next stage of our life together.
xiii
Table of Contents
Abstract ........................................................................................................................... i
Keywords ........................................................................................................................v
List of Publications ...................................................................................................... vii
Statement of Original Authorship ................................................................................. ix
Acknowledgements ....................................................................................................... xi
Table of Contents ........................................................................................................ xiii
List of Figures ............................................................................................................ xvii
List of Tables .............................................................................................................. xxi
List of Abbreviations ................................................................................................ xxiii
Chapter 1: Introduction .............................................................................. 27
1.1 Research Problem ...............................................................................................29
1.2 Objectives ...........................................................................................................31
1.3 Hypotheses .........................................................................................................32
Chapter 2: Background literature .............................................................. 33
2.1 Structure and function of the Cornea ..................................................................35
2.2 Limbus ................................................................................................................36
2.3 Mesenchymal stromal cells (MSC) ....................................................................37
Effect of MSC on corneal repair ........................................................................39
2.4 Limbal Mesenchymal Stromal cells (L-MSCs) ..................................................39
2.5 Limbal tissue dysfunction...................................................................................41
Overview ............................................................................................................41
Current management of limbal dysfunction .......................................................41
Potential strategies for incorporating cultured L-MSC ......................................42
2.6 Silk fibroin in corneal tissue engineering ...........................................................43
2.7 Strategies for enhancing cell attachment to silk fibroin .....................................44
xiv
2.8 Summary and outline of studies ......................................................................... 46
Chapter 3: Research Study One ................................................................. 47
3.1 Introduction ........................................................................................................ 51
3.2 Materials and Methods ....................................................................................... 53
Sourcing of tissue............................................................................................... 53
Preparation of human cadaveric limbal stromal tissue ...................................... 53
Optimization of explant attachment method ...................................................... 53
Optimization of growth medium for initiating explant cultures ........................ 54
Long-term expansion of cultures in different media .......................................... 55
Resazurin assay as an indicator of cells proliferation and viability. .................. 55
Comparison of isolation techniques explant vs suspension ............................... 56
Immunocytochemistry ....................................................................................... 56
Flow cytometry .................................................................................................. 57
Statistical analysis .............................................................................................. 57
3.3 Results ................................................................................................................ 58
Optimization of explant attachment method ...................................................... 58
Optimization of growth medium for initiating explant cultures ........................ 58
The effects of culture medium on the growth and phenotype of established
cultures ............................................................................................................... 61
Comparison of stromal cell cultures established by explant and collagenase
method ............................................................................................................... 67
3.4 Discussion .......................................................................................................... 68
3.5 Conclusion ......................................................................................................... 71
Chapter 4: Research Study Two ................................................................. 73
4.1 Introduction ........................................................................................................ 77
4.2 Materials and methods ....................................................................................... 81
Animal research ethics ....................................................................................... 81
Human research ethics ....................................................................................... 81
xv
Isolation and cultivation of rabbit L-MSC (RLMSC) ........................................81
Isolation and cultivation of human limbal epithelial (HLE) cells ......................82
Establishment of cultures on human amniotic membrane (HAM) .....................83
Sourcing and general care of rabbits ..................................................................83
Monitoring of serum C-reactive protein levels ...................................................84
Anesthesia ..........................................................................................................84
Wounding of rabbits ...........................................................................................85
Application of cultures to ocular surface ............................................................85
Post-operative care .............................................................................................86
Clinical assessments ...........................................................................................87
Analysis of clinical images .................................................................................87
General histology ...............................................................................................88
Immunostaining ..................................................................................................88
4.3 Results ................................................................................................................91
Construction and analysis of treatment cultures .................................................91
Baseline response to wounding (epithelial debridement without suturing) ........91
Effect of treatment on serum C-reactive protein levels ......................................94
Effect of treatment on re-epithelialization ..........................................................94
Effect of treatment of neovascularization ...........................................................94
Histological analyses ........................................................................................100
4.4 Discussion ........................................................................................................109
Chapter 5: Research Study Three ............................................................ 113
5.1 Introduction ......................................................................................................117
5.2 Materials and methods ......................................................................................120
Materials and consumables for manufacturing of fibroin membranes .............120
Degumming of standard cocoon silk ................................................................120
Generation of fibroin solutions .........................................................................120
xvi
Preparation of standard SF membranes and coating of tissue culture plastic
121
Preparation of PEG-treated, HRP-crosslinked RGD-fibroin membranes ........ 121
Isolation and cultivation of cells ...................................................................... 122
Cell attachment and growth assay.................................................................... 122
Establishment of L-MSC/HLE co-cultures on fibroin membranes.................. 122
Analysis of co-culture 3D structure ................................................................. 123
In vivo testing of co-cultures on RGD-Fibroin membranes ............................. 123
5.3 Results .............................................................................................................. 125
Comparison of L-MSC attachment to fibroin versus recombinant RGD-
fibroin .............................................................................................................. 125
Comparison of HLE cell attachment to fibroin versus recombinant RGD-
fibroin .............................................................................................................. 131
Optimization of HLE/L-MSC co-cultures on recombinant RGD-fibroin ........ 131
Feasibility of engrafting HLE/L-MSC to the ocular surface using fibroin ...... 131
5.4 Discussion ........................................................................................................ 140
Chapter 6: General Discussion ................................................................. 143
6.1 Potential clinical applications of limbal stromal cells. .................................... 144
6.2 Limbal stromal cell phenotype......................................................................... 145
6.3 Administration of limbal stromal cells to the eye. ........................................... 147
6.4 Conclusion ....................................................................................................... 148
Bibliography .................................................................................................. 149
xvii
List of Figures
Figure 2-1 LE = limbal epithelium. BL = Bowman's layer. BV = blood vessels.
Asterisk = termination point for Bowman's layer. ....................................... 35
Figure 3-1 Demonstration of mesenchymal cell outgrowth from pieces of limbal
stromal tissue explanted into culture dishes. ................................................ 59
Figure 3-2 Comparison of cell outgrowth achieved when using different methods
to promote attachment of explanted stromal tissue...................................... 60
Figure 3-3 Visual comparison of cellular outgrowth from limbal tissue explants
using different culture media. ...................................................................... 62
Figure 3-4 Quantitative comparison of cellular outgrowth from limbal tissue
explants using different culture media. ........................................................ 63
Figure 3-5 Visual comparison of stromal cell culture expansion in different
culture media. ............................................................................................... 64
Figure 3-6 Effect of different culture media on the phenotype of L-MSC. ............... 66
Figure 4-1 Representative images of histological sections (H&E stained)................ 92
Figure 4-2 Demonstration of method used to mechanically debride the corneal
epithelium. ................................................................................................... 93
Figure 4-3 Comparison of serum CRP levels between cohorts of treated rabbits. .... 95
Figure 4-4 Time course of re-epithelialization as measured under cobalt lamp
illumination after fluorescein staining. ........................................................ 96
Figure 4-5 Gross appearance of rabbit eyes at 12 weeks under bright light
illumination. ................................................................................................. 97
Figure 4-6 Appearance of rabbit eyes at 12 weeks under cobalt lamp illumination
after fluorescein staining .............................................................................. 98
Figure 4-7 Comparison of corneal neovascularization observed between animals
after 12 weeks .............................................................................................. 99
xviii
Figure 4-8 Normal structure and profile of keratin expression for rabbit cornea
and conjunctiva. ......................................................................................... 101
Figure 4-9 Basic histology of rabbit eyes at 12 weeks as revealed by staining of
sections with hematoxylin and eosin (H&E) and periodic acid-Schiff
stain (PAS). ................................................................................................ 102
Figure 4-10 Immuno-histochemical staining of rabbit eyes at 12 weeks to
demonstrate typical presence of corneal (K3) and conjunctival (K13)
epithelium. .................................................................................................. 103
Figure 4-11 Screening of potential human-specific antibodies by
immunostaining of human corneal-limbal epithelial cells (HLE). ............ 105
Figure 4-12 Screening of potential human-specific antibodies by
immunostaining of rabbit corneal-limbal epithelial cells (RLE). .............. 106
Figure 4-13 Immuno-histochemical staining of human cadaveric eyes to
demonstrate the specific reactivity of human-specific antibody (anti-
HNA mab 235-1) on human corneal/limbal tissue sections. ..................... 107
Figure 4-14 Immuno-histochemical staining of rabbit eyes with anti-HNA mab
235-1 at 12 weeks to trace the presence/absence of grafted human
cultured epithelial cells. ............................................................................. 108
Figure 5-1 Visual comparison of L-MSC attachment to tissue culture plastic
(TCP), TCP coated with Bombyx mori silk fibroin (Fibroin), or TCP
coated with recombinant fibroin incorporating the RGD-cell binding
motif (RGD-fibroin). .................................................................................. 127
Figure 5-2 Quantification of L-MSC attachment to tissue culture plastic (TCP)
coated with Bombyx mori silk fibroin (Fibroin) or recombinant fibroin
incorporating the RGD-cell binding motif (RGD-fibroin). ....................... 128
Figure 5-3 Visual comparison of L-MSC cultures established in the presence of
serum (10% v/v FBS) on tissue culture plastic (TCP), TCP coated with
Bombyx mori silk fibroin (Fibroin), or TCP coated with recombinant
fibroin incorporating the RGD-cell binding motif (RGD-fibroin). ............ 129
Figure 5-4 Quantification of L-MSC growth in cultures established in the
presence of serum (10% v/v FBS) on tissue culture plastic (TCP) coated
xix
with Bombyx mori silk fibroin (Fibroin) or recombinant fibroin
incorporating the RGD-cell binding motif (RGD-fibroin). ....................... 130
Figure 5-5 Visual comparison of HLE cell attachment to tissue culture plastic
(TCP), TCP coated with Bombyx mori silk fibroin (Fibroin), or TCP
coated with recombinant fibroin incorporating the RGD-cell binding
motif (RGD-fibroin)................................................................................... 133
Figure 5-6 Quantification of HLE attachment to tissue culture plastic (TCP),
TCP coated with Bombyx mori silk fibroin (Fibroin), or TCP coated
with recombinant fibroin incorporating the RGD-cell binding motif
(RGD-fibroin). ........................................................................................... 134
Figure 5-7 Confocal fluorescence micrographs demonstrating the basic
morphology of HLE cells grown on free-standing membranes (~10
cm²) prepared from standard fibroin (Fibroin), compared to membranes
prepared from recombinant fibroin incorporating the RGD-cell binding
motif (RGD-fibroin)................................................................................... 135
Figure 5-8 Phase contrast microscopy images of HLE cultures established on
membranes prepared from RGD fibroin solution, compared to
membranes prepared from RGD fibroin solution treated with a porogen
(low molecular weight poly(ethylene) oxide or PEO) and a cross-
linking agent (horseradish peroxidase or HRP) prior to casting. ............... 136
Figure 5-9 Confocal fluorescence microscopy images demonstrating the relative
stratification of HLE cultures grown on RGD Fibroin/PEO/HRP
membranes, in the absence and presence of L-MSC (cultivated on the
opposing membrane surface; not shown). ................................................. 137
Figure 5-10 Fitting a two-dimensional fibroin membrane to the domed surface
of a rabbit cornea. ...................................................................................... 138
Figure 5-11 Post surgery examination of rabbit eye treated with a co-culture of
human limbal epithelial cells and rabbit mesenchymal stromal cells
grown on RGD-Fibroin/PEG/HRP ............................................................ 139
xx
xxi
List of Tables
Table 3-1 Details of culture media ............................................................................. 55
Table 3-2 Quantitative comparison of L-MSC growth in different media over 16
days. ............................................................................................................. 65
Table 3-3 Effect of L-MSC isolation technique on culture phenotype. ..................... 67
Table 4-1 Prior studies of corneal tissue response to L-MSC when applied in
vivo. .............................................................................................................. 79
Table 4-2 Summary of study design .......................................................................... 86
Table 4-3 Summary of clinical data for wounded and treated animals .................... 104
xxii
xxiii
List of Abbreviations
AF Attachment Factor
α-sma Alpha smooth muscle actin
bFGF Basic fibroblast growth factor
BM-MSC Bone marrow mesenchymal stromal/stem cells
BMSF Bombyx mori silk fibroin
BL Bowman’s layer
CD34 Cluster of differentiation cell surface antigen 34
CD90 Cluster of differentiation cell surface antigen 90
DMEM Dulbecco’s modified Eagle’s medium
ECM Extracellular matrix
EGF Epidermal growth factor
FBS Foetal bovine serum
FGF Fibroblast growth factor
H&E Hematoxylin and eosin stain
HAM Human amniotic membrane
HLE Human limbal epithelial cells
HLA Human leukocyte antigen
HRP Horseradish peroxidase
ICC Immunocytochemistry
IHBI Institute of Health and Biomedical Innovation
K3 Keratin 3
K13 Keratin 13
KSR Knockout Serum Replacement
xxiv
LIF Leukemia inhibitory factor
L-MSCs Limbal mesenchymal stromal/stem cells
LSCD Limbal stem cell deficiency
MSC Mesenchymal stromal/stem cells
NEAA Non-essential amino acids
NGS Normal goat serum
PAS Periodic acid-Schiff stain
PBS Phosphate buffered saline
PEG Poly(ethylene glycol)
QUT Queensland University of Technology
RGD Arginine-Glycine-Aspartic acid
RLS Rabbit limbal stroma
SCM Stem Cell Medium
SSM Standard serum-supplemented medium
TCP Tissue culture plastic
TX Treatment
xxv
27
Chapter 1: Introduction
28
29
1.1 RESEARCH PROBLEM
On a global scale, corneal diseases are a major cause of vision impairment and in
severe cases can cause blindness. Within Australia, the bulk of serious corneal
disorders can be treated effectively by using a donor corneal transplant (over 1000
people each year) (Williams et al. 2015). Donor corneal tissue, however, despite its
relative success as a therapy, still carries the inherent risk of immune rejection. This
problem is especially evident in the case of diseases involving the peripheral region of
the cornea, or corneal limbus, since this region contains blood vessels and a higher
number of immune cells compared to the central cornea. Nevertheless, the epithelial
progenitor cells responsible for maintaining the surface of the corneal epithelium are
also concentrated within the corneal limbus (Schermer et al. 1986). Injuries and
diseases affecting the limbus can therefore lead to a condition known as limbal stem
deficiency (LSCD) (Jawaheer et al. 2017); characterised by the loss of corneal
epithelial cells and replacement with conjunctival tissue. If left un-treated, LSCD
eventually leads to a chronic inflammatory condition involving pain, vision loss,
persistent epithelial defects, corneal vascularisation and scarring. Thus, while LSCD
is a relatively rare condition within Australia (Bobba et al. 2017) and other developed
countries, the question of how to restore structure and function to the corneal limbus
can, on an individual basis, be a significant and challenging problem.
Over the last 20 twenty years, strategies for repairing the corneal limbus have logically
been focussed on implanting healthy limbal epithelial progenitor cells to replace those
lost through disease and trauma (Holland 2015). Given the high rejection rate for donor
tissue grafts involving the limbus (Williams et al. 2015), however, an autologous
transplant is preferred. The autologous tissue sample may be either grown in the
laboratory to expand the number of epithelial cells available for transplant (Pellegrini
et al. 1997) or may be dissected and implanted as tissue fragments without prior
cultivation (Sangwan et al. 2012). Autologous treatments are nonetheless limited to
conditions where the required amount of healthy tissue can be safely acquired without
inducing LSCD at the biopsy site. Alternative sources of tissue are therefore required
in the case of patients with disease or trauma affecting both eyes (i.e. bi-lateral LSCD).
To this end, a variety of different progenitor cells including those sourced from oral
mucosa, bone marrow and adipose tissue, have all been considered as a potential
30
therapy for the treatment of LSCD (Nakamura et al. 2015). The focus of this thesis,
however, is on exploring the potential therapeutic value of mesenchymal progenitor
cells isolated from the limbal stroma of donor eye tissue.
In the course of studying the biology of the corneal-limbus, a variety of research groups
have demonstrated the existence of mesenchymal progenitor cells in cultures derived
from limbal tissue biopsies (Branch et al. 2012; Bray et al. 2014; Garfias et al. 2012;
Polisetty et al. 2008). In studies to date, these so-called limbal mesenchymal stromal
cells, or L-MSC, have been shown to encourage the growth of limbal epithelial cells
in culture (Ainscough et al. 2011; Bray et al. 2014; Nakatsu et al. 2014) and to display
similar immunological properties to MSC derived from bone marrow and other tissues
(Bray et al. 2014; Garfias et al. 2012). Notably, cultured L-MSC express low levels of
the cell surface antigen HLA-DR, which suggests their suitability for allogeneic
transplantation (Polisetty et al. 2008). Moreover, L-MSC display immunosuppressive
properties similar to those associated with MSC derived from other tissues (Bray et al.
2014; Garfias et al. 2012). Banked cultures of donor L-MSC might therefore provide
a valuable tool for treating a variety of corneal disorders in a similar way to that
envisaged for MSC derived from other tissues including bone marrow. Nevertheless,
there is uncertainty concerning optimal methods for L-MSC isolation and cultivation
(Sidney et al. 2015a). Moreover, there is little data concerning the effects of L-MSC
when applied to the ocular surface in vivo.
A separate but related issue of importance is the question of how cells such as L-MSC
should be applied to the ocular surface. Historically human donor amniotic membrane
(HAM) has been the most commonly used vehicle for implanting cell cultures onto the
ocular surface (Schwab et al. 2006). HAM, however, displays variable properties both
within and between donors and, within Australia, is presently only available at
significant cost from overseas suppliers. As a result, a number of alternatives to HAM
have been studied including fibrin glue (Pellegrini et al. 1997) and contact lenses (Di
Girolamo et al. 2009). The present study, however, will evaluate the potential value of
membranes prepared from the silk structural protein, fibroin.
In previous studies it has been shown that fibroin membranes support the attachment
and growth of a variety of ocular cell types (Chirila et al. 2008; Madden et al. 2011;
31
Shadforth et al. 2012). Nevertheless, at the time of commencing this project, it has yet
to be determined if fibroin membranes provide a suitable vehicle for applying cells to
the ocular surface. In particular, the lack of recognisable cell-adhesion sites within the
commonly used fibroin, derived from cocoons of the domesticated silkworm Bombyx
mori, suggests that the performance of this material would benefit from inclusion of a
classic cell-adhesion binding motif such as the amino acid sequence arginine-glycine-
aspartic acid (RGD).
1.2 OBJECTIVES
Consideration of the above clinical problem and questions has led to the following
research objectives. With regard to developing a protocol for cultivating donor L-MSC
by tissue banks, a number of parameters including cell isolation technique and choice
of culture medium have been investigated for their potential influence on yield and
purity. The resulting optimised strategy is subsequently used when preparing stocks of
L-MSC used in later studies. With regard to furthering knowledge of L-MSC action in
vivo, the effects of allogeneic rabbit L-MSC are investigated using a rabbit model of
ocular surface injury. Notably, the rabbit L-MSC are administered while cultivated
upon a sheet of HAM and are investigated for effects on corneal wound healing when
applied alone, as well as when applied in conjunction with a culture of human limbal
epithelial cells (HLE). Finally, the potential value of membranes prepared from fibroin
that has been genetically fused with the cell-adhesion motif RGD, is evaluated as a
potential alternative to the use of HAM.
Thus, the specific aims of this study are:
1) To optimise conditions for the routine isolation and cultivation of limbal
mesenchymal stromal cells (L-MSC) from human cadaveric eye tissue.
2) To evaluate the effect of L-MSC on corneal wound healing in vivo.
3) To investigate the feasibility of using fibroin membranes, containing RGD
sequences, as a vehicle for the co-application of L-MSC and epithelial cells to the
ocular surface.
32
1.3 HYPOTHESES
1. With regard to techniques for L-MSC isolation and culture, the central hypothesis
is that, the establishment of cultures from pieces of intact limbal stroma seeded into
culture will result in L-MSC cultures of superior purity to those generated from
enzymatically digested tissue. Moreover, the resulting phenotype of passaged
cultures will be altered through use of media designed to encourage retention of the
progenitor cell phenotype.
2. With regard to testing the effects of L-MSC in vivo, the central hypothesis is that
administration of donor rabbit L-MSC in conjunction with human limbal epithelial
(HLE) cells cultivated on HAM, will lead to improved wound healing compared to
the current standard therapy (HAM with HLE alone).
3. Finally, with regard to optimising fibroin membranes for cell implantation, it is
hypothesised that optimal cell attachment and growth will be achieved through use
of recombinant silk fibroin containing the cell adhesion motif RGD. Moreover, the
optimised fibroin membranes will prove to be a feasible choice of vehicle for
implanting co-cultures of HLE and L-MSC to the ocular surface.
33
Chapter 2: Background literature
34
35
2.1 STRUCTURE AND FUNCTION OF THE CORNEA
The cornea is composed of five definable layers: the corneal epithelium, Bowman’s
layer, corneal stroma, Descemet’s membrane, and corneal endothelium (Figure 2-1).
Recently an additional layer termed “Dua’s layer” has been described as residing
between the corneal stroma and Descemet’s membrane, however, this theory has yet
to gain wider acceptance (Dua et al. 2013; McKee et al. 2014). Together, the layers of
the cornea produce a tissue that is specialized for refracting and transmitting light onto
the retina, while also serving as a protective barrier for the eye (DelMonte and Kim
2011)
Figure 2-1 LE = limbal epithelium. BL = Bowman's layer. BV = blood vessels. Asterisk = termination point for Bowman's layer.
The corneal epithelium is composed of 5-7 layers of stratified squamous, epithelial
cells. Mature corneal epithelial cells in humans can be distinguished from other types
of epithelial tissue by expression of the keratin pair, K3/K12, which becomes more
concentrated towards the surface (Schermer et al. 1986). These epithelial cells tightly
adhere to one another and to the underlying stroma. As such, they present an effective
barrier to the external environment, prevent the entry of pathogens and limit water loss
from the cornea. In addition, with the aid of the tear film, the corneal epithelium
contributes to the transparency and refractive power of the cornea (DelMonte and Kim
2011). Cells of the corneal epithelium have a relatively short lifespan, however, (7-10
days) and so must be continually replenished to maintain the structural and functional
integrity of the cornea.
36
In the central cornea, the corneal stroma resides posterior to the corneal epithelium,
and is separated from the epithelium by an acellular layer called Bowman’s layer (BL).
Corneal stroma, which accounts for approximately 80% of the thickness of the cornea
at the molecular level, is characterised by three main groups of proteins, specifically,
collagens (including types I & III), proteoglycans (including keratocan and lumican)
and various glycoproteins. Collagen fibers and other extracellular matrix (ECM)
components of the stroma are highly organised into alternating orthogonal sheets
(DelMonte and Kim 2011). The major cell type located in a healthy corneal stroma is
the keratocyte (characterised by expression of CD34) (Fini 1999; Sidney et al. 2014).
In the case of injury, these keratocytes become activated and transform into activated
wound repair fibroblasts (CD34-/CD90+) and myo-fibroblasts (CD34-/CD90+/-sma+)
(Fini 1999).
The most posterior layer of the cornea, the corneal endothelium, is a single layer of
cobblestone-shaped epithelial cells that face the aqueous humour (DelMonte and Kim
2011). This layer is separated from the stroma by a thickened basement membrane
known as Descemet’s membrane. Corneal endothelial cells contribute to maintenance
of corneal transparency via a process known as electro-osmosis, which is mediated in
part via the action of numerous ion channels including a Na+/K+ ATPase. Unlike cells
of the corneal epithelium and stroma, the corneal endothelium displays little
proliferative capacity in vivo, however, there is evidence for the existence of
endothelial progenitor cells located towards the peripheral margin in children and
younger adults (DelMonte and Kim 2011; Walshe and Harkin 2014).
2.2 LIMBUS
The corneal limbus is defined as the peripheral boundary between the cornea and
adjacent sclera covered with conjunctival tissue. Bowman’s layer terminates at the
limbus and the basement membrane displays prominent folds into the stroma resulting
in the formation of epithelial crypts (Dua et al. 2005). In humans, the progenitor cells
for replenishing the corneal epithelium are concentrated primarily within the basal
layer of the limbal epithelium, and especially within the limbal crypts (Schermer et al.
1986). While there are no markers specific for these progenitor cells, they are generally
37
identified through the absence of K3/K12 and the presence of various non-specific
epithelial progenitor cell markers including the transcription factor Np63
(Pellegrini et al. 2001).The uppermost layer of the limbal epithelium expresses K3 in
conjunction with K13 (Li et al. 2016). The adjoining conjunctival epithelium also
expresses K13 (Ramirez-Miranda et al. 2011), contains mucin secreting goblet cells
and is less stratified than the corneal epithelium.
The limbal stroma can be distinguished from that of the adjacent cornea by the
presence of blood vessels; a feature that may well contribute to the micro-environment
required to support the adjacent epithelial progenitor cells (Ljubimov and Saghizadeh
2015). In the absence of Bowman’s layer at the limbus, there is opportunity for closer
interaction between basal epithelial cells and adjacent stromal cells including
melanocytes, immune cells and presumptive cells of keratocyte/fibroblast lineage
(DelMonte and Kim 2011; Hashmani et al. 2013). As for the vascular tissue, it is
considered that this increased contact with limbal stromal cells via secretion of ECM
components, cell adhesion molecules and growth factors might also contribute to the
control and maintenance of the limbal epithelial stem cell niche (Casaroli-Marano et
al. 2015; Ljubimov and Saghizadeh 2015). It has recently been demonstrated that
cultures established from the limbal stroma display characteristics of mesenchymal
stromal/stem cells (Polisetty et al. 2008).
2.3 MESENCHYMAL STROMAL CELLS (MSC)
Human mesenchymal stromal/stem cells (MSCs) are a heterogenous population of
fibroblast-like cells isolated from many adult tissues including bone marrow, adipose
tissue, peripheral blood, foetal liver, skeletal muscle, placenta, amniotic fluid and more
recently the limbal stroma of the human cornea (Harkin et al. 2015a). MSCs exhibit a
high capacity for self-renewal and typically display the capacity to differentiate into
cell types associated with mesodermal-derived tissues (e.g. bone, cartilage, muscle) as
well as other tissues (nervous, skin, liver and lung) (Branch et al. 2012; Salem and
Thiemermann 2009; Yao and Bai 2013).
38
In recent years, evidence has emerged that the therapeutic properties of MSCs can be
attributed to dynamic, paracrine interactions between MSCs and host cells. In
particular, the MSC secretome, has been shown to modulate several processes in vitro
and in vivo, such as cell proliferation, survival, differentiation, immunomodulation,
angiogenesis and stimulation of adjacent tissue cells. As a consequence, the MSC
secretome is now considered a promising source of therapeutic agents for wound
healing and tissue regeneration (Gebler et al. 2012).
MSC are an excellent candidate for cell therapy because (a) human MSC are easily
accessible; (b) the isolation of MSCs is straightforward and (c) the cell numbers can
be expanded to clinical scale in a relatively short period of time. Furthermore, MSC
can be stored with minimal loss of potency and human trials thus far have shown no
adverse reactions to allogeneic versus autologous MSC transplants (Cejka et al. 2016;
Phinney and Prockop 2007; Reinshagen et al. 2011; Salem and Thiemermann 2009;
Williams and Hare 2011).
The endogenous role for MSC is typically considered to be a contribution to
maintenance of specific stem cell niches (classically hematopoietic stem cells). As a
consequence, MSCs play key roles in regulating organ homeostasis, wound healing,
and successful aging (Williams and Hare 2011). A key mechanism by which resident
MSC help maintain structural homeostasis is by modulating local inflammation and
immune cell activity at sites of injury or infection. Specifically, they exert
immunomodulatory effects on cells of both innate and adaptive immune systems,
including inhibition of macrophage function, dendritic cell maturation and activation
(Gebler et al. 2012; Lanza et al. 2012; Maxson et al. 2012). Because of these features,
MSCs are of great interest to those seeking to develop novel therapies for immune-
mediated disorders, such as graft-vs-host diseases, autoimmune diseases and
neurodegenerative disorders (Da Silva Meirelles et al. 2009).
According to the International Society for Cell and Gene Therapy (ISCT), MSCs are
defined by three criteria: 1) MSCs must adhere to the plastic tissue culture dish under
the normal tissue culture conditions; 2) several specific markers must be expressed
including CD90, CD73 and CD105. By contrast, the following cell surface markers
should be absent; CD34, CD45, CD14, CD11b, CD79alpha, CD19 and HLA-DR), and
39
3) MSC typically display evidence of osteogenic, chondrogenic and adipogenic
differentiation when grown in various induction media in vitro (Dominici et al. 2006).
Effect of MSC on corneal repair
As described above, MSC have drawn attention as potential therapeutic agents thanks
to their immunomodulatory and anti-inflammatory properties. Initially, MSC isolated
from bone marrow (BM-MSC) were considered as a therapy for corneal tissue repair,
however, an increasing number of research studies have demonstrated that MSC (or
their secretomes) derived from other tissues including adipose tissue, umbilical cord,
and dental pulp, may also have value (Harkin et al. 2016). Generally, MSC have been
shown to decrease markers of corneal inflammation and increase markers of healing
such as re-epithelialisation. Moreover, MSC and/or their products have been
successfully used to improve retention of corneal allografts in rodent models (Oh et al.
2012). Nevertheless, the optimal dosage, timing and route of MSC administration has
yet to be determined for the treatment of corneal disorders. As an alternative
mechanism of action, there have been reports of MSC trans-differentiation into various
corneal cell types. Evidence for this theory is quite strong with regard to MSC
differentiation into keratocytes, but inconclusive for corneal epithelial cells and
endothelial cells (Harkin et al. 2015b). The strongest evidence for epithelial
differentiation resides with demonstration of increased mRNA expression for keratin
3 when MSC are exposed to conditioned medium derived from corneal epithelial cells.
Nevertheless, it has yet to be demonstrated that MSC from any tissue source are
capable of forming a tissue that is anatomically and functionally indistinguishable
from normal corneal epithelium.
2.4 LIMBAL MESENCHYMAL STROMAL CELLS (L-MSCS)
Several investigations have confirmed the presence of MSC-like cells in cultures
derived from the corneal-limbal stroma ( Polisetty et al., 2008; Branch et al., 2012;
Hashmani et al., 2013; Bray et al., 2014). The MSC phenotype has been defined in
large part by possession of cell surface markers as defined by the International Society
for Cell and Gene Therapy (ISCT; Branch et al., 2012), but these cells have also been
shown to display immunosuppressive properties in vitro (Bray et al. 2014; Garfias et
al. 2012). While the existence of such cells is interesting, the clinical relevance of this
40
discovery, unlike that for their limbal epithelial equivalents, has yet to be established.
As such, L-MSC cultures may be considered a “potential therapy” for the treatment of
an as yet undefined condition. Nevertheless, a number of observations suggest that
banked cultures of allogeneic L-MSC might one day become just as useful as the
banked supplies of donor tissue from which they are derived. Firstly, multiple studies
have determined that L-MSC or their equivalents (limbal fibroblasts; Ainscough et al
2011) encourage the growth of corneal epithelial cells from limbal epithelial
progenitor cells in vitro (Ainscough et al. 2011; Bray et al. 2014; Nakatsu et al. 2014).
Furthermore, L-MSC or their equivalents derived from corneal stromal have been
shown to improve stromal wound healing in rodent models (Basu et al. 2014). At very
least, therefore, it is conceivable that L-MSC might provide a useful adjunct therapy
to cultured limbal epithelium used in the treatment of limbal stem cell deficiency (refer
below to section 2.5.1). Such therapies will, however, require development of
optimised protocols for use by cell and tissue banks. To this end, prior studies have
established that the effects of stromal cells on epithelial growth in vitro, drop off
substantially when cultures are established from tissue biopsied from merely a few
millimetres into the adjacent scleral tissue (Ainscough et al. 2011). Moreover, this
negative phenotype was found to be associated with higher expression of the
myofibroblast marker -smooth muscle actin (-sma) (Ainscough et al. 2011). Since
-sma expression is upregulated during cultivation of stromal cells in the presence of
serum, it might well be argued that L-MSC should be cultivated under serum-free
conditions. Additionally, cultures maintained under serum-free conditions or using
commercial serum-replacements have been reported to display a less differentiated
keratocyte phenotype characterised in part by expression of CD34 (Sidney et al. 2015a;
Sidney and Hopkinson 2017). For example, a study in 2015 demonstrated that L-MSCs
could be expanded in a serum-free medium called “Stem Cell Medium (SCM)”
supplemented with a Knockout Serum Replacement and other growth factors such as
basic fibroblast growth factors (bFGF) and leukemia inhibitory factor (LIF) (Sidney et
al. 2015a). The study suggested that the cells maintained a less differentiated,
progenitor cell phenotype under serum-free conditions compared to the ones cultured
in the standard serum supplemented medium (SSM) (Sidney et al. 2015a). Ultimately,
the required L-MSC phenotype might need to be different according to the required
clinical application. For example, while it may be advantageous to utilise cells with a
progenitor-keratocyte phenotype (i.e. CD90-/CD34+) for optimal repair of the cornea
41
stroma, those with an MSC/fibroblast phenotype (CD90+/CD34-) might be necessary
for optimal epithelial cell growth.
2.5 LIMBAL TISSUE DYSFUNCTION
Overview
A wide range of conditions including mechanical force (trauma), microbial infections,
severe dry-eye syndrome, tumours, surgery, thermal or chemical burns, metabolic
abnormalities and immune system disorders may cause damage to, and a loss of limbal
function (Sati et al. 2015). Relatively small injuries are repaired by the limbal epithelial
progenitor cells, as they proliferate, migrate and differentiate into mature limbal-
corneal epithelial cells (DelMonte and Kim 2011). However, in cases where the limbal
epithelial stem cell population is substantially reduced (referred to as limbal stem cells
deficiency; LSCD), the corneal epithelium eventually breaks down resulting in
inflammation and scarring of the ocular surface. This chronic condition subsequently
leads to conjunctivalization of the corneal surface, corneal vascularization and
ultimately vision loss (Tseng 1989). Depending upon the extent and depth of injury,
providing an external source of epithelial as well as stromal stem/progenitor cells may
be essential to treat this disease. Nevertheless, strategies to date have been logically
focussed on replacing the epithelial cells necessary to restore a smooth and transparent
ocular surface.
Current management of limbal dysfunction
For severe ocular surface disorders, surgical replacement of diseased or damaged
tissue and restoration of epithelial progenitor cells is required. For this purpose, either
autologous or allogeneic donor tissue must be transplanted. Either small pieces of
autologous limbus can be transplanted directly to the affected site, or the tissue sample
can be used to generate a sheet of limbal-corneal epithelial cells in vitro, prior to
transfer to the ocular surface (Baylis et al. 2011; Marchini et al. 2012; Nakamura et al.
2006; Pellegrini et al. 1997; Shortt et al. 2007; Spinelli et al. 2010). Typically, the
cultured tissue is grown on either a sheet of devitalised human amniotic membrane
(HAM) (Shimazaki et al. 2002) or fibrin glue (Pellegrini et al. 1997), but other
materials including contact lenses (Di Girolamo et al. 2009) have also been used to
42
support cultures during growth and transfer. Autologous tissue grafts have the
advantage of avoiding immune rejection. There is however a risk of inducing LSCD if
the biopsy taken from the healthy limbus is too large. In cases where no healthy patient
tissue is available (i.e. bilateral LSCD) donor/allogeneic limbal tissue transplant have
been attempted, however, the majority of transplants fail due to immunological
rejection within 1-2 years (Williams et al. 2015). As an alternative, therefore,
autologous biopsies of oral mucosal epithelium have been used as a substitute for
limbal tissue (Inatomi et al. 2006; Ng and Yung 2015; Schalek et al. 2013). A
temperature-sensitive cell-culture detachment technique also has been reported to
avoid the requirement for using human amniotic membrane as a carrier for when
transplanting cultured autologous oral mucosal epithelial cells (Burillon et al. 2012).
Alternatively, an expanded culture of allogeneic donor limbal epithelium is used for
transplant (Buznyk et al. 2015; Qi et al. 2013).
Potential strategies for incorporating cultured L-MSC
Since limbal stromal cells reside closely to the basal limbal epithelial cells in vivo, it
seems logical to mimic this relationship when considering options for performing co-
implantation of both cell types. This rationale is further supported by the
demonstration of increased colony forming efficiency for limbal epithelial cells when
grown in direct contact with L-MSC in vitro (Bray et al. 2014). Thus, an ideal
replacement for the limbal stem cell niche might well be a semi-permeable scaffold
that supports epithelial cell growth on one side and stromal cell growth on the other.
Sheets of donor HAM could therefore be used as a substitute for the natural basement
membrane that separates basal epithelial cells of the limbal stroma, from an underlying
culture of L-MSC. Nevertheless, a variety of simpler approaches have mostly thus far
been trialled in animal models when looking to test the effects of L-MSC. These
studies will be reviewed in more detail in Chapter 4. For example, in one study, three
methods for applying allogenic limbal MSC in suspension were studied in a rat model
of alkali injured ocular surface. This study reported that topical and subconjunctival
application were more effective than intraperitoneally injection (Acar et al. 2015).
Nevertheless, transplantation of cultured L-MSC while attached to amniotic
membrane or synthetic scaffolds has also been attempted (Holan et al. 2015; Yao et al.
2012; Yao and Bai 2013). There do not appear to have been any prior studies of the
43
effects of limbal epithelial cells when combined with L-MSC, by cultivation on
opposing sides of HAM. A major component of this project will therefore be to
evaluate such a treatment. Nevertheless, a potential alternative to HAM will also be
studied composed of the silk structural protein fibroin.
2.6 SILK FIBROIN IN CORNEAL TISSUE ENGINEERING
While a variety of materials have been used as a vehicle for transplanting cells to the
ocular surface, donor HAM has been used most widely (Schwab et al. 2006). HAM
has anti-inflammatory properties and is widely considered to be a safe and effective
treatment for the acute management of LSCD (Konomi et al. 2013). However, some
disadvantages associated with this biomaterial exist including variability in supply,
high cost, its potential risk of transmitting diseases and its opacity, particularly for the
ocular surface usage has prompted the search for an alternative (Bray et al. 2011;
Meller et al. 2000). These limitations have prompted the evaluation of alternative
materials including compressed collagen-based materials (Levis et al. 2015), contact
lenses (Di Girolamo et al. 2009), fibrous scaffolds prepared from synthetic polymers
(Holan and Javorkova 2013) and membranes prepared from the silk structural protein,
fibroin (Harkin et al. 2011).
Silk Fibroin (SF) is the chief component of silkworm cocoon silk (Altman et al. 2003).
The most commonly used source is cocoons of the domesticated silkworm Bombyx
mori (BMSF). Using relatively simple extraction techniques, fibroin protein can be
readily isolated and prepared as an aqueous solution of hydrolysed protein fragments.
Given the fragmentation that occurs during isolation, the fibroin solutions are related
to silk fibers as gelatine is to collagen fibers. Despite fragmentation, a range of
different structures can be fabricated from these hydrolysed solutions including porous
sponges (e.g. formed by freeze drying), transparent membranes and 3D scaffolds of
electro-spun fibres (Harkin et al. 2011). Since purified fibroin is water soluble, the
resulting structures need to be further stabilised by exposure to either water vapour
(“water annealing”) or ethanol; which promotes conversion to a more water insoluble
-sheet structure (Altman et al. 2003). The resulting structures are generally stronger
than equivalents generated from collagen and have the advantage of being cheaper to
44
prepare, with a lower risk of potential contamination with pathogens. Fibroin-based
materials also generally degrade quite slowly in vivo (>60 days) which may offer
potential advantages depending upon the tissue and condition of interest. In addition,
fibroin-based materials have been reported to display evidence of biocompatibility (as
judged by relative absence of inflammatory or immune response) when implanted into
the corneal stroma (Higa et al. 2011) or sub-retinal space animals (Maya-Vetencourt
et al. 2017). It remains to be seen, however, whether it is feasible to attach fibroin
membranes to the ocular surface in a similar way to how HAM has been routinely
used. Moreover, at the time of starting this project, there have been no reported
attempts at using fibroin as a vehicle for corneal cell implantation. As a minimal
requirement, fibroin will be required to support the attachment and growth of corneal
cells.
2.7 STRATEGIES FOR ENHANCING CELL ATTACHMENT TO SILK
FIBROIN
Since silk fibroin, unlike collagen, does not naturally contain any recognisable cell-
binding motifs, cell attachment is usually facilitated through the addition of other
factors. For example, it has been shown that fibroin supports the attachment and
growth of many different types of cells, when grown in the presence of serum or ECM
proteins (Bray et al. 2011; Harkin et al. 2011; Lawrence et al. 2009). In 2008 a study
conducted by our group first demonstrated that fibroin membranes support the
attachment and growth of human limbal epithelial (HLE) cells and thus may support
implantation of these cells (Chirila et al. 2008). A variety of other ocular cell types
have subsequently been grown on fibroin membranes including corneal endothelial
cells (Madden et al. 2011) and retinal pigment epithelial cells (Shadforth et al. 2012).
In both cases, optimal cell attachment has required pre-coating of the fibroin
membranes with purified ECM proteins including vitronectin and type IV collagen.
An alternative form of fibroin scaffold prepared from partially purified silk fibres has
been shown to support the growth of L-MSC when grown in serum-supplemented
growth medium (Bray et al. 2012a), however the suitability of fibroin membranes has
yet to be evaluated.
While cell attachment to fibroin can be facilitated through addition of ECM proteins,
from a clinical manufacturing perspective, however, it would be advantageous to avoid
45
the use of ECM proteins since they are generally sourced from animal or human tissues
and as such carry significant expense and risk of disease transmission (Schwab et al.
2006). An alternative strategy, therefore, has been to incorporate synthetic peptides
that are known to mimic the cell-binding sites found with ECM proteins including the
amino acid sequence; arginine-glycine-aspartic acid or RGD. Two basic strategies
have been evaluated thus far; either using fibroin derived from alternative species of
silkworm that are known to contain “theoretically active” RGD sequences
(Hogerheyde et al. 2014), or the modification of fibroin by coating with RGD peptides
(Bray et al. 2013). Neither strategy, however, has so far led to significant increases in
cell attachment and growth when studying limbal epithelial cells. Nevertheless, an
alternative strategy has recently become available through the generation of a
genetically engineered form of fibroin than contains copies of the RGD sequence fused
within one of the main fibroin gene products (Kambe et al. 2010a; 2010b). This
material has been reported to enhance the attachment and growth of chondrocytes and
so may also be suitable for other cells of mesenchymal lineage including L-MSC. The
response of limbal epithelial cells to this recombinant fibroin is also presently
unknown.
46
2.8 SUMMARY AND OUTLINE OF STUDIES
A smooth and transparent corneal surface is essential for vision. This corneal surface
is primarily maintained through the actions of epithelial progenitor cells residing
within the corneal limbus. The potential accessory role of limbal mesenchymal
progenitor cells, however, is presently unclear. Moreover, a standard protocol for
cultivating limbal stromal cells has yet to be established.
Cultures derived from limbal stroma display a mesenchymal stromal/stem cell
(MSC) phenotype when established in serum-supplemented growth medium. These
limbal MSC encourage the growth of limbal epithelial cells in vitro, but their effects
on epithelial growth in vivo are unclear.
Membranes prepared from silk fibroin may provide a useful vehicle for the co-
implantation of limbal epithelial cells and L-MSC, but cell adhesion is unlikely to be
optimal since fibroin derived from the domesticated silkworm Bombyx mori lacks
recognised cell-bining motifs.
The following experimental chapters will therefore; explore different approaches to
establishing L-MSC cultures, examine the effects of L-MSC applied to the wounded
ocular surface, and investigate the potential benefits of the RGD-cell binding motif on
the use of fibroin membranes as a vehicle for the co-transplantation of limbal epithelial
cells and L-MSC.
47
Chapter 3: Research Study One
OPTIMISATION OF CULTURE CONDITIONS FOR THE
ESTABLISHMENT OF LIMBAL MESENCHYMAL STROMAL CELL
CULTURES
48
49
Statement of contribution
The following people contributed to data presented in this chapter.
Following initial screening of antibodies by myself, the final assessment of stromal
cell phenotype by multiple channel flow cytometry in Table 3-2 was performed by
staff from Mater Pathology Services (a NATA-accredited testing laboratory).
50
51
3.1 INTRODUCTION
While the therapeutic value of L-MSC is presently unclear, further evaluation of their
safety and efficacy will be facilitated through the development of standardised
manufacturing protocols. The basic question posed at the commencement of this first
study, therefore, is; what advice would be given to tissue banks seeking to efficiently
establish master stocks of L-MSC with high yield and purity?
In studies to date, methods for the isolation and cultivation of L-MSC have mostly
relied upon use of serum-supplemented growth medium (Bray et al. 2014; Polisetty et
al. 2008). The inclusion of serum is useful for two main reasons; it supplies cell
attachment factors in the form of vitronectin and fibronectin and provides growth
factors that promote proliferation. Nevertheless, since prolonged exposure of stromal
cells to serum is associated with myofibroblast formation (Bray et al. 2012b), such
conditions may well be suboptimal for subsequent effects on epithelial cell growth.
Thus, it could well be argued that optimal culture conditions will be those that
discourage differentiation and encourage retention of progenitor cell characteristics.
To this end, Sidney et al (2015a) have recently reported improved retention of
progenitor cell state when limbal stromal cells are cultivated in a serum-free growth
medium referred to as Stem Cell Medium. More specifically, the authors observed
increased expression of the keratocyte marker CD34. It was therefore decided to test
SCM with the view to confirming the observation of Sidney et al. (2015a).
A second feature of techniques commonly used to cultivate L-MSC is the practice of
establishing cultures from tissue samples that have been dissociated via enzymatic
digestion in collagenase (Bray et al. 2012b; Garfias et al. 2012). While digestion
provides an efficient means for quickly releasing cells from the limbal stroma, this
technique will theoretically release a range of different cell types in addition to those
that give rise to those with an MSC phenotype. As such, a number of contaminating
cells including melanocytes and immune cells may also be present. From a
manufacturing point of view, the requirement for collagenase is also a disadvantage
since a defined or clinical grade of enzyme would eventually be required in order to
comply with requirements of good manufacturing practice (GMP), thus adding to
overall costs.
52
As an alternative to use of collagenase digests, I therefore developed a protocol based
upon the use of limbal tissue explants seeded into culture dishes. In doing so, the
rationale was that the type of cells that preferentially emigrated out of the tissue
fragments would be more likely to be those associated with stromal wound healing in
vivo. More specifically, it was anticipated that exposing tissue to serum-supplemented
growth medium would recreate the normal processes of stromal cell activation,
migration, proliferation and differentiation as seen in vivo during inflammation (Fini
1999). Moreover, once the cultures had been established and expanded, the cells could
subsequently be reverted to a more progenitor-like state (i.e. higher CD34), by placing
back under culture conditions as developed by Sidney et al (2015a).
As a starting point to this study, however, it was first necessary to establish a reliable
technique for attaching pieces of limbal stromal to culture dishes. Three different
methods are examined in this preliminary study; (1) direct attachment using serum-
coated culture plastic, (2) culture plastic pre-treated with a commercial gelatine-based
product called Attachment factor, and (3) indirect attachment of tissue fragments by
immersion in a gel composed of medical grade porcine type I collagen. Having decided
on the optimal attachment method, a comparison is subsequently made of standard
serum-supplemented medium compared to Stem Cell Medium (Sidney et al 2015) for
initiating cultures from explants, as well as for subsequent expansion. Finally, a
comparison of cultures established from tissue explants is made with those established
from collagenase digests of limbal stroma. Throughout these studies, the phenotype of
cultures is analysed using a combination of immunocytochemistry and flow cytometry
for key stromal cell markers including CD34, CD90 and -sma.
53
3.2 MATERIALS AND METHODS
Sourcing of tissue
Samples of cadaveric human corneal limbus were obtained in the form of corneal-
scleral rims discarded following routine surgery at the Queensland Eye Hospital.
Access to tissue samples was enabled through an existing agreement with the QEH
with accompanying clearance from QUT’s Human Research Ethics Committee
(Approval No. 0800000807).
Preparation of human cadaveric limbal stromal tissue
Each sample was washed in Hanks Balanced Salt Solution (HBSS) three times
separately. The epithelial and endothelial tissue layers were subsequently removed by
mechanical dissection (scraping with curved watchmaker forceps) following digestion
for 1 h in a 2.5 mg/mL Dispase II (Gibco cat#:17105-041) solution at 37 °C.
Optimization of explant attachment method
Three alternative methods of attaching tissue were examined. For each treatment, 5
separate limbal tissues explants of a reproducible size were excised using a 2 mm
diameter punch biopsy (Kai medical). The three different methods for explant adhesion
tested were:
1- Serum coated tissue culture plastic (TCP): Explants were placed in TCP 6-
well plates pre-coated with FBS and allowed to stand at room temperature for
1 h. Following this, 500 µL of FBS was gently added to individual culture
wells, then incubated at 37 °C in 5% CO₂/air overnight. The following day,
500 µL of SSM was added to each well then plates were incubated once again
at 37 °C in 5% CO₂/air.
2- Attachment factor (AF): For this treatment, a commercial gelatine based (life
technologies Ref: S006-100) was used as an explant adhesive. Briefly, explants
were placed in TCP wells pre-coated with AF as per the manufacturers’
instructions and allowed to stand at room temperature for 1 h. Following this,
500 µL of medium was added to each well and incubated at 37 °C, in 5%
54
CO₂/air. After 24 h, an additional 500 µL medium was added to each well then
plates were incubated once again at 37 °C in 5%/air CO₂.
3- Immersion in collagen gel type I (Col): A collagen solution [Cellmatrix Type
1-P (Nitta Gelatin Inc.)] was prepared as per manufacturer’s instruction, then
30 µL of this solution was placed on the top of each biopsy. Cultures were
incubated at 37 °C in 5% CO₂/air for 30 minutes so that the collagen
polymerised. Subsequently, 500 µL of serum-supplemented medium (SSM)
[the media composition has been stated in table 3-1] was added to each well
and plates were incubated at 37 °C in 5% CO₂/air overnight. The following
day, an additional 1000 µL of SSM was added to each well and after 2 days, a
further 2 mL medium was added. Throughout the culture period, media was
partially replaced every 2 or 3 days with SSM containing DMEM high glucose
(Life Technologies 10313-021) with 10% (vol/vol) Foetal Bovine Serum
(FBS), 2 mmol/L-glutamine (Life Technologies 25030-081) and 1% Penicillin
Streptomycin (Life Technologies 15140-122. Culture success was assessed
based on the number of explants/6-well plate displaying any outgrowth after 7,
10 and 14 days.
Optimization of growth medium for initiating explant cultures
Explants were prepared as per 3.2.2 and 3.2.3 (part 3) and incubated in either SSM or
SCM1 with media partially replaced every 2 days. Two different versions of the SCM
medium were tested owing to uncertainty regarding which base medium had been used
in the original Sidney et al (2015a) study. A resazurin assay was conducted to measure
the metabolic activity of cell cultures at 20 days.
55
Table 3-1 Details of culture media
Medium Basal medium Supplements Reference
SSM DMEM (Life
Technologies)
High glucose
10313-021
10% (vol/vol) FBS (HyClone, SH 30084.03)
2 mmol/L-glutamine, 100 unit/mL penicillin,
100µL/mL streptomycin.
Bray et al.
(2012b)
SCM1 Knockout
DMEM/F12
(Life
Technologies
12660-021)
2 mmol/L-glutamine
20% (vol/vol) Knockout SR (Life Technologies
108 28-010), 1% (vol/vol) non-essential amino
acids, 4 ng/mL b-FGF (Invitrogen 13256029), 5
ng/mL h-LIF (Invitrogen PHC9484), 100 unit/mL
penicillin, 100 µL/mL streptomycin)
Sidney et al.
(2015a)
SCM2 DMEM
High glucose
(Life
Technologies)
High glucose
10313-021
2 mmol/L-glutamine, 20% (vol/vol) Knockout SR
(Life Technologies 108 28-010), 1% (vol/vol) non-
essential amino acids, 4 ng/mL b-FGF (Invitrogen
13256029), 5 ng/mL hLIF (Invitrogen PHC9484),
100 unit/mL penicillin, 100 µL/mL streptomycin.
Long-term expansion of cultures in different media
L-MSC explant cultures were established as described previously in SSM and
expanded to passage 3. Cells were then cultured with either SSM, SCM1 or SCM2
media. After 16 days all cultures were harvested by rinsing with Phosphate buffered
saline (PBS) and Versene [0.02% (w/v) EDTA in Phosphate Buffered Saline] (Gibco
Cat. No. 15040066) followed by incubation for 5-10 minutes in TrypLE select enzyme
(Gibco Cat. No. 12563011) and seeded coverslips at a density of 15,000 cells/cm².
After 2 days, cultures were fixed with 10% formalin and processed for
immunocytochemistry (ICC) to determine cell phenotype.
Resazurin assay as an indicator of cells proliferation and viability.
Resazurin assay was performed as per the manufacturers’ instructions to estimate the
metabolic activity of cultured cells. Briefly, resazurin (Sigma, R7017) reagent was
added to the primary cultures (explants immersed in collagen), as well as the negative
controls (no cells) for each medium in two different culture media to a final
concentration of 700 µM. After 4 h incubation at room temperature, 800 µL of the
medium was transferred to a new 24-well plate, and the absorbance1 of each well was
1 Measuring either fluorescence or absorbance is optional based upon the manufacturer protocol.
56
measured on a spectrometer at 570nm and 600 nm. Resorufin (reduced form of
resazurin) was calculated using the following formulation.
% Reduced = 1 +(117,216) A570 – (80,586) A600
(155,677) A`600 – (14,652) A`570 × 100
A 600 = absorbance of test wells at 600nm
A 570 = absorbance of test wells at 570nm
A`600 = absorbance of negative control wells at 600nm
A`570 = absorbance of negative control wells at 570nm.
117,216 = molar extinction coefficient of resazurin in the oxidized form at 600nm
80,586 = molar extinction coefficient of resazurin in the oxidized form at 570nm
14,562 = molar extinction coefficient of resazurin in the reduced form at 600nm
155,677 = molar extinction coefficient of resazurin in the reduced form at 570nm
Comparison of isolation techniques explant vs suspension
As described previously in section 3.2.2, limbal stromal tissue was prepared from 5
different unique donors. Specifically, using a 2 mm biopsy punch, 16-18 equal size
biopsies were obtained from each limbal rim with equal number of biopsies used for
each isolation method (8-9). Explants were placed on TCP as described previously in
20 µL collagen solution. Alternatively, excised tissue was processed for collagenase
digestion. Briefly, excised limbal stroma was digested for 12-48 h in 2 mL of 1 mg/mL
type I collagenase (Gibco cat No. 1700.017 340.00 units/mg). The resulting digest was
then gently pipetted to loosen remaining tissue clumps, before being washed and re-
suspended in SSM. For cell cultures generated via either of the two techniques (explant
or enzyme digestion) culture medium was partially replaced with SSM every 2-3 days.
In order to obtain enough cells for phenotyping with flow cytometry and
immunostaining with α-sma (at least 1.5x10⁶), cultures were expanded up to passage
Immunocytochemistry
Immunostaining was performed to determine the expression of CD34, CD90 and α-
smooth muscle actin in L-MSC cultures. Cells were seeded at a density of 15,000
cells/cm² on glass coverslips and cultured with SSM at 37oC in 5% CO2/air. After 24
h, cells were washed three times with PBS before fixation with 10% formalin for 15
57
minutes. Following this, cells were rinsed briefly with PBS then permeabilized (only
for staining with α-sma) with 0.1% (vol/vol) Triton X-100 (Sigma) and blocked (for
CD34, CD90 and α-sma) with 1% Normal Goat Serum (NGS) for 1 h at room
temperature. Subsequently cultures were incubated in 1:100 dilution primary
antibodies CD34 (Dako M7165), CD90 (BD-555593) or α-sma (Dako M0851) with
1% NGS in PBS at room temperature for 1 h. Cultures were then washed 3 times (15-
30 minutes each wash) then 1:100 dilution secondary antibody Alexa Fluor 488 goat
anti-mouse immunoglobulin G (Life Technologies Ref: A11001) in PBS was applied
to each culture for 1h at room temperature or overnight at 4°C. Primary antibody was
not applied to the negative controls. All cultures were counterstained with 2µg/ml
Hoechst 33342 nuclear stain in PBS to reveal total cell number. The stained cultures
on glass coverslips, were mounted on glass slides in 100% glycerol and sealed. Images
of stained cultures were obtained using a Nikon TE2000-U fluorescence microscope
equipped with a CoolSNAP cooled CCD camera and labelled montages created using
Adobe Photoshop.
Flow cytometry
Cells were obtained through either of the two isolation techniques explant in collagen
versus collagenase digestion were grown up to passage 3-4 in SSM. Cells were
harvested when 85-90% confluent, by a brief rinsing with PBS, and versene, and being
treated with TrypLE for 5 minutes. Initial tests of reactivity were confirmed for each
antigen separately using antibodies sourced from Miltenyi-Biotec and the
MACSQuant 10 Analyser. Each cell sample was subsequently taken to the
immunology pathology laboratory department at the Adult Mater Hospital (a NATA
accredited laboratory) and characterised by flow cytometry using the same standard
panel of markers as listed in Table 3.3.
Statistical analysis
Data were analyzed by use of Graph-Pad Prism version 7.1. A non-parametric one-
way ANOVA test (Kruskal-Wallis) was used to compare explant growth conditions
(Figure 3-2). A two-tailed T test was used for comparing two culture media (Figure
3-4).
58
3.3 RESULTS
Optimization of explant attachment method
An essential step in establishing a culture of stromal cells from a limbal tissue explant
is to ensure that the pieces of tissue are securely attached to the culture dish. Three
alternative methods of attaching tissue were therefore examined; serum in the form of
100% FBS, a commercially available gelatine-based reagent known as Attachment
Factor and a collagen gel. All three methods supported the attachment of tissue;
however, considerable care must be taken during the first 24-48 h to ensure tissue
remains attached (refer to the methods section 3.2.3).
While all attachment methods supported the establishment of mesenchymal cell
cultures (Figure 3.1) the timing and efficiency of cell outgrowth varied between
techniques (Figure 3.2). When examined over a 14-day culture period, collagen gels
consistently supported the greatest number of explants with cell outgrowth (up to a
maximum of 6 per plate) and was found to be significantly more efficient than
Attachment Factor at 14-days.
Optimization of growth medium for initiating explant cultures
Having demonstrated that collagen gels provide the most efficient method for
attaching pieces of limbal stroma, the effects of two types of culture media were
examined on mesenchymal cell outgrowth; standard serum supplemented growth
medium (SSM) and a medium reported previously to maintain the stromal cells in a
progenitor cell state (Stem Cell Medium or SCM (Sidney et al. 2015a) have referred
to as SCM1. While both media supported outgrowth of mesenchymal cells, a
noticeable difference in cell numbers emerging from the tissue were evident when
examined using an Olympus TS-100 inverted phase contrast microscope equipped
with a 10x objective lens (Figure 3-3). Examination of metabolic activity at 20-days
(by resazurin assay) within paired cultures for each donor, confirmed that significantly
greater growth was achieved using SSM (p <0.001, by paired two-tailed T test; Figure
3-4).
59
Figure 3-1 Representative images of mesenchymal cell outgrowth from pieces of
limbal stromal tissue explanted into culture dishes.
Tissue explants were attached to tissue culture plastic by using either foetal bovine
serum (100% FBS), a commercial gelatine-based adhesive (Attachment Factor or AF)
or collagen gel. The phase contrast images display the appearance of cultures after
approximately 10 days growth in standard serum supplemented growth medium.
60
Figure 3-2 Comparison of cell outgrowth achieved when using different methods
to promote attachment of explanted stromal tissue.
Pieces of tissue were attached to the bottom of each well of a 6-well culture plate using
either 100% FBS (serum), Attachment Factor (AF) or a collagen gel (Col) (total of 18
explants). The resulting cultures were analysed by phase contrast optics for evidence
of cell outgrowth over two weeks culture in serum supplemented growth medium. Line
graphs demonstrate the mean +/- SEM for the number of explants displaying visible
evidence of cell outgrowth per plate from 5 separate experiments. Asterisk indicates a
significant difference in outgrowth observed between use of Attachment Factor and
Collagen gel (p<0.5). There was no significant difference between use of Collagen
gels and the serum attachment method however.
61
The effects of culture medium on the growth and phenotype of
established cultures
Given the superior properties of SSM for establishing cultures, three unique cultures
were established from separate donors and expanded to passage 2 before being stored
in liquid nitrogen. A comparison of growth and cell phenotype for cultures maintaining
in either SSM or two versions of SCM (Table 3-1) were subsequently made. Cells
seeded into SSM (4,000/cm2) grew to confluency within 7 days and were able to be
passaged twice over a period of 16 days. The suspected poor growth within the two
SCM type media was confirmed by cell counts performed on day 16 which were
typically lower than the original number of cells that had been seeded into culture
Table 3-2. In contrast, cells seeded into either of the SCM (using either Knockout
DMEM/F12 or DMEM) failed to achieve confluency and gradually formed clusters
over the 16-day period (Figure 3-5). The poor growth of stromal cells within each of
the SCM type media led to insufficient numbers being generated to support an analysis
of cell phenotype by flow cytometry. Immunocytochemistry, however, could still be
performed on cells that had been harvested after 16 days. Cultures were seeded onto
glass cover slips and maintained in their respective growth medium for a further 24-h
prior to immunostaining for CD34, CD90 and alpha-smooth muscle actin (-sma).
Cells grown in SSM were >99% positive for CD90. Approximately <10% of cells
displayed positive staining for -sma within stress fibers. CD34 was rarely observed
except for one donor where approximately 4% of cells displayed clear evidence of
staining (Figure 3-6).
The presence of cell-aggregates for cultures maintained in either SCM type medium,
made it difficult to clearly observe the results of staining by conventional fluorescence
microscopy. The analysis was therefore limited to cells that were more clearly visible
outside of the aggregates. The majority of cells (>95%) grown in either SCM medium
were positively stained for CD90. Neither staining for CD34 nor -sma could be
clearly observed above background. Some degree of fluorescence was observed within
aggregates for both CD34 and -sma, but this was not investigated further.
62
Figure 3-3 Visual comparison of cellular outgrowth from limbal tissue explants
using different culture media.
Tissue explants of approximately equal size (~2 mm diameter) were attached to culture
plates using collagen gels and subsequently incubated in either standard serum-
supplemented growth medium (SSM; containing 10% FBS) or a medium designed to
maintain the stromal cells in a progenitor cell state (Stem Cell Medium or SCM;
containing 20% “Knock-out” Serum Replacement or KSR). Phase contrast
micrographs display typical appearance of cellular outgrowth observed after
approximately 10 days.
63
Figure 3-4 Quantitative comparison of cellular outgrowth from limbal tissue
explants using different culture media.
Tissue explants of approximately equal size (~2 mm diameter) were grown within
collagen gels for 20 days using either standard serum-supplemented growth medium
(SSM; containing 10% FBS) or a medium designed to maintain stromal cells in a
progenitor cell state (Stem Cell Medium or SCM; containing 20% “Knock-out” Serum
Replacement; KSR). The relative metabolic activity within each culture was
subsequently used as an indicator of cellular outgrowth (by measuring the % reduction
of Resazurin). Bars represent the mean +/- SEM for seven different tissue donors, with
at least 3 cultures being established per donor in each medium. Asterisk indicates a
significantly lower response to cultivation in SCM (p <0.001, by paired two-tailed T
test, n = 7) 2..
2 Since the resazurin assay is a measure of metabolic activity rather than a direct
measure of cell numbers care must be taken when interpreting this data. Nevertheless,
the higher levels of metabolic activity observed above are consistent with the
differences directly observed by phase contrast microscopy in Figure 3.3.
64
Figure 3-5 Representative images of stromal cell culture expansion in different
culture media.
Cultures initiated from tissue explants and expanded in serum-supplemented medium
to passage 3 (SSM; 10% FBS in DMEM), were subsequently further passaged in this
same medium or Stem Cell Medium (as used in Figure 3-3 and Figure 3-4) prepared
using either Knock-out DMEM/F12 medium (SCM1) or DMEM (SCM2) as the base
medium. Phase contrast micrographs demonstrate typical morphology of cultures on
days 7 and 16 respectively. Asterisk is used to highlight that the culture maintained in
SSM was passaged twice during this time period, while those maintained in SCM1 and
SCM2 are the original cultures that failed to reach confluency.
65
Table 3-2 Quantitative comparison of L-MSC growth in different media over 16
days. Cultures initiated from tissue explants and expanded in serum-supplemented medium
to passage 3 (SSM; 10% FBS in DMEM), were subsequently further passaged in this
same medium or Stem Cell Medium (as used in Figure 3-3 and Figure 3-4) prepared
using either Knock-out DMEM/F12 medium (SCM1) or DMEM (SCM2) as the base
medium. Cell counts at harvest indicate value range obtained for three tissue donors.
Day/Medium SSM SCM1 SCM2
0
Seeded:
~3 x 105 cells
per 75 cm2 flask.
Seeded:
~3 x 105 cells
per 75 cm2 flask.
Seeded:
~3 x 105 cells
per 75 cm2 flask.
5-7 ~90% confluent
Harvested:
1.2 to 6 x 106 cells.
Re-seeded:
~3 x 105 cells
per 75 cm2 flask.
14 ~90% confluent
Harvested:
1.1 to 1.5 x 106 cells.
Re-seeded:
~3 x 105 cells
per 75 cm2 flask.
16 Harvested:
0.5 to 1.5 x 105 cells
per 75 cm2 flask per
donor.
Harvested:
0.7 to 3.5 x 105 cells
per 75 cm2 flask per
donor.
66
Figure 3-6 Representative images of the effect of different culture media on the
phenotype of L-MSC.
Cultures initiated from tissue explants and expanded in serum-supplemented medium
to passage 4 (SSM; 10% FBS in DMEM), were subsequently grown for a further 16
days in this same medium or Stem Cell Medium (as used in Figure 3-3 and Figure 3-4)
prepared using either Knock-out DMEM/F12 medium (SCM1) or DMEM (SCM2) as
the base medium. The cultures were then passaged onto glass cover slips (15,000
cells/cm2) and grown overnight prior to fixation and immunostaining for CD34, CD90
or -sma. Negative control (Neg.Con.) was performed by omission of primary
antibody step.
67
Comparison of stromal cell cultures established by explant and
collagenase method
The phenotype of cultures established using either the optimal explant technique or
from collagenase-digested tissue, were examined in parallel for five consecutive tissue
donors. The resulting cultures (p3 or p4) were analysed by flow cytometry using a
panel of antibodies typically employed to identify mesenchymal stromal cells and
potential contaminating cells. As demonstrated in Table 3-3, a consistent profile of
positive staining was observed across all 5 donors, with >99% of cells expressing
CD73, CD90 and CD105, and <1% of cells routinely expressing CD14, CD34, CD45,
CD79a or HLA-DR. A further analysis of phenotype by immunocytochemistry
revealed that less than 10% of cells contained evidence of staining for -sma within
stress fibers. Thus, no significance difference was discerned between cultures that had
been initiated using either technique.
Table 3-3 Effect of L-MSC isolation technique on culture phenotype.
Cells were isolated from the limbal stroma of 5 sequential tissue donors using the
explant technique (Ex.) in parallel with a conventional collagenase digestion technique
(Col.). All resulting cultures were maintained and passaged 3 to 4 times in standard
serum supplemented medium before being analysed by flow cytometry or
immunocytochemistry (for -sma).
Antigen
% Immunoreactivity by Flow Cytometry or ICC*
Donor 1 Donor 2 Donor 3 Donor 4 Donor 5
Ex. Col. Ex. Col. Ex. Col. Ex. Col. Ex. Col.
CD14 <1 <1 <1 <1 <1 <1 <1 <1 1.1 1.7
CD34 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1
CD45 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1
CD73 >99 >99 >99 >99 >99 >99 98 99 >99 >99
CD79a <1 <1 <1 <1 <1 <1 <1 <1 <1 <1
CD90 >99 >99 >99 >99 >99 >99 >99 >99 >99 >99
CD105 99 98 >99 >99 >99 >99 97 97 >99 >99
HLA-DR <1 <1 <1 <1 <1 <1 <1 <1 <1 <1
-sma 2.8 0.9 5 4 10 5 0 9 1 9.5
*ICC = immunocytochemistry.
68
3.4 DISCUSSION
A key question in the development of L-MSC as a potential therapy for treating corneal
disease is to decide which culture conditions are appropriate for the initiation and
subsequent expansion of cultures prior to clinical use. In particular, there is debate over
the relative merits of using serum-supplemented growth medium, compared to
alternative growth medium that are either serum-free or use serum-replacements
(Sidney et al. 2015a).
On the one hand, use of serum provides an effective source cell attachment factors and
mitogens that are necessary to initiate and drive the expansion of cultures. The
resulting cultures are known to express markers that are typical of MSC (Polisetty et
al. 2008), promote growth of epithelial cells (Bray et al. 2012a) and display useful
immunological properties (Bray et al. 2014). Use of serum, however, drives stromal
cell differentiation, including the formation of myo-fibroblasts which, despite being a
part of the normal wound healing process, are also associated with poor epithelial cell
growth and corneal scarring (Ainscough et al. 2011).
On the other hand, it would be beneficial not to have animal derived products in the
culture media owing to potential for contamination with infectious materials.
Moreover, the use of serum-free media or media containing serum-replacements has
been demonstrated to promote expression of markers that are associated with less
differentiated cells, similar to the keratocytes found within corneas under normal
conditions (Sidney et al. 2015a)3. It can therefore be equally argued that these growth
conditions are more appropriate since they are more likely to support production of
cells that more closely mimic healthy corneal tissue. Nevertheless, less is known about
the effects of these more keratocyte-like cells on epithelial cells and whether they
display similar immunological properties to MSC.
3 Whilst we have used CD 34 as a marker for keratocyte, it is apparent from the work of others (Sydney
et al. 2015.a) that they consider the expression of CD34 to be indicative of a progenitor cell status.
69
In the present study, it is argued that fragments of intact tissue when seeded into culture
dishes will provide a more appropriate way to initiate stromal cell cultures since this
will effectively mimic the processes of activation, migration and proliferation during
wound healing in vivo. The results indicate that suspension of tissue fragments
(explants) within collagen gels provides a more effective method of initiating cultures
than by use of serum or gelatine as tissue adhesives. Media containing either serum
(SSM) or serum-replacement (SCM) both supported cellular outgrowth, but it was
significantly greater when using serum-supplemented medium. On this basis it was
decided to initiate cultures in collagen gels using serum-supplemented medium, while
retaining the option to expand cultures in SCM type media. Unfortunately, neither of
the two SCM-type media were found to support culture expansion and so this option
was not investigated further. Finally, it was demonstrated that cultures established
using the explant method are actually indistinguishable from those established from
enzymatically dissociated tissue after three to four passages. On this basis, it can be
concluded that the use of tissue explants suspended in collagen gels offers no
advantage over use of digested tissue. Moreover, it would appear that SCM-type media
are unsuitable for expanding cultures of limbal stromal cells. Nevertheless, a closer
analysis of the issue suggests that a combination of techniques may actually have
benefits after all.
During the design of experiments, focus was placed on comparing standard serum-
supplemented medium (SSM) with the culture medium found previously (SCM)
(Sidney et al 2015) to support greatest expression of markers associated with corneal
stromal progenitor cells including CD34. Nevertheless, these prior studies were
conducted using culture plastic that had been pre-coated with a commercial gelatine-
based product. This difference, therefore, is very likely to explain the poorer growth
observed in this chapter when established cultures were subsequently transferred into
SCM without pre-treatment of culture plastic. Nevertheless, the initial finding of
superior growth when cultures are initiated in SSM still holds since the cultures
established in SCM had the benefit of being suspended in collagen.
The absence of added ECM factors is also likely to explain the clumping observed for
cultures that had been transferred to either SCM-type medium (Figure 3-5 and Figure
70
3-6). While the formation of the cell clusters could also be an indicator of a less
differentiated state, further studies including an analysis of staining for CD34 by
confocal microscopy would be required to confirm this. In any case, the clumping
presently made it difficult to interpret results when performing routine
immunocytochemistry, since the cell clumps displays a higher degree of background
auto-fluorescence. With these issues in mind, it will be necessary to repeat these
studies using gelatin-coated culture plastic before a definitive conclusion can be made
regarding the relative effects of SSM compared with SCM on the phenotype of
cultured limbal stromal cells. Nevertheless, based upon the culture initiation studies in
collagen gels, it could be concluded that superior migration and proliferation is
observed in serum-supplemented growth medium.
From the foregoing, it would seem difficult to instruct tissue banks to use collagen gels
when initiating cultures from limbal stroma, however, it remains the method of choice
for our laboratory for a number of practical reasons. During my initial training period,
I experienced significant difficulty in using collagenase as a tool for initiating cultures.
This difficulty was eventually overcome by purchasing a new batch of enzyme from
the manufacturer. Since similar variations in activity between batches of collagenase
could also be experienced by tissue banks, it would seem advisable to use a technique
that avoid this issue. By comparison, the gels prepared from medical grade collagen
consistently supported stromal cell outgrowth. Moreover, the availability of this
reagent as a medical grade product offers the significant advantage of being available
in a formulation that would comply with requirements for material used in clinical
manufacturing.
71
3.5 CONCLUSION
In conclusion, collagen gels are an effective tool for initiating cultures from limbal
stromal tissue. Moreover, optimal outgrowth from tissue fragments is observed when
using standard serum supplemented growth medium. Nevertheless, the use of an
explant technique for initiating cultures does not appear to offer any advantages over
use of collagen-treated tissue since the resulting cultures are indistinguishable from
each other when using a panel designed to test MSC identity. It remains possible,
however, that differences may be observed through the use of SCM-type media (with
ECM coating) that are designed to suppress stromal cell differentiation. Ultimately, it
will be necessary to perform a more detailed study of the effects of limbal stromal cells
cultured under different conditions, using a variety of functional assays including
epithelial growth assays and immunological assays.
72
73
Chapter 4: Research Study Two
EVALUATION OF L-MSC SAFETY AND EFFICACY USING A RABBIT
MODEL OF OCULAR SURFACE TRAUMA
74
75
Statement of contribution
Owing to the scale and ethical considerations associated with this study, input was
necessary from a variety of experienced researchers. The majority of HLE cultures
were established by Ms Rebecca Dawson owing to her expertise in growing human
epithelial cells for clinical use. Likewise, all surgical procedures and clinical
assessments (slit lamp) were conducted by an experienced ophthalmic surgeon (Dr
Fiona Li). Anesthesia was conducted and monitored by either Professor Damien
Harkin or Dr Cora Lau (University of Queensland Biological Resources). Professor
Harkin also assisted with blood collections, clinical photography and histology.
Routine post-operative care of animals was provided by staff from the Herston Medical
Research Centre, with weekly assessments being conducted by Dr Li, Professor Harkin
and myself. Serum CRP measurements were conducted by an experienced clinical
biochemist (Mr Steven Weier).
76
77
4.1 INTRODUCTION
Wound healing within the eye, as for other organs and tissues, is often dependent upon
the activation, proliferation and differentiation of resident progenitor cells. In the case
of the ocular surface, progenitor cells for replenishing the corneal and conjunctival
epithelia are known to be concentrated within the corneal limbus (Schermer et al.
1986) and the conjunctival fornix (Stewart et al. 2015) respectively. This knowledge
has been successfully exploited for the treatment of patients with severe injuries of the
ocular surface. For example, a biopsy of healthy limbal tissue can be safely acquired
and used as a source of corneal epithelial cells for the treatment of limbal stem cell
deficiency (LSCD) (Pellegrini et al. 1997; Sangwan et al. 2012; Schwab 1999; Tsai et
al. 2000; Tsubota et al. 1995). In such cases, successful outcomes have been logically
attributed to the presence of epithelial progenitor cells present within the tissue biopsy.
Nevertheless, limbal biopsies also contain stromal progenitor cells with the potential
to affect tissue repair (Basu et al. 2014).
Limbal stromal progenitor cells have been described as fibroblasts (Ainscough et al.
2011; Dravida et al. 2005; Massie et al. 2014), but can be defined as a type of
mesenchymal stromal cell (MSC) by virtue of their immuno-phenotype and immuno-
regulatory properties (Branch et al. 2012; Bray et al. 2014; Bray et al. 2012b; Garfias
et al. 2012; Polisetty et al. 2008). While studies have demonstrated the ability of limbal
fibroblasts/MSC to encourage limbal epithelial cell growth in vitro (Ainscough et al.
2011; Kureshi et al. 2015; Massie et al 2014; O'Callaghan et al. 2016), there has also
been interest in developing therapies based upon their application in vivo (Acar et al.
2015; Basu et al. 2014; Eslani et al. 2017; Holan et al. 2015; Syed-Picard et al. 2016).
These studies have been encouraged by multiple reports of clinical efficacy for MSC
derived from more traditional tissue sources including bone marrow and adipose tissue
(Harkin et al. 2015a).
Prior reports of limbal MSC activity in vivo, are currently limited to a handful of
studies in rodents and rabbits (Table 4-1). Significantly, three out of the five studies
have examined the effects of L-MSC on wounds caused by methods typically used to
induce LSCD. Despite significant variations in wound models and methods of
administration, a consistent pattern of improved stromal healing has been observed as
78
indicated by increased corneal transparency and reductions in edema, and/or corneal
neovascularization. The effects of L-MSC on the ocular surface, however, are less
clear with only two studies having examined the effects of L-MSC on re-
epithelialization and the resulting epithelial phenotype only having been examined in
one case. Moreover, the effects of L-MSC have yet to be examined in conjunction with
a cultured limbal epithelial cell transplant which, given the severity of wounds being
studied, is surprising.
In order to develop a better understanding of the potential role for L-MSC in repairing
the ocular surface, I have had the opportunity as part of a team to investigate the effects
of allogeneic rabbit L-MSC (RLMSC) when applied alone or in conjunction with
human limbal epithelial (HLE) cells cultivated on human amniotic membrane (HAM).
As in previous studies, the majority of epithelial tissue has been removed from both
the cornea and limbus immediately prior to treatment. A mechanical method of
epithelial debridement is used in order to create a more defined wound than that caused
by chemicals. Re-epithelialization of the cornea, enabled through either the implanted
HLE and/or any retained rabbit epithelial cells (including the adjacent conjunctiva), is
monitored weekly for up to 12 weeks by slit lamp, with the resulting epithelial
phenotype being examined using a variety of histological techniques. In particular, the
relative presence of keratins 3 and 13 were examined as an indicator of corneal
epithelial cells (Schermer et al. 1986) and conjunctival epithelial cells (Ramirez-
Miranda et al. 2011), respectively. Moreover, the fate of applied HLE cells has been
examined by immunohistochemistry. These results demonstrate that while L-MSC
consistently encourage re-epithelialization of the ocular surface, the effect of these
cells on corneal neovascularization varies dramatically according to whether or not
these stromal progenitor cells have been pre-conditioned by growth in the presence of
corneal epithelial cells.
79
Table 4-1 Prior studies of corneal tissue response to L-MSC when applied in vivo.
Study Species Wound model
Treatment Key Outcomes
L-MSC
Donor
Host Agent Area Age Formulation Route Epi. Trans. Edema Fibrosis CNV
Basu et
al., 2014
Hum. M Mechanical
Algerbrush II
Central
cornea
including
basement
membrane
<1 h Suspended in
fibrin glue.
Topical NE
NE
Acar et
al., 2015
Rat Rat Chemical
Alkali burn
Cornea &
limbus
24 h Suspended in
medium.
- Topical
- Conj. inj.
- i.p. inj.
Ph: NE
NE NE
Holan et
al., 2015
Rab. Rab. Chemical
Alkali burn
Cornea &
limbus
<1 h Attached to
fibrous
scaffold
prepared
from PLA.
Topical Ph: K3+
NE
Eslani et
al., 2017
Hum.
M
M Mechanical
Algerbrush II
Cornea &
limbus
<1 h Suspended in
fibrin glue.
Topical NE NE NE NE
Syed-
Picard et
al., 2018
Hum. M Mechanical
27G needle
Stromal
<1 h Cultivated
sheet.
Stromal
implant
NE 1-wk:
5-wk:
Norm.
NE NE No
Abbreviations: Species, Hum. = Human, M = mouse, Rab. = rabbit. Treatment, Conj. Inj = sub-conjunctival injection, i.p. = intraperitoneal
injection. Key Outcomes, Epi. = epithelialization, Trans = transparency, CNV = corneal neovascularization, NE = not examined, Ph = cell
phenotype, K3 = keratin 3, Norm. = normal., = significant increase, = significant decrease, No = not present.
81
4.2 MATERIALS AND METHODS
Animal research ethics
All procedures involving rabbits were conducted in accordance with the ‘Animal Care
and Protection Act’ (Queensland State Government, Australia, 2001), ‘Australian
Code for the Care and Use of Animals for Scientific Purposes’ (8th Edition, 2013) and
the ‘ARVO Statement for Use of Animals in Ophthalmic and Vision Research’. The
project was conducted with the approval of the University Animal Ethics Committee
at the Queensland University of Technology (UAEC approval number 1200000575).
Human research ethics
Studies involving the use of human corneal tissue acquired from cadaveric donors were
conducted with donor/next-of-kin consent and Human Research Ethics Committee
(HREC) approval received from the Metro South Hospital and Health Service (HREC
approval number: HREC/07/QPAH/048) and the Queensland University of
Technology (HREC approval number: 0800000807).
Isolation and cultivation of rabbit L-MSC (RLMSC)
For the purpose of establishing a master stock of RLMSC, cadaveric corneal tissue
was used. The tissue was obtained within 30 minutes post-mortem, from a male rabbit
used in a non-ophthalmic surgical training workshop (the use of male cells was
strategically done to support subsequent tracking of cell fate using fluorescence in situ
hybridization (FISH) for the rabbit Y chromosome by others within the research team.
This data is unavailable at time of submitting this thesis). While the formation of the
cell clusters could be an indicator of a less differentiated state, it seems more likely
that this occurred due to lack of cell attachment factors such as serum or gelatine. In
any case, to confirm whether or not the cells cultured in SCM are less differentiated a
confocal microscopy of the cell clumps stained with CD 34 could be performed in
future. While the formation of the cell clusters could be an indicator of a less
differentiated state, it seems more likely that this occurred due to lack of cell
attachment factors such as serum or gelatine. In any case, to confirm whether or not
the cells cultured in SCM are less differentiated a confocal microscopy of the cell
82
clumps stained with CD34 could be performed in future. Following excision, the
anterior eye segment was washed twice in Hanks’ balanced salt solution (HBSS),
incubated in 2.5 mg/mL Dispase II (Gibco Cat. No. 17105-041) for 1.5 h at 37 C and
scraped with a scalpel blade to remove the epithelial and endothelial tissue layers. A 2
mm diameter trephine blade was then used to obtain several punch biopsies of limbal
stroma. The biopsies were attached to the bottom of tissue culture dishes using 30 µL
of type I collagen gel (1 mg/mL) and subsequently submerged in stromal cell growth
medium (SSM as in chapter 3) consisting of DMEM with high glucose (Life
Technologies Cat. No. 10313-021), 10% (v/v) fetal bovine serum (FBS), 2 mM L-
glutamine (Life Technologies Cat. No. 25030-081) and 1% penicillin/streptomycin
solution (Life Technologies Cat. No. 15140-122). The resulting primary culture was
subsequently harvested by rinsing with Versene (Gibco Cat. No. 15040066) followed
by incubation for 5-10 minutes in TrypLE select enzyme (Gibco Cat. No. 12563011).
Subsequent culture expansion was conducted in SSM to passage 2 before resuspension
in 90% FBS/10% DMSO and storage at 2 x 106/mL in liquid nitrogen.
Isolation and cultivation of human limbal epithelial (HLE) cells
Samples of cadaveric human eye tissue were typically supplied in the form of surgical
off-cuts suspended in Optisol corneal storage medium at 4 C and processed within 10
days post-mortem. Each sample was washed 3 times for 5 min in PBS, cut into quarters
and trimmed using a scalpel and digested for 1 h in 2.5 mg/mL Dispase II dissolved in
DMEM medium. Epithelial cells were subsequently harvested from the limbal margin
of the cornea by scraping and aspirating with a pipette tip. The harvested epithelial
cells were subsequently washed and re-suspended in epithelial cell growth medium
(with centrifugation at 300 g for 5 min) before being seeded into a 25 cm2 culture flask
pre-seeded with 106 growth-arrested (using 2 x 25 Gy) murine 3t3 cells (ATCC;
CCL92). The epithelial culture medium consisted of DMEM (Life Technologies Cat.
No. 10313-021) combined in a 3:1 ratio with Hams F12 (Life Technologies 11765-
062) and supplemented with 10% FBS (HyClone, SH 30084.03), 2 mM L-glutamine,
10 ng/mL recombinant human EGF (Invitrogen PHG0311), 5.6 µg/mL isoproterenol
(Sigma Cat. No. I6504), 180 µg/mL adenine hydrochloride hydrate (Sigma Cat. No.
A9795), 5 µg/mL transferrin (Sigma Cat. No. T1147), 1% non-essential amino acids
(Life Technologies Cat. No. 11140-050), 1.36 ng/mL tri-iodo-L-thyronine sodium salt
83
(T3) (Sigma-Aldrich Cat. No. T6397), 1 µg/mL insulin (Sigma Cat. No. I6634), 0.4
µg/mL hydrocortisone (Sigma Cat. No. H4001) and 1% penicillin/streptomycin
solution (Life Technologies Cat. No. 15140-122). Cultures were harvested using
Versene and TrypLE as described above for stromal cells and typically seeded onto
human amniotic membrane after being passaged twice.
Establishment of cultures on human amniotic membrane (HAM)
Human amniotic membrane (HAM) was supplied attached to nitrocellulose backing
paper and frozen in 50% glycerol/50% balanced salt solution by the New Zealand
National Eye Bank (Auckland, New Zealand). With the exception of 1 piece (refer to
Table 4-2), all remaining pieces of HAM were procured from the same donor. Prior to
seeding of cells, each piece of HAM was thawed, washed 3 times for 5-min in Hanks’
balanced salt solution and mounted within a custom-made cell culture chamber
(Ludowici chamber) (Harkin et al. 2017). Once securely mounted within the chamber,
the majority of the nitrocellulose backing paper was carefully peeled away using
watchmaker forceps to facilitate visualization of HAM structure and the subsequently
established cultures using phase contrast microscopy. Prior to seeding of cells, the
upper HAM surface (the epithelium side) was treated with Versene followed by 0.05%
trypsin/1 mM EDTA (5-7 min at 37 C) in an effort to loosen any remaining amniotic
epithelial cells. After adding 1 mL of epithelial growth medium, the majority of
amniotic cells were removed by gentle trituration across the membrane surface using
a 1 mL pipette. If necessary, the process was repeated until the majority of epithelial
cells (approximately greater than 75%) had been removed (as assessed by phase
contrast microscopy). Human limbal epithelial (HLE) cells were seeded onto the upper
HAM surface at a density of 105/cm2. Rabbit limbal mesenchymal stromal cells
(RLMSC) were applied to the lower membrane surface at a density of 0.5 x 105/cm2.
In the case of co-cultures, the RLMSC were seeded approximately 48 h prior to
addition of the HLE cells. All cultures were prepared in duplicate and were maintained
for approximately 10-12 days in epithelial culture medium prior to use.
Sourcing and general care of rabbits
Female New Zealand White rabbits (2.5-3.0 kg) were sourced from either a
commercial rabbit breeding facility or from the University of Queensland Biological
84
Resources rabbit breeding facility. Routine health checks were performed prior to
commencing studies and a radio-frequency identification microchip (MyChip, Provet
Pty Ltd Australia) was implanted subcutaneously into the scruff of the neck. The
rabbits were initially housed in floor pens in groups of up to 4, but post-operatively
were maintained in individual rabbit cages (Tecniplast Australia Pty Ltd, Australia).
Straw bedding, shredded paper and environmental enrichment (cardboard boxes and
plastic toys) were provided. The food supply consisted of a commercial, laboratory-
grade, high-fiber, and low-starch, pelleted rabbit diet (Specialty Feeds, Western
Australia) supplemented with fresh fruit and vegetables. Food and water were supplied
ad libitum and levels checked daily.
Monitoring of serum C-reactive protein levels
Samples of whole blood were obtained from each rabbit immediately prior to
wounding (day 0) and on days 1, 3, 7 and 84 (12 weeks) following
wounding/treatment. Blood was obtained via 24-gauge cannula (BD Insyte, Cat. No.
381212) inserted into a lateral ear vein. A cream containing 25 mg/g lignocaine and
25 mg/g prilocaine (Emla; Astra Zeneca) was applied topically to lateral ear veins
approximately 1 h prior to bleeding to anesthetize the area. During each collection,
rabbits were firmly wrapped in a blanket, with eyes shielded, and placed on a warming
mat. Approximately 2-3 mL of blood was collected directly into an SST II Advance
blood collection tube with lid removed (BD Vacutainer, Cat. No. 367956) and allowed
to clot for approximately 30 minutes at room temperature. The resulting serum was
retrieved following centrifugation and stored at -80 C until testing. Levels of CRP in
each serum sample were subsequently determined using a commercial ELISA kit,
according to manufacturer’s instructions (ICL Inc., Cat. No. E-15CRP).
Anesthesia
Rabbits were pre-medicated with 50 µg/kg buprenorphine (Temgesic® 300µg/mL,
Jurox Pty Ltd, Australia) subcutaneously, approximately 20 minutes prior to general
anesthesia. Anesthetic induction was performed using an injectable combination of
15 mL/kg ketamine (Ilium Ketamil® 100mg/mL, Troy Laboratories Australia Pty Ltd)
and 0.25 mg/kg medetomidine (Domitor® 1 mg/mL, Pfizer Animal Health, NSW
Australia). A 24G intravenous cannula (Optiva® 24G IV Catheter Radiopaque, Medex
85
Medical Ltd, Great Britain) was introduced into the marginal ear vein to allow for
intravenous surgical maintenance fluid therapy. General anesthesia was maintained
via a size 1 mask (Vetquip, Castle Hill, NSW Australia) under 1-2% isoflurane
(Attane, Bayer Australia) through an Isoflurane Tec 3 vaporizer fitted to a MQV1100
Anesthetic Machine (Mediquip Pty Ltd, Australia). Topical anesthesia in the form of
2-3 drops of 0.5% proparacaine (Alcaine® 0.5% eye drops, Alcon Laboratories Pty
Ltd Australia) was also employed.
Wounding of rabbits
An experienced ophthalmic surgeon (Dr Fiona Li) performed all the procedures with
the aid of a speculum and surgical microscope. All surgical instruments and swabs
were either purchased sterile or sterilized prior to surgery by autoclave. The handle of
the AlgerBrush II debridement tool was decontaminated by spraying with 70%
ethanol. An area measuring approximately 50-100 cm2 around each eye was
decontaminated using a sterile surgical swab doused in 10% w/v povidone-iodine
(Betadine® Antiseptic Solution, Mundipharma B.V., Netherlands). A sterile field was
created using a nylon surgical drape containing a circular hole measuring
approximately 25 cm2. Each rabbit’s eye was proptosed prior to surgery by placing a
sterile glove with cross-shaped slit and applying light downwards pressure with aid of
a scalpel blade handle. Epithelial debridement was preceded by a 360° conjunctival
peritomy, approximately 1.5-mm beyond the limbus, with dissection towards the
limbus. Debridement then commenced initially with 360° superficial limbal
keratectomy using an AlgerBrush II fitted with 2.5-mm round-ended, diamond-dusted
burr (Rumex International Cat. No. 16-051-2.5B). The same device was subsequently
applied in a circular manner with light pressure across the corneal surface. Fluorescein
staining under cobalt illumination was performed in order to ensure that the majority
of epithelium had been removed.
Application of cultures to ocular surface
After removal from transport medium, each culture was positioned so that the central
area came into contact with the ocular surface. Care was taken to ensure that the side
containing epithelial cells (whenever present) was facing upwards and the side
containing stromal cells (whenever present) was facing downwards in order to
86
replicate the normal anatomical distribution for each cell type. The periphery of each
culture was then slowly and gradually released from the culture chamber by carefully
cutting with iris scissors. Further trimming of the HAM was performed until a
peripheral flap of approximately 3-5 mm was overlying the sclera. Eight
discontinuous, superficial and regularly spaced sutures (10.0 Vicryl) were then
inserted to secure the HAM to the sclera. The peripheral edge of the HAM including
sutures were subsequently covered with a circular conjunctival flap using eight
additional sutures. Transport medium was applied drop-wise to the surface of the
HAM every 5-10 minutes in an effort to reduce potential drying of the culture. Finally,
the rabbit’s nictitating membrane was secured to the lower temporal side eyelid using
a 4.0 nylon suture and a central tarsorrhaphy performed.
Table 4-2 Summary of study design
Cohort Rabbit Treatment (Tx)
HLE Donor HAM Donor RLMSC Donor
No Tx
A - - -
B - - -
C - - -
HLE
HAM
D HD1 p1 HD7 -
E HD2 p2 HD7 -
F HD3 p2 HD7 -
HLE
HAM
RLMSC
G HD4 p2 HD7 RD1 p3
H HD5 p2 HD7 RD1 p4
I HD6 p2 HD7 RD1 p4
HAM
RLMSC
J - HD7 RD1 p4
K - HD7 RD1 p4
L - HD8 RD1 p4
Post-operative care
During recovery from anesthesia each animal was supplied with oxygen and fitted with
a 10-cm diameter soft cat recovery Elizabethan collar to reduce further trauma to the
injured eye by incidental cleaning or brushing against objects. All animals were awake
within 1-h post-surgery and responsive to food and water within 2 to 3 h. Post-
operative pain management was performed using a multi-modal analgesic protocol.
This consisted of alternating morning and afternoon subcutaneous injections of 0.05
87
mg/kg meloxicam (Metacam® 5 mg/mL, Boehringer Ingelheim Vetmedica, Inc.) and
50 mg/kg buprenorphine (Temgesic® 300 mg/mL, Jurox Pty Ltd, Australia) until the
morning of the 5th day. In addition, a combination eye ointment preparation consisting
of 5 mg/g neomycin sulfate, 5000 IU/g polymixin B sulfate, 2.5 mg/g prednisolone
and 50 mg/g sulfacetamide sodium (Amacin® Eye and Ear Ointment, Jurox Pty Ltd,
Australia) was supplied twice daily throughout the entire post-operative period, as well
as after each clinical examination. The suture securing the nictitating membrane to the
lower temporal eyelid was removed after 7 days. After 12 weeks, each animal was
euthanized by slow intravenous injection with 325 mg/kg of sodium pentobarbital.
Clinical assessments
Clinical assessments were performed weekly for up 12 weeks. Two to three drops of
0.5% proparacaine (Alcaine® 0.5% eye drops, Alcon Laboratories Pty Ltd Australia)
were inserted into each eye approximately 5 min prior to each assessment. A speculum
was sometimes inserted to provide a clearer view of the corneal margins and a sterile
cotton-bud used to retract the nictitating membrane temporarily if required. Each
examination commenced by taking a photograph to record the presentation of the
ocular surface. A Canon EOS 6D digital SLR camera equipped with a Canon macro
lens (EF 100 mm 1:2.8 L IS USM)) and Canon Macro Ring Lite MR-14EX II flash
was used (Camera settings: ISO 400, 1/100, f13). A slit lamp examination was
subsequently performed to assess changes in corneal structure including stromal
edema and epithelial integrity using a Keeler Classic portable slit lamp. A fluorescein
paper strip soaked in saline was inserted beneath the upper eyelid and held with light
pressure from a gloved hand for approximately 1 min prior to examination under the
slit lamp's cobalt lamp. A yellow lens filter and blue flash filter were applied during
photography (Camera setting: ISO 1600, 1/60, f8.0) under cobalt lamp illumination.
Analysis of clinical images
The approximate size of epithelial defects in each wounded eye at a given time point
was determined using ImageJ (Version 1.48v; National Institutes of Health, USA)
image analysis software. Briefly, the relative size of each defect was measured by
tracing images of fluorescein-stained eyes with a computer mouse (using the freehand
measure function) and then expressing these values as a percentage of the total corneal
88
area (as defined by an ellipse outlining the approximate corneal margin). The time
course of changes in percentage defect for each animal was plotted using Prism 6
(Graph Pad) and analyzed using a two-way ANOVA followed by Tukey’s post-hoc
test. Relative differences in the degree of corneal neovascularization were determined
on standard clinical images (normal illumination) obtained at 12 weeks. ImageJ was
again used to obtain approximate measurements of corneal area displaying blood
vessels as a percentage of total corneal area. Individual values for each animal were
plotted using Prism 6 and analyzed using by Kruskal-Wallis test followed by Dunn’s
multiple comparisons test.
General histology
Prior to retrieving eyes from deceased animals, the orientation of tissue was labelled
by applying a marker pen to the superior sclera/conjunctiva. Excised tissue in the form
of whole enucleated eyes was typically fixed overnight in neutral buffered formalin
followed by transfer to 70% ethanol. The anterior cap from each eye was subsequently
removed with aid of iris scissors and processed into paraffin. Prior to embedding, three
cuts were made along the superior-inferior axis resulting in four strips of corneal tissue.
The first cut was made directly through the centre of each cornea resulting in two hemi-
corneas of approximately equal size. Each tissue piece was subsequently cut again
resulting in a “longer central” and “shorter peripheral” segment of cornea. During
embedding the opposing cut surfaces were placed face-down within the mold. After
subsequent facing, each section removed off the block therefore contained four tissue
sections; two spanning the entire cornea and limbus from along the central superior-
inferior axis and two similarly orientated sections from the mid-temporal and mid-
nasal peripheral cornea. Approximately a dozen sections were mounted and examined
for each block. Three whole sections acquired from regular spaced intervals were
initially examined for general morphology after staining with Ehrlich’s hematoxylin
and eosin and adjacent sections were stained for goblet cells using periodic acid,
Schiff’s reagent and Mayer’s hematoxylin.
Immunostaining
Immunostaining was subsequently performed using primary antibodies selective for
keratin 3, keratin 13 or human nuclear antigen (HNA). An immuno-peroxidase method
89
was used for detection of keratins in tissue sections and an immunofluorescence
method was used for detection of HNA in either cell cultures (optimization of antibody
selection) and tissue sections.
Epitope retrieval (ER) was performed prior to immune-detection of keratins by
immersing deparaffinized slides in CINtec® Histology Kit (Roche, Cat. No. 9511)
epitope retrieval solution (1-mm EDTA/10 mM Tris buffer, pH 9.0) for 10 minutes at
85 ºC. The Coplin jar containing slides in ER solution was then placed for a further 20
minutes at room temperature during which time the temperature dropped to
approximately 55 ºC. After rinsing in staining buffer (10 mM Tris buffered saline with
0.025% Triton X-100) the slides were transferred to a staining rack placed within a
humidified container. Endogenous peroxidases were inactivated by treatment with
0.3% hydrogen peroxide for 10 minutes. After further rinsing in buffer the slides were
incubated for 1h at room temperature in buffer containing primary antibodies to either
keratin 3 (a 1:300 dilution of mouse monoclonal AE5 obtained from Millipore Pty Ltd,
Cat. No. CBL218) or keratin 13 (a 1:300 dilution of mouse monoclonal AE8, Abcam
Pty Ltd, Cat. No. ab16112). Binding of primary antibodies was subsequently detected
using a horseradish peroxidase/polymer-conjugated goat-anti-mouse detection system
(a component of CINtec® Histology Kit, Roche, Cat. No. 9511). Negative controls
were performed by excluding the primary antibody incubation step. Positive controls
consisted of non-wounded tissue sections stained with antibodies to either keratin 3 or
keratin 13. The chromogen used was diaminobenzidine (DAB). Nuclear
counterstaining was performed by treatment for 5 minutes with Gill’s hematoxylin
solution (United Biosciences, Carindale, Queensland, Cat. No. G1-1L), followed by
rinsing in Scott’s tap water substitute (United Biosciences, Carindale, Queensland,
Cat. No. SCOT-1L). After dehydration through graded alcohols and clearing in xylene,
the slides were mounted in plastic mounting medium and imaged using an Olympus
BX41 microscope equipped with a 20x/0.8 NA UPlanApo oil-immersion lens and
Nikon Ri1 digital camera. Images were acquired using NIS Elements version 4 and all
post-acquisition image modifications was undertaken using Adobe Photoshop CS5
(Version 12.0, Adobe Systems Inc.). Image modifications consisted of (in order) initial
re-sizing (to 8 x 6 cm at 300 dpi), cropping, montage creation, threshold optimization
using levels function, and labelling.
90
Prior to investigating the fate of implanted human cells, a preliminary study was
conducted using three commercial antibodies with potential selective specificity for
human versus rabbit cells; the anti-mitochondrial antibody 113-1 (Merck Millipore
Cat. No. MAB1273), the anti-HNA clone 235-1 (Merck Millipore Cat. No. MAB1281)
and the anti-HNA clone 3E1.3 (Merck Millipore Cat. No. MAB4383). These
antibodies were screened using early passage (p3) cultures of corneal-limbal epithelial
cells established from human and rabbit limbal tissue in 24-well culture plates. Each
culture was fixed for 10-min in neutral buffered formalin, permeabilised by treatment
with 0.3% Triton/PBS (2 x 5 min) and blocked by incubation for 30-min at room
temperature in 2% normal goat serum/PBS. Each primary antibody was subsequently
applied at a 1:100 dilution in PBS containing 1% NGS and incubated overnight at 4
C. After four washed in PBS, the secondary antibody (Alexa 488-conjugated goat-
anti-mouse IgG) was applied at 1:100 dilution in PBS containing 1% NGS and
incubated in the dark for 1 h at room temperature. Given the success of this staining
protocol, the same protocol was subsequently used to stain deparaffinised tissue
sections. Additional controls consisted of sections obtained from normal and human
tissue. Imaging of immunofluorescence was conducted using a Nikon TE-2000
equipped with a CoolSNAP ES cooled CCD camera and NIS Elements (F package).
Modification of images was conducted as described above for bright field images of
immune-peroxidase stained tissue sections.
91
4.3 RESULTS
Construction and analysis of treatment cultures
Nine pairs of duplicate cultures were established on HAM throughout this study. All
HAM samples were acquired from the same human donor with the exception of the
last pair of cultures seeded with RLMSC alone (due to a supply issue). Each pair of
duplicate cultures containing HLE (with or without RLMSC) were prepared from a
unique human tissue donor and seeded onto HAM at either passage 1 or 2. All cultures
containing RLMSC were established using cells from the same donor rabbit and same
passage number (p4) (With the exception of rabbit G which was treated with RD1 P3).
In the case of cultures prepared from HLE alone on HAM, one of the duplicate cultures
developed a hole during the cultivation period and thus was unable for further analysis.
For all other sets, however, a duplicate culture was available for confirmation of
culture integrity by routine histology. Examination of sections after staining with
hematoxylin and eosin revealed a disorganized and stratified epithelium of
approximately 5 layers for all HAM samples seeded with HLE (Figure 4-1). In
contrast, RLMSC cultures were noticeably more stratified when grown in the presence
of HLE.
Baseline response to wounding (epithelial debridement without
suturing)
Examination of eyes by fluorescein staining immediately after wounding suggested
that the majority of epithelial cells had been removed from the cornea and limbus
(Figure 4-2). Gradual re-epithelialization was observed over the subsequent 12 weeks,
but all animals failed to heal completely over this time period. Corneal
neovascularization was evident within 3-4 weeks with approximately 3-4 quadrants
becoming involved by 12 weeks. Some corneal opacity was retained at 12 weeks and
the ocular surface remained rough. Histology at 12 weeks demonstrated a mixed
phenotype of K3 and K13 positive epithelial cells in two animals with the third
displaying evidence of mature conjunctival epithelium (K13 with PAS+ goblet cells).
Serum C-reactive protein (CRP) levels generally peaked within 24 h of wounding and
declined to baseline levels by 72 h, before rising again to moderately elevated levels
by 12 weeks.
92
Figure 4-1 Representative images of histological sections (H&E stained)
Obtained from spare cultures of human limbal epithelial (HLE) cells and/or rabbit
limbal mesenchymal stromal cells (RLMSC) grown on denuded human amniotic
membrane (HAM).
93
Figure 4-2 Demonstration of method used to mechanically debride the corneal
epithelium.
(A) Photograph displaying application of the Algerbrush II rotating burr tool to the
corneal-limbus of a proptosed rabbit eye (in vivo). (B) Gross evaluation of epithelial
debridement via fluorescein staining immediately following application of the
Algerbrush II instrument. The relatively even distribution of green fluorescence
observed under cobalt lamp illumination (Keeler handheld slit lamp) suggests that the
majority of epithelial cells have been removed. A similar level of debridement was
achieved for all animals used in this study.
94
Effect of treatment on serum C-reactive protein levels
The majority of treated animals (8 out of 9) generally displayed a similar profile of
changes in serum CRP levels to the non-treated cohort, with an initial peak being
observed at 24 h after wounding, followed by a decline within 3-7 days, and
moderately elevated levels being retained at 12 weeks. Animals that received co-
cultures of RLMSC and HLE on HAM, however, displayed a significantly greater
increase in serum CRP levels at 24h compared to all other cohorts (p<0.0001 by two-
way ANOVA followed by Tukey’s multiple comparisons test). In addition, animals
that received HLE alone on HAM displayed significantly lower serum CRP levels after
1 week when compared to non-treated animals (p<0.005) (Figure 4-3).
Effect of treatment on re-epithelialization
All treated cohorts displayed a gradual increase in re-epithelialization over the 12
weeks of observation as monitored by fluorescein staining under cobalt illumination
(Figure 4-4). The fastest rates of re-epithelialization were observed in cohorts treated
with RLMSC, with the greatest overall healing being observed in animals receiving
both HLE and RLMSC on HAM. Analysis of data by two-way ANOVA followed by
Tukey’s multiple comparison test confirmed significant differences between each
treatment cohort.
Effect of treatment of neovascularization
All animals developed varying degrees of corneal neovascularization over the 12
weeks of observation (Figure 4-5 and Figure 4-7). The greatest level of
neovascularization was observed in animals receiving cultures of RLMSC alone on
HAM, which was found to be significantly higher than for animals wounded without
treatment (p<0.05 by Kruskal-Wallis test followed by Dunn’s multiple comparisons
test).
95
Figure 4-3 Comparison of serum CRP levels between cohorts of treated rabbits.
Line graphs indicate the mean +/- SEM of values (mg/L) for each cohort of 3 rabbits.
Single asterisk indicates significant (p < 0.005) difference to animals wounded without
treatment (No Tx). Analysis of data using a two-way ANOVA (double asterisk)
indicates a significant difference (p < 0.0001) between animals treated with co-cultures
(HLE-HAM-RLMSC) compared to all other cohorts.
96
Figure 4-4 Time course of re-epithelialization as measured under cobalt lamp
illumination after fluorescein staining.
Line graphs indicate the mean +/- SEM of values for gradual increase in re-
epithelialization for each cohort of 3 rabbits over the 12 weeks. Analysis of data using
a two-way ANOVA reveals significant differences between all treatment groups (p
0.05 or less for each comparison).
97
Figure 4-5 Gross appearance of rabbit eyes at 12 weeks under bright light
illumination.
Labels ‘A’ through ‘L’ indicate identity of each rabbit as summarized in Table 4-2.
Treatment groups as described above consisted of controls (No Tx), human limbal
epithelial cells grown on human amniotic membrane (HLE-HAM), HLE and rabbit
mesenchymal stromal cells grown on HAM (HLE-HAM-RLMSC), or HAM with
RLMSC alone (HAM-RLMSC).
98
Figure 4-6 Appearance of rabbit eyes at 12 weeks under cobalt lamp illumination
after fluorescein staining
Labels ‘A’ through ‘L’ indicate identity of each rabbit as summarized in Table 4-2.
Treatment groups as described above consisted of controls (No Tx), human limbal
epithelial cells grown on human amniotic membrane (HLE-HAM), HLE and rabbit
mesenchymal stromal cells grown on HAM (HLE-HAM-RLMSC), or HAM with
RLMSC alone (HAM-RLMSC).
99
Figure 4-7 Comparison of corneal neovascularization observed between animals
after 12 weeks
Line and error bars indicate the mean +/- SEM for each treatment cohort. The % CNV
for each animal (A through L) was calculated based upon estimated measures of
corneal area with blood vessels using ImageJ. Asterisk indicates a significant
difference in CNV for animals receiving HAM with RLMSC cultured on the
underlying surface, compared to animals that had been wounded without subsequent
treatment (p < 0.05).
100
Histological analyses
Examination of control (non-wounded) tissue demonstrated the expected normal
structure for corneal and conjunctival tissue respectively (Figure 4-8). In brief, the
cornea displayed a stratified epithelium that was devoid of goblet cells (GC) and
stromal blood vessels (BV). Moreover, the corneal epithelium displayed positive
immunostaining for K3 and was negative for K13. Conversely, the conjunctival
epithelium displayed positive staining for K13, but was negative for K3. The
uppermost layer of the limbal epithelium, however, stained positively for both markers
(data not shown).
Examination of H&E stained sections of wounded eyes confirmed the development of
corneal vascularization with the largest and best developed vessels being observed in
animals that received treatment with HAM seeded with RLMSC alone (Figure 4-9).
The presence of goblet cells (as confirmed by treatment with periodic acid followed
by Schiff reagent) was also most consistently observed in animals treated with RLMSC
alone. Strongest immunostaining for K13 (a marker for superior limbal and
conjunctival epithelial cells) was also observed in this cohort (Figure 4-10). In contrast,
the clearest example of immunostaining for K3 (a marker for superior limbal and
corneal epithelial cells) was observed in two animals treated with both HLE and
RLMSC on HAM. Conversely, no staining for K3 was observed for animals treated
with HLE alone on AM. The epithelia that had partially regenerated in animals
wounded without treatment expressed both K3 and K13 or K13 alone in the presence
of PAS-stained goblet cells (Figure 4-10 and Figure 4-9).
Given the remarkable staining for K3 in two animals receiving HLE in conjunction
with RLMSC, the potential presence of human cells was investigated by
immunostaining using an antibody to human nuclear antigen. Fixed cultures of human
and rabbit corneal epithelial cells were initially screened by immunofluorescence to
confirm the specificity of this antibody (Figure 4-11 and Figure 4-12). Subsequent
staining of tissue sections by immunofluorescence indicated that the epithelium
regenerated in animals treated with HLE combined with RLMSC was not of human
origin (Figure 4-13 and Figure 4-14).
101
Figure 4-8 Normal structure and profile of keratin expression for rabbit cornea
and conjunctiva.
Goblet cells (GC) within conjunctiva are highlighted by staining using periodic acid
Schiff reagent (PAS) method. Antibodies to K3 and K13 are specific to corneal
epithelium and conjunctival epithelium respectively. Control (Con.) displays results
when neither primary antibody is used.
102
Figure 4-9 Basic histology of rabbit eyes at 12 weeks as revealed by staining of
sections with hematoxylin and eosin (H&E) and periodic acid-Schiff stain (PAS).
Labels ‘A’ through ‘L’ indicate identity of each rabbit as summarized in Table 4-2
Treatment groups as described above consisted of controls (No Tx), human limbal
epithelial cells grown on human amniotic membrane (HLE-HAM), HLE and rabbit
mesenchymal stromal cells grown on HAM (HLE-HAM-RLMSC), or HAM with
RLMSC alone (HAM-RLMSC). Arrows highlight presence of goblet cells (GC) which
are especially evident after staining with the PAS.
103
Figure 4-10 Immuno-histochemical staining of rabbit eyes at 12 weeks to
demonstrate typical presence of corneal (K3) and conjunctival (K13) epithelium.
Labels ‘A’ through ‘L’ indicate identity of each rabbit as summarized in Table 4-2.
Treatment groups as described above consisted of controls (No Tx), human limbal
epithelial cells grown on human amniotic membrane (HLE-HAM), HLE and rabbit
mesenchymal stromal cells grown on HAM (HLE-HAM-RLMSC), or HAM with
RLMSC alone (HAM-RLMSC).
104
Table 4-3 Summary of clinical data for wounded and treated animals
Cohort Tx Final Assess Histology
HLE
Donor
HAM
Donor
RLMS
C
Donor
Rabbit CRP
(mg/L)
%
Defect
%
CNV
PAS K3 K
1
3
No Tx - - - A 35.8 35.6 15.5 - + +
- - - B 24.1 32.9 24.9 + - +
- - - C 37.8 2.25 14.5 - + +
HLE
HAM
HD1
p1 HD7 -
D 20.1 26.9 37.7 + - +
HD2
p2 HD7 -
E 18.6 26.7 42.8 + - +
HD3
p2 HD7 -
F 48.8 38.8 35.2 - - +
HLE
HAM
RLMS
C
HD4
p2 HD7
RD1
p3
G 38.2 0.0 26.2 - + +
HD5
p2 HD7
RD1
p4
H 27.1 10.25 24.4 - + +
HD6
p2 HD7
RD1
p4
I 10.8 26.8 37.7 + - +
HAM
RLMS
C
-
HD7
RD1
p4
J 17.6 0.0 95.9 + - +
-
HD7
RD1
p4
K 13.8
3.9 89.6 + - +
-
HD8
RD1
p4
L 36.5 8.5 69.0 + - +
105
Figure 4-11 Screening of potential human-specific antibodies by immunostaining
of human corneal-limbal epithelial cells (HLE).
Third passage cultures of HLE (p3) displayed reactivity towards all three monoclonal
antibodies (mab) tested; an anti-mitochondrial antibodies (mab) tested; an anti-
mitochondrial antibody (113-1) and two antibodies to human nuclear antigen (HNA;
235-1 and 3E1.3). The 3E1.3 antibody to HNA bound more selectively to nucleoli than
mab 235-1.
106
Figure 4-12 Screening of potential human-specific antibodies by immunostaining
of rabbit corneal-limbal epithelial cells (RLE).
Third passage cultures of RLE (p3) displayed reactivity towards anti-mitochondrial
mab 113-1 and anti-HNA mab 3E1.3, but not towards the anti-HNA mab 235-1 (bright
areas corresponded to non-specific debris).
107
Figure 4-13 Representative images of immuno-histochemical staining of human
cadaveric eyes to demonstrate the specific reactivity of human-specific antibody
(anti-HNA mab 235-1) on human corneal/limbal tissue sections.
Human corneal/limbal sections displayed reactivity towards anti-HNA mab 235-1 only
when both primary and secondary antibody were applied.
108
Figure 4-14 Representative images of immuno-histochemical staining of rabbit
eyes with anti-HNA mab 235-1 at 12 weeks to trace the presence/absence of
grafted human cultured epithelial cells.
Neither wounded and non-treated nor wounded treated rabbit corneal/limbal sections
showed reactivity towards anti-HNA mab 235-1.
109
4.4 DISCUSSION
Autologous transplants of corneal-limbal tissue have been widely demonstrated as an
effective treatment for ocular surface disease (Sangwan et al. 2012; Vazirani et al.
2016). While the efficacy of these transplants is logically related to the presence of
epithelial progenitor cells, the potential contributions of other cell types present within
the transplanted tissue remains unclear. In particular, the presence of mesenchymal
stromal cells (MSC) in cultures established from limbal tissue biopsies (L-MSC) in
vitro suggests that these cells might be exploited to improve clinical outcomes. In
particular, L-MSC have been shown to encourage the growth of corneal epithelial cells
derived from limbal tissue biopsies (Bray et al. 2014). Cultured L-MSC may therefore
be used to encourage re-epithelialization in vivo by facilitating the implantation and
growth of transplanted epithelial cells, while also encouraging the growth of any
healthy epithelial cells remaining within the tissue. Moreover, the immunosuppressive
properties of L-MSC might be exploited to improve the efficacy of epithelial cells
derived from donor tissue, as would be expected to be necessary in cases where total
LSCD in suspected in both eyes.
In the present study, the ocular surface of rabbits has been debrided using a rotating
burr tool (Algerbrush II). Prior analyses of wounds created using this method indicate
that this is generally an efficient way to remove epithelial cells from both the cornea
and limbus with minimal damage to the underlying stroma (Li et al. 2016). While the
level of epithelial debridement can be checked by fluorescein staining (Figure 4-2), it
remains possible that small islands of epithelial cells are retained. The epithelial
wounds created in this study should therefore be regarded as extensive, but by no
means should be considered as a model of total LSCD. The goal of the study was
therefore to investigate the impact of cultured L-MSC when applied to extensive,
freshly-created epithelial wounds, rather than chronic wounds with an established
LSCD phenotype. Interestingly, the outcomes observed were found to be highly
dependent upon whether the L-MSC had previously been cultivated in the presence of
epithelial cells.
Overall, the results from this study illustrate that when allogeneic cultures of rabbit L-
MSC are applied to the ocular surface in the absence of cultivated epithelial cells, the
110
rate of re-epithelialization is significantly improved, but the epithelium originates from
the peripheral conjunctival tissue (Table 4-3). Moreover, the enhanced
conjunctivalization is associated with a significant increase in corneal
neovascularization. In contrast, when the L-MSC are supplied in the presence of
epithelial cells cultivated from the limbus, there is less conjunctivalization of the
ocular surface and an associated decrease in corneal neovascularization. Moreover, the
marked improvement in epithelial phenotype observed for two animals (G & H)
suggests that the rabbit L-MSC may have encouraged the implantation and retention
of human epithelial cells. The failure to detect human cells in these animals, however,
indicates that a different mechanism was involved.
Given that the regenerated epithelium is derived from rabbit cells (either from
remnants of corneal-limbal epithelium or surrounding conjunctiva tissue) the
improved outcomes for animals G & H might have arisen through two processes.
Firstly, the factors secreted by the co-cultures may provide a more potent trigger for
stimulating remnants of corneal epithelium. Alternatively, pre-cultivation of L-MSC
in the presence of epithelial cells may have conditioned these cells to enable an
enhanced healing response when subsequently applied to the ocular surface. While
these two theories are not mutually exclusive, the enhanced stratification of stromal
cultures observed in the presence of epithelial cells suggests an effect of HLE on L-
MSC biology. It is also possible that the L-MSC may have in turn altered the biology
of the applied HLE. Notably, application of co-cultures was associated with a
significantly higher level of serum C-reactive protein (CRP) at 24 h after wounding.
While CRP is a rather non-specific marker of acute inflammation, its production by
the liver is signalled by interleukin-6 (IL-6), which has itself been shown to be
upregulated in co-cultures of HLE and limbal fibroblasts compared to HLE cultured
under control conditions (Notara et al. 2010).
In drawing comparisons with the work of others, the present findings agree with those
of Acar et al. (2015) and Holan et al. (2015) in so far as both studies reported increased
re-epithelialization of acute wounds when treated with L-MSC. Nevertheless, the
majority of prior studies have demonstrated a decrease in CNV when L-MSC are
applied to the ocular surface (Table 4-1). Differences in methodology including
wounding method, treatment method and animal model may well account for this. In
111
particular, the use of HAM as a carrier for L-MSC in conjunction with HLE is novel
to the present study.
Given prior reports of cultivated HLE being applied to the ocular surface of rabbits, it
was somewhat surprising to observe that animals treated with HLE alone on HAM
displayed little to no difference compared with wounded/non-treated animals apart
from a significant reduction in serum CRP levels after 7 days. This effect may well
have been due to the amniotic membrane alone, but further studies would be required
to confirm this. Given the relatively small size of the cohorts used (n=3) is was equally
surprising that significant differences were observed between the cohorts examined.
In view of the time required to complete each cohort (3 months) it could be argued that
the results may have been influenced by improvements to the team’s techniques over
time. Nevertheless, since the worst outcomes were achieved with the last cohort
examined (RL-MSC alone on HAM) the differences seem unlikely to be affected by
some kind of learning curve.
In conclusion, the results from this chapter provide further evidence of a potential
clinical application for L-MSC. In particular, the greatest benefits were observed when
the stromal cells were applied in conjunction with a culture of human corneal-limbal
epithelial cells. Nevertheless, since no human epithelial cells could be detected
following treatment, it appears that the effects of L-MSC might be mediated in part by
pre-conditioning of the stromal cells in culture by the epithelial cells prior to their
application to the ocular surface. Given reports of stromal cell capacity to express K3
(Sidney et al. 2015b) it remains possible that the epithelial-conditioned rabbit L-MSC
may have contributed to the improved K3 expression observed in two animals. Since
the L-MSC used were acquired from a male rabbit, there is potential to examine the
tissue samples further by in situ hybridization (ISH) using a probe for the rabbit Y
chromosome. The required kit has recently been purchased and so it remains possible
that an answer to this question may be obtained prior to submitting this study for
publication. Nevertheless, care must clearly be taken when applying L-MSC to the
ocular surface since marked corneal vascularization appears to be induced if no
preconditioning is used.
112
113
Chapter 5: Research Study Three
OPTIMIZATION OF A FIBROIN-BASED SUBSTRATE FOR DELIVERING
L-MSC AND EPITHELAL CELLS TO THE OCULAR SURFACE
114
115
Statement of contribution
The fibroin-coated plates and fibroin membranes used in this study were manufactured
by Dr Shuko Suzuki (Queensland Eye Institute). A few cultures of human limbal
epithelial cells (HLE) were established by Ms Rebecca Dawson. Professor Damien
Harkin assisted with studies involving confocal fluorescence microscopy. The clinical
component of this study was conducted with necessary input from others as described
for Chapter 4.
116
117
5.1 INTRODUCTION
As demonstrated in the preceding chapter, amniotic membrane (AM) provides an
effective mechanism for applying cultured cells to the ocular surface. Not surprisingly,
therefore, the majority of clinical studies have utilized AM as a tool for delivering
cultured limbal epithelial for the treatment of ocular surface disease. Nevertheless, a
number of significant problems can be encountered when using AM. First and
foremost, AM can be difficult to obtain. For example, while the mechanics of donor
AM procurement and banking are fairly straight forward and well understood, the
regulatory costs associated with maintaining a licensed service can be prohibitive. This
is currently the situation in Australia and as such AM must be purchased from overseas
at significant cost. Once supplied, however, a number of technical problems can be
encountered. For example, the physical properties of AM can vary significantly both
within and between donor batches, including the ease at which the amniotic epithelial
cells can be removed prior to seeding of cells. Moreover, the fibrous nature of the thick
basement membrane can often impair visualization of cells cultivated upon the AM
(Maharajan et al. 2007). As a result, a number of alternatives to AM have been
explored including fibrin gels (Pellegrini et al. 1997) and contact lenses (Di Girolamo
et al. 2009). The purpose of the present chapter, however, is to further explore the
potential of membranes prepared from the silk protein fibroin.
Silk fibroin is responsible for the mechanical properties of silk fibers and can be readily
extracted from a number of different species including cocoons produced by the
domesticated silkworm Bombyx mori (Chirila et al. 2008). The techniques employed
to isolate fibroin from cocoon silk result in a mixture of protein fragments produced
via hydrolysis of the native fibroin protein. A range of structures can be readily
fabricated from these aqueous solutions of fibroin including porous sponges,
electrospun fibers, coated-films and freestanding membranes (Harkin et al. 2011). In
particular, fibroin membranes have been widely explored as a potential substrate for
ocular cell types owing to their relatively high stability compared with other materials
such as collagen, and their high transparency.
Chirila et al. (2008) first proposed use of fibroin membranes for reconstructing the
ocular surface by demonstrating the attachment and growth of corneal-limbal epithelial
118
cells. Subsequent studies have confirmed the suitability of fibroin membranes as a
substrate for corneal stromal cells (Bray et al. 2012a; Gil et al. 2010a; Gil et al. 2010b;
Lawrence et al. 2009) and corneal endothelial cells (Madden et al. 2011). Nevertheless,
in the absence of naturally occurring cell-adhesion motifs, attachment of cells to
fibroin derived from Bombyx mori is facilitated through the use of serum-
supplemented growth medium and often requires further optimization through coating
to purified extracellular matrix components including collagens or vitronectin
(Madden et al. 2011; Shadforth et al. 2012). As an alternative to this approach, a
number of groups have explored altering the adhesive properties of fibroin via
incorporation of the classical cell-binding motif abbreviated to RGD (arginine-
glycine-asparagine). Studies by Gil et al. (2010a, 2010b) for example, demonstrated
faster growth of corneal stromal cells when using fibroin membranes that had been
chemically bound to exogenous RGD containing peptide. Building upon this idea,
others have attempted to exploit the presence of naturally occurring RGD-containing
sequences that are present within certain non-domesticated species of silkworm
including Antheraea pernyi. Nevertheless, freestanding membranes are technically
more difficult to produce from APSF (Hogerheyde et al. 2014) and no benefits were
observed when blends of BMSF and APSF are used as a substrate for corneal-limbal
epithelial cells (Bray et al. 2013). Prior attempts at exploiting the use of RGD-
containing peptides have therefore produced mixed results. At alternative option,
however, has recently emerged in the form of Bombyx mori silkworms that have been
genetically engineered to produce fibroin containing the RGD-binding motif (Kambe,
et al. 2010a, 2010b).
Fibroin is normally produced from three gene products resulting in a heavy chain, a
light chain and chaperone protein (fibrohexamerin), in the ratio of 6:6:1 (Kambe et al.
2010b). In order to produce fibroin containing the RGD sequence, Kambe et al.
(2010a, 2010b), genetically engineered Bombyx mori silkworms to produce fibroin
light chains fused directly to two sequential RGDS sequences (L-RGDS x 2, or LRF).
While materials produced from the resulting mix of fibroin protein supported similar
levels of chondrocyte attachment (up to 24 h), significant increases in the expression
of cell-adhesion molecules and extracellular matrix proteins were observed. Thus, the
recombinant RGD silk fibroin (RGD-SF) produced by Kambe et al. (2010b) appears
119
to be more beneficial for long-term cultures rather than for initial cell attachment per
se.
In the course of the preceding chapter, where co-cultures of epithelial cells and stromal
cells were established on AM, a number of additional test cultures were established in
parallel on membranes prepared from conventional fibroin. These preliminary studies
demonstrated that while conventional fibroin can be used to support cell attachment,
the pattern of stromal cell growth was typically patchy and confluent sheets of
epithelial cells tended to detach after approximately two weeks in cultures. At this
same time, the conventional fibroin membranes were of insufficient strength to support
suturing. Moreover, evidence from other studies performed within the group (Suzuki
et al. 2015) indicated that optimal communication between epithelial cells and stromal
cells (seeded on opposing membrane surfaces) would require use of membranes that
had been rendered more permeable by incorporation of low molecular weight
poly(ethylene oxide) (300 Da) as a porogen during the casting phase.
Thus, the present study was designed to evaluate the properties of an advanced
formulation of fibroin membrane that exploited the improved permeability of PEO-
Fibroin membranes, while also examining the potential benefits of RGD-Fibroin
obtained from Kambe and co-workers. In addition, an attempt was made to strengthen
the fibroin membranes by use of horseradish peroxidase as a cross-linking agent. To
begin, the potential benefits of the RGD-fibroin alone were examined for effects on
the adhesion and long-term morphology of L-MSC and HLE in culture. The long-term
growth of L-MSC and HLE co-cultures was subsequently examined using membranes
prepared form RGD-fibroin, PEO and HRP. Finally, the feasibility of applying such
cultures to the ocular surface was examined in the same rabbit model as presented in
the preceding chapter.
120
5.2 MATERIALS AND METHODS
Materials and consumables for manufacturing of fibroin membranes
Standard fibroin was sourced in the form of Bombyx mori silk cocoons supplied by
Tajima Shoji C. Ltd. (Yokohama, Japan), all cut in half and with the pupae removed.
Recombinant RGD silk fibroin (RGD-fibroin) as originally described by Kambe et al.
(2010) was provided in the form of degummed fibres by the National Agriculture and
Food Research Organization (NARO, Tsukuba, Japan). Horseradish peroxidase (HRP)
Type VI (supplied as lyophilized powder; Lot#SLBL4932V), hydrogen peroxide
(30%) and poly(ethylene glycol) (PEG, MW 300 Da) were all supplied by Sigma-
Aldrich (St Louis, MO, USA). Minisart®-GF pre-filters (0.7 μm) and Minisart® filters
(0.2 μm) were supplied by Sartorius Stedim Biotech (Göttingen, Germany). The
dialysis cassettes Slide-A-Lyzer® (MWCO 3.5 kDa) were supplied by Thermo
Scientific (Rockford, IL, USA). Water of high purity (Milli-Q) was used.
Degumming of standard cocoon silk
Approximately 2.5 g of cut cocoon pieces (approximately 1 cm2 in size) were placed
in 1 L boiling solution of sodium carbonate (0.02 M) for 1 h to remove sericin (i.e.
degumming step). The degummed fibres were subsequently washed in 1 L of water at
60 °C for 20 min, three times in succession with squeezing to remove the excess liquid
between each wash. This was followed by drying the fibres in a fume hood for at least
12 h.
Generation of fibroin solutions
Dried silk fibres containing either standard fibroin or RGD-fibroin were dissolved in
9.3 M lithium bromide at 60 °C for 4 h, and the solution was transferred into a dialysis
cassette (MWCO 3.5 kDa) and dialyzed for 3 days with 6 water exchanges. Finally,
each fibroin solution was filtered through two connected syringe filters (porosities of
0.7 and 0.2 µm) and stored at 4 °C. The resulting solutions, with a concentration of
about 3% w/v fibroin (as determined by gravimetric analysis), were diluted to required
concentration by adding water.
121
Preparation of standard SF membranes and coating of tissue culture
plastic
The freestanding, standard SF membranes were prepared by casting the 1.78% w/v
fibroin solution in a custom-made casting table where the supporting glass plate was
pre-coated with a polyolefin polymer (Topas®) film. The blade height was set in order
to generate an approximate dry thickness of 6 μm for the resulting fibroin membranes.
Alternatively, the fibroin solution was poured into a 24-well tissue culture plate to
create fibroin coatings (using 2% w/v fibroin solution, 256 µL/well). After drying at
room temperature, the membranes and fibroin coated plates were water-annealed in a
vacuum chamber at −80 kPa for 6 h at room temperature in the presence of a container
filled with water. The freestanding membranes were then peeled off from the
supporting Topas® film. Subsequently, the fibroin-coated plates and membranes were
sterilized by applying 75% ethanol for 1 h followed by one rinse with PBS and two
rinses with serum free culture medium.
Preparation of PEG-treated, HRP-crosslinked RGD-fibroin
membranes
PEG was introduced as a porogen to increase membrane permeability as has been
established in multiple published studies (Higa et al. 2011; Suzuki et al. 2015). Since
this treatment makes the membranes weaker (Suzuki et al. 2015) HRP induced self-
crosslinking of fibroin, through catalysed tyrosine groups forming dityrosine linkages,
was performed to increase the strength of membranes. Stock solutions of HRP (150
U/mL) and H2O2 (0.3%) were first prepared. Then PEG was slowly blended into the
1.78% RGD-fibroin solution at a PEG/fibroin ratio of 2:1 (by weight), followed by
addition of equal volumes of each HRP and H2O2 solution. The final concentration of
HRP in the mixture was 1.1 U for 1 mg of protein (RGD-fibroin). The mixture was
then cast into a Topas® pre-coated petri dish covered with a lid and stored at 40 °C for
2 h to form a gel. The volume was set to obtain 1.81 mg fibroin/cm2. The resulting gel
was dried in a fan-driven oven at room temperature for at least 12 h. After drying, the
membranes were soaked in water (1L/dish) for 3 days with two water exchanges per
day to remove the PEG. The membranes were subsequently treated with 3% H2O2
solution for 10 min at room temperature to quench any residual HRP activity, followed
122
by rinsing with water three times. The membranes were peeled off from the underlying
Topas® film in a dish and stored in water at 4 °C prior to use.
Isolation and cultivation of cells
Techniques used for the initial isolation and cultivation of L-MSC and human limbal
epithelial cells (HLE) were as described in Sections 4.2.3 and 4.2.4.
Cell attachment and growth assay
Cells were applied to culture surfaces at a density of either 15000 cells/cm² (for L-
MSC) or 25,000 cells/cm² (limbal epithelial cells) and incubated for 90-min prior to
analysis. The subsequent morphology and numbers of attached cells was examined
under both serum-free as well as serum-supplemented growth conditions. Test culture
surfaces included tissue culture plastic (TCP), TCP coated with standard fibroin
(Fibroin) and TCP-coated with recombinant fibroin containing RGD (RGD-fibroin).
After 90-min, cultures were rinsed briefly 3 times with HBSS to remove any non-
attached cells. Photographs of each culture were taken using an Olympus TS-100
inverted phase contrast microscope equipped with a 10x objective lens. Following
photography, the buffer was removed, and each plate stored frozen at -80 °C until
analysis of dsDNA content in each well using a PicoGreen assay kit. For longer term
tests of cell growth on each surface (for between 6 to 10 days), the cultures were
maintained in their regular serum-supplemented growth medium until analyses as
described above.
Establishment of L-MSC/HLE co-cultures on fibroin membranes
Membranes fabricated from either standard fibroin, RGD-fibroin or PEG/HRP/RGD-
fibroin were mounted in custom designed cell culture chambers (Ludowici chamber;
(Harkin et al. 2017) as used for mounting of fibroin membrane (Section 4.2.5). L-MSC
were seeded at a density of 0.5 x 105 cells/cm² on one side of each membrane and
cultivated for 2-3 days before the addition of human limbal epithelial cells (HLE).
After 2-3 days, the chamber was inverted and HLE cells seeded onto the opposite side
of each membrane at a density of 105 cells/cm². Cultures were subsequently maintained
in epithelial growth medium (Section 4.2.4). The resulting cultures were grown for
123
approximately 12 days prior to either analysis of structure or application to the ocular
surface (feasibility study).
Analysis of co-culture 3D structure
For analysis of structure, the co-cultures were fixed in 10% buffered formalin and
subsequently stored in PBS. A 4 mm trephine punch was subsequently used to sample
cultures for further analysis. Samples were treated with 0.1% Triton X-100 in PBS for
30 min before staining with rhodamine phalloidin (1:100 dilution in PBS) and 1 µM
Hoechst 33342 overnight at 4 °C. The stained samples were subsequently washed
extensively in fresh PBS (3 to four washes of 30 minutes each) and mounted for
confocal microscopy under glass coverslip in a 1:1 mixture of PBS and glycerol. A
Nikon A1 confocal system equipped with a Plan Apo 20x/0.75 N.A. lens was used to
construct stacks of approximately 45-50 XY images, with a Z step size of between 0.5-
1 µm and the pinhole set to approximately 1 airy unit.
In vivo testing of co-cultures on RGD-Fibroin membranes
The feasibility of applying fibroin membranes to the ocular surface of rabbits was
initially trialed using normal eyes of deceased rabbits acquired in the course of studies
reported in Chapter 4. Given the difficulty encountered with applying a flat membrane
to a domed ocular surface, a “petal-wrap” strategy was adapted from that used in the
confectionary food industry (Demaine et al. 2009). Incorporating knowledge of the
approximate radius of curvature for the adult rabbit cornea (Bozkir et al. 1997), four
triangular-shape cut-outs were excised from each membrane culture immediately prior
to application Figure 5-10. The modified structure was then applied to the denuded
ocular surface (epithelial side facing up) as performed previously for human amniotic
membrane (HAM; Section 4.2.10) with the following modifications. The eight initial
sutures were placed as two near each corner of the trimmed triangular-shaped flaps of
fibroin membrane. The secured membrane edges were then, as previously, secured
using a circular conjunctival flap with 8 additional sutures. As an added precaution, a
contact lens with high oxygen permeability (Air Optix Aqua Night & Day/lotrafilcon
A; Dk/t = 175) was applied to the surface of the implanted culture, prior to performing
the partial tarsorrhaphy) to further protect the epithelial cells. This strategy was
subsequently tested in 1 live animal (with animal research ethics approval) using the
124
same surgical and monitoring protocols as described in Chapter 4. The treated animal
was monitored for up to 7 weeks.
125
5.3 RESULTS
Comparison of L-MSC attachment to fibroin versus recombinant
RGD-fibroin
The relative attachment of L-MSC to standard fibroin, compared to a recombinant
formulation of fibroin incorporating the RGD cell-binding motif, was assessed both
visually and by quantification of dsDNA as a proxy for cell numbers. For the purpose
of these binding studies, each formulation of fibroin was applied as a coated film on
tissue culture plastic (TCP). Non-coated TCP was used as a positive control and all
tests were performed over 90 minutes in the presence of serum (10% FBS) as well as
in serum-fee culture medium.
Visual assessment by phase contrast microscopy revealed that while similar numbers
of L-MSC appeared to be present under all conditions tested, the majority of cells that
had attached to regular fibroin under serum-free conditions were noticeably less spread
than those in any other wells (Figure 5-1). In the presence of serum, however, all
surfaces appeared to perform equally well at promoting cell spreading. Quantification
of the dsDNA content within each well (for five L-MSC donors, with each surface
being tested in quadruplicate for each donor) confirmed that cell attachment over the
initial 90 minutes of culture was not influenced by the choice of fibroin used (Figure
5-2).
Given some of the trends observed over 90 minutes, it was decided to perform a longer-
term study by cultivating L-MSC on each formulation of fibroin. Non-coated TCP was
once again used as a control. Visual assessment of cultures after 6 days in complete
culture medium (containing 10% FBS) demonstrated that while the L-MSC had grown
to confluency on TCP, cultures established at the same seeding density on either
formulation of fibroin were approximately only 20-40% confluent (Figure 5-3). By 10
days, the confluency of L-MSC on RGD-fibroin more closely resembled that observed
on TCP, with far fewer clusters observed. In contrast, however, L-MSC grown for 10
days on regular fibroin remained sub-confluent and the aggregated clumps of cells
became more apparent. Quantification of the dsDNA content within each well (for 4
L-MSC donors, with each surface being tested in quadruplicate for each donor)
indicated significantly more cells growing on RGD-fibroin compared to regular fibroin
126
after 6 days, but no significant difference between culture surfaces was observed by
day 10, since the cells cultured on the regular fibroin had sufficient time to achieve the
same level of confluency as on RGD-fibroin by 10 days (Figure 5-4).
127
Figure 5-1 Visual comparison of L-MSC attachment to tissue culture plastic
(TCP), TCP coated with Bombyx mori silk fibroin (Fibroin), or TCP coated with
recombinant fibroin incorporating the RGD-cell binding motif (RGD-fibroin).
Cells were seeded at a density of 15,000 cells/cm² in 24-well culture plates and
incubated for 90 minutes in the absence (No Serum) or presence of foetal bovine serum
(10% v/v) in culture medium. Phase contrast images display the typical appearance of
cells after 90 minutes incubation followed by three rinses in phosphate buffered saline.
128
Figure 5-2 Quantification of L-MSC attachment to tissue culture plastic (TCP)
coated with Bombyx mori silk fibroin (Fibroin) or recombinant fibroin
incorporating the RGD-cell binding motif (RGD-fibroin).
Cells were seeded at a density of 15,000 cells/cm² in 24-well cultures and incubated
for 90 minutes in the absence (No Serum) or presence of foetal bovine serum (10%
v/v) in culture medium. Each well was briefly rinsed three times with phosphate
buffered saline before analysis of dsDNA content using the PicoGreen assay. Bars
represent the mean +/- SEM for L-MSC cultures derived from five unique donors.
Each test condition for a given donor was tested in quadruplicate. No significant
difference between culture conditions was detected by a non-parametric one-way
ANOVA (Friedman test with Dunn’s multiple comparisons test; n = 5).
129
Figure 5-3 Representative images of comparison of L-MSC cultures established
in the presence of serum (10% v/v FBS) on tissue culture plastic (TCP), TCP
coated with Bombyx mori silk fibroin (Fibroin), or TCP coated with recombinant
fibroin incorporating the RGD-cell binding motif (RGD-fibroin).
Cells were seeded at a density of 15,000 cells/cm² in 24-well culture plates and
photographed after 6 and 10 days respectively. White arrows indicate the presence of
cell clumps that became more apparent in cultures established on TCP coated with
regular fibroin.
130
Figure 5-4 Quantification of L-MSC growth in cultures established in the
presence of serum (10% v/v FBS) on tissue culture plastic (TCP) coated with
Bombyx mori silk fibroin (Fibroin) or recombinant fibroin incorporating the
RGD-cell binding motif (RGD-fibroin).
Cells were seeded at a density of 15,000 cells/cm² in 24-well culture plates and
analysed for dsDNA content using the PicoGreen assay after either 6 days (Part A) or
10 days (Part B). Bars in each graph represent the mean +/- SEM for data obtained
using L-MSC cultures established from 4 unique donors. Each test condition for a
given donor was tested in quadruplicate. Asterisk in Part A indicates a significant
difference (p < 0.05; n = 4) between cultures established on the two formulations of
fibroin at 6 days using a non-parametric one-way ANOVA (Friedman test with Dunn’s
multiple comparisons test).
131
Comparison of HLE cell attachment to fibroin versus recombinant
RGD-fibroin
Visual assessment of HLE cell attachment indicated that these cells were also
noticeably less spread after 90 minutes incubation on standard fibroin (Figure 5-5).
Unlike for L-MSC, however, the presence of serum seemed to have no effect on the
spreading of HLE attached to regular fibroin. Quantification of the dsDNA content
within each well (for four HLE donors, with each surface being tested in quadruplicate
for each donor) indicated similar numbers of cells attached to each surface, with the
exception of fibroin under serum free conditions that supported the attachment of
significantly less cells than for the positive control (TCP in the presence of serum
(Figure 5-6).
Optimization of HLE/L-MSC co-cultures on recombinant RGD-fibroin
In an effort to facilitate the communication of HLE and L-MSC grown together on
opposing surfaces of a freestanding RGD-fibroin membrane, a comparison was made
with co-cultures established on RGD-membranes prepared using a low molecular
weight poly(ethylene) oxide (300 Da PEO; as a porogen) and horseradish peroxidase
(HRP; as a cross-linking agent).
Feasibility of engrafting HLE/L-MSC to the ocular surface using
fibroin
RGD-fibroin membranes prepared using PEO/HRP were approximately 3-fold thicker
and more rigid than those prepared without the porogen and cross-linking agent. These
modified membranes were easier to handle when wet but became noticeably more
brittle when dried than standard RGD-fibroin membranes.
Cultures prepared on RGD-fibroin/PEO/HRP membranes retained a more flat
architecture following their release from the mounting ring. The flatter architecture
facilitated visualization of the attached cells by phase contrast microscopy (Figure 5-8)
In the presence of L-MSC grown on the opposing membrane surface, the HLE grown
on the upper surface adopted a more compact morphology. Similar areas were
occasionally observed when co-cultures were prepared on standard RGD-fibroin
132
membranes, but the tendency of these less rigid membranes to fold when released from
their mounting ring, made observations more difficult.
The influence of L-MSC on HLE grown on RGD-fibroin/PEO/HRP membranes was
subsequently examined by confocal fluorescence microscopy (Figure 5-9).
Visualization of staining with rhodamine phalloidin demonstrated that HLE cultures
grown in the presence of an underlying layer of L-MSC were 2 to 4-fold more stratified
than those grown in the absence of the stromal cells.
Encouraged by this result, an attempt was made to apply co-cultures of cells to the
ocular surface of rabbits. Initial studies were conducted using a rabbit cadaver (Figure
5-10). Use of a “petal wrap” technique improved conformity of the fibroin membrane
to the ocular surface and sutures could be successfully inserted without tearing the
membrane. Based upon these observations, a further test of feasibility was conducted
using one live rabbit (Figure 5-11). As in Chapter 4, the corneal epithelium, including
limbus, was debrided using the Algerbrush II tool. During implantation of the co-
culture, one of the petal segments broke away, but the remaining three segments were
successfully sutured to the ocular surface. For added protection, a contact lens was
inserted prior to suturing the eyelids together. One week later the cornea appeared clear
with no signs of infection or discharge. Some clumped material was evident beneath
the contact lens which was subsequently identified back in the laboratory to be the
fibroin membrane. The lens itself had become shifted slightly towards the superior
fornix but was still in place. The lens was therefore re-centred before returning the
animal to housing. Subsequent examinations on days 14 and 21 post-surgery revealed
remarkable signs of recovery with nearly complete re-epithelialisation by day 21.
Nevertheless, a prominent epithelial defect returned by 28 days. The epithelial defect
progressively became worse over the next three weeks until covering over half the
corneal surface. Corneal vascularisation was evident by day 28 and also became more
pronounced over the next 3 weeks. A decision was therefore made to euthanize the
animal and collect the tissue for histology after 7 weeks. Histology confirmed the
presence of the large epithelial defect with the surrounding tissue being of conjunctival
phenotype as indicated by the presence of goblet cells (data not shown).
133
Figure 5-5 Visual comparison of HLE cell attachment to tissue culture plastic
(TCP), TCP coated with Bombyx mori silk fibroin (Fibroin), or TCP coated with
recombinant fibroin incorporating the RGD-cell binding motif (RGD-fibroin).
Cells were seeded at a density of 25,000 cells/cm² in 24-well culture plates and
incubated for 90 minutes in the absence (No Serum) or presence of foetal bovine serum
(10% v/v) in culture medium. Phase contrast images display the typical appearance of
cells after 90 minutes incubation followed by three rinses in phosphate buffered saline.
134
Figure 5-6 Quantification of HLE attachment to tissue culture plastic (TCP), TCP
coated with Bombyx mori silk fibroin (Fibroin), or TCP coated with recombinant
fibroin incorporating the RGD-cell binding motif (RGD-fibroin).
Cells were seeded at a density of 25,000 cells/cm² in 24-well cultures and incubated
for 90 minutes in the absence (No Serum) or presence of foetal bovine serum (10%
v/v) in culture medium. Each well was briefly rinsed three times with phosphate
buffered saline before analysis of dsDNA content using the PicoGreen assay. Bars
represent the mean +/- SEM for HLE cultures derived from four unique donors. Each
test condition for a given donor was tested in quadruplicate. Asterisk indicates a
significant difference (p < 0.05) between Fibroin-coated TCP (Fibroin) compared to
TCP with serum-supplemented culture medium (TCP + FBS) using a non-parametric
one-way ANOVA (Friedman test with Dunn’s multiple comparisons test; n = 4).
135
Figure 5-7 Confocal fluorescence micrographs demonstrating the basic
morphology of HLE cells grown on free-standing membranes (~10 cm²) prepared
from standard fibroin (Fibroin), compared to membranes prepared from
recombinant fibroin incorporating the RGD-cell binding motif (RGD-fibroin).
Cultures were maintained for approximately 2 weeks prior to fixation in neutral
buffered formalin. A 6 mm diameter circle was subsequently excised from the middle
of each membrane and stained with rhodamine phalloidin and Hoechst nuclear stain.
Each set of images (left to right) display representative confocal sections extracted
from a Z-stack of approximately 45 XY sections generated with 1 µm step size (1 airy
unit). The corresponding YZ view is displayed at bottom of each set of images.
136
Figure 5-8 Phase contrast microscopy images of HLE cultures established on
membranes prepared from RGD fibroin solution, compared to membranes
prepared from RGD fibroin solution treated with a porogen (low molecular
weight poly(ethylene) oxide or PEO) and a cross-linking agent (horseradish
peroxidase or HRP) prior to casting.
Cultures were maintained for 12 days in the presence and absence of L-MSC grown
on the opposing membrane surface. All membranes were removed from culture
chambers prior to photography. Some folding is evident within images of cultures
grown on the RGD-Fibroin membranes (white arrows).
137
Figure 5-9 Confocal fluorescence microscopy images demonstrating the relative
stratification of HLE cultures grown on RGD Fibroin/PEO/HRP membranes, in
the absence and presence of L-MSC (cultivated on the opposing membrane
surface; not shown).
Parallel cultures prepared from the same donor’s HLE cells were maintained for 12
days prior to fixation and staining with rhodamine phalloidin (red/pink) and Hoechst
nuclear stain (blue). Fibroin membrane (labelled “F” with arrows) is visible within the
Z profile views (at bottom and right hand side of each image) due to a combination of
auto-fluorescence (blue channel) and some residual excess phalloidin (red channel).
138
Figure 5-10 Fitting a two-dimensional fibroin membrane to the domed surface of
a rabbit cornea.
(A) Schematic for the rabbit eye where “r” represents the radius of curvature for the
cornea. (B) “Petal-wrap” design for mapping a 2D material to cover the surface of a
sphere (adapted from (Demaine et al. 2009). This design is printed to desired
magnification where “r” is matched to the approximate radius of curvature for a given
rabbit’s cornea (approximately 7.3 mm based on findings of (Bozkir et al. 1997).A
mounted fibroin membrane is subsequently placed over the printed image and four
triangles excised using a sterile ophthalmic blade with aid of a dissecting microscope.
(C) Typical appearance of fibroin membrane following removal of the four triangular
segments. (D) Example of preliminary attempt to apply a fibroin membrane to the
ocular surface of a deceased rabbit. Two “petals” of membrane (indicated by arrows)
have been trimmed to approximately 3 mm beyond the limbus. The gaps between each
petal are larger than usual owing to a more crude attempt at cutting without the aid of
a dissecting microscope.
139
Figure 5-11 Post surgery examination of rabbit eye treated with a co-culture of
human limbal epithelial cells and rabbit mesenchymal stromal cells grown on
RGD-Fibroin/PEG/HRP
Examinations were done at day 14, 21, 28 and 49 under bright light illumination (left
side) and fluorescein stained under cobalt lamp illumination (right side).
140
5.4 DISCUSSION
While the suitability of silk fibroin membranes as a substrate for corneal cell growth
has been well established (Bray et al. 2011; Chirila et al. 2008; Madden et al. 2011;
Shadforth et al. 2012), an optimal formulation has yet to be determined for clinical
use. Assuming that it is to be used in a similar manner to amniotic membranes, an
optimal formulation should be strong enough to enable the use of sutures, while also
being sufficiently permeable to support the diffusion of nutrients along with epithelial-
stromal cell communication. To address this issue, the present study has explored the
potential of silkworms that have been genetically modified to secrete fibroin light
chains fused to two copies of the RGD-cell binding motif Kambe et al. (2010b). In
addition, poly(ethylene) glycol (PEG) and horseradish peroxidase (HRP) have been
used as tools for optimizing the permeability and strength of the resulting RGD-fibroin
membranes, respectively (Chirila et al. 2017). While the outcomes from this study are
mixed, the resulting membranes prepared from RGD-Fibroin/PEG/HRP display a
number of positive qualities that make them an attractive candidate for further study.
It was hypothesized that the use of fibroin that had been genetically modified to contain
the RGD-cell binding motif would result in significant improvements in cell
attachment and growth. While no significant improvement was observed in the
attachment of either L-MSC or HLE over 90 minutes, L-MSC grown on RGD-fibroin
membranes achieved a more even confluency over 10 days, compared to those seeded
on standard fibroin membranes. The lack of effect on HLE growth is consistent with
the findings of others, using membranes prepared from fibroin known to naturally
contain RGD peptide (APSF) (Bray et al. 2013). Thus, HLE seem to be quite
unreceptive to any attempt at adding RGD to fibroin membranes. On the other hand,
stromal cells and their derivatives (including chondrocytes), based upon present and
prior data (Gil et al. 2010a; Gil et al. 2010b; Kambe et al. 2010b), are more responsive.
Based upon the time-course of benefits associated with inclusion of RGD (6-10 days),
it seems likely that the effects on L-MSC may be due to enhanced ECM production as
has been reported for chondrocytes (Kambe et al. 2010b) and transformed cultures of
human corneal stromal cells (Gil et al. 2010a). Nevertheless, further studies are
required to confirm this.
141
Poly(ethylene glycol) (PEG) has been extensively used as a tool for increasing the
permeability of fibroin films and membranes (Higa et al. 2011; Suzuki et al. 2015).
During casting and subsequent drying, the PEG molecules coalesce to form
hydrophilic regions that are subsequently removed during washing in water. While the
higher molecular weight forms of this compound (>20,000 Da; poly(ethylene oxide))
can be used to promote pore formation, the lower molecular weight PEG (e.g. 300 Da),
results in increased permeability as measured by the movement of various dyes,
polymers, proteins and gases. Nevertheless, freestanding membranes prepared from
fibroin-PEG are too fragile to support clinical applications such as those aimed for in
this study (Suzuki et al. 2015). To address this issue, the plant derived cross-linking
agent genipin has been used to increase the stability fibroin-PEG membranes (Suzuki
et al. 2015). Horseradish peroxidase, however, was used in the present study since it
is faster and more effective than genipin for crosslinking fibroin (Chirila et al. 2017).
As an added benefit, the combined treatments resulted in freestanding membranes that,
while thicker, were flatter and easier to handle than standard RGD-fibroin membranes.
Moreover, the resulting membranes remained sufficiently permeable to support
increased stratification in response to an underlying layer of L-MSC (Figure 5-9).
As a final test of their clinical suitability, an attempt was made to apply co-cultures of
HLE/L-MSC growing on RGD-Fibroin/PEG/HRP membranes to the ocular surface.
Unlike amniotic membrane, flat sheets of fibroin membrane conform poorly to the
domed architecture of the ocular surface. A petal-wrap design adopted from the
confectionary industry was therefore trialled (Demaine et al. 2009). Based upon
published values for radius of curvature in rabbit corneas, four triangular segments
were removed from the completed culture on fibroin membrane immediately prior to
surgery. While the resulting construct could be sutured and mapped better to the ocular
surface, the subsequent detachment and temporary healing indicates poor transfer of
cells to the ocular surface. At time of preparing this thesis, however, more encouraging
results have been achieved when rabbit limbal epithelial cells grown on PEG-treated
fibroin membranes are placed face down upon the wounded ocular surface of rabbits
(Li et al. 2017). Presumably, the more direct contact between the epithelial cells and
the ocular surface enables more efficient cell transfer in a similar manner to how
cultures grown on contact lenses are used (Di Girolamo et al. 2009). Thus, fibroin
membranes may ultimately be a better replacement for contact lenses than for amniotic
142
membrane. In which case, the more challenging task of replacing amniotic membrane
will remain.
Unlike the consistent pattern of wound healing observed in the in-vivo study in the
fourth chapter of this thesis, in this study a return of a prominent epithelial defect was
observed after such a remarkably faster re-epithelialization (nearly complete healing
by 21 days). One potential explanation would be that wound healing occurred within
the first 2-3 weeks was temporary. This result could be due to an insufficient number
of cells transferred, insufficient time for the cells to securely attach to the ocular
surface. These and other issues will be addressed in the following General Discussion
chapter.
143
Chapter 6: General Discussion
This study has originated from a desire to assist patients who have lost significant
function to their corneal limbus in one or both eyes. More specifically, it is proposed
that improved methods for the treatment of limbal tissue dysfunction will be achieved
through the development of novel cell and tissue-based therapies. Ideally, the goal
would therefore be to provide patients with a bioengineered tissue construct that
perfectly mimics the structure and function of the normal corneal-limbus. Going a step
further, one can imagine being able to provide in the future a treatment tailored to the
needs of an individual patient by using some form of advanced imaging technique to
scan the affected tissue, which could then be combined with 3D printing to create a
customised tissue replacement. Like all new treatments, however, it will take time to
develop an understanding of all the various components required to achieve efficacy.
To this end, multiple studies including clinical trials over the last twenty years have
established the importance of limbal epithelial cells as the source of progenitor cells
for regenerating the corneal epithelium (Nakamura et al. 2015). The potential benefits
of including other cell types, however, remains unclear. The role of this study,
therefore, has been to explore the potential of progenitor cells isolated from the
adjacent connective tissue within the corneal limbus. More specifically, these cells
have been examined when grown under conditions that encourage adoption of a
mesenchymal stromal cells (MSC) phenotype.
In the preceding chapters it has been demonstrated that cultures of human limbal
mesenchymal stromal cells (L-MSC) can be equally well generated from either tissue
explants seeded in collagen gels or from collagenase-digests of limbal tissue. Growth
medium supplemented with serum provided the best method for initiating cultures,
however, the benefits of using serum-free medium or medium supplemented with
serum-substitutes is an issue that requires further investigation. Subsequent testing of
L-MSC in a rabbit corneal injury model indicates that substantial differences in clinical
outcomes may be achieved according to whether the stromal cell cultures have been
pre-grown in either the presence or absence of limbal epithelial cells. A strategy for
co-culturing of L-MSC and limbal epithelial cells was subsequently developed using
144
membranes prepared from silk fibroin as a scaffold. The performance of these co-
cultures was significantly improved through the use of fibroin that had been genetically
altered to contain the cell-adhesion motif RGD. Further refinement of these fibroin
membranes will however most likely be necessary in order to achieve similar levels of
clinical success as is present achieved using donor human amniotic membrane.
Following further reflection on the significance and limitations of experiments
presented in this thesis, there are three key topics deserving of more general discussion;
(1) which clinical conditions are likely to benefit most from the use of cells isolated
from donor limbal stroma, (2) what is the most appropriate phenotype required for
efficacy, and (3) what is likely to be the most appropriate method for administering
these cells for clinical impact?
6.1 POTENTIAL CLINICAL APPLICATIONS OF LIMBAL STROMAL
CELLS.
Unlike limbal epithelial cells, L-MSC are not currently recognised as a cell therapy for
the treatment of corneal disease. The immunological properties of L-MSC, however,
along with their ability to reduce corneal scarring and encourage growth of limbal
epithelial cells, suggest that they have potential as a biological therapy for aiding
corneal repair (Basu et al. 2014; Bray et al. 2014). Studies presented in Chapter 4 of
this thesis (Figure 4-6) support this case since the application of L-MSC to the ocular
surface of rabbits consistently encouraged a faster rate of re-epithelialization (Figure
5-9). Moreover, studies in chapter 5 demonstrate that L-MSCs encourage better
stratification of HLE. The treatment of corneal ulcers and persistent epithelial defects
could therefore well benefit from topical treatment with L-MSC. Looking more
deeply, it is possible that L-MSC could also have applications in the treatment of
stromal disease. Evidence of this has already be found in a rodent model by (Basu et
al. 2014) who, when using a similar wounding technique to that used presently, noted
increased transparency and reduced scarring in the presence of L-MSC suspended in
fibrin gel. L-MSC might also prove useful for the treatment of a range of corneal
dystrophies involving the stroma such as lattice corneal dystrophies. Indeed, since
there is already interest in using MSC derived from non-corneal tissues to replace
145
keratocytes (Zhang et al. 2015), it could be argued that L-MSC represent a more tissue-
appropriate source of cells.
Considering the deepest layer of the cornea, there has been interest in using corneal
stromal cells as a source of corneal endothelial cells (Hatou et al. 2013). There exists
a developmental basis for this idea given that cells of corneal stroma and the corneal
endothelium are both derived from neural crest cells during development (Tuft and
Coster 1990). Attempts have therefore been made to investigate the steps required to
de-differentiate corneal stromal back to a neural crest phenotype, before subsequently
converting them to endothelial cells (Hatou et al. 2013). Finally, it is possible that the
immunological properties of L-MSC could be exploited to aid the retention of corneal
allografts. Such studies are already well advanced for bone marrow-derived MSC in
animal models where it has been shown that the stromal cells extend the survival of
intentionally mismatched corneal transplants via a mechanism that involves secretion
of the regulatory molecule TSG-6 (Lee et al. 2014; Oh et al. 2010). Given the extent
of this progress, however, it seems likely that administration of purified factors such
as TSG-6 will provide a more convenient and potentially cheaper therapy than those
based upon the banking of MSC from corneal-limbus or other tissues.
6.2 LIMBAL STROMAL CELL PHENOTYPE.
In the present study, the technique for growing limbal stromal cells has encouraged the
development of a phenotype associated with mesenchymal stromal cells. Meanwhile,
others have been investigating ways to maintain limbal stromal cells under conditions
that promote their retention of characteristics that more closely resemble the
keratocytes found in normal tissue (Sidney et al. 2015a). The case for either strategy
can be argued equally. Adoption of the L-MSC phenotype has the advantage of
providing a connection with the wider body of MSC literature. Moreover, from a more
practical perspective it builds upon my research group’s existing foundation of work
where data has been acquired using cells validated as MSC (Bray et al. 2012b; 2014).
Nevertheless, there is insufficient data available at this time to know whether either
the MSC/fibroblast or keratocyte strategy is better than the other. For example, while
some positive effects have been presently found for L-MSC applied in vivo, it is
possible that similar or even better results may be achieved if limbal stromal cells were
146
cultivated on HAM using a keratocyte-type growth medium. As mentioned earlier, the
appropriate choice of cells may ultimately depend upon the nature of the condition
being treated. In any case, there is certainly value in continuing to investigate the
mechanisms that control the phenotype of limbal stromal cells in culture.
At the outset of Chapter 3, it was envisaged that the use of a serum-free growth medium
(SCM) such as that used by Sidney et al. (2015a) would result in a shift in stromal cell
phenotype towards that of keratocytes. Given the different substrate used presently for
culture expansion studies (standard rather than gelatine-coated tissue plastic) the
findings of Sidney et al. (2015a) could not be replicated. Nevertheless, recent studies
by the Nottingham-based group (Sidney et al. 2015b) have produced some interesting
follow-up observations. More specifically, this group have addressed the feasibility of
converting cultures of L-MSC back to a more keratocyte phenotype. In brief, these
studies have shown that, with the inclusion of additional growth factor supplements
(retinoic acid, bFGF and TGF-3) to the authors’ original SCM formulation, it is
possible to restore expression of key cellular markers (including CD34) and ECM
molecules associated with the keratocyte phenotype. Nevertheless, this reversion is
only partial since MSC markers were retained (CD73, CD90 and CD105). Potentially,
complete reversion to a keratocyte phenotype might yet be achieved following
implantation and wound resolution.
Returning to the MSC context, there are a number of emerging concepts within this
field that may yet prove relevant to the development of L-MSC-based therapies. To
begin, it is recognised that MSC cultures derived from bone marrow and other tissues
are heterogenous (Phinney 2012). Efficacy may therefore be related to one or more
sub-populations of cells rather than the whole. Additionally, some sub-populations
may actually have detrimental effects, as is anticipated in the case of myofibroblasts.
Concerns around which stromal cell phenotype is best for certain applications will
therefore need to take-into-account that more than one population is present. This
concept is not new to studies of corneal stromal cells since it has already been
discussed in the literature (Sidney et al. 2014) and attempts have been made to isolate
sub-populations by flow cytometry (Funderburgh et al. 2016). The emphasis, however,
has so far been based on the isolation of less differentiated cells which may or may not
be appropriate depending upon the intended application. If similar approaches are to
147
be used for particular subsets of differentiated cells, then this will likely require a better
understanding of specific cell surface markers.
An equally important issue to have emerged from the broader MSC literature is the
concept of “licensing”. Through studies of the immunological properties of MSC, it
has emerged that preconditioning of the stromal cells by treatment with cytokines can
be as critical as dosage and timing of administration for achieving efficacy. Thus, it is
likely that some form of licensing may also be necessary to achieve optimal results
using L-MSC. Indeed, as shown in Chapter 4, it is possible that preconditioning of L-
MSC by factors released from cultured limbal epithelial cells, may provide a
mechanism for switching these cells between a “pro” and “anti” angiogenic state.
Suspension of L-MSC in fibrin glue may provide similar benefits since Basu et al
(2014) noted a decrease in corneal vascularisation, whereas the present studies noted
a dramatic increase in corneal neovascularisation for L-MSC cultured alone on HAM.
A closer examination of L-MSC under these various conditions will therefore be
required in order see whether there are differences in the secretion of angiogenic
inducers (e.g. vascular endothelial growth factor or VEGF) and inhibitors (e.g.
pigment epithelium derived factor; PEDF) when grown on fibrin, HAM, or in the
presence of limbal epithelial cells.
6.3 ADMINISTRATION OF LIMBAL STROMAL CELLS TO THE EYE.
If materials such as HAM and fibrin glue significantly alter the phenotype and
biological effects of L-MSC, then careful consideration will need to be given to how
these cells are physically administered to the ocular surface. In animal studies to date,
L-MSC have been administered either as cell suspensions (intravenously, sub-
conjunctivally, and topically) or while adhered to a scaffold (Acar et al. 2015; Holan
et al. 2015). It is not clear, however, how either the stromal cell phenotype or secretome
may be affected under these different conditions. Moreover, if the local tissue
environment itself alters stromal cell function, then it could well be necessary to utilise
either exosomal vesicles or conditioned medium derived from L-MSC that have been
maintained under optimal conditions in vitro. From a manufacturing perspective, it
will best to have knowledge of what factor, or limited combination of L-MSC-derived
148
factors, are required for efficacy since therapies based upon purified factors should
theoretically be easier to standardise, define and validate prior to release.
Assuming that some form of scaffold is required for L-MSC, a number of lessons have
been gained from the current study. Firstly, the unique properties of HAM are very
difficult to emulate using silk fibroin. Despite being able to optimise the attachment
and growth of L-MSC/epithelial co-cultures using RGD, the resulting membranes still
display difficulty conforming to the ocular surface and remaining sutured in place. It
is therefore significant that better results than those presented in this thesis have
recently been achieved when cultures of limbal epithelial cells are simply placed face-
down upon the wounded ocular surface (Li et al. 2017). Presumably this strategy has
resulted in far better transfer of epithelial cells to the ocular surface, than when the
epithelial cells are facing outwards. It is likewise possible than many of the epithelial
cells cultured on HAM in Chapter 4 may well have become dislodged and especially
owing to the complications posed by rabbits having a nictitans. If, however, scaffolds
are simply to be used in this way, then there are perhaps far easier strategies than going
to the trouble of procuring HAM or fabricating membranes from silk fibroin. For
example, contacts lenses have already been shown to provide a useful vehicle for the
delivery of epithelial cells (Di Girolamo et al. 2009) and thus presumably would also
support growth of L-MSC cells.
6.4 CONCLUSION
In the course of my PhD, I have examined the potential benefit of using L-MSC for
corneal wound healing, and also explored the feasibility of a new formulation of silk
fibroin as a tool for transferring these cells onto the ocular surface. The results of this
study demonstrate the benefit of exploiting L-MSCs to support the growth of the
epithelial cells and in vivo (Figure 4-5). The feasibility of using RGD/HRP/PEG-
Fibroin as a substrate for cell culture In vitro was also confirmed (Figure 5-9). The
outcomes of this thesis provide a foundation for the use of the L-MSCs and the new
RGD silk fibroin scaffold and offers a stepping stone for future studies in cellular
therapy for LSCD.
149
Bibliography
Acar, U., Pinarli, F. A., Acar, D. E., Beyazyildiz, E., Sobaci, G., Ozgermen, B. B.,
Sonmez, A. A., and Delibasi, T. 2015. “Effect of Allogeneic Limbal
Mesenchymal Stem Cell Therapy in Corneal Healing: Role of Administration
Route,” Ophthalmic Research (53:2), pp. 82–89.
Ainscough, L. S., Linn, M. L., Barnard, Z., Schwab, I. R., and Harkin, D. G. 2011.
“Effects of Fibroblast Origin and Phenotype on the Proliferative Potential of
Limbal Epithelial Progenitor Cells,” Experimental Eye Research (92:1), pp. 10–
19.
Altman, G. H., Diaz, F., Jakuba, C., Calabro, T., Horan, R. L., Chen, J., Lu, H.,
Richmond, J., and Kaplan, D. L. 2003. “Silk-Based Biomaterials,” Biomaterials
(24:3), pp. 401–416.
Basu, S., Hertsenberg, A. J., Funderburgh, M. L., Burrow, M. K., Mann, M. M., Du,
Y., Lathrop, K. L., Syed-Picard, F. N., Adams, S. M., Birk, D. E., and
Funderburgh, J. L. 2014. “Human limbal biopsy-derived stromal stem cells
prevent corneal scarring,” Science translational medicine (6:266), pp. 266-172
Baylis, O., Figueiredo, F., Henein, C., Lako, M., and Ahmad, S. 2011. “13 Years of
Cultured Limbal Epithelial Cell Therapy: A Review of the Outcomes,” Journal
of Cellular Biochemistry (112:4), pp. 993–1002.
Bobba, S., Di Girolamo, N., Mills, R., Daniell, M., Chan, E., Harkin, D. G., Cronin,
B. G., Crawford, G., McGhee, C., and Watson, S. 2017. “Nature and Incidence
of Severe Limbal Stem Cell Deficiency in Australia and New Zealand,” Clinical
& Experimental Ophthalmology (45:2), pp. 174–181.
Bozkir, G., Bozkir, M., Dogan, H., Aycan, K., and Güler, B. 1997. “Measurements of
Axial Length and Radius of Corneal Curvature in the Rabbit Eye,” Acta Medica
Okayama (51:1), pp. 9–11.
Branch, M. J., Hashmani, K., Dhillon, P., Jones, D. R. E., Dua, H. S., and Hopkinson,
A. 2012. “Mesenchymal stem cells in the human corneal limbal stroma,”
Investigative ophthalmology & visual science (53:9), pp. 5109–5116.
150
Bray, L. J., George, K. a., Ainscough, S. L., Hutmacher, D. W., Chirila, T. V., and
Harkin, D. G. 2011. “Human Corneal Epithelial Equivalents Constructed on
Bombyx Mori Silk Fibroin Membranes,” Biomaterials (32:22), pp. 5086–5091.
Bray, L. J., George, K. a., Hutmacher, D. W., Chirila, T. V., and Harkin, D. G. 2012a.
“A Dual-Layer Silk Fibroin Scaffold for Reconstructing the Human Corneal
Limbus,” Biomaterials (33:13), pp. 3529–3538.
Bray, L. J., Heazlewood, C. F., Atkinson, K., Hutmacher, D. W., and Harkin, D. G.
2012b. “Evaluation of Methods for Cultivating Limbal Mesenchymal Stromal
Cells,” Cytotherapy (14), pp. 936–947.
Bray, L. J., Heazlewood, C. F., Munster, D. J., Hutmacher, D. W., Atkinson, K., and
Harkin, D. G. 2014. “Immunosuppressive properties of mesenchymal stromal cell
cultures derived from the limbus of human and rabbit corneas,” Cytotherapy
(16:1), pp. 64–73.
Bray, L. J., Suzuki, S., Harkin, D. G., and Chirila, T. 2013. “Incorporation of
Exogenous RGD Peptide and Inter-Species Blending as Strategies for Enhancing
Human Corneal Limbal Epithelial Cell Growth on Bombyx Mori Silk Fibroin
Membranes,” Journal of Functional Biomaterials (4:2), pp. 74–88.
Burillon, C., Huot, L., Justin, V., Nataf, S., Chapuis, F., Decullier, E., and Damour, O.
2012. “Cultured Autologous Oral Mucosal Epithelial Cell Sheet (CAOMECS)
Transplantation for the Treatment of Corneal Limbal Epithelial Stem Cell
Deficiency,” Investigative Ophthalmology and Visual Science (53:3), pp. 1325–
1331.
Buznyk, O., Pasyechnikova, N., Islam, M. M., Iakymenko, S., Fagerholm, P., and
Griffith, M. 2015. “Bioengineered Corneas Grafted as Alternatives to Human
Donor Corneas in Three High-Risk Patients,” Clinical and Translational Science
(8:5), pp. 558-62.
Casaroli-Marano, P, R., Nieto-Nicolau, N., Martínez-Conesa, E. M., Edel, M., and
B.Álvarez-Palomo, A. 2015. “Potential Role of Induced Pluripotent Stem Cells
(IPSCs) for Cell-Based Therapy of the Ocular Surface,” Journal of Clinical
Medicine (4:2), pp. 318–342.
Cejka, C., Holan, V., Trosan, P., Zajicova, A., Javorkova, E., and Cejkova, J. 2016.
151
“The Favorable Effect of Mesenchymal Stem Cell Treatment on the Antioxidant
Protective Mechanism in the Corneal Epithelium and Renewal of Corneal Optical
Properties Changed after Alkali Burns,” Oxidative Medicine and Cellular
Longevity (2016), Hindawi Publishing Corporation.
Chirila, T., Barnard, Z., Zainuddin, Harkin, D. G., Schwab, I. R., and Hirst, L. 2008.
“Bombyx mori silk fibroin membranes as potential substrata for epithelial
constructs used in the management of ocular surface disorders,” Tissue
engineering.Part A (14:7), p. 1203.
Chirila, T. V., Suzuki, S., and Papolla, C. 2017. “A Comparative Investigation of
Bombyx Mori Silk Fibroin Hydrogels Generated by Chemical and Enzymatic
Cross-Linking,” Biotechnology and Applied Biochemistry, pp. 1–11.
DelMonte, D. W., and Kim, T. 2011. “Anatomy and physiology of the cornea,”
Journal of Cataract & Refractive Surgery (37:3), pp. 588–598.
Demaine, E. D., Demaine, M. L., Iacono, J., and Langerman, S. 2009. “Wrapping
Spheres with Flat Paper,” Computational Geometry: Theory and Applications
(42:8), pp. 748–757.
Dominici, M., Le Blanc, K., Mueller, I., Slaper-Cortenbach, I., Marini, F. C., Krause,
D. S., Deans, R. J., Keating, A., Prockop, D. J., and Horwitz, E. M. 2006.
“Minimal Criteria for Defining Multipotent Mesenchymal Stromal Cells. The
International Society for Cellular Therapy Position Statement,” Cytotherapy
(8:4), pp. 315–317.
Dravida, S., Pal, R., Khanna, A., Tipnis, S. P., Ravindran, G., and Khan, F. 2005. “The
Transdifferentiation Potential of Limbal Fibroblast-like Cells,” Developmental
Brain Research (160:2), pp. 239–251.
Dua, H. S., Faraj, L. A., Said, D. G., Gray, T., and Lowe, J. 2013. “Human Corneal
Anatomy Redefined: A Novel Pre-Descemet’s Layer (Dua’s Layer),”
Ophthalmology (120:9), pp. 1778–1785.
Dua, H. S., Shanmuganathan, V. A., Powell-Richards, A. O., Tighe, P. J., and Joseph,
A. 2005. “Limbal Epithelial Crypts: A Novel Anatomical Structure and a Putative
Limbal Stem Cell Niche.,” The British Journal of Ophthalmology (89:5), pp.
529–32.
152
Fini, M. 1999. “Keratocyte and Fibroblast Phenotypes in the Repairing Cornea,”
Progress in Retinal and Eye Research (18:4), pp. 529–551.
Funderburgh, J. L., Funderburgh, M. L., and Du, Y. 2016. “Stem Cells in the Limbal
Stroma,” The Ocular Surface (14:2), pp. 113-20.
Garfias, Y., Nieves-Hernandez, J., Garcia-Mejia, M., Estrada-Reyes, C., and Jimenez-
Martinez, M. C. 2012. “Stem Cells Isolated from the Human Stromal Limbus
Possess Immunosuppressant Properties.,” Molecular Vision (18:May 2012), pp.
2087–95.
Gebler, A., Zabel, O., and Seliger, B. 2012. “The Immunomodulatory Capacity of
Mesenchymal Stem Cells,” Trends in Molecular Medicine (18:2), pp. 128–134.
Gil, E. S., Mandal, B. B., Park, S.-H., Marchant, J. K., Omenetto, F. G., and Kaplan,
D. L. 2010a. “Helicoidal multi-lamellar features of RGD-functionalized silk
biomaterials for corneal tissue engineering,” Biomaterials (31:34), pp. 8953–
8963.
Gil, E. S., Park, S.-H., Marchant, J., Omenetto, F., and Kaplan, D. L. 2010b. “Response
of human corneal fibroblasts on silk film surface patterns,” Macromolecular
Bioscience (10:6), pp. 664–673.
Di Girolamo, N., Bosch, M., Zamora, K., Coroneo, M. T., Wakefield, D., and Watson,
S. L. 2009. “A Contact Lens-Based Technique for Expansion and Transplantation
of Autologous Epithelial Progenitors for Ocular Surface Reconstruction,”
Transplantation (87:10), pp. 1571–1578.
Harkin, D. G., Dunphy, S. E., Shadforth, A. M. A., Dawson, R. A., Walshe, J., and
Zakaria, N. 2017. “Mounting of Biomaterials for Use in Ophthalmic Cell
Therapies,” Cell Transplantation (26:11), pp. 1717–1732.
Harkin, D. G., Foyn, L., Bray, L. J., Sutherlan, A. J., Li, F. J., and Cronin, B. G. 2015b.
“Concise Reviews : Can Mesenchymal Stromal Cells Differentiate into Corneal
Cells ? A Systematic Review of Published Data,” Stem Cells, pp. 785–791.
Harkin, D. G., George, K. A., Madden, P. W., Schwab, I. R., Hutmacher, D. W., and
Chirila, T. V. 2011. “Silk Fibroin in Ocular Tissue Reconstruction,” Biomaterials
(32:10), pp. 2445–2458.
Harkin, D. G., Sutherland, A. J., Bray, L. J., Foyn, L., and Li, F. J. 2015a. “The Use
153
of MSCs in the Treatment of Diseases of the Cornea”. Chapter 36, in The Biology
and Therapeutic Application of Mesenchymal Cells, Kerry Atkinson, Wiley
Harkin, D. G., Sutherland, A. J., Bray, L. J., Foyn, L., Li, F. J., and Cronin, B. G. 2016.
“The Use of Mesenchymal Stromal Cells in the Treatment of Diseases of the
Cornea,” in The Biology and Therapeutic Application of Mesenchymal Cells, pp.
524–543.
Hashmani, K., Branch, M. J., Sidney, L. E., Dhillon, P. S., Verma, M., McIntosh, O.
D., Hopkinson, A., and Dua, H. S. 2013. “Characterization of Corneal Stromal
Stem Cells with the Potential for Epithelial Transdifferentiation.,” Stem Cell
Research & Therapy (4:3), Stem Cell Research & Therapy, p. 75.
Hatou, S., Yoshida, S., Higa, K., Miyashita, H., Inagaki, E., Okano, H., Tsubota, K.,
and Shimmura, S. 2013. “Functional Corneal Endothelium Derived from Corneal
Stroma Stem Cells of Neural Crest Origin by Retinoic Acid and Wnt/β-Catenin
Signaling,” Stem Cells and Development (22:5), pp. 828–839.
Higa, K., Takeshima, N., Moro, F., Kawakita, T., Kawashima, M., Demura, M.,
Shimazaki, J., Asakura, T., Tsubota, K., and Shimmura, S. 2011. “Porous Silk
Fibroin Film as a Transparent Carrier for Cultivated Corneal Epithelial Sheets,”
Journal of Biomaterials Science, Polymer Edition (22:17), pp. 2261–2276.
Hogerheyde, T. a., Suzuki, S., Stephenson, S. a, Richardson, N. a, Chirila, T. V,
Harkin, D. G., and Bray, L. J. 2014. “Assessment of Freestanding Membranes
Prepared from Antheraea Pernyi Silk Fibroin as a Potential Vehicle for Corneal
Epithelial Cell Transplantation.,” Biomedical Materials (Bristol, England) (9:2),
pp 1-9.
Holan, V., and Javorkova, E. 2013. “Mesenchymal Stem Cells, Nanofiber Scaffolds
and Ocular Surface Reconstruction.,” Stem Cell Reviews (9:5), pp. 609–19.
Holan, V., Trosan, P., Cejka, C., E.Javorkova, A., Zajicova, B., Hermankova, M.,
Chudickova, J., and Cejkova. 2015. “A Comparative Study of the Therapeutic
Potential of Mesenchymal Stem Cells and Limbal Epithelial Stem Cells for
Ocular Surface Reconstruction,” Tissue Engineering and Regenerative Medicine,
pp. 1–11.
Holland, E. J. 2015. Management of Limbal Stem Cell Deficiency : A Historical
154
Perspective , Past , Present , and Future, (34:10), pp. 9–15.
Inatomi, T., Nakamura, T., Kojyo, M., Koizumi, N., Sotozono, C., and Kinoshita, S.
2006. “Ocular Surface Reconstruction with Combination of Cultivated
Autologous Oral Mucosal Epithelial Transplantation and Penetrating
Keratoplasty.,” American Journal of Ophthalmology (142:5), pp. 757–64.
Jawaheer, L., Anijeet, D., and Ramaesh, K. 2017. “Diagnostic Criteria for Limbal
Stem Cell Deficiency—a Systematic Literature Review,” Survey of
Ophthalmology (62:4), pp. 522–532.
Kambe, Y., Takeda, Y., Yamamoto, K., Kojima, K., Tamada, Y., and Tomita, N.
2010a. “Effect of RGDS-Expressing Fibroin Dose on Initial Adhesive Force of a
Single Chondrocyte,” Bio-Medical Materials and Engineering (20:6), pp. 309–
316.
Kambe, Y., Yamamoto, K., Kojima, K., Tamada, Y., and Tomita, N. 2010b. “Effects
of RGDS Sequence Genetically Interfused in the Silk Fibroin Light Chain Protein
on Chondrocyte Adhesion and Cartilage Synthesis,” Biomaterials (31:29), pp.
7503–7511.
Konomi, K., Satake, Y., Shimmura, S., Tsubota, K., and Shimazaki, J. 2013. “Long-
Term Results of Amniotic Membrane Transplantation for Partial Limbal
Deficiency,” Cornea (32:8), pp. 1110–1115.
Lanza, F., Campioni, D., Mauro, E., Pasini, A., and Rizzo, R. 2012.
“Immunosuppressive Properties of Mesenchymal Stromal Cells,” Advances in
Stem Cell Research, pp. 281–301.
Lawrence, B. D., Marchant, J. K., Pindrus, M. A., Omenetto, F. G., and Kaplan, D. L.
2009. “Silk Film Biomaterials for Cornea Tissue Engineering,” Biomaterials
(30:7), pp. 1299–1308.
Lee, R. H., Yu, J. M., Foskett, A. M., Peltier, G., Reneau, J. C., Bazhanov, N., Oh, J.
Y., and Prockop, D. J. 2014. “TSG-6 as a Biomarker to Predict Efficacy of Human
Mesenchymal Stem/progenitor Cells (hMSCs) in Modulating Sterile
Inflammation in Vivo.,” Proceedings of the National Academy of Sciences of the
United States of America (111:47), pp. 16766–71.
Levis, H. J., Kureshi, A. K., Massie, I., Morgan, L., Vernon, A. J., and Daniels, J. T.
155
2015. “Tissue Engineering the Cornea : The Evolution of RAFT,” Journal of
Functional Biomaterials, pp. 50–65.
Li, F. J., Nili, E., Lau, C., Richardson, N. A., Walshe, J., Barnett, N. L., Cronin, B. G.,
Hirst, L. W., Schwab, I. R., Chirila, T. V., and Harkin, D. G. 2016. “Evaluation
of the AlgerBrush II Rotating Burr as a Tool for Inducing Ocular Surface Failure
in the New Zealand White Rabbit,” Experimental Eye Research (147), pp. 1–11.
Li, Y., Yang, Y., Yang, L., Zeng, Y., Gao, X., and Xu, H. 2017. “Poly(ethylene
Glycol)-Modified Silk Fibroin Membrane as a Carrier for Limbal Epithelial Stem
Cell Transplantation in a Rabbit LSCD Model,” Stem Cell Research and Therapy
(8:1), Stem Cell Research & Therapy, pp. 1–19.
Ljubimov, A. V., and Saghizadeh, M. 2015. “Progress in Corneal Wound Healing,”
Progress in Retinal and Eye Research (49), pp. 17–45.
Madden, P. W., Lai, J. N. X., George, K. A., Giovenco, T., Harkin, D. G., and Chirila,
T. V. 2011. “Human Corneal Endothelial Cell Growth on a Silk Fibroin
Membrane,” Biomaterials (32:17), pp. 4076–4084.
Maharajan, V. S., Shanmuganathan, V., Currie, A., Hopkinson, A., Powell-Richards,
A., and Dua, H. S. 2007. “Amniotic Membrane Transplantation for Ocular
Surface Reconstruction: Indications and Outcomes,” Clinical & Experimental
Ophthalmology (35:2), p. 140-147.
Marchini, G., Pedrotti, E., Pedrotti, M., Barbaro, V., Di Iorio, E., Ferrari, S., Bertolin,
M., Ferrari, B., Passilongo, M., Fasolo, A., and Ponzin, D. 2012. “Long-Term
Effectiveness of Autologous Cultured Limbal Stem Cell Grafts in Patients with
Limbal Stem Cell Deficiency due to Chemical Burns,” Clinical & Experimental
Ophthalmology (40:3), pp. 255–267.
Massie, I., Levis, H. J., and Daniels, J. T. 2014. “Response of Human Limbal Epithelial
Cells to Wounding on 3D RAFT Tissue Equivalents: Effect of Airlifting and
Human Limbal Fibroblasts,” Experimental Eye Research (127), pp. 196–205.
Maxson, S., Lopez, E. A., Yoo, D., Danilkovitch-Miagkova, A., and LeRoux, M. A.
2012. “Concise Review: Role of Mesenchymal Stem Cells in Wound Repair,”
STEM CELLS Translational Medicine (1:2), pp. 142–149.
Maya-Vetencourt, J. F., Ghezzi, D., Antognazza, M. R., Colombo, E., Mete, M.,
156
Feyen, P., Desii, A., Buschiazzo, A., Di Paolo, M., Di Marco, S., Ticconi, F.,
Emionite, L., Shmal, D., Marini, C., Donelli, I., Freddi, G., MacCarone, R., Bisti,
S., Sambuceti, G., Pertile, G., Lanzani, G., and Benfenati, F. 2017. “A Fully
Organic Retinal Prosthesis Restores Vision in a Rat Model of Degenerative
Blindness,” Nature Materials (16:6), pp. 681–689.
McKee, H. D., Irion, L. C. D., Carley, F. M., Brahma, A. K., Jafarinasab, M. R.,
Rahmati-Kamel, M., Kanavi, M. R., and Feizi, S. 2014. “Re: Dua et Al.: Human
Corneal Anatomy Redefined: A Novel Pre-Descemet Layer (Dua’s Layer)
(Ophthalmology 2013;120:1778-85),” Ophthalmology (121:5), pp. 2010–2011.
Meller, D., Pires, R. T., Mack, R. J., Figueiredo, F., Heiligenhaus, A., Park, W. C.,
Prabhasawat, P., John, T., McLeod, S. D., Steuhl, K. P., and Tseng, S. C. 2000.
“Amniotic Membrane Transplantation for Acute Chemical or Thermal Burns.,”
Ophthalmology (107:5), p. 980–9; discussion 990.
Nakamura, T., Inatomi, T., Sotozono, C., Ang, L. P. K., Koizumi, N., Yokoi, N., and
Kinoshita, S. 2006. “Transplantation of Autologous Serum-Derived Cultivated
Corneal Epithelial Equivalents for the Treatment of Severe Ocular Surface
Disease.,” Ophthalmology (113:10), pp. 1765–72.
Nakamura, T., Inatomi, T., Sotozono, C., Koizumi, N., and Kinoshita, S. 2015. “Ocular
Surface Reconstruction Using Stem Cell and Tissue Engineering,” Progress in
Retinal and Eye Research (51), pp. 187–207.
Nakatsu, M. N., González, S., Mei, H., and Deng, S. X. 2014. “Human Limbal
Mesenchymal Cells Support the Growth of Human Corneal Epithelial
Stem/progenitor Cells,” Investigative Ophthalmology & Visual Science (55:10),
pp. 6953–6959.
Ng, T. K., and Yung, J. S. 2015. SM Gr up SM Ophthalmology Research Progress and
Human Clinical Trials of Mesenchymal Stem Cells in Ophthalmology : A Mini
Review, (1:1), pp. 1–9.
Notara, M., Shortt, A. J., Galatowicz, G., Calder, V., and Daniels, J. T. 2010. “IL6 and
the Human Limbal Stem Cell Niche: A Mediator of Epithelial-Stromal
Interaction,” Stem Cell Research (5:3), pp. 188–200.
Oh, J. Y., Lee, R. H., Yu, J. M., Ko, J. H., Lee, H. J., Ko, A. Y., Roddy, G. W., and
157
Prockop, D. J. 2012. “Intravenous Mesenchymal Stem Cells Prevented Rejection
of Allogeneic Corneal Transplants by Aborting the Early Inflammatory
Response,” Molecular Therapy (20:11), pp. 2143–2152.
Oh, J. Y., Roddy, G. W., Choi, H., Lee, R. H., Ylostalo, J. H., Rosa, R. H., and
Prockop, D. J. 2010. “Anti-Inflammatory Protein TSG-6 Reduces Inflammatory
Damage to the Cornea Following Chemical and Mechanical Injury,” Proceedings
of the National Academy of Sciences (107:39), pp. 16875–16880.
Pellegrini, G., Dellambra, E., Golisano, O., Martinelli, E., Fantozzi, I., Bondanza, S.,
Ponzin, D., McKeon, F., and De Luca, M. 2001. “P63 Identifies Keratinocyte
Stem Cells,” Proceedings of the National Academy of Sciences (98:6), pp. 3156–
3161.
Pellegrini, G., Traverso, C. E., Franzi, A. T., Zingirian, M., Cancedda, R., and De
Luca, M. 1997. “Long-Term Restoration of Damaged Corneal Surfaces with
Autologous Cultivated Corneal Epithelium,” Lancet (349:9057), pp. 990–993.
Phinney, D. G. 2012. “Functional Heterogeneity of Mesenchymal Stem Cells:
Implications for Cell Therapy,” Journal of Cellular Biochemistry (113:9), pp.
2806–2812.
Phinney, D. G., and Prockop, D. J. 2007. “Concise Review: Mesenchymal
Stem/multipotent Stromal Cells: The State of Transdifferentiation and Modes of
Tissue Repair--Current Views.,” Stem Cells (25:11), pp. 2896–2902.
Polisetty, N., Fatima, A., Madhira, S. L., Sangwan, V. S., and Vemuganti, G. K. 2008.
“Mesenchymal cells from limbal stroma of human eye,” Molecular Vision (14),
p. 431.
Qi, X., Xie, L., Cheng, J., Zhai, H., and Zhou, Q. 2013. “Characteristics of Immune
Rejection after Allogeneic Cultivated Limbal Epithelial Transplantation.,”
Ophthalmology (120:5), pp. 931–6.
Ramirez-Miranda, A., Nakatsu, M. N., Zarei-Ghanavati, S., Nguyen, C. V, and Deng,
S. X. 2011. “Keratin 13 Is a More Specific Marker of Conjunctival Epithelium
than Keratin 19.,” Molecular Vision (17:May), pp. 1652–1661.
Reinshagen, H., Auw‐Haedrich, C., Sorg, R. V, Boehringer, D., Eberwein, P.,
Schwartzkopff, J., Sundmacher, R., and Reinhard, T. 2011. “Corneal surface
158
reconstruction using adult mesenchymal stem cells in experimental limbal stem
cell deficiency in rabbits,” Acta Ophthalmologica (89:8), pp. 741–748.
Rowsey, T. G., and Karamichos, D. 2017. “The Role of Lipids in Corneal Diseases
and Dystrophies: A Systematic Review,” Clinical and Translational Medicine
(6:1), p. 30.
Salem, H. K., and Thiemermann, C. 2009. “Mesenchymal Stromal Cells ; Current
Understanding and Clinical Status,” Stem Cells (28:3), pp 585-96.
Sangwan, V. S., Basu, S., MacNeil, S., and Balasubramanian, D. 2012. “Simple
Limbal Epithelial Transplantation (SLET): A Novel Surgical Technique for the
Treatment of Unilateral Limbal Stem Cell Deficiency,” British Journal of
Ophthalmology (96:7), pp. 931–934.
Sati, A., Shukla, S., Lal, I., and Sangwan, V. S. 2015. “Treating Limbal Stem Cell
Deficiency: Current and Emerging Therapies,” Expert Opinion on Orphan Drugs
(3:6), pp. 619–631.
Schalek, P., Otruba, L., Hornáčková, Z., and Hahn, A. 2013. “Mucosal maxillary cysts:
long-term subjective outcomes after surgical treatment,” European Archives of
Oto-Rhino-Laryngology (270:8), pp. 2263–2266.
Schermer, A., Galvin, S., and Sun, T. T. 1986. “Differentiation-Related Expression of
a Major 64K Corneal Keratin in Vivo and in Culture Suggests Limbal Location
of Corneal Epithelial Stem Cells,” Journal of Cell Biology (103:1), pp. 49–62.
Schwab, I. R. 1999. “Cultured Corneal Epithelia for Ocular Surface Disease.,”
Transactions of the American Ophthalmological Society (97), pp. 891–986.
Schwab, I. R., Johnson, N. T., and Harkin, D. G. 2006. “Inherent Risks Associated
with Manufacture of Bioengineered Ocular Surface Tissue.,” Archives of
Ophthalmology (124:12), pp. 1734–1740.
Shadforth, A. M. a, George, K. a., Kwan, A. S., Chirila, T. V., and Harkin, D. G. 2012.
“The Cultivation of Human Retinal Pigment Epithelial Cells on Bombyx Mori
Silk Fibroin,” Biomaterials (33:16), pp. 4110–4117.
Shimazaki, J., Aiba, M., Goto, E., Kato, N., Shimmura, S., and Tsubota, K. 2002.
“Transplantation of Human Limbal Epithelium Cultivated on Amniotic
Membrane for the Treatment of Severe Ocular Surface disorders1 1The Authors
159
Do Not Have Any Proprietary Interest in the Products Mentioned or Used in This
Study.,” Ophthalmology (109:7), pp. 1285–1290.
Shortt, A. J., Secker, G. A., Munro, P. M., Khaw, P. T., Tuft, S. J., and Daniels, J. T.
2007. “Characterization of the Limbal Epithelial Stem Cell Niche: Novel Imaging
Techniques Permit in Vivo Observation and Targeted Biopsy of Limbal
Epithelial Stem Cells.,” Stem Cells (Dayton, Ohio) (25:6), pp. 1402–9.
Sidney, L., Branch, M., Dua, H., and Hopkinson, A. 2015a. “Effect of Culture Medium
on Propagation and Phenotype of Corneal Stroma-Derived Stem Cells,”
Cytotherapy (17:12), pp. 1706.
Sidney, L. E., Branch, M. J., Dunphy, S. E., Dua, H. S., and Hopkinson, A. 2014.
“Concise Review: Evidence for CD34 as a Common Marker for Diverse
Progenitors,” Stem Cells (32:6), pp. 1380–1389.
Sidney, L. E., and Hopkinson, A. 2017. “Corneal Keratocyte Transition to
Mesenchymal Stem Cell Phenotype and Reversal Using Serum-Free Medium
Supplemented with Fibroblast Growth Factor-2, Transforming Growth Factor-
??3 and Retinoic Acid,” Journal of Tissue Engineering and Regenerative
Medicine (4:March 2017), pp. 203–215.
Sidney, L. E., McIntosh, O. D., and Hopkinson, A. 2015b. “Phenotypic Change and
Induction of Cytokeratin Expression during in Vitro Culture of Corneal Stromal
Cells,” Investigative Ophthalmology and Visual Science (56:12), pp. 7225–7235.
Da Silva Meirelles, L., Fontes, A. M., Covas, D. T., and Caplan, A. I. 2009.
“Mechanisms Involved in the Therapeutic Properties of Mesenchymal Stem
Cells,” Cytokine and Growth Factor Reviews (20:5–6), pp. 419–427.
Spinelli, A., Luca, M. De, Pellegrini, G., and Ph, D. 2010. Limbal Stem-Cell Therapy
and Long-Term Corneal Regeneration, pp. 147–155.
Stewart, R. M. K., Sheridan, C. M., Hiscott, P. S., Czanner, G., and Kaye, S. B. 2015.
“Human Conjunctival Stem Cells Are Predominantly Located in the Medial
Canthal and Inferior Forniceal Areas,” Investigative Ophthalmology & Visual
Science (56:3), pp. 2021–2030.
Suzuki, S., Dawson, R., Chirila, T., Shadforth, A., Hogerheyde, T., Edwards, G., and
Harkin, D. 2015. “Treatment of Silk Fibroin with Poly(ethylene Glycol) for the
160
Enhancement of Corneal Epithelial Cell Growth,” Journal of Functional
Biomaterials (6:2), pp. 345–366.
Tsai, R. J.-F., Li, L.-M., and Chen, J. K. 2000. Reconstruction of Damaged Corneas
by Transplantation of Autologous Limbal Epithelial Cells, (182), pp. 96–97.
Tseng, S. C. G. 1989. “Concept and Application of Limbal Stem Cells,” Eye (3:2), pp.
141–157.
Tsubota, K., Toda, I., Saito, H., Shinozaki, N., and Shimazaki, J. 1995.
“Reconstruction of the Corneal Epithelium by Limbal Allograft Transplantation
for Severe Ocular Surface Disorders,” Ophthalmology (102:10), pp. 1486–1496.
Tuft, S. J., and Coster, D. J. 1990. “The Corneal Endothelium,” Eye (4:3), pp. 389–
424.
Vazirani, J., Ali, M. H., Sharma, N., Gupta, N., Mittal, V., Atallah, M., Amescua, G.,
Chowdhury, T., Abdala-Figuerola, A., Ramirez-Miranda, A., Navas, A., Graue-
Hernández, E. O., and Chodosh, J. 2016. “Autologous Simple Limbal Epithelial
Transplantation for Unilateral Limbal Stem Cell Deficiency: Multicentre
Results,” British Journal of Ophthalmology (100:10), pp. 1416–1420.
Walshe, J., and Harkin, D. G. 2014. “Serial Explant Culture Provides Novel Insights
into the Potential Location and Phenotype of Corneal Endothelial Progenitor
Cells,” Experimental Eye Research (127), pp. 9–13.
Williams, A. R., and Hare, J. M. 2011. “Mesenchymal Stem Cells: Biology,
Pathophysiology, Translational Findings, and Therapeutic Implications for
Cardiac Disease,” Circulation Research (109:8), pp. 923–940.
Williams, K., Keane, M., Galettis, R., Jones, V., Mills, R., and Coster, D. 2015. The
Australian Corneal Graft Registry 2015 Report, (August), pp. 1–409.
Yao, L., and Bai, H. 2013. “Review: mesenchymal stem cells and corneal
reconstruction,” Molecular vision Journal Article (19), p. 2237.
Yao, L., Li, Z., Su, W., Li, Y., Lin, M., Zhang, W., Liu, Y., Wan, Q., and Liang, D.
2012. “Role of Mesenchymal Stem Cells on Cornea Wound Healing Induced by
Acute Alkali Burn.,” PloS One (7:2), pp. 1-7.
Zhang, L., Coulson-Thomas, V. J., Ferreira, T. G., and Kao, W. W. Y. 2015.
161
“Mesenchymal Stem Cells for Treating Ocular Surface Diseases,” BMC
Ophthalmology (15:1), BMC Ophthalmology.
