(Methods in Molecular Biology 2145) Mark Ahearne - Corneal Regeneration - Methods and Protocols-Springer US - Humana (2020)

Methods in
Molecular Biology 2145
Mark Ahearne Editor
Corneal
Regeneration
Methods and Protocols
METHODS IN MOLECULAR BIOLOGY
Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, UK
For further volumes:

http://www.springer.com/series/7651
For over 35 years, biological scientists have come to rely on the research protocols and
methodologies in the critically acclaimed Methods in Molecular Biology series. The series was
the first to introduce the step-by-step protocols approach that has become the standard in all
biomedical protocol publishing. Each protocol is provided in readily-reproducible step-by-
step fashion, opening with an introductory overview, a list of the materials and reagents
needed to complete the experiment, and followed by a detailed procedure that is supported
with a helpful notes section offering tips and tricks of the trade as well as troubleshooting
advice. These hallmark features were introduced by series editor Dr. John Walker and
constitute the key ingredient in each and every volume of the Methods in Molecular Biology
series. Tested and trusted, comprehensive and reliable, all protocols from the series are
indexed in PubMed.
Corneal Regeneration
Methods and Protocols
Edited by
Mark Ahearne
Trinity Centre for Biomedical Engineering, Trinity College Dublin, Dublin, Ireland
Editor
Mark Ahearne
Trinity Centre for Biomedical
Engineering
Trinity College Dublin
Dublin, Ireland
ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-0716-0598-1 ISBN 978-1-0716-0599-8 (eBook)
https://doi.org/10.1007/978-1-0716-0599-8
© Springer Science+Business Media, LLC, part of Springer Nature 2020

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply,
even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations
and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to
be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer
Nature.
The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.
Preface
Corneal blindness is one of the most common causes of blindness worldwide. To treat this
condition, there has been much interest in trying to develop new therapies to regenerate or
repair damaged and diseased corneal tissue. This book will provide a detailed overview of
several laboratory techniques that are used to develop regenerative therapies to help treat
corneal blindness. These include how to optimize cell culture conditions, how to apply
gene-editing techniques, how to prepare different types of scaffold for corneal regeneration,
and how to evaluate the success of these therapies using in vitro and in vivo models and cell
and material characterization techniques. This book will be of interest to new and experi-
enced laboratory researchers working on different aspects of corneal regeneration as well as
ophthalmologists and patients interested in learning more about the latest techniques and
technology.
Dublin, Ireland Mark Ahearne
v
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
1 Isolation and Culture of Corneal Stromal Stem Cells . . . . . . . . . . . . . . . . . . . . . . . . 1

Richard M. Nagymihaly, Morten C. Moe, and Goran Petrovski
2 In Vitro Expansion of Corneal Endothelial Cells for Transplantation . . . . . . . . . . 17
Kim Santerre, Isabelle Xu, Mathieu Thériault, and Stéphanie Proulx
3 Primary Culture of Cornea-Limbal Epithelial Cells In Vitro . . . . . . . . . . . . . . . . . . 29
Finbarr O’Sullivan
4 Optimization of Human Limbal Stem Cell Culture by Replating
a Single Limbal Explant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Marina Lo pez-Paniagua, Teresa Nieto-Miguel, Sara Galindo,
Laura Garcı́a-Posadas, Ana de la Mata, Rosa M. Corrales,
Margarita Calonge, and Yolanda Diebold
5 A Guide to the Development of Human CorneaOrganoids
from Induced Pluripotent Stem Cells in Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
James W. Foster, Karl J. Wahlin, and Shukti Chakravarti
6 Gene Editing for Corneal Stromal Regeneration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Tara Moore, Connie Chao-Shern, Larry DeDionisio,
Kathleen A. Christie, and M. Andrew Nesbit
7 Preparation and Administration of Adeno-associated Virus Vectors
for Corneal Gene Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Liujiang Song, Jacquelyn J. Bower, and Matthew L. Hirsch
8 The Self-assembly Approach as a Tool for the Tissue Engineering
of a Bi-lamellar Human Cornea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Gaëtan Le-Bel, Pascale Desjardins, Camille Couture,
Lucie Germain, and Sylvain L. Guérin
9 Formation of Corneal Stromal-Like Assemblies Using Human Corneal
Fibroblasts and Macromolecular Crowding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Mehmet Gürdal, Gülinnaz Ercan, and Dimitrios I. Zeugolis
10 Preparation of Dried Amniotic Membrane for Corneal Repair . . . . . . . . . . . . . . . . 143
Andrew Hopkinson, Emily R. Britchford, and Laura E. Sidney
11 Fabrication of Corneal Extracellular Matrix-Derived Hydrogels . . . . . . . . . . . . . . 159
Mark Ahearne and Julia Fernández-Pérez
12 Synthesis and Application of Collagens for Assembling a Corneal Implant . . . . . 169
Elle Edin, Fiona Simpson, and May Griffith
13 Development and Validation of a 3D In Vitro Model to Study
the Chemotactic Behavior of Corneal Stromal Fibroblasts . . . . . . . . . . . . . . . . . . . 185
Evrim Ceren Kabak, Julia Fernández-Pérez, and Mark Ahearne
vii
viii Contents
14 Femtosecond Laser-Assisted Surgery for Implantation

of Bioengineered Corneal Stroma to Promote Corneal Regeneration. . . . . . . . . . 197
Neil Lagali and Mehrdad Rafat
15 The Use of Animal Models to Assess Engineered Corneal Tissue . . . . . . . . . . . . . 215
Robert Thomas Brady and Peter W. Madden
16 X-Ray Diffraction Imaging of Corneal Ultrastructure . . . . . . . . . . . . . . . . . . . . . . . 231
Keith M. Meek, Andrew J. Quantock, Sally Hayes, and James Bell
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Contributors
MARK AHEARNE • Department of Mechanical and Manufacturing Engineering, School of

Engineering, Trinity College Dublin, The University of Dublin, Dublin, Ireland; Trinity
Centre for Biomedical Engineering, Trinity Biomedical Sciences Institute, Trinity College
Dublin, The University of Dublin, Dublin, Ireland
JAMES BELL • Structural Biophysics Group, School of Optometry and Vision Sciences, Cardiff
University, Cardiff, UK
JACQUELYN J. BOWER • Lineberger Comprehensive Cancer Center, University of North
Carolina at Chapel Hill, Chapel Hill, NC, USA
ROBERT THOMAS BRADY • Department of Ophthalmology, Mater Misericordiae University
Hospital, Dublin, Ireland
EMILY R. BRITCHFORD • Academic Ophthalmology, Division of Clinical Neuroscience, School
of Medicine, University of Nottingham, Nottingham, UK; NuVision Biotherapies Ltd,
MediCity, Nottingham, UK
MARGARITA CALONGE • Grupo de Superficie Ocular, Instituto Universitario de
Oftalmobiologı́a Aplicada (IOBA), Universidad de Valladolid, Valladolid, Spain; Centro
de Investigacion Biomédica en Red de Bioingenierı́a, Biomateriales y Nanomedicina
(CIBER-BBN). Instituto de Salud Carlos III, Madrid, Spain
SHUKTI CHAKRAVARTI • Department of Ophthalmology and Pathology, NYU Langone
Health, Alexandria Life Sciences Center, New York, NY, USA
CONNIE CHAO-SHERN • Biomedical Sciences Research Institute, Ulster University, Coleraine,
Northern Ireland, UK; Avellino Lab USA, Inc., Menlo Park, CA, USA
KATHLEEN A. CHRISTIE • Biomedical Sciences Research Institute, Ulster University,
Coleraine, Northern Ireland, UK
ROSA M. CORRALES • Grupo de Superficie Ocular, Instituto Universitario de Oftalmobiologı́a
Aplicada (IOBA), Universidad de Valladolid, Valladolid, Spain; Centro de Investigacion
Biomédica en Red de Bioingenierı́a, Biomateriales y Nanomedicina (CIBER-BBN).
Instituto de Salud Carlos III, Madrid, Spain
CAMILLE COUTURE • Centre universitaire d’ophtalmologie - recherche (CUO-Recherche) et,
Université Laval, Québec, QC, Canada; Centre de recherche en organogénèse expérimentale
de l’Université Laval/LOEX, Centre de recherche du CHU de Québec-Université Laval,
Université Laval, Québec, QC, Canada; Département de chirurgie, Faculté de médecine,
Université Laval, Québec, QC, Canada; Département d’ophtalmologie, Faculté de Mé
decine, Université Laval, Québec, QC, Canada
LARRY DEDIONISIO • Avellino Lab USA, Inc., Menlo Park, CA, USA
PASCALE DESJARDINS • Centre universitaire d’ophtalmologie - recherche (CUO-Recherche) et,
Université Laval, QCQuébec, Canada; Département d’ophtalmologie, Faculté de Médecine,
Université Laval, Québec, QC, Canada
YOLANDA DIEBOLD • Grupo de Superficie Ocular, Instituto Universitario de Oftalmobiologı́a
ix
x Contributors
ELLE EDIN • Maisonneuve-Rosemont Hospital Research Centre, Montréal, QC, Canada;

Department of Ophthalmology and Institute of Biomedical Engineering, Université de
Montréal, Montréal, QC, Canada
GÜLINNAZ ERCAN • Faculty of Medicine, Department of Medical Biochemistry, Ege
University, Izmir, Turkey; Department of Stem Cell, Institute of Health Sciences, Ege
University, Izmir, Turkey
JULIA FERNÁNDEZ-PÉREZ • Department of Mechanical and Manufacturing Engineering,
School of Engineering, Trinity College Dublin, The University of Dublin, Dublin, Ireland;
Trinity Centre for Biomedical Engineering, Trinity Biomedical Sciences Institute, Trinity
College Dublin, The University of Dublin, Dublin, Ireland
JAMES W. FOSTER • The Wilmer Eye Institute, Johns Hopkins University, Baltimore, MD,
USA
SARA GALINDO • Grupo de Superficie Ocular, Instituto Universitario de Oftalmobiologı́a
LAURA GARCÍA-POSADAS • Grupo de Superficie Ocular, Instituto Universitario de
Oftalmobiologı́a Aplicada (IOBA), Universidad de Valladolid, Valladolid, Spain
LUCIE GERMAIN • Centre universitaire d’ophtalmologie - recherche (CUO-Recherche) et,
Université Laval, QCQuébec, Canada; Département de chirurgie, Faculté de médecine,
MAY GRIFFITH • Maisonneuve-Rosemont Hospital Research Centre, Montréal, QC, Canada;
SYLVAIN L. GUÉRIN • Centre universitaire d’ophtalmologie - recherche (CUO-Recherche) et,
MEHMET GÜRDAL • Regenerative, Modular & Developmental Engineering Laboratory
(REMODEL), National University of Ireland Galway (NUI Galway), Galway, Ireland;
Science Foundation Ireland (SFI) Centre for Research in Medical Devices (CÚRAM),
National University of Ireland Galway (NUI Galway), Galway, Ireland; Faculty of
Medicine, Department of Medical Biochemistry, Ege University, Izmir, Turkey
SALLY HAYES • Structural Biophysics Group, School of Optometry and Vision Sciences, Cardiff
University, Cardiff, UK
MATTHEW L. HIRSCH • Gene Therapy Center, University of North Carolina at Chapel Hill,
Chapel Hill, NC, USA; Department of Ophthalmology, University of North Carolina,
Chapel Hill, NC, USA
ANDREW HOPKINSON • Academic Ophthalmology, Division of Clinical Neuroscience, School of
Medicine, University of Nottingham, Nottingham, UK; NuVision Biotherapies Ltd,
MediCity, Nottingham, UK
EVRIM CEREN KABAK • Trinity Centre for Biomedical Engineering, Trinity Biomedical
Sciences Institute, Trinity College Dublin, The University of Dublin, Dublin, Ireland;
Biotechnology Unit, Nobel Pharmaceuticals A.S, Gebze, Kocaeli, Turkey
Contributors xi
NEIL LAGALI • Division of Ophthalmology, Institute for Biomedical and Clinical Sciences,
Linköping University, Linköping, Sweden
GAËTAN LE-BEL • Centre universitaire d’ophtalmologie - recherche (CUO-Recherche) et,
Université Laval, Québec, QC, Canada; Département de chirurgie, Faculté de médecine,
MARINA LÓPEZ-PANIAGUA • Grupo de Superficie Ocular, Instituto Universitario de
PETER W. MADDEN • Department of Mechanical and Manufacturing Engineering, School of
Engineering, Trinity College Dublin, University of Dublin, Dublin, Ireland; Trinity
Centre for Biomedical Engineering, Trinity Biomedical Science Institute, Trinity College
Dublin, University of Dublin, Dublin, Ireland
ANA DE LA MATA • Grupo de Superficie Ocular, Instituto Universitario de Oftalmobiologı́a
KEITH M. MEEK • Structural Biophysics Group, School of Optometry and Vision Sciences,
Cardiff University, Cardiff, UK
MORTEN C. MOE • Department of Ophthalmology, Center for Eye Research, Oslo University
Hospital, Oslo, Norway; Faculty of Medicine, Institute of Clinical Medicine, University of
Oslo, Oslo, Norway
TARA MOORE • Biomedical Sciences Research Institute, Ulster University, Coleraine,
Northern Ireland, UK; Avellino Lab USA, Inc., Menlo Park, CA, USA
RICHARD M. NAGYMIHALY • Department of Ophthalmology, Center for Eye Research, Oslo
University Hospital, Oslo, Norway
M. ANDREW NESBIT • Biomedical Sciences Research Institute, Ulster University, Coleraine,
Northern Ireland, UK
TERESA NIETO-MIGUEL • Grupo de Superficie Ocular, Instituto Universitario de
FINBARR O’SULLIVAN • National Institute for Cellular Biotechnology and SSPC-SFI, Centre
for Pharmaceuticals, Dublin City University, Dublin, Ireland
GORAN PETROVSKI • Department of Ophthalmology, Center for Eye Research, Oslo University
Hospital, Oslo, Norway; Faculty of Medicine, Institute of Clinical Medicine, University of
Oslo, Oslo, Norway
STÉPHANIE PROULX • Centre de recherche du Centre hospitalier universitaire (CHU) de Qué
bec—Université Laval, axe médecine régénératrice, Hôpital du Saint-Sacrement, Québec,
QC, Canada; Département d’Ophtalmologie et d’oto-rhino-laryngologie-chirurgie cervico-
faciale, Faculté de médecine, Université Laval, Québec, QC, Canada; Centre de recherche
en organogénèse expérimentale de l’Université Laval/LOEX, Québec, QC, Canada
ANDREW J. QUANTOCK • Structural Biophysics Group, School of Optometry and Vision
Sciences, Cardiff University, Cardiff, UK
xii Contributors
MEHRDAD RAFAT • Department of Biomedical Engineering, Linköping University,

Linköping, Sweden; LinkoCare Life Sciences AB, Linköping, Sweden
KIM SANTERRE • Centre de recherche du Centre hospitalier universitaire (CHU) de Québec—
Université Laval, axe médecine régénératrice, Hôpital du Saint-Sacrement, Québec, QC,
Canada; Département d’Ophtalmologie et d’oto-rhino-laryngologie-chirurgie cervico-
LAURA E. SIDNEY • Academic Ophthalmology, Division of Clinical Neuroscience, School of
Medicine, University of Nottingham, Nottingham, UK
FIONA SIMPSON • Maisonneuve-Rosemont Hospital Research Centre, Montréal, QC, Canada;
LIUJIANG SONG • Gene Therapy Center, University of North Carolina at Chapel Hill, Chapel
Hill, NC, USA; Department of Ophthalmology, University of North Carolina, Chapel
Hill, NC, USA
MATHIEU THÉRIAULT • Centre de recherche du Centre hospitalier universitaire (CHU) de
Québec—Université Laval, axe médecine régénératrice, Hôpital du Saint-Sacrement, Qué
bec, QC, Canada; Département d’Ophtalmologie et d’oto-rhino-laryngologie-chirurgie
cervico-faciale, Faculté de médecine, Université Laval, Québec, QC, Canada; Centre de
recherche en organogénèse expérimentale de l’Université Laval/LOEX, Québec, QC,
Canada
KARL J. WAHLIN • Shiley Eye Institute, UC San Diego, La Jolla, CA, USA
ISABELLE XU • Centre de recherche du Centre hospitalier universitaire (CHU) de Québec—
Université Laval, axe médecine régénératrice, Hôpital du Saint-Sacrement, Québec, QC,
Canada; Département d’Ophtalmologie et d’oto-rhino-laryngologie-chirurgie cervico-
DIMITRIOS I. ZEUGOLIS • Regenerative, Modular & Developmental Engineering Laboratory
(REMODEL), National University of Ireland Galway (NUI Galway), Galway, Ireland;
Science Foundation Ireland (SFI) Centre for Research in Medical Devices (CÚRAM),
National University of Ireland Galway (NUI Galway), Galway, Ireland
Chapter 1
Isolation and Culture of Corneal Stromal Stem Cells

Richard M. Nagymihaly, Morten C. Moe, and Goran Petrovski
Abstract
An increasing body of evidence authenticates the benefit of corneal stroma-derived stem cells (CSSCs) in
tissue engineering and regeneration oriented research, and potentially in the development of clinically
relevant cellular therapies. Postmortem corneal tissue obtained from otherwise discarded material after
keratoplasties is oftentimes the source of the cells for ex vivo research. Relatively easy to isolate and cultivate
as well as inexpensive to culture, CSSCs now represent a well-described cell type with attributes of
mesenchymal stem cells (MSCs). These include differentiation- and immunosuppressive potential, as well
as a favorable capacity to expand in vitro. Here, we in detail describe two straightforward methods to isolate
and establish CSSC cultures ex vivo.
Key words Cornea, Corneal stroma, Keratocytes, Corneal fibroblasts, Corneal stromal stem cells,
Mesenchymal stem cells, Isolation, Cell culturing
1 Introduction
The corneal stroma takes 90% of the volume and mass of the
cornea. It is characterized by a unique dome-shape and transpar-
ency, a result of its components, mostly collagen I and bridging
molecules—proteoglycans. The complex tissue layer is 0.5–0.7 mm
thick and consists of approximately 200 layers of collagen lamellae,
orthogonally stacked on each other. Embedded between these
layers lie the keratocytes, the most abundant cells of the stroma.
The keratocytes are scarcely distributed across the stroma, concen-
trated in the anterior part (20,000–25,000 cells/mm2) (Fig. 1),
decreasing in the middle and increasing in the posterior part again
[1]. These cells remain quiescent in healthy conditions, displaying a
dendritic-like morphology; however, upon activation by either dis-
ease or injury, the keratocytes may differentiate into fibroblasts or
myofibroblasts [2]. The function of the latter two is to assist
stromal wound healing through the secretion of matrix metallo-
proteinases (MMPs) and deposition of de novo extracellular matrix
(ECM) components, in order to remodel the tissue [3, 4].
Mark Ahearne (ed.), Corneal Regeneration: Methods and Protocols, Methods in Molecular Biology, vol. 2145,
https://doi.org/10.1007/978-1-0716-0599-8_1, © Springer Science+Business Media, LLC, part of Springer Nature 2020
1
2 Richard M. Nagymihaly et al.
Fig. 1 Corneal cross-sectional structure. Arrows mark the nuclei of keratocytes embedded between the layers
of collagen in the corneal stroma (hematoxylin and eosin staining)
Cells isolated from the corneal stroma are known to develop a

fibroblastic/myofibroblastic phenotype in vitro, when cultured in
the presence of serum as a growth supplement [5–8]. Moreover,
these cultured corneal stroma-derived cells have been attributed
mesenchymal-stem cell like properties, such as trilineage differenti-
ation potential into bone, cartilage, and adipose tissue, as well as
immunomodulatory/immunosuppressive potential [9]. Early
communications by research groups identified precursors of the
corneal epithelium, the limbal stem cells as the multipotent stem
cells of the stroma; however later, as more evidence and results were
generated, it became apparent that in culture, the two cell types
represent distinct cell groups, with distinct functions and potential
[9, 10]. The relative ease of access to tissue and having such unique
characteristics helped the corneal stroma-derived stem cells
(CSSCs) attain a reputation in tissue engineering, regenerative
medicine, and cellular therapy and become a popular target for
corneal research [10–13].
Research groups worldwide use the explant or enzymatic
method to obtain the cells from postmortem whole corneas graded
unsuitable for corneal transplantation or discarded material after
keratoplasties. The CSSCs are now a well-studied phenomenon,
and in this chapter, we are going to discuss the means of extraction
and expansion to establish primary cell cultures from human cor-
neal postmortem tissue.
Corneal Stroma-Derived Stem Cells 3
2 Materials
The equipment and solutions listed here are required to carry out
the work detailed in this chapter:
2.1 Lab Equipment 1. Cell culture Petri dish (vented, with lid) 60 15 mm, sterile.
and Instruments 2. Cell culture Petri dish (vented, with lid) 110 17 mm, sterile.
3. Cell strainer, 70 μm, individually packed, sterile.
4. Conjunctival scissors (curved).
5. Corneal Trephine Blade, 8 mm.
6. Corneal Trephine Blade, 18 mm.
7. CO2 incubator.
8. Cryogenic tube with cap, 1.8 mL, self-standing, sterile.
9. Electronic pipette controller and serological pipette tips (5, 10,
25 mL).
10. Filter unit, 0.2 μm, individually packed, sterile.
11. Hettich Universal 30F centrifuge (4 150 g).
12. Laminar flow cabinet.
13. Liquid nitrogen tank.
14. Nalgene 5100-0001 PC/HDPE Mr. Frosty Cryogenic Freez-
ing Container.
15. Precisa Balance XT 120 A scale.
16. Rocking shaker, 2D rocking.
17. Sterile, individually packed disposable scalpels.
18. Sterile, individually packed 12- or 24-well non-treated cell
culturing plates.
19. Sterile, individually packed 60 mL enteral syringes.
20. Sterile water, bottled 1 L.
21. Sterilized, metallic Iris forceps (preferably non-toothed).
22. Ultralow temperature (ULT) freezer (~ 80 C).
23. Upright brightfield microscope.
24. Water bath, capable of warming to 37 C.
25. 1 mL micropipette and suitable pipette tips (100–1000 μL).
26. 15 mL Centrifuge tubes, conical bottom, with screw cap,
sterile.
27. 25 cm2 Cell culture flasks, cell culture treated, sterile.
28. 50 mL Centrifuge tubes, screw cap, sterile.
2.2 Solutions 1. 5% Betadine povidone-iodine ophthalmic solution.

2. Dulbecco’s Modified Eagle Medium (DMEM) Low glucose
with Glutamax™: 1 g/L D-glucose, 110 mg/L sodium pyru-
vate, phenol red, sterile.
3. Antibiotic Antimycotic Solution (100): 10,000 units/mL
penicillin, 10 mg/mL streptomycin, 25 μg/mL amphotericin
B. Sterile-filtered using 0.22 μm filter unit and syringe.
4. Dulbecco’s Phosphate Buffered Saline (DPBS) 1: Calcium
chloride and magnesium chloride free, sterile.
5. Culture medium: DMEM, 10% (v/v%) fetal bovine serum
(FBS) cell culture tested and heat inactivated, 1% (v/v%) anti-
biotic antimycotic solution (100) (see Note 1). Use a 0.22 μm
filter unit and a 60 mL syringe to sterile-filter the culture
medium aliquots in clean, sterile 50 mL tubes. Store in
50 mL aliquots at 4 C for up to a maximum of 3 weeks.
6. Collagenase solution: 0.5 mg/mL collagenase from clostrid-
ium histolyticum (type I, 0.25–1.0 FALGPA unit/mg solid,
125 CDU/mg solid) in serum-free medium.
7. Freezing medium: 10% (v/v%) dimethyl sulfoxide (DMSO),
90% (v/v%) culture medium. Always prepare fresh before
using.
8. Trypan Blue, sterile-filtered.
9. FNC Coating Mix®: 10 μg/mL fibronectin, 35 μg/mL colla-
gen type 1, 1 mg/mL bovine serum albumin (BSA), 200 μg/
mL potassium chloride, 1 μg/mL phenol red, 1.7 mg/mL
D-glucose, 4.8 mg/mL HEPES buffer, 7 mg/mL sodium
chloride, 1.7 mg/mL sodium phosphate.
3 Methods
All procedures detailed here are in accordance with the directives of

the Helsinki Declaration and all tissue harvesting was approved by
the Regional Committee for Medical and Health Research Ethics in
Norway.
3.1 Preparations Corneal tissue can be obtained from postmortem whole corneas
graded unsuitable for corneal transplantation or discarded material
after keratoplasties. Prior to enucleation, eyes are disinfected with
5% Betadine at the local Department of Pathology, then transferred
to the ophthalmology department and processed for eye banking.
Briefly, eyes are rinsed in sterile water and an 18 mm trephine is
centered around the cornea and a groove is cut into the sclera and
incised in a laminar flow cabinet. Conjunctival scissors are used to
cut around the corneal disc. Adhering tissue from the iris is
Fig. 2 Corneal button with the scleral collar, obtained after a DMEK surgery in a
Petri dish. Note the suture placed in at the 12 o’clock position
removed using forceps, and then the corneal button is placed in

sterile saline. A suture is usually placed in the sclera to mark the
12 o’ clock position. Following assessment in the microscope and
scoring the corneal endothelium, the tissue is placed in organ
culture medium until transplantation in a 31 C incubator. The
leftover corneal tissue is most commonly obtained after Descemet’s
membrane endothelial keratoplasty (DMEK) and are received
endothelium-free by the research laboratory. When the intact cor-
neal button is collected, the endothelium is removed from the
posterior stroma using an 8 mm trephine and peeled away by
forceps under a stereomicroscope. In the following section, two
methods to isolate CSSCs are described: (1) an explanted tissue
technique and (2) a technique using enzymatic digestion.
3.2 Isolation All steps are performed in a laminar flow hood (see Note 2)
of CSSCs (Explant
1. Remove the corneal button from the storage or transportation
Technique)
medium and place it in a sterile 110 17 mm Petri dish with
2–3 mL serum-free culturing medium (see Note 3) (Fig. 2).
Keep both sides of the tissue hydrated.
2. It is crucial to remove the corneal epithelium before the rest of
the steps are performed. A clean, sterile scalpel is the best
option to use for this (see Note 5). Grab the tissue using an
Iris forceps and scrape away the epithelium. Repeat the scraping
from other angles. Rinse off the floating epithelium pieces with
more medium, and then move the corneal button to a clean
Petri dish.
Fig. 3 Images showing the removal of the sclera (a) and the leftover corneal stroma, devoid of the epithelium
and endothelium (b)
Fig. 4 Pieces of stroma in a 12-well plate with culturing medium
3. Cut around and remove the scleral collar using surgical scissors
(Fig. 3a). At this point the “naked” stroma is ready for cell
isolation (Fig. 3b).
4. Hold the tissue on one side with the forceps and use the blade
to chop the stroma into pieces of approximately 2 2 mm.
Add 1–2 mL serum-free medium to the Petri dish (see Note 3)
to prevent the tissue from desiccation (see Note 4).
5. Pick up the tissue pieces using forceps and place 1–2 pieces per
well in a 12-well plate (see Note 6), as depicted in Fig. 4. Make
sure to prevent the tissue from drying out by adding a few
drops of culture medium in the wells (see Note 4).
6. When all the pieces have been placed in the wells, add
200–250 μL culture medium to the plate to barely cover the
tissue explants (see Note 7), and then place the 12-well plate in
the incubator. Do not disturb the plate for the next 24 h.
7. After the first day, carefully add pre-warmed culture medium
up to 1–1.5 mL to the wells with the tissue pieces (see Note 4).
Exercise caution not to touch or lift the attached tissue. Cul-
ture cells by refreshing the medium every 3–4 days. Observe
Fig. 5 Corneal stroma explant (∗) in the top right corner of the image is shown (a), with a dense cell layer
around it (b), at 4 and 10 magnification, respectively
growth of adherent elongated or spindle-shaped cells from

attached tissue bits around days 6–7 under the microscope
(Fig. 5) (see Note 8).
3.3 Isolation Steps 1, 2, 3, and 6 are performed in a laminar flow hood.

of CSSCs (Collagenase
1. Remove the corneal button from the organ culture bottle or
Digestion Technique)
other storage container and clear away the corneal endothelium
and epithelium, followed by the scleral collar as already
described.
2. Take a sterile 60 15 mm Petri dish and add 1 mL of the
0.5 mg/mL collagenase solution. Add the endothelium- and
epithelium-free corneal stromal tissue to the dish and cut it into
four quadrants using the surgical blade and the forceps. Use the
scalpel to chop up the four quadrants into a mince, similarly as
seen in Fig. 6. When done, pipette 4 mL of collagenase solution
into the Petri dish, then cover and label accordingly.
3. Decontaminate a 2D rocker/shaker device thoroughly using
70% ethanol solution. Place the device into the incubator along
with the Petri dish containing the corneal tissue in collagenase
solution, as shown in Fig. 7. Turn on the rocker device and set
it to gentle rocking. Close the door of the incubator and
incubate overnight at 37 C (see Note 9).
4. Take out the Petri dish from the incubator. Observe the sus-
pension/solution looking for leftover tissue pieces. If neces-
sary, the undigested parts can be exposed to a higher
collagenase concentration (see Note 10). Transfer the digest
to a clean 15 mL tube, through a strainer with at least 70 μm
pore size (see Note 11), as shown in Fig. 8. Include additional
culture medium to neutralize enzymatic activity (see Note 4).
Centrifuge the tube at 100 g for 10 min.
Fig. 6 Image showing minced corneal tissue in a plastic Petri dish
Fig. 7 2D Rocker/shaker in the 37 C incubator with the Petri dish containing the
minced corneal tissue in a collagenase solution
Fig. 8 Image shows straining of the digest of stromal tissue

Fig. 9 Colonies of cells obtained by digestion of the corneal stroma at 4 and 10 magnifications,
respectively
5. Observe the cell pellet and aspirate the supernatant carefully,

using a micropipette. Gently triturate pellet in 1 mL
pre-warmed culture medium. The cells are now ready for seed-
ing. Fill a 25 cm2 flask with 4 mL pre-warmed culture medium
(see Note 6). Use a 1 mL micropipette to seed cells in the
25 cm2 culture dish and then place it in the incubator. Do
not disturb the plate for the next 24 h.
6. Refresh culture medium (pre-warmed) between 24 and 48 h
after seeding. Observe colonies of adherent cells under the
microscope (Fig. 9).
3.4 Sub-culturing Steps 2, 4, and 6 are performed in a laminar flow hood.

CSSCs
1. Observe and estimate the confluency of cells under an upright
microscope. When the cells reach 60–80% confluency (see Note
12) in the culture flask, they may be transferred to a larger
container. Place aliquots of culture medium and DPBS in a
37 C water bath 20 min before commencing the transfer
process (see Note 4).
2. Aspirate and discard culture medium using an electronic
pipette controller (motorized pipette) and a sterile serological
pipette. Add 0.2 mL pre-warmed DPBS per cm2 culture sur-
face area (see Note 3) to the cells and wash for 10–15 s. Discard
the buffer and add enough TrypLE (see Note 13) to cover the
surface of the cell culture dish, usually 0.02 mL/cm2, and place
the flask into the incubator at 37 C.
3. Incubate the flask for 2–3 min and observe cell detachment
under the upright microscope. Tap the container gently to
assist cell detachment. Add culture medium, after cells sepa-
rated from the plastic surface, to neutralize the enzymatic
effect.
Fig. 10 Counting CSSCs in a Bürker chamber
4. Collect cell suspension using the pipette controller into a ster-

ile, 15 mL centrifuged tube through a 70 μm cell strainer to
filter out protein clots or extracellular matrix elements. Centri-
fuge at 100 g for 10 min at room temperature.
5. Observe the cell pellet and decant or aspirate the supernatant.
Resuspend and carefully triturate cells in 1–2 mL culture
medium. Determine cell number and viability in a Bürker
chamber and staining by trypan blue, by mixing 50 μL cell
suspension with 50 μL stain (see Note 14) (Fig. 10).
6. Fill a clean cell culture dish of the desired size with the recom-
mended volume of pre-warmed culture medium (see Note 15).
Use culture medium to adjust cell density of the suspension to
achieve at least 2000–5000 cells/cm2 coverage of the cultur-
ing surface area. Seed cells in the new container, close the lid,
and incubate at 37 C. CSSCs take between 4–6 h to attach.
Refresh medium every 3–4 days (Fig. 11).
3.5 Freezing CSSC All steps are performed in a laminar flow hood.
Aliquots
1. Observe cell density in the desired culture dish under the
microscope. Upon reaching 60–80% confluency, cells may be
stored frozen.
2. Repeat steps 2 and 3 described in Subheading 3.4
(Sub-culturing CSSCs) to detach the cells from the culture
surface.
3. Adjust cell density to approximately 1,000,000 cells per mL
medium suspension. Divide the cell suspension into aliquots
that contain 500,000 to 1,000,000 cells in 15 mL centrifuge
tubes (see Note 16) and centrifuge at 150 g for 5 min at
room temperature.
4. Decant the supernatant and resuspend cells in 0.5 mL to 1 mL
freezing medium. Make sure to triturate cell pellet thoroughly
but gently. Transfer an aliquot of cells in freezing medium to a
Fig. 11 Refreshing the culture medium
Fig. 12 Freezing CSSC aliquots
freezing tube (Fig. 12). Repeat the process with all aliquots of
cells (see Note 17). Place all tubes in a freezing container.
Transfer the freezing container to an ultra-freezer ( 86 C),
for overnight cooling, as soon as possible.
5. The next day, remove the freezing tubes from the container and
transfer them to a liquid nitrogen tank. Frozen cells can be
safely stored in liquid nitrogen for extended time [14].
3.6 Thawing CSSC Steps 3 and 4 are to be performed in a laminar flow hood.
Aliquots
1. Pre-warm medium in a water bath for at least 20 min before
thawing cells. Fill the desired cell culture flask(s) with the
pre-warmed medium.
2. Remove cell aliquot(s) from liquid nitrogen and thaw by hand-
holding tubes for 1–2 min in the water bath at 37 C.
3. Open the lid of the freezing tube and carefully triturate the cell
suspension using a 1 mL micropipette. Adjust cell number to
the desired density with culture medium before seeding in the
flasks with the pre-warmed medium to achieve coverage of
approximately 2000–5000 cells/cm2 culturing surface (see
Note 18). Placed cell culture flask(s) in the incubator.
4. The next day, observe cell attachment under the microscope
and refresh culture medium to remove any remaining DMSO
from the culture. Cells may now be used for any type of
investigation.
4 Notes
1. DMEM with low glucose content is the most commonly used

culture medium for CSSCs; although other types of media have
been used [6, 15]. Generally, 10% FBS is added to the medium;
however, different research laboratories use various concentra-
tions ranging from 2% to 20% FBS. Notably, a few studies have
used animal-free medium to successfully expand the cells; how-
ever, without serum in the culture medium the cells fail to
exhibit mesenchymal stem cell (MSC) like characteristics and
adopt a more keratocyte-like resting state [7, 16–18]. Cells
require a source of glutamine and this essential amino acid
should be added to the culture medium at a 20 mM final
concentration, if it has not already been incorporated in the
basal medium.
2. For all steps performed in a laminar flow hood, disinfect sur-
faces using a 70% ethanol solution, or equivalent surface disin-
fectant solution.
3. Other sterile, physiological buffers and solutions will suffice,
such as HBSS.
4. Make sure to warm the culture medium aliquots and other
washing buffers in a 37 C water bath before applying it to
the cells.
5. Alternatively, or complementary to the scraping, the corneal
epithelium can be removed using digestion in dispase.
6. Coating the surface of the cell culture dish is an optional step.
The authors did not observe decreased cell attachment, when
Fig. 13 Epithelial contamination in a culture with explanted stromal tissue
coating was omitted. Alternatively, a coating solution, contain-

ing fibronectin and collagen I, may be used. For details, refer to
the Materials section.
7. It is not necessary to cover the surface of the wells with any type
of attachment agent; however, it is encouraged to add enough
culturing medium to barely cover the tissue bits. More liquid
will float the tissue and prevent it from attachment. This may
reduce the success of establishing cultures. Supplementary
medium may be added following the first 24 h.
8. When establishing explant cultures using corneal tissue from
rims, occasionally cells displaying epithelial morphology may
emerge in the culture wells at passage 0 (Fig. 13). Possibly,
these contaminating cells originate from limbal epithelial cell
niches, located peripherally in the basal epithelial layer. Wells or
cultures contaminated by epithelial cells should not be used to
enrich CSSC populations.
9. Alternatively, a 50 mL tube may be used, if a proper centrifuge
and a necessary setup are available.
10. It is possible to use a stronger collagenase solution, up to
1.5–2 mg/mL, for a shorter incubation time (4–6 h). The
rocking platform speeds up the digestion process; although
when not available, overnight digestion is encouraged. Success
of the digestion process may be assessed macroscopically. When
leftover, undigested tissue pieces are present in the solution/
suspension, a stronger concentration of 2 mg/mL collagenase
may be used for 1–2 h after the overnight incubation to
completely digest the tissue and maximize cell yield.
11. The authors’ personal experience is that when isolating cells
from corneal rims or from corneas stored in organ culture for
extended times (>3 weeks), the cell yield is lower, while
straining the digest may cause additional cell loss and should be
avoided.
12. Mesenchymal stem cell cultures should be prevented from
reaching high confluency, as cells in overgrown cultures tend
to transdifferentiate and lose the MSC phenotype.
13. Alternatively, trypsin can be used; however, the enzyme is
known for its harshness on the cell membrane. Notably, the
authors discovered no adverse effects on cell morphology,
when exposure to trypsin was used to detach the cells for a
maximum of 5 minutes.
14. When determining the cell count, remember to double the
number, to correct for the dilution with trypan blue. Blue
cells in the chamber are dead and should be excluded from
the calculations.
15. Refer to the manufacturer of the culture dish. Generally,
0.2 mL medium per cm2 of culturing surface is sufficient.
16. Alternatively, the whole container (well or flask) of cells can be
frozen as one aliquot; however, freezing cells in extreme high
numbers is advised against.
17. Make sure that everything is prepared before suspending the
cells in freezing medium. DMSO is necessary to ensure that ice
crystal formation in the cells is blocked; however, the substance
is toxic. In order to minimize the toxic effect, work effectively
and quickly. Do not expose the cells to freezing medium at
room temperature for extended periods of time (>10 min).
18. Seeding cell density may vary depending on the application or
intended use.
Acknowledgements
The work presented was supported by the South-Eastern Norway

Regional Health Authority and by the Oslo University Hospital,
Norway.
References
1. Forrester J, Dick AD, McMenamin PG, 49:17–45. https://doi.org/10.1016/j.pre

Roberts F, Pearlman E (2016) Chapter 1- teyeres.2015.07.002
anatomy of the eye and orbit. In: Forrester J, 3. Wilson SE, Chaurasia SS, Medeiros FW (2007)
Dick AD, McMenamin PG, Roberts F, Pearl- Apoptosis in the initiation, modulation and
man E (eds) The eye, 4th edn. Elsevier, termination of the corneal wound healing
Amsterdam, Pages 1–102.e2. https://www. response. Exp Eye Res 85(3):305–311.
sciencedirect.com/science/article/pii/ https://doi.org/10.1016/j.exer.2007.06.009
B9780702055546000010 4. Kureshi AK, Funderburgh JL, Daniels JT
2. Ljubimov AV, Saghizadeh M (2015) Progress (2014) Human corneal stromal stem cells
in corneal wound healing. Prog Retin Eye Res exhibit survival capacity following isolation
from stored organ-culture corneas. Invest 12. Szabo DJ, Nagymihaly R, Vereb Z,
Ophthalmol Vis Sci 55(11):7583–7588. Josifovska N, Noer A, Liskova P, Facsko A,
https://doi.org/10.1167/iovs.14-14448 Moe MC, Petrovski G (2018) Ex vivo 3d
5. Funderburgh JL, Mann MM, Funderburgh human corneal stroma model for schnyder cor-
ML (2003) Keratocyte phenotype mediates neal dystrophy - role of autophagy in its patho-
proteoglycan structure: a role for fibroblasts genesis and resolution. Histol Histopathol 33
in corneal fibrosis. J Biol Chem 278 (5):455–462. https://doi.org/10.14670/
(46):45629–45637. https://doi.org/10. HH-11-928
1074/jbc.M303292200 13. Shojaati G, Khandaker I, Sylakowski K, Fun-
6. Sidney LE, Branch MJ, Dua HS, Hopkinson A derburgh ML, Du Y, Funderburgh JL (2018)
(2015) Effect of culture medium on propaga- Compressed collagen enhances stem cell ther-
tion and phenotype of corneal stroma-derived apy for corneal scarring. Stem Cells Transl Med
stem cells. Cytotherapy 17(12):1706–1722. 7(6):487–494. https://doi.org/10.1002/
https://doi.org/10.1016/j.jcyt.2015.08.003 sctm.17-0258
7. Lynch AP, O’Sullivan F, Ahearne M (2016) 14. Kumar A, Xu Y, Yang E, Du Y (2018) Stemness
The effect of growth factor supplementation and regenerative potential of corneal stromal
on corneal stromal cell phenotype in vitro stem cells and their secretome after long-term
using a serum-free media. Exp Eye Res storage: Implications for ocular regeneration.
151:26–37. https://doi.org/10.1016/j.exer. Invest Ophthalmol Vis Sci 59(8):3728–3738.
2016.07.015 https://doi.org/10.1167/iovs.18-23824
8. Nagymihaly R, Vereb Z, Facsko A, Moe MC, 15. Matthyssen S, Ni Dhubhghaill S, Van
Petrovski G (2017) Effect of isolation tech- Gerwen V, Zakaria N (2017) Xeno-free culti-
nique and location on the phenotype of vation of mesenchymal stem cells from the cor-
human corneal stroma-derived cells. Stem neal stroma. Invest Ophthalmol Vis Sci 58
Cells Int 2017:9275248. https://doi.org/10. (5):2659–2665. https://doi.org/10.1167/
1155/2017/9275248 iovs.17-21676
9. Vereb Z, Poliska S, Albert R, Olstad OK, 16. Wilson SL, Yang Y, El Haj AJ (2014) Corneal
Boratko A, Csortos C, Moe MC, Facsko A, stromal cell plasticity: In vitro regulation of cell
Petrovski G (2016) Role of human corneal phenotype through cell-cell interactions in a
stroma-derived mesenchymal-like stem cells in three-dimensional model. Tissue Eng Part A
corneal immunity and wound healing. Sci Rep 20(1–2):225–238. https://doi.org/10.1089/
6:26227. https://doi.org/10.1038/ ten.TEA.2013.0167
srep26227 17. Sidney LE, Hopkinson A (2018) Corneal ker-
10. Szabo DJ, Noer A, Nagymihaly R, atocyte transition to mesenchymal stem cell
Josifovska N, Andjelic S, Vereb Z, Facsko A, phenotype and reversal using serum-free
Moe MC, Petrovski G (2015) Long-term cul- medium supplemented with fibroblast growth
tures of human cornea limbal explants form 3d factor-2, transforming growth factor-beta3 and
structures ex vivo - implications for tissue engi- retinoic acid. J Tissue Eng Regen Med 12(1):
neering and clinical applications. PLoS One 10 e203–e215. https://doi.org/10.1002/term.
(11):e0143053. https://doi.org/10.1371/ 2316
journal.pone.0143053 18. Foster JW, Gouveia RM, Connon CJ (2015)
11. Ahearne M, Wilson SL, Liu KK, Rauz S, El Haj Low-glucose enhances keratocyte-
AJ, Yang Y (2010) Influence of cell and colla- characteristic phenotype from corneal stromal
gen concentration on the cell-matrix mechani- cells in serum-free conditions. Sci Rep
cal relationship in a corneal stroma wound 5:10839. https://doi.org/10.1038/
healing model. Exp Eye Res 91(5):584–591. srep10839
https://doi.org/10.1016/j.exer.2010.07.013
Chapter 2
In Vitro Expansion of Corneal Endothelial Cells

for Transplantation
Kim Santerre, Isabelle Xu, Mathieu Thériault, and Stéphanie Proulx
Abstract
The corneal endothelium forms a leaky barrier between the corneal stroma and the aqueous humor of the
anterior chamber. This cell monolayer maintains the corneal stroma in a state of relative dehydration, a
process called deturgescence, which is required in order to obtain corneal stromal transparency. Endothelial
dysfunctions lead to visual impairment that ultimately can only be treated surgically via the corneal
transplantation of a functional endothelium. Shortages of corneas suitable for transplantation has motivated
research toward new alternatives involving in vitro corneal endothelial cell (CEC) expansion.
This chapter describes current methods that allow isolate and culture CECs. In brief, Descemet mem-
brane is peeled out of the cornea and digested in order to obtain CECs. Cells are then seeded and cultured.
Key words Cornea, Corneal endothelium, Cell culture, Cell expansion
1 Introduction
The corneal endothelium consists of a monolayer of corneal endo-

thelial cells (CECs), which lie on a thick basal membrane named the
Descemet’s membrane. It is located at the posterior side of the
cornea and faces the aqueous humor of the anterior chamber. CECs
play a key role in maintaining the corneal stroma partially dehy-
drated, which is required for corneal transparency. This is achieved
through a “pump/leak balance,” where the leaky endothelial bar-
rier prevents the cornea from swelling by slowing the intake of
liquid into the stroma and pumping it back to the anterior chamber
[1]. CECs do not proliferate in vivo. This is thought to be caused
by a strong contact inhibition mechanism and the presence of
transforming growth factor (TGF)-β in the aqueous humor,
which maintain a strong expression of the cyclin-dependent kinase
inhibitor p27 within the cells [2, 3]. Upon damage to the corneal
endothelium, the remaining cells will migrate and enlarge in order
17
18 Kim Santerre et al.
to cover the damaged zone and reform the endothelial barrier

[4]. Endothelial cell density in healthy adult ranges from 2500 to
3000 cells/mm2 [5]. The endothelium remains however functional
down to a breaking point of 400–700 cells/mm2 [5]. A dysfunc-
tional endothelium will result in corneal edema, serious vision
impairments, and even blindness [6].
Endothelium keratoplasty (EK) is currently the only available
treatment for endothelial dysfunctions or trauma. Both Descemet
stripping automated endothelial keratoplasty (DSEAK) and Desce-
met membrane endothelial keratoplasty (DMEK) can be per-
formed with cadaveric tissues suitable for transplantation [7]. In
North America, approximately 25,000 EKs are performed each
year [8]. With a growing and aging population, this number will
inevitably increase the demand for the limited available corneas
eligible for transplantation. New alternatives were therefore needed
to address these limitations. A possible solution would be to
increase the treating potential of each corneal tissue by isolating
and expanding the number of cells. At present, investigated strate-
gies include the creation of tissue-engineered corneal endothelium
and the intracameral injection of a cell suspension [9, 10]. These
promising alternatives to conventional corneal transplantation both
require cell expansion.
This chapter details the methodology and the materials used to
successfully isolate CECs with the peel-and-digest method and to
expand CECs in vitro, including media changes, cell passages, cell
count, cryopreservation, and cell thawing.
2 Materials
2.1 Labware 1. Filtration unit: The filtration unit is mounted with a 47 mm

diameter and 0.22 μm filter set.
2. 15 mL and 50 mL tubes.
3. Low-binding 0.22-μm disposable filter.
4. 1.5 mL tube. Sterilize by autoclaving.
5. 60 15 mm Tissue culture dish.
6. Sterile gauze.
7. 22-scalpel blade.
8. Dissecting curved scissor.
9. Parafilm.
10. Dissecting curved forceps.
11. Dissecting straight forceps.
In Vitro Expansion of Corneal Endothelial Cells 19
12. Plastic culture dishes: 24-well culture plate with 1.9 cm2 sur-
face, 12-well culture plate with 3.8 cm2 surface, 6-well culture
plate with 9.5 cm2 surface, 25 cm2 culture flask.
13. Flint glass disposable Pasteur pipets. Sterilize by autoclaving.
14. Hemocytometer.
15. Glass coverslips.
16. Hand tally counter.
17. Nalgene freezing container.
18. Cryogenic vials.
19. Sterile plastic pipette: 1 mL pipette, 2 mL pipette, 5 mL
pipette, 10 mL pipette, and 25 mL pipette.
2.2 Culture Media 1. Phosphate buffered saline (PBS): 127 mM NaCl, 2.7 mM KCl,
6.5 mM Na2HPO4, 1.5 mM KH2PO4. Yields a 10 stock
2.2.1 Solutions
solution. Store at room temperature. Working solution is
done by mixing 10 PBS with pyrogenic ultrapure water and
sterilized by filtration through a filtration unit.
2. Collagenase A: Collagenase A (0.15 units/mg) is dissolved in
expansion medium or sterile 1 PBS at a final concentration of
1 mg/mL (0.15 units/mL). Solution is sterilized by filtration
through a 0.22 μm low-binding disposable filter. Collagenase A
must be used within a day.
3. Trypsin/EDTA: Trypsin/EDTA 0.05%-0.53 mM is aliquoted
in sterile 15 mL tubes and stored at 20 C (long-term) or
4 C (short-term) (see Note 1).
4. Dimethyl sulfoxide (DMSO).
5. FNC-coating mix.
6. Trypan blue solution 0.4%: Aliquot in 1.5 mL tubes.
7. HCl 0.01 N.
8. Anhydrous ethyl alcohol (ethanol) 99%.
2.2.2 Base Media 1. DMEM: Dulbecco’s Modified Eagle Minimal Essential

Medium. Store at 4 C.
2. Opti-MEM: Complete with 0.20 g/L calcium chloride dehy-
drate and 2.4/L sodium bicarbonate. Adjust pH to 7.1. Steril-
ize through a filtration unit. Store at 4 C.
2.2.3 Additives 1. Epidermal growth factor (EGF): Dissolve one vial (100 μg) of
EGF in 2 mL of HCl 0.01 N. Add the mix to 38 mL of
expansion medium. Sterilize by filtration through a 0.22 μm
low-binding disposable filter. Aliquot in 1.5 sterile tubes and
store at 20 C (see Note 1). Thaw at room temperature.
2. Penicillin/Streptomycin: Aliquot penicillin/streptomycin

100 in sterile 15 mL tubes and store at 20 C (long-term
preservation) or 4 C (short-term preservation) (see Note 1).
Aliquots are thawed at room temperature or in 37 C water
bath prior to their use.
3. Chondroitin sulfate: Dissolve chondroitin sulfate A sodium salt
from bovine trachea in expansion medium to yield a 1.6%
solution. Sterilize through a filtration unit. Aliquot in 15 mL
tubes. Store at 20 C (long-term preservation) or 4 C
(short-term preservation) (see Note 1). Thaw at room temper-
ature or in a 37 C water bath.
4. Ascorbic acid: Dissolve 2 mg/mL of L-ascorbic acid in expan-
sion medium. Sterilize by filtration through a 0.22 μm
low-binding disposable filter. Store at 4 C and use within
a day.
5. Fetal bovine serum (FBS): Thaw in cold water (1–2 days).
Inactivate in hot water (56 C) for 30 min. Distribute in
single use aliquots (50 mL tubes) and store at 20 C
(see Note 1).
2.2.4 Complete Media 1. Extraction medium: DMEM supplemented with 10% FBS and
penicillin/streptomycin. Store at 4 C.
2. Expansion medium: Opti-MEM supplemented with 8% FBS,
5 ng/mL EGF, penicillin/streptomycin (working concentra-
tion: 1). Store at 4 C.
3. Complete expansion medium: Expansion medium completed
with 20 μg/mL ascorbic acid and 0.08% chondroitin sulfate.
Complete expansion medium must be used within a day.
4. Freezing medium: Freezing medium is composed of 10%
DMSO in FBS. Freezing medium is kept on ice or stored at
4 C. Freezing medium must be used within a day.
5. Optisol-GS.
2.3 Donor Tissues 1. Human globes or corneas unsuitable for transplantation

(see Note 2) are normally provided by an Eye Bank. In our
work, tissues are provided by the Centre Universitaire d’Oph-
talmologie (CUO) Eye Bank, Hôpital Saint-Sacrement,
Québec, Canada. Globes are kept at 4 C until dissection.
Corneas are kept in Optisol-GS or Extraction medium at
4 C until Descemet membrane peeling.
3 Methods
3.1 Corneal 1. Surround the eye specimen with a folded sterile gauze to avoid
Endothelial Cells touching it (Fig. 1a) and place it in a 100 mm sterile Petri dish.
Isolation 2. Make a first opening in the sclera with a 22-scalpel blade
3.1.1 Globe Dissection (Fig. 1b). Use a curved scissor to carve out the cornea (Fig. 1c).
3. Remove the iris and the lens (Fig. 1d).
4. Store the cornea in extraction medium endothelial side facing-
up in a 60 mm Petri dish (Fig. 1e) and seal it with parafilm until
Descemet peeling. Corneas must be used within a few days and
kept at 4 C.
3.1.2 Descemet 1. Place the cornea with the endothelial side facing up in a 60 mm
Membrane Peeling Petri dish, with enough extraction medium to cover the cornea
(Fig. 1e). Under a binocular microscope (see Note 3), use
dissecting curved forceps to maintain the cornea in place
(Fig. 1f).
2. Make a scratch at the periphery of Descemet membrane and
gently unroll strips from the stroma using dissecting straight
forceps (Fig. 1g, h).
3. Add the strips to 11 mL of expansion medium and incubate at
37 C overnight.
Fig. 1 Steps from eye globe dissection to Descemet membrane peeling. (a–d) Macroscopic images showing
the successive steps of corneal removal from a human eye globe; (e, f) Macroscopic images of a human
cornea with the endothelium facing up; (g, h) Stereomicroscope images showing the peeling of Descemet’s
membrane
3.1.3 Collagenase A CECs isolated with Collagenase A can form aggregates. To gener-
Digestion ate single cells, additional steps can be performed (see Note 4).
Alternatively, isolation can be done using EDTA (see Note 5).
1. Centrifuge Descemet membrane strips at 300 g for 10 min
and remove supernatant.
2. Incubate Descemet membrane strips in 1 mL of collagenase A
1 mg/mL for 2–4 h at 37 C.
3. Centrifuge digested Descemet membrane strips at 300 g
10 min and remove supernatant.
4. Dissolve cell pellet in complete expansion medium (see Note
4).
5. Seed CECs on a FNC-coated plastic culture dish (see Notes 6
and 7).
3.2 Cell Expansion Cell morphology (fibroblast-like or endothelial morphology;

Fig. 2) and confluence are assessed under a phase contrast micro-
scope (see Note 8).
3.2.1 Medium Change Medium is changed every 2–3 days (see Note 9).
1. Warm complete expansion medium in a 37 C water bath.
2. Remove cell medium with a sterile glass pipette branched to a
vacuum.
3. Add fresh warmed complete expansion medium to cells with a
sterile plastic pipette. Volume added will vary depending on
culture dish used.
3.2.2 CECs Passages Prior to cell passages, trypsin/EDTA must be warmed in a 37 C

(Subculture) water bath.
Fig. 2 Phase contrast images of cultured corneal endothelial cells. CECs with an endothelial, a fibroblastic, or a
mixed morphology are shown. Scale bar ¼ 200 μm
1. Remove medium from confluent or nearly confluent CECs

(see Note 10) with a sterile glass pipette.
2. Rinse CECs with warm trypsin/EDTA or PBS.
3. Add trypsin/EDTA to cells. Monitor cell detachment every
5 min under a phase contrast microscope.
4. Add expansion medium or FBS when CECs are detached to
neutralized trypsin/EDTA. Do up-and-downs to remove cells
that are still present on the plastic dish.
5. Collect cells with a sterile plastic pipette and put them in a
15 mL tube. Rinse plastic dish with expansion medium and add
it to the 15 mL tube.
6. Centrifuge the tube at 300 g for 10 min and remove
supernatant.
7. Resuspend cells in 1 mL of warm complete expansion medium.
Take an aliquot (50 μL) for cell counting (see Subheading 3.3).
8. Seed cells on plastic culture dishes at a density of 20,000 cells/
cm2.
3.3 Cell Count This step can be performed under non-sterile conditions. Prior to
its use, clean the hemocytometer and glass coverslip with ethanol
99%, rinse with water and dry (see Note 11).
1. Mix 1:1 cells with trypan blue colorant (see Note 12).
2. Fill both chambers of the hemocytometer underneath the cov-
erslip, allowing cell suspension to be drawn out by capillary
action.
3. Place the hemocytomer under a phase contrast microscope.
Focus on the grid lines with a 10 objective.
4. Count cells manually with a hand tally counter from five
squares for each chamber. Viable cells are unstained while
dead cells are blue.
5. Using the following formula, calculate cell concentration
(cells/mL). Here the dilution factor is 2 and the surface equals
to 104:
Total cell counted

Cell concentration ¼ Dilution factor
Number of squares
Surface
6. Viability can be assessed with the formula:
Live cells
%viability ¼ 100
Live cells þ Dead cells
3.4 Cryopreservation Nalgene freezing container must be filled with 99% ethanol and
store at 20 C prior its use.
1. Follow steps 1–7 from Subheading 3.2.2.
2. Centrifuge tubes and discard supernatant.
3. Resuspend cells at the desired concentration in freezing
medium and put the tube on ice (see Note 13).
4. Aliquot in cryogenic vials on ice and place the vials in the
Nalgene freezing container.
5. Store the Nalgene freezing container at 80 C overnight.
6. Transfer the cryogenic vials in liquid nitrogen for long-term
preservation.
3.5 Cells Thawing 1. Place cryogenic tubes in a 37 C water bath. Do not let the cell
suspension completely thaw; a small ice pellet has to remain.
2. Add the cells to 9 mL of cold expansion medium.
3. Centrifuge tubes and remove supernatant.
4. Dissolve cell pellet in 1 mL of complete expansion medium.
Take an aliquot (50 μl) for cell counting and assessing viability
(see Subheading 3.3).
5. Seed cells on plastic culture dish at a density of 20,000 cells/
cm2.
4 Notes
1. Avoid freeze-thaw cycle that could damage serum or additives.

2. Protocols need to be approved by the institution’s committee
for the protection of human subjects. Young donors and short
postmortem delays usually yield better CECs cultures
3. Descemet peeling must be done under sterile conditions next
to a Bunsen burner to avoid bacterial contamination. Forceps
are sterilized after each dissection.
4. Additional steps for CEC isolation can be performed to sepa-
rate aggregates into single cells. Before seeding CECs, 1 mL of
trypsin/EDTA or TrypLE reagent is added for 5–10 min
[11]. Trypsin/EDTA or TrypLE are inactivated with the
same amount of expansion medium. Cells are centrifuged,
supernatant is removed, and cells are resuspended in complete
expansion medium before being seeded on plastic culture dish.
5. CECs isolation can be performed with 0.02% EDTA solution
instead of collagenase A [12]. Descemet strips are incubated in
1 mL of EDTA for 45 min followed by up-and-down pipetting
using a glass pipette. Cells are centrifuged and resuspended in
complete expansion medium to seed them on FNC-coated

plastic culture dishes. Afterward, Descemet strips without
cells attached can be removed from culture under a binocular
microscope with sterile straight forceps. Alternatively, they can
be removed at the next media change through aspiration with a
sterile glass pipette branched to a vacuum.
6. FNC mix is added to plastic culture for 1 min at room temper-
ature before rinsing. This allows to obtain FNC-coated culture
plastic dishes. FNC-coated plastic culture has been shown to
improve endothelial morphology [13].
7. A pair of corneas (diameter of ~11 mm) allows to isolate
enough cells for the seeding of one well of a 6-well plate.
8. For clinical purpose, cell quality controls should be performed
on CECs at the final passage. The quality of the CECs can be
established in many ways, such as using morphometric analysis
and cell surface markers expression [10, 14–21]. Cell surface
markers associated with an endothelial phenotype include
CD56 [15], CD166 [19], CD200 [17], CD248 [15], cox-
sackie adenovirus receptor (CAR) [15], collagen type VIII
[14, 16], N-cadherin [16], TGF-β2 [16], SLC4A11 [14],
and glypican 4 [17], while endothelial with a fibroblastic-like
phenotype have been shown to express CD24 [18], CD26
[18], CD44 [19], CD105 [18], CD109 [15], and others [16]
9. When CECs reach confluency, the complete expansion
medium can be switched for a maturation medium for
7–28 days [22, 23]. The dual media approach has been
shown to improve endothelial phenotype throughout passages.
Basic maturation medium composition: Opti-MEM, 8% FBS and
1 penicillin/streptomycin.
10. Confluency is assessed by the percentage of CECs covering
plastic culture.
11. Cell count can also be performed with an automated cell
counter.
12. Trypan blue coloration is cytotoxic. Cell counting should be
done in less than 5 min.
13. DMSO is a toxic oxidative agent at temperatures above 10 C.
Working with cells in contact with a solution containing
DMSO must be done quickly and on ice.
Acknowledgements
Procurement of eyes for research was possible thanks to a partner-

ship with Héma-Québec, the CUO Eye Bank, and a “Fonds de
recherche du Québec – Santé (FRQ-S)” Vision Health Research
Network (VHRN) Infrastructure Program (S.P.). I.X. was a recipi-
ent of Master Training Awards from Université Laval (Wilbrod-
Bhérer), the Fondation du CHU de Québec, the LOEX Center,
and the VHRN.
References
1. Bonanno JA (2003) Identity and regulation of 10. Kinoshita S, Koizumi N, Ueno M,
ion transport mechanisms in the corneal endo- Okumura N, Imai K, Tanaka H, Yamamoto Y,
thelium. Prog Retin Eye Res 22(1):69–94 Nakamura T, Inatomi T, Bush J, Toda M,
2. Joyce NC, Harris DL, Mello DM (2002) Hagiya M, Yokota I, Teramukai S,
Mechanisms of mitotic inhibition in corneal Sotozono C, Hamuro J (2018) Injection of
endothelium: Contact inhibition and cultured cells with a rock inhibitor for bullous
tgf-beta2. Invest Ophthalmol Vis Sci 43 keratopathy. N Engl J Med 378
(7):2152–2159 (11):995–1003. https://doi.org/10.1056/
3. Yoshida K, Kase S, Nakayama K, Nagahama H, NEJMoa1712770
Harada T, Ikeda H, Harada C, Imaki J, 11. Li W, Sabater AL, Chen YT, Hayashida Y, Chen
Ohgami K, Shiratori K, Ilieva IB, Ohno S, SY, He H, Tseng SC (2007) A novel method of
Nishi S, Nakayama KI (2004) Involvement of isolation, preservation, and expansion of
p27kip1 in the proliferation of the developing human corneal endothelial cells. Invest
corneal endothelium. Invest Ophthalmol Vis Ophthalmol Vis Sci 48(2):614–620. https://
Sci 45(7):2163–2167 doi.org/10.1167/iovs.06-1126
4. Matsuda M, Sawa M, Edelhauser HF, Bartels 12. Senoo T, Obara Y, Joyce NC (2000) Edta: a
SP, Neufeld AH, Kenyon KR (1985) Cellular promoter of proliferation in human corneal
migration and morphology in corneal endothe- endothelium. Invest Ophthalmol Vis Sci 41
lial wound repair. Invest Ophthalmol Vis Sci 26 (10):2930–2935
(4):443–449 13. Engler C, Kelliher C, Speck CL, Jun AS (2009)
5. Edelhauser HF (2006) The balance between Assessment of attachment factors for primary
corneal transparency and edema: The proctor cultured human corneal endothelial cells. Cor-
lecture. Invest Ophthalmol Vis Sci 47 nea 28(9):1050–1054. https://doi.org/10.
(5):1754–1767. https://doi.org/10.1167/ 1097/ICO.0b013e3181a165a3
iovs.05-1139 14. Chng Z, Peh GS, Herath WB, Cheng TY, Ang
6. American Academy of Opthalmology A (1997) HP, Toh KP, Robson P, Mehta JS, Colman A
Corneal endothelial photography - three~year (2013) High throughput gene expression anal-
revision. Ophthalmology 104(8):1360–1365 ysis identifies reliable expression markers of
7. Talajic JC, Straiko MD, Terry MA (2013) Des- human corneal endothelial cells. PLoS One 8
cemet’s stripping automated endothelial kera- (7):e67546. https://doi.org/10.1371/jour
toplasty: then and now. Int Ophthalmol Clin nal.pone.0067546
53(2):1–20. https://doi.org/10.1097/IIO. 15. Bartakova A, Alvarez-Delfin K, Weisman AD,
0b013e31827eb6ba Salero E, Raffa GA, Merkhofer RM Jr, Kunze-
8. EBAA (2016) 2015 eye banking statistical vitzky NJ, Goldberg JL (2016) Novel identity
report. Washington, DC and functional markers for human corneal
endothelial cells. Invest Ophthalmol Vis Sci
9. Proulx S, Audet C, Uwamaliya J, 57(6):2749–2762. https://doi.org/10.1167/
Deschambeault A, Carrier P, Giasson CJ, iovs.15-18826
Brunette I, Germain L (2009) Tissue engineer-
ing of feline corneal endothelium using a devi- 16. Ueno M, Asada K, Toda M, Schlotzer-
talized human cornea as carrier. Tissue Eng Schrehardt U, Nagata K, Montoya M,
Part A 15(7):1709–1718. https://doi.org/ Sotozono C, Kinoshita S, Hamuro J (2016)
10.1089/ten.tea.2008.0208 Gene signature-based development of elisa
assays for reproducible qualification of cultured
human corneal endothelial cells. Invest 20. Okumura N, Ishida N, Kakutani K, Hongo A,
Ophthalmol Vis Sci 57(10):4295–4305. Hiwa S, Hiroyasu T, Koizumi N (2017) Devel-
https://doi.org/10.1167/iovs.16-19806 opment of cell analysis software for cultivated
17. Cheong YK, Ngoh ZX, Peh GS, Ang HP, Seah corneal endothelial cells. Cornea 36
XY, Chng Z, Colman A, Mehta JS, Sun W (11):1387–1394. https://doi.org/10.1097/
(2013) Identification of cell surface markers ICO.0000000000001317
glypican-4 and cd200 that differentiate 21. Toda M, Ueno M, Hiraga A, Asada K,
human corneal endothelium from stromal Montoya M, Sotozono C, Kinoshita S,
fibroblasts. Invest Ophthalmol Vis Sci 54 Hamuro J (2017) Production of homogeneous
(7):4538–4547. https://doi.org/10.1167/ cultured human corneal endothelial cells indis-
iovs.13-11754 pensable for innovative cell therapy. Invest
18. Hamuro J, Toda M, Asada K, Hiraga A, Ophthalmol Vis Sci 58(4):2011–2020.
Schlotzer-Schrehardt U, Montoya M, https://doi.org/10.1167/iovs.16-20703
Sotozono C, Ueno M, Kinoshita S (2016) 22. Peh GS, Chng Z, Ang HP, Cheng TY,
Cell homogeneity indispensable for regenera- Adnan K, Seah XY, George BL, Toh KP, Tan
tive medicine by cultured human corneal endo- DT, Yam GH, Colman A, Mehta JS (2015)
thelial cells. Invest Ophthalmol Vis Sci 57 Propagation of human corneal endothelial
(11):4749–4761. https://doi.org/10.1167/ cells: a novel dual media approach. Cell Trans-
iovs.16-19770 plant 24(2):287–304. https://doi.org/10.
19. Okumura N, Hirano H, Numata R, 3727/096368913X675719
Nakahara M, Ueno M, Hamuro J, 23. Beaulieu Leclerc V, Roy O, Santerre K, Proulx
Kinoshita S, Koizumi N (2014) Cell surface S (2018) Tgf-beta1 promotes cell barrier func-
markers of functional phenotypic corneal tion upon maturation of corneal endothelial
endothelial cells. Invest Ophthalmol Vis Sci cells. Sci Rep 8(1):4438. https://doi.org/10.
55(11):7610–7618. https://doi.org/10. 1038/s41598-018-22821-9
1167/iovs.14-14980
Chapter 3
Primary Culture of Cornea-Limbal Epithelial Cells In Vitro

Finbarr O’Sullivan
Abstract
The cultivation of corneal-limbal cells in vitro represents an excellent means to generate models to study
cornea function and disease processes. These in vitro expanded cornea-limbal epithelial cell cultures are rich
in stem cells for cornea, and hence can be used as a cell therapy for cornea-limbal deficiency. This chapter
details the primary culture of these cornea-limbal cells, which can be used as model for further studies of the
cornea surface.
Key words Cornea, Limbal, Primary culture, Feeder cells, Human amniotic membrane, Explants
1 Introduction
The cornea epithelium accounts for approximately 10% of the total

cornea thickness. It rests on a basement membrane or basal lamina,
in close contact with Bowman’s membrane, separating it from the
mesenchymal stroma. The cornea epithelium is composed of three
groups of cells, a superficial squamous cells with a flat polygonal
morphology of about 2–3 cells in thickness, an intermediate cell
layer of 2–3 cells, called wing cells, and a final basal layer of epithe-
lium with a cuboidal/columnar morphology. At the border
between the transparent cornea and the opaque sclera is the limbus.
This region was first proposed in the early 1970s as having a role in
cornea renewal. The epithelium of the limbus consists of multilay-
ered cells rich in minute vessels with the presence of Langerhans
and melanocyte cells (which produce and secrete pigments) and a
gradient of epithelial cells that includes progenitors of the corneal
epithelium. The “X, Y, Z” hypothesis of corneal epithelial mainte-
nance, developed by Thoft and Friend, proposes that corneal epi-
thelial stem cells are located in the limbus [1]. It is these cells that
are responsible for producing transient amplifying cells that prolif-
erate and migrate toward the center of the cornea in a centripetal
manner, replacing epithelial cells lost from the surface of the cor-
nea. As this region has a rich stem cells population that differentiate
29
30 Finbarr O’Sullivan
into cornea epithelia cells and has a greater proliferative potential

when compared to central cornea, it is this area of the cornea that is
selected preferentially for the generation of in vitro cultures.
Limbal-corneal cultures can be generated either by the out-
growth from tissue explants or the dissociation of the tissue by
enzymatic means. For both methods contradictory reports exist as
to which method is superior [2, 3]. In truth both methods have
their advantages and disadvantages and some researchers have
found only minimal differences in culture outcomes. Explant cul-
tures involve plating a piece of tissue approximately 1 mm by 1 mm
in size on a tissue culture surface and allowing cell outgrowth to
occur. As no harsh enzymes are being used, cell damage should be
minimal, improving the chances of viable cell outgrowth. In addi-
tion to avoiding cell loss during tissue processing the explant tech-
nique maintains the stem cell niche of limbal tissue. This niche
effect is reflected in the culture obtained itself as studies have
shown a decrease in stem cell markers in the cells the greater the
distance from the explant [4].
The alternative method to explant cultures is to dissociate the
cells from tissue by enzymatic means to generate a suspension
culture. One of the main advantages this method has is that it
allows the purification of a homogeneous cell population. The
common enzymes used in cell dissociation methods for limbal
corneal epithelial cultures are dispase and trypsin [2, 5, 6]. Dispase
is considered the more gentle of the two enzymatic methods and
this is reflected in the long incubation times, that range from 1 h to
18 h [5, 7]. However, prolonged incubation with dispase can still
result in cellular damage [8].
In addition to these two methods for initiating a primary cell
culture of limbal-cornea epithelial cells, a myriad of cultivation
methods exists. These methods range of different matrices, feeder
cell systems to advanced carrier systems and serum-free media that
seek to replicate the stem cell niche. Of the various matrices used
for limbal-cornea epithelial cell cultures, human amniotic mem-
brane (HAM) has probably proved to be the most popular. The
reasons for this popularity are due in part to its long history in
ophthalmic surgery, having low or no immunogenicity and it also is
a convenient carrier to use clinically for expanded cell therapy
[9]. In addition to low immunogenicity, a number of growth
factors associated with epithelial growth and proliferation have
been found on HAM [10]. Perhaps even more importantly HAM
has been shown as being a good match for the basement membrane
where limbal-corneal epithelium resides, with a similar constitution
of collagens type IV and VII, fibronectin, and laminins such as α3,
β1, and β2 [11, 12]. Furthermore, there is some evidence the ratio
of components and the bio-mechanical properties of HAM are a
close match to that found in the limbal niche [13]. However, HAM
does require the use of donor tissue, so supply can be scarce and
Primary Culture of Cornea-Limbal Epithelial Cells 31
even when tissue is sourced from accredited tissue banks there is

always a small concern regarding viral disease transmission.
The in vitro culture of some cell types, particularly those more
fastidious cells such as stem cells, often requires the co-culture of
reproductively inactive cells known as feeder cells. The first
reported use of feeder cells in cell culture dates back to 1956 by
Puck and colleagues [14]. Feeder cells help by simulating the cell
microenvironment by secreting extracellular metabolites that main-
tain the stem cells. This microenvironment is a combination of
different mechanisms that have not been fully defined. Theses
mechanisms are believed to include cell-to-cell and cell-to-extracel-
lular matrix (ECM) interaction [15] in addition to the production
of soluble factors promoting growth and inhibit apoptosis
[16, 17]. The positive impact of fibroblast feeder cells, such as the
NIH 3T3 cell line or limbal-cornea fibroblast cells, on the long-
term survival and propagation of limbal-cornea epithelial cells has
been demonstrated in numerous studies [17–19].
One of the most common media for culturing the cells used is
based on the original work of Rheinwald and Green in the 1970s
for keratinocyte cell culture. This consists of a medium of minimal
essential medium (MEM) and Ham’s F-12 (3: 1) supplement with
hydrocortisone, insulin, triiodothyronine, adenine, cholera toxin,
and EGF [20]. Each of these supplements has a positive role to play
in the promotion of cellular growth and the prevention of differen-
tiation [20–22]. As this media contains the undefined component
of serum and in order to improve the defined nature and perfor-
mance of the media developed by Rheinwald and Green, various
serum-free medias have been developed. Over the last 30 years
various studies have been conducted into the development of a
defined serum-free media to successfully culture limbal-cornea epi-
thelial cells in vitro [17]. These have included the addition of such
additives as bovine pituitary extract to B-27 serum-free supplement
growth-factor all with the aim to eliminate the requirement of
serum [23–25]. Commercial serum-free media that have been
refined for the cultivation of keratinocytes such as EpiLife are
now available [26].
2 Materials
2.1 Preparation 1. Recombinant epidermal growth factor (rhEGF) stock solution:

of Media Supplements 100 μg/mL in 0.1% (w/v) bovine serum albumin (BSA),
10 mM acetic acid solution. Prepare a 10 mM acetic acid
solution from a stock of glacial acetic acid by slowly adding
57 μL to 25 mL of deionized water and adjust the final volume
to 100 mL with deionized water using a graduated cylinder.
Add 5 mg of BSA slowly to 5 mL of 10 mM acetic acid. In a
laminar flow cabinet, observing aseptic techniques use a syringe
fitted with a 0.22 μm filter to sterilize the solution. Add 200 μg

of rhEGF powder to 2 mL of this sterile solution to yield a
stock solution with 100 μg/mL rhEGF. Dispense the solution
as 10 μL aliquots into sterile micro-centrifuge tubes and store
at 20 C, avoiding repeated freeze thaw cycles.
2. Hydrocortisone Stock Solution: 0.83 μM in basal media. In a
laminar flow cabinet, observing aseptic technique, use a syringe
fitted with a 0.22 μm filter to sterilize 5 mL of 95% (v/v)
ethanol. Add 250 μL of the sterilized ethanol to 1 mg of
hydrocortisone (sterile). When dissolved, add 250 μL of hydro-
cortisone solution to 2.25 mL of basal media. Dispense the
8.3 μM stock of hydrocortisone as 100 μL aliquots into sterile
micro-centrifuge tubes and store at 20 C; avoid repeated
freeze thaw cycles.
3. Triiodothyronine Stock Solution: 2 μM in PBS-A. In a laminar
flow cabinet, observing aseptic technique dissolve 1 mg of
triiodothyronine in 1 mL of sterile PBS-A to generate a
1.5 mM solution. Add 26.6 μL of 1.5 mM triiodothyronine
solution to 19.973 mL of sterile PBS-A to yield a stock solution
of 2 μM triiodothyronine. Using good aseptic technique, use a
syringe fitted with a 0.22 μm filter to sterilize the triiodothyro-
nine stock solution and dispense as 100 μL aliquots into sterile
micro-centrifuge tubes and store at 20 C; avoid repeated
freeze thaw cycles.
4. Cholera Toxin A Subunit Stock Solution: 1 mg/mL in sterile
ultrapure water (UPW). Using good aseptic technique, in a
laminar flow add 1 mL of sterile UHP water to 1 mg of cholera
toxin A subunit (sterile). Dispense as 100 μL aliquots in sterile
micro-centrifuge tubes and store at +4 C. Do not freeze
solution.
2.2 Culture Media 1. Corneal Epithelial Growth Media: 67.5% (v/v) 1 Dulbecco’s
modified Eagle media (DMEM), 22.5% (v/v) 1 Ham’s F-12
nutrient media, 10% (v/v) fetal calf serum (FCS), 10 ng/mL
rhEGF, 0.4 μg/mL hydrocortisone solution, 2 109 M tri-
iodothyronine solution, 100 ng/mL cholera toxin A subunit
solution. Store corneal epithelial growth media at 4 C for up
to 2 weeks.
2. NIH 3T3 Growth Media: 95% (v/v) DMEM, 5% (v/v) fetal
calf serum. Store NIH 3T3 growth media at 4 C for up to
3 weeks.
2.3 Other solutions 1. PHEM Buffer: 60 mM 1,4-piperazinediethanesulfonic acid

(PIPES), 25 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfo-
nic acid (HEPES), 10 mM egtazic acid (EGTA), 2 mM MgCl2.
To a glass beaker add 9.072 g of PIPES, 13.254 g of HEPES,
11.902 g of EGTA, 0.203 g of MgCl2. Add 400 mL of UHP

water to dissolve. Adjust the pH to 6.9 with 10 M KOH
(approximately 7 mL). Transfer to a graduated cylinder and
make to the 500 mL. This can be kept at room temperature for
3 weeks.
2. HANKS Buffered Saline Solution (HBSS).
3. Trypsin-EDTA solution.
4. Phosphate-buffered saline A (PBS-A).
2.4 Immuno- 1. Methanol.

fluorescence Reagents 2. Normal goat serum.
3. Blocking buffer: 5% (v/v) normal goat serum in PBS-A.
4. Tween-20: 0.25% (v/v) in PHEM.
5. Mouse anti-cytokeratin 3.
6. Rabbit anti-cytokeratin 12.
7. Mouse anti-ABC-G2.
8. Rabbit anti-ΔNP63.
9. Goat anti-mouse IgG Alexa Fluor 488.
10. Goat anti-rabbit IgG Alexa Fluor 555.
11. DAPI (40 , 6- diamidino-2-phenylindole).
3 Methods
3.1 Option A: 1. Culture NIH 3T3 fibroblast cells in DMEM supplemented

Irradiation of NIH 3T3 with 5% (v/v) fetal calf serum, until 60–70% confluence is
Cells reached.
2. Trypsinize the cells and wash in HBSS. Resuspend the NIH
3T3 cells in HBSS and place vial in gamma irradiator (see Note
1). Set the gamma irradiator for 60 Gray (see Note 2).
3. When irradiation is complete, centrifuge the cells at 100 g for
5 min.
4. Resuspend the cells in growth media, perform a cell count,
plate the irradiated NIH-3T3 cells at a density of
2.4 104 cells/cm2, and allow 24 h for attachment to occur
(see Note 3).
3.2 Option B: 1. Thaw overnight at 4 C cyro-preserved human amniotic mem-

Preparation brane (HAM). Carefully remove the HAM from its freezing
of Denuded (HAM) medium and carefully peel from the nitrocellulose backing
paper with sterile forceps on a sterile 90 mm dish.
2. Stretch the HAM out on the dish, make sure to place the
epithelial side face up, and add 20 mL of 0.25% trypsin-
EDTA solution. Cover the dish and place in a 37 C incubator

for 15–20 min.
3. Following the 15–20 min incubation gently scrubbed the
HAM surface with a cell scraper to remove the amniotic epi-
thelium (see Note 4).
4. Remove the trypsin solution with the detached cells. Wash the
HAM twice with PBS-A to remove scraped cells.
5. Carefully turn the membrane over to the endothelial side and
scrape the HAM with a cell scraper to remove mucus. Wash
twice with PBS-A. The resulting membrane is now termed as
denuded human amniotic membrane (dHAM).
6. Depending on the size available, trim the dHAM to pieces of
2–2.5 cm2 in area. Wrap the dHAM pieces around a 1.5 cm2
glass slide that has a thickness of approximately 2 mm. Ensure
that the HAM is tucked in under the glass slide (see Note 5).
Place in a well of a 6-well plate or a 35 mm2 dish and slowly add
growth media to just cover the dHAM. The dish is now ready
to receive limbal explants.
3.3 Generation 1. In a sterile 90 mm dish place the cornea-scleral tissue and rinse
of Limbal-Cornea with growth media to remove any loose material.
Explants 2. Using a scalpel cut the cornea-sclera into four quadrants and
remove any excess tissue such as sclera, iris, cornea to isolate the
cornea-scleral rim.
3. Locate the limbal ring by identifying a pale pink-brown ring in
the endothelial side of the cornea-scleral rim. With the scale
blade gently scrape off the endothelial layer.
4. Place a sterile glass microscope slide into a 90 mm dish; this will
act as a hard surface for cutting the ring out on. Cut away all
excess sclera and cornea to isolate the limbal ring.
5. Using a firm chopping motion cut the limbal ring into
1 1 mm2 explants. Place the 1 1 mm2 explants epithelial
side down in the center of well. The well will either be option A
pre-seeded irradiated NIH-3T3 feeder cells or option
B dHAM.
6. Place approximately 6–7 explants in a ring centripetally in the
center to the irradiated NIH-3T3 feeder cells or dHAM. Media
level should be low so that the growth surface is moist, but the
explants should not be floating.
7. Cover the cell culture dish and carefully place in a 37 C CO2
incubator. Allow the explants to settle for 1–2 h before remov-
ing from the incubator and very slowly and carefully adding
1 mL of growth media.
8. The following day refeed explant cultures with growth media.

Examine the explant culture for adhesion and cell outgrowth.
Outgrowth should become visible by day 2–5 as a sheet of
cobble-like cells growing from the explant.
9. Refeed the culture on alternative days until day 15–20; at this
stage the culture should be confluent.
3.4 Confirmation To confirm the identity of the cells that have grown from the
of Limbal-Cornea Stem explant conduct immunofluorescence for specific markers of
Cells limbal-cornea cells.
1. Wash the cell culture sample gently for 5 min, three times, with
PBS-A to remove traces of growth media and any debris.
2. Fix the cell culture sample using ice-cold methanol for 10 min.
3. Block the cell culture sample with the blocking buffer for 1 h at
room temperature.
4. Remove the blocking buffer and wash the cell culture sample
once with PHEM buffer for 5 min and once with 0.25% (v/v)
Tween-20 in PHEM buffer for 5 min.
5. Remove the buffer and add primary antibody diluted appropri-
ately in 1% (v/v) normal goat serum in PHEM buffer. Primary
antibodies considered suitable are mouse anti-cytokeratin
3, rabbit anti-cytokeratin 12, mouse anti-ABC-G2, and rabbit
anti-ΔNP63. Incubate the cell culture sample and primary
antibody overnight at 4 C.
6. Remove the primary antibody the following morning and wash
the cell culture sample for 5 min, three times, with 0.25% (v/v)
Tween-20 in PHEM buffer.
7. Add the appropriate secondary antibody, e.g., Goat anti-mouse
IgG Alexa Fluor 488 or Goat anti-rabbit IgG Alexa Fluor
555, diluted in 1% (v/v) normal goat serum in PHEM buffer.
Incubate at room temperature for 1 h in the dark (see Note 6).
8. Remove the secondary antibody solutions and wash the cell
culture sample for 5 min, three times, with 0.25% (v/v) Tween-
20 in PHEM buffer.
9. Counterstain the nuclei with DAPI for 2 min, then wash twice
PHEM Buffer and mount the cell culture sample with an anti-
fade mountant and coverslip the sample.
10. Limbal-cornea epithelial stem cells should show bright nuclear
staining for ΔNP63, ABC-G2, while differentiating limbal-
cornea epithelial cells will show weak staining for these markers
and will be strongly positive for cytokeratin 3 and 12 (see
Note 7).
4 Notes
1. Avoid using media with phenol red during the irradiation

process as free radical can be generated that will lead to cell
death.
2. Batches of irradiated NIH-3 T3 cells can be prepared and
stored in liquid nitrogen to avoid having to irradiate fresh
cells each time. Cells can be cryopreserved using a 10%
DMSO—90% FCS Solution. DMSO should be added slowly
to cells to avoid shock and cells cooled at 1 C/min.
3. Mitomycin C is an alternative means to generate feeder cells
using a concentration of 10 μg/mL for 2–4 h. Care must be
taken to ensure no Mitomycin C is present after preparation of
the feeder cells, as it can negatively impact on cell growth.
4. When scrapping the surface of the HAM care must be taken not
to press too hard as this might tear the underlying basement
membrane. This method removes of 90–100% of epithelium
from HAM.
5. It is important to use a glass slide with sufficient weight under
which the amniotic membrane is tucked. Use of a light cover-
slip will not hold the amniotic membrane in place during the
duration of the cell culture.
6. The fluorescent secondary antibodies are light sensitive and
should be prepared out of direct light in dim conditions.
Once the secondary has been added, all incubation steps should
be in the dark to avoid “quenching” of the fluorescent signal.
7. Although a variety of putative limbal-cornea stem cell markers
have been proposed, the expression of no maker is truly defini-
tive. The general consensus is that the co-expression of a num-
ber of makers represents the best means of identifying cells as
be of being limbal-corneal epithelium.
Acknowledgements
This work is supported by Science Foundation Ireland cofunded by

ERDF, grant no 12/RC/2275_P2.
References
1. Thoft RA, Friend J (1983) The X, Y, Z hypoth- 3. Ghoubay-Benallaoua D, Basli E, Goldschmidt
esis of corneal epithelial maintenance. Invest P et al (2011) Human epithelial cell cultures
Ophthalmol Vis Sci 24:1442–1443 from superficial limbal explants. Mol Vis
2. Koizumi N, Cooper LJ, Fullwood NJ et al 17:341–354
(2002) An evaluation of cultivated corneal lim- 4. Kolli S, Lako M, Figueiredo F et al (2008) Loss
bal epithelial cells, using cell-suspension cul- of corneal epithelial stem cell properties in out-
ture. Investig Ophthalmol Vis Sci growths from human limbal explants cultured
43:2114–2121 on intact amniotic membrane. Regen Med
3:329–342. https://doi.org/10.2217/ 16. Burroughs J, Gupta P, Blazar BR, Verfaillie CM

17460751.3.3.329 (1994) Diffusible factors from the murine cell
5. Liu S, Li J, Wang C et al (2006) Human limbal line M2-10B4 support human in vitro hemato-
progenitor cell characteristics are maintained in poiesis. Exp Hematol 22:1095–1101
tissue culture. Ann Acad Med Singap 35:80–86 17. Tseng SC, Kruse FE, Merritt J, Li DQ (1996)
6. Pellegrini G, Traverso CE, Franzi AT et al Comparison between serum-free and
(1997) Long-term restoration of damaged cor- fibroblast-cocultured single-cell clonal culture
neal surfaces with autologous cultivated cor- systems: evidence showing that epithelial anti-
neal epithelium. Lancet 349:990–993. apoptotic activity is present in 3T3 fibroblast-
https://doi.org/10.1016/S0140-6736(96) conditioned media. Curr Eye Res 15:973–984.
11188-0 https://doi.org/10.3109/
7. Espana EM, Romano AC, Kawakita T et al 02713689609017643
(2003) Novel enzymatic isolation of an entire 18. Kim MK, Lee JL, Oh JY et al (2008) Efficient
viable human limbal epithelial sheet. Investig cultivation conditions for human limbal epithe-
Ophthalmol Vis Sci 44:4275–4281. https:// lial cells. J Korean Med Sci 23:864. https://
doi.org/10.1167/iovs.03-0089 doi.org/10.3346/jkms.2008.23.5.864
8. Spurr SJ, Gipson IK (1985) Isolation of cor- 19. Pellegrini G, Golisano O, Paterna P et al
neal epithelium with Dispase II or EDTA. (1999) Location and clonal analysis of stem
Effects on the basement membrane zone. cells and their differentiated progeny in the
Invest Ophthalmol Vis Sci 26:818–827 human ocular surface. J Cell Biol
9. Kubo M, Sonoda Y, Muramatsu R, Usui M 145:769–782. https://doi.org/10.1083/jcb.
(2001) Immunogenicity of human amniotic 145.4.769
membrane in experimental xenotransplanta- 20. Rheinwald JG, Green H (1975) Serial cultiva-
tion. Invest Ophthalmol Vis Sci 42:1539–1546 tion of strains of human epidermal keratino-
10. Koizumi N, Inatomi T, Sotozono C, et al (2000) cytes: the formation of keratinizing colonies
Growth factors in amniotic membrane Growth from single cells. Cell 6:331–343. https://doi.
factor mRNA and protein in preserved human org/10.1016/s0092-8674(75)80001-8
amniotic membrane. Curr Eye Res. 20(3) : 21. Sun TT, Green H (1977) Cultured epithelial
173–177 cells of cornea, conjunctiva and skin: absence of
11. Moriyama T, Asahina I, Ishii M et al (1994) marked intrinsic divergence of their differen-
Basement membrane assembly and differentia- tiated states. Nature 269:489–493. https://
tion of cultured corneal cells: importance of doi.org/10.1038/269489a0
culture environment and endothelial cell inter- 22. Yaeger PC, Stiles CD, Rollins BJ (1991)
action. Exp Cell Res 4:415–427 Human keratinocyte growth-promoting activ-
12. Fukuda K, Chikama T, Nakamura M, Nishida ity on the surface of fibroblasts. J Cell Physiol
T (1999) Differential distribution of subchains 149:110–116. https://doi.org/10.1002/jcp.
of the basement membrane components type 1041490114
IV collagen and laminin among the amniotic 23. Yokoo S, Yamagami S, Yanagi Y et al (2005)
membrane, cornea, and conjunctiva. Cornea Human corneal endothelial cell precursors
18:73–79. https://doi.org/10.1097/ isolated by sphere-forming assay. Invest
00003226-199901000-00013 Ophthalmol Vis Sci 46:1626–1631. https://
13. Chen B, Jones RR, Mi S et al (2012) The doi.org/10.1167/iovs.04-1263
mechanical properties of amniotic membrane 24. Brejchova K, Trosan P, Studeny P et al (2018)
influence its effect as a biomaterial for ocular Characterization and comparison of human
surface repair. Soft Matter 8:8379. https://doi. limbal explant cultures grown under defined
org/10.1039/c2sm26175h and xeno-free conditions. Exp Eye Res
14. Puck TT, Marcus PI, CieciuraI SJ (1956) 176:20–28. https://doi.org/10.1016/j.exer.
Clonal growth of mammalian cells in vitro; 2018.06.019
growth characteristics of colonies from single 25. Zamudio A, Wang Z, Chung S-H, Wolosin JM
HeLa cells with and without a feeder layer. J (2016) Inhibition of TGFβ cell signaling for
Exp Med 103:273–283. https://doi.org/10. limbal explant culture in serumless, defined
1084/jem.103.2.273 xeno-free conditions. Exp Eye Res
15. Ehmann UK, Stevenson MA, Calderwood SK, 145:48–57. https://doi.org/10.1016/j.exer.
DeVries JT (1998) Physical connections 2015.10.021
between feeder cells and recipient normal 26. Lekhanont K, Choubtum L, Chuck RS et al
mammary epithelial cells. Exp Cell Res (2009) A serum- and feeder-free technique of
243:76–86. https://doi.org/10.1006/excr. culturing human corneal epithelial stem cells on
1998.4054 amniotic membrane. Mol Vis 15:1294–1302
Chapter 4
Optimization of Human Limbal Stem Cell Culture by

Replating a Single Limbal Explant
Marina López-Paniagua, Teresa Nieto-Miguel, Sara Galindo,
Laura Garcı́a-Posadas, Ana de la Mata, Rosa M. Corrales,
Margarita Calonge, and Yolanda Diebold
Abstract
Cultured limbal epithelial stem cell transplantation is a clinical procedure used to regenerate the corneal
epithelium in patients with limbal stem cell deficiency. The protocols used to expand limbal epithelial cells
in vitro need to be optimized, since the scarcity of human ocular tissue donors is limiting the potential use of
this procedure. Here, we describe a method to consecutively expand a single human limbal explant. With
this method it is possible to obtain up to three limbal epithelial primary cultures from the same explant, thus
increasing the efficiency of the in vitro cell culture.
Key words Cornea, Corneal epithelium, Corneoscleral samples, Limbus, Limbal explant, Limbal
stem cells, Limbal primary cultures, In vitro expansion
1 Introduction
Limbal epithelial stem cells (LESCs) are responsible for corneal

epithelium renewal [1, 2]. Corneal epithelium integrity is para-
mount for maintaining the corneal transparency required for opti-
mal vision. However, several diseases, as well as chemical or physical
insults, can compromise the integrity of the corneal epithelium,
causing wounds and/or opacities. In many of these cases, corneal
transplantation is a solution that may be used to restore sight.
However, in the case of the destruction or dysfunction of LESCs
or their niche, the outcome of corneal transplantation is usually
poor [3]. In these cases, it is necessary to replace the missing stem
cells to regenerate the corneal epithelium. To accomplish this,
clinicians carry out cultured limbal epithelial transplantation
(CLET). This therapy was first developed by Pellegrini et al. in
1997 [4] and has been reproduced in various clinical trials [5–
11]. Since a limiting factor of this method is obtaining of enough
39
40 Marina López-Paniagua et al.
LESCs, several protocols have been proposed to optimize this

procedure in an attempt to increase its performance [12–14]. In
this protocol we propose a method based on the consecutive
expansion of a single human limbal explant obtained from cadaveric
donors. Using this protocol, it is possible to expand the same
explant up to three times, allowing clinicians to obtain larger num-
bers of limbal epithelial cells that retain stem cell properties
[15, 16]. These cultured LESCs can be used for either research or
clinical purposes.
2 Materials
All the solution preparations should be performed in sterile condi-

tions in a laminar flow hood irradiated with ultraviolet light and
cleaned with ethanol 70% (v/v) in water. Surgical instruments
should be sterilized by steam sterilization (autoclaved) prior to use.
Complete culture media should be kept at 4 C and
pre-warmed to 37 C in a thermostatic bath prior to use.
2.1 Medium 1. Shipping and culture medium (see Notes 1 and 2): Dulbecco’s
and Solutions modified Eagle medium/Ham’s F12 solution (DMEM/F12)
(1:1), 2.5 ng/mL epidermal growth factor (EGF), 10 μg/mL
insulin, 5.5 μg/mL transferrin, 5 ng/mL sodium-selenite,
0.01 μg/mL hydrocortisone, 0.5% dimethyl sulfoxide
(DMSO), 132.5 ng/mL cholera toxin, 50 μg/mL gentamicin,
2.5 μg/mL amphotericin B, and 5% fetal bovine serum (FBS).
2. Wash solution: Hanks Balanced Salt Solution (HBSS), 50 μg/
mL gentamicin, and 2.5 μg/mL amphotericin B.
3. Biosafe culture medium (see Note 3): DMEM/F12 (1:1),
2.5 ng/mL EGF, 10 μg/mL insulin, 5.5 μg/mL transferrin,
5 ng/mL sodium-selenite, 0.5 μg/mL hydrocortisone, 1 μM
isoproterenol, 0.18 mM adenine, 2 nM triiodotironina, 50 μg/
mL gentamicin, 2.5 μg/mL amphotericin B, and 10% human
serum (HS).
4. Fetal bovine serum (FBS).
2.2 Plastic and Glass 1. Plastic or glass sterile container for the shipment of corneoscl-
eral tissues.
2. 35 mm Petri dishes.
3. 5 mL disposable serological pipettes.
4. Micropipette tips (100–1000 μL and 20–200 μL).
5. Tissue culture 12-well polystyrene plates (3.8 cm2/well).
Consecutive Culture of a Human Limbal Explant 41
2.3 Surgical 1. Fine-tipped tweezers.

Instruments 2. Blunt-tipped tweezers.
3. Scissors.
4. Crescent knife.
5. 7.5 mm trephine.
6. Scalpel.
7. Surgical compass.
2.4 Equipment 1. Laminar flow hood.

2. Autoclave sterilizer.
3. Thermostatic bath.
4. 100–1000 μL micropipette.
5. 20–200 μL micropipette.
6. Cell culture CO2 incubator.
7. Inverted phase contrast microscope.
3 Methods
All the steps described in this section must be carried out in a

laminar-flow hood previously exposed to ultraviolet radiation and
cleaned with ethanol 70% (v/v) in water to maintain sterile
conditions.
3.1 Preservation Human corneoscleral samples can be collected in a different geo-

and Shipment graphic location from which they are processed. In such cases,
of Cadaveric Human samples should be stored and shipped under specific conditions
Corneoscleral Tissues until their processing:
1. Fill a sterile container with shipping medium (see Notes 4 and 5).
2. Place the cadaveric human corneoscleral tissue (see Note 6)
into the container filled with shipping medium and send it to
the laboratory where tissues are going to be processed. The
shipment temperature should be maintained at 4 C (see
Note 7).
3.2 Limbal Explant 1. Remove the corneoscleral tissue from the container used for
Preparation the shipment. Remove the sample from the scleral area using
blunt-tipped tweezers in order to avoid damaging the limbal
region.
2. Place the corneoscleral tissue (Fig. 1a) in a 35 mm Petri dish
and add 4 mL of wash solution (see Note 8 and Fig. 1b).
Fig. 1 Limbal explant isolation. Corneoscleral tissues (a) are rinsed with wash solution (b). Excess of
conjunctiva, iris, and corneal endothelium are removed (c and d). Subsequently, central cornea is removed
from the corneoscleral button using a trephine (e–i). Finally, a corneoscleral ring is cut into 1–3 mm2 limbal
explants (j and k), which are plated into 12-well polystyrene plates (l)
3. Remove the wash solution and add other 4 mL of the same

solution to perform a second wash. Repeat this step twice for a
total of three washes.
4. Excise the conjunctival tissue from the corneoscleral sample
using fine-tipped tweezers and scissors. To do this, grab a
small portion of the conjunctival tissue with the tweezers and
cut it using scissors. Repeat this procedure until all the con-
junctival tissue is removed (Fig. 1c).
5. Remove both the wash solution and the remnants of conjunc-
tiva from the Petri dish using a 5 mL pipette (see Note 9).
6. Add 4 mL of wash solution to the Petri dish and turn over the
corneoscleral sample so that the corneal endothelium and the
iris face upward. Take a sample from the scleral area using
blunt-tipped tweezers to avoid damaging the limbal region.
7. Remove the iris and the corneal endothelium. Hold the cor-
neoscleral sample using the fine-tipped tweezers and scratch
the superior surface of the sample with a crescent knife (see
Note 10 and Fig. 1d).
8. Repeat step 5 to remove remnants of the iris and corneal
endothelium from the Petri dish.
9. Add 4 mL of wash solution to the Petri dish, turn over the

corneoscleral sample, and excise the scleral tissue until only
1–2 mm of sclera remain around the limbal ring. To perform
this procedure, take different small pieces of sclera with fine-
tipped tweezers and remove them using scissors.
10. Repeat step 5 to remove scleral debris and add 4 mL of wash
solution to the Petri dish.
11. Excise the central corneal button from the corneoscleral sam-
ple using a trephine that is 7.5 mm in diameter (see Note 11).
Place the corneoscleral sample with the corneal stroma facing
up and the corneal epithelium facing down over the concave
surface of the trephine and press the corneoscleral sample with
the trephine knife (see Notes 12 and 13, and Fig. 1e–h). This
will remove the central cornea allowing you to obtain the
corneoscleral ring, which includes scleral, limbal, and corneal
tissues (Fig. 1g).
12. Place the corneoscleral ring with the epithelium facing down
on the lid of the Petri dish (Fig. 1i).
13. Prepare limbal explants. Hold the corneoscleral ring with fine-
tipped tweezers and cut limbal portions of 1–3 mm2 using a
scalpel (see Notes 14 and 15, Figs. 1j, k and 2).
3.3 Initial Limbal 1. Place the limbal explants with the limbal epithelium facing up
Primary Culture (LPC) in a 12-well plate (a single explant per well of 3.8 cm2) (Fig. 1l).
Fig. 2 Optimal size of limbal explants. Diagram showing optimal limbal explant
characteristics (see Note 15)
Fig. 3 Cell culture growth from a single limbal explant. (a–e) Representative phase contrast microphotographs
of cell outgrowths from a single limbal explant replated for generating consecutive limbal primary cultures
(LPC). (f–j) Representative phase contrast microphotographs of confluent LPC samples. LPC0-3 shows
homogeneous, cuboidal, and epithelial-like morphology, while LPC4 shows elongated and fibroblast-like
morphology. Scale bar: 100 μm
Keep the plate uncovered and inside the laminar-flow hood for
30 min (see Note 16).
2. Add 50 μL of FBS to each well and incubate overnight at 37 C,
5% CO2, and 95% relative humidity in a cell culture incubator
(see Notes 3, 17, and 18).
3. Add 500 μl of culture medium to each well and incubate the
plate at 37 C, 5% CO2, and 95% relative humidity.
4. Monitor each limbal explant daily under an inverted phase
contrast microscope and carefully change the culture medium
every 2–3 days (see Notes 19 and 20).
5. Remove the limbal explant from its well when cells migrating
from the edges of the limbal explant form a ring of cells around
it. In order to perform this step, take the explant from the
scleral area with fine-tipped tweezers and use other similar
tweezers to break the connection between the cell outgrowth
and the limbal tissue by making a quick upward movement (see
Note 21).
6. Once an explant is removed from a well, maintain the cell
outgrowth under the same culture conditions until the primary
culture reaches confluence (see Note 22 and Fig. 3), at which
time designate it “LPC0.”
3.4 Consecutive LPC 1. Place the removed limbal explant into a new 3.8 cm2 polysty-
rene well, as in step 1 from Subheading 3.3.
2. Follow steps 2–6 from Subheading 3.3 in order to obtain
consecutive LPCs (see Notes 23–25, and Figs. 3 and 4).
Fig. 4 Experimental timeline for generating consecutive human limbal epithelial stem cell cultures by replating
a single limbal explant. In order to establish each limbal primary culture (LPC) the explant is maintained in
culture until a cell outgrowth has surrounded it. Then, the explant is removed and replated in order to obtain
the next LPC. After removing the explant, cells are maintained in culture until they reach confluence, at which
time the culture is considered an established LPC. The total time needed from limbal plating to reach a
confluent LPC0 is about 26 days, while the time required to obtain confluent LPC1 and LCP2 cultures is around
30 and 24 days, respectively. In addition, time elapsed between when the first limbal explant is plated into a
well of a 12-well plate until a confluent LPC2 is obtained is about 42 days (see Note 25)
4 Notes
1. The shipping and culture media can be stored at 4 C for

1 month. If they have not been used during this period of
time, they should be discarded.
2. To carry out this protocol, we recommend using this composi-
tion for the shipping medium. Nevertheless, another composi-
tion could also be valid for sending corneoscleral samples
intended to perform this protocol.
3. This protocol could also be carried out using a biosafe culture
medium.
4. Fill the bottle/container with shipping medium completely to
ensure that the ocular sample never gets dry and damaged. We
strongly recommend not using low volumes of shipping
medium in order to prevent, as much as possible, the potential
consequences of evaporation or freezing due to changes in
environmental temperature.
5. Each bottle/container should be used to store and ship only
one corneoscleral sample.
6. We recommend collecting corneoscleral tissues from the eye-
ball leaving at least 3 mm of scleral tissue surrounding the
corneal area. We suggest marking the superior area in order
to orient the tissue sample once is out from the eyeball, since
limbal epithelial stem cells are mainly located in the superior

and the inferior limbal ring areas [17]. The marking can be
done by leaving a small projection of scleral tissue in the
superior area.
7. Do not keep ocular tissues under storage conditions for more
than 5 days.
8. At the end of this step, the corneoscleral tissue must be placed
with corneal epithelium facing up.
9. We recommend avoiding the use of a smaller pipette, since
remnants of conjunctival tissue could obstruct the pipette and
prevent the proper collection and removal of tissue debris.
10. In order to avoid damaging the limbus, it is important to take
the corneoscleral sample from the scleral region without touch-
ing the limbal or the corneal areas.
11. First divide the trephine into its two parts by separating the
part containing the blade from the part with the concave
surface where the corneoscleral sample should be placed
(Fig. 1e–h).
12. To cut a regular limbal ring, make sure that the corneoscleral
tissue is perfectly centered on the concave surface of the tre-
phine (Fig. 1e).
13. To cut the central corneal button press the corneoscleral sam-
ple with the blade only once (Fig. 1f). If you press the tissue
with the blade two or more times, it is possible that the tissue
could move over the concave surface of the trephine and,
consequently, produce an irregular limbal ring.
14. First, we suggest dividing the corneoscleral ring into two equal
portions, using a scalpel. Keep one of them in a Petri dish with
4 mL of wash solution to prevent tissue drying and damaging.
Second, we recommend cutting the other half portion into two
equal pieces, again submerging one of them in the same wash
solution. Once you have ¼ of the corneoscleral ring on the lid
of the Petri dish, divide this tissue portion into 6 equal sections,
in order to obtain limbal explants of 1–3 mm2 each. Divide
each ¼ of the corneoscleral portion into 6 equal limbal explants
this way (Fig. 1j, k).
15. The size of each limbal explant can be verified using a surgical
compass.
16. The 12-well plate with the limbal explants should be placed
without its cover at the back of the laminar-flow hood. In this
way, the air flow will favor the evaporation of the remaining
wash solution and, consequently, the successful attachment of
the explant to the polystyrene.
17. If using the biosafe culture medium to carry out this protocol,
50 μL of HS instead of FBS must be added in the step.
18. The FBS (or HS in cases where biosafe culture medium is used,
see Note 17) must be added to the top of the limbal explants
using a micropipette (20–200 μL) that is maintained perpen-
dicular to the limbal explant as serum is added.
19. A micropipette (100–1000 μL) should be used to routinely
change the culture medium. The tip of the micropipette should
be positioned next to the wall of the well during culture
medium changes to prevent touching, and possibly hoisting,
limbal tissue. In addition, fresh culture medium should be
slowly added to the well in order to ensure that limbal explants
remain in the same position.
20. We suggest making a mark in the cover of the plate indicating
which wells contain migrating cells growing from limbal
explants. This will allow you to add 1 mL of culture medium
only to marked wells when changing the culture medium. The
outgrowth of the different limbal explants may vary, meaning
that within the same plate you may have to add 0.5 mL of
culture medium/well (if no outgrowth is observed) or 1 mL of
culture medium/well (when outgrowth is observed).
21. It is possible that some cells may detach from the substratum
when the limbal explant is removed from the well.
22. We consider an LPC successful when it reaches more than 80%
confluence.
23. The same limbal explant can be consecutively cultivated up to
seven times, obtaining confluent LPCs from LPC0 to LPC6.
However, only the first three LPCs (LPC0-LPC2) will main-
tain an epithelial cell morphology and a limbal epithelial cell
phenotype [15]. If using the biosafe culture medium to carry
out the protocol (see Note 3), only two consecutive LPCs
(LPC0 and LPC1) will maintain a limbal epithelial cell
phenotype [16].
24. The percentage of confluent LPCs obtained from a corneoscl-
eral sample can vary depending on the tissue donor age, since
there is a decreasing trend in the growth potential of limbal
tissues with increasing donor age [15, 18, 19].
25. In case of using the biosafe culture medium for performing this
protocol, note that both time elapsed between limbal explant
plating to limbal explant removal and time elapsed between
limbal explant removal to reach a confluent LPC might be

slightly different from times mentioned here [16].
Acknowledgments
Financial Support: Ministerio de Economı́a y Competitividad and

Fondo Europeo de Desarrollo Regional, Spain (SAF2015-63594-R
MINECO/FEDER, EU). Centro de Investigación Biomédica en
Red de Bioingenierı́a, Biomateriales y Nanomedicina (CIBER-
BBN). Instituto de Salud Carlos III, Spain. Centro en Red de
Medicina Regenerativa y Terapia Celular de Castilla y León, Spain.
References
1. Cotsarelis G, Cheng SZ, Dong G, Sun TT, non-xenogenic, cultivated limbal stem cell
Lavker RM (1989) Existence of slow-cycling transplantation. J Transl Med 12:58. https://
limbal epithelial basal cells that can be prefer- doi.org/10.1186/1479-5876-12-58
entially stimulated to proliferate: implications 9. Zhao Y, Ma L (2015) Systematic review and
on epithelial stem cells. Cell 57:201–209 meta-analysis on transplantation of ex vivo
2. Notara M, Alatza A, Gilfillan J, Harris AR, cultivated limbal epithelial stem cell on amni-
Levis HJ, Schrader S, Vernon A, Daniels JT otic membrane in limbal stem cell deficiency.
(2010) In sickness and in health: corneal epi- Cornea 34:592–600. https://doi.org/10.
thelial stem cell biology, pathology and ther- 1097/ICO.0000000000000398
apy. Exp Eye Res 90:188–195 10. Ramı́rez BE, Sánchez A, Herreras JM,
3. Garg P, Krishna PV, Stratis AK, Gopinathan U Fernández I, Garcı́a-Sancho J, Nieto-Miguel T,
(2005) The value of corneal transplantation in Calonge M (2015) Stem cell therapy for cor-
reducing blindness. Eye 19:1106–1114. neal epithelium regeneration following good
https://doi.org/10.1038/sj.eye.6701968 manufacturing and clinical procedures. Biomed
4. Pellegrini G, Traverso CE, Franzi AT, Res Int 2015:408495. https://doi.org/10.
Zingirian M, Cancedda R, De Luca M (1997) 1155/2015/408495
Long-term restoration of damaged corneal sur- 11. Calonge M, Pérez I, Galindo S, Nieto-Miguel-
faces with autologous cultivated corneal epi- T, López-Paniagua M, Fernández I,
thelium. Lancet 349:990–993. https://doi. Alberca M, Garcı́a-Sancho J, Sánchez A, Her-
org/10.1016/S0140-6736(96)11188-0 reras JM (2019) A proof-of-concept clinical
5. Shortt AJ, Secker GA, Rajan MS, Meligonis G, trial using mesenchymal stem cells for the treat-
Dart JK, Tuft SJ, Daniels JT (2008) Ex vivo ment of corneal epithelial stem cell deficiency.
expansion and transplantation of limbal epithe- Transl Res 206:18–40. https://doi.org/10.
lial stem cells. Ophthalmology 1016/J.TRSL.2018.11.003
115:1989–1997. https://doi.org/10.1016/J. 12. Lindberg K, Brown ME, Chaves HV, Kenyon
OPHTHA.2008.04.039 KR, Rheinwald JG (1993) In vitro propagation
6. Rama P, Matuska S, Paganoni G, Spinelli A, De of human ocular surface epithelial cells for
Luca M, Pellegrini G (2010) Limbal stem-cell transplantation. Investig Ophthalmol Vis Sci
therapy and long-term corneal regeneration. N 34:2672–2679
Engl J Med 363:147–155. https://doi.org/ 13. Shortt AJ, Secker GA, Notara MD, Limb GA,
10.1056/NEJMoa0905955 Khaw PT, Tuft SJ, Daniels JT (2007) Trans-
7. Baylis O, Figueiredo F, Henein C, Lako M, plantation of ex vivo cultured limbal epithelial
Ahmad S (2011) 13 years of cultured limbal stem cells: a review of techniques and clinical
epithelial cell therapy: a review of the out- results. Surv Ophthalmol 52:483–502.
comes. J Cell Biochem 112(4):993–1002 https://doi.org/10.1016/J.
8. Zakaria N, Possemiers T, Dhubhghaill SN, SURVOPHTHAL.2007.06.013
Leysen I, Rozema J, Koppen C, Timmermans 14. Mariappan I, Maddileti S, Savy S, Tiwari S,
JP, Berneman Z, Tassignon MJ (2014) Results Gaddipati S, Fatima A, Sangwan VS,
of a phase I/II clinical trial: standardized, Balasubramanian D, Vemuganti GK (2010) In
vitro culture and expansion of human limbal Roald B, Lyberg T (2009) Comparison of the
epithelial cells. Nat Protoc 5:1470–1479. histology, gene expression profile, and pheno-
https://doi.org/10.1038/nprot.2010.115 type of cultured human limbal epithelial cells
15. López-Paniagua M, Nieto-Miguel T, de la from different limbal regions. Investig
Mata A, Galindo S, Herreras JM, Corrales Ophthalmol Vis Sci 50:5165–5172. https://
RM, Calonge M (2013) Consecutive expan- doi.org/10.1167/iovs.08-2884
sion of limbal epithelial stem cells from a single 18. Notara M, Shortt AJ, O’Callaghan AR, Daniels
limbal biopsy. Curr Eye Res 38:537–549. JT (2013) The impact of age on the physical
https://doi.org/10.3109/02713683.2013. and cellular properties of the human limbal
767350 stem cell niche. Age (Omaha) 35:289–300.
16. López-Paniagua M, Nieto-Miguel T, de la https://doi.org/10.1007/s11357-011-9359-
Mata A, Galindo S, Herreras JM, Corrales 5
RM, Calonge M (2017) Successful consecutive 19. James SE, Rowe A, Ilari L, Daya S, Martin R
expansion of limbal explants using a biosafe (2001) The potential for eye bank limbal rings
culture medium under feeder layer-free condi- to generate cultured corneal epithelial allo-
tions. Curr Eye Res 42:685–695. https://doi. grafts. Cornea 20:488–494. https://doi.org/
org/10.1080/02713683.2016.1250278 10.1097/00003226-200107000-00010
17. Utheim TP, Raeder S, Olstad OK, Utheim ØA,
de la Paz M, Cheng R, Huynh TT, Messelt E,
Chapter 5
A Guide to the Development of Human CorneaOrganoids

from Induced Pluripotent Stem Cells in Culture
James W. Foster, Karl J. Wahlin, and Shukti Chakravarti
Abstract
The cornea is the outermost transparent and refractive barrier surface of the eye necessary for vision.
Development of the cornea involves the coordinated production of extracellular matrix, epithelial differen-
tiation, and endothelial cell expansion to produce a highly transparent tissue. Here we describe the
production of multilayered three-dimensional organoids from human-induced pluripotent stem cells.
These organoids have the potential for multiple downstream applications which are currently unattainable
using traditional in vitro techniques.
Key words Cornea, Organoids, iPSC, Corneal epithelium, Corneal stroma
1 Introduction
The cornea is the outermost barrier tissue of the eye and may be
viewed as a specialized skin to the eye. From a tissue engineering
point of view, made up of three major resident cell types, it is a
relatively simple tissue. The outer most layer of the cornea consists
of a stratified epithelium, a central stroma of orthogonally stacked
collagen fibril lamellae interspersed with flattened cells, the kerato-
cytes, and an innermost single cell layered endothelium. The self-
renewal potential is highest in the basal epithelial cells of the stra-
tified epithelium, decreasing dramatically in the stromal and endo-
thelial cells [1–3].
In development, the cornea arises from the cranial ectoderm
and neural crest cells which subsequently differentiate into the Pax6
positive ocular surface ectoderm (giving rise to the corneal epithe-
lium) and the neuroectoderm (keratocytes and endothelial cells)
[4, 5]. These cell types must then coordinate to produce the highly
aligned collagenous extracellular matrix necessary for the unob-
structed passage of incident light [6].
The generation of induced pluripotent stem cells (iPSC) has
revolutionized the field of regenerative medicine and has allowed
51
52 James W. Foster et al.
for investigations into development and disease and opened new

avenues for therapeutic intervention [7]. In the eye, the focus has
often been on the development of the retina and in this the devel-
opment of optic cups has been a major advance [8–12]. Since these
early investigations the field has progressed to produce ever more
specific cell types [10, 13–15]. In the cornea field, there has been
progress in the generation of iPSC-derived corneal epithelial cells
[16], keratocytes [17], endothelial cells [18], and all cell types in
2D culture [19]. Thus, we sought to leverage the development of
iPSC-derived corneal progenitors in a self-assembly model of cor-
neal development to produce stem cell-derived corneal organoids
[20]. Here we describe these techniques in detail.
2 Materials
2.1 IMR 90.4 iPSC IMR-90.4 iPSC cells are maintained by clonal propagation on
Cell Line Culture growth factor-reduced Matrigel in mTeSR1 medium under hypoxia
(10% CO2, 5% O2) in a copper-lined incubator.
2.2 Small Molecules 1. 10 mM all-trans-retinoic acid (ATRA): 50 mg ATRA, 16 mL of

and Supplements dimethyl sulfoxide (DMSO). Store at 80 C for up to
4 months.
2. 10 mM ( )-Blebbistatin: 1 mg blebbistatin, 340 μL DMSO.
Make 10 aliquots and store at 80 C.
3. E6 supplement: 100 mL water, 7.5 g NaHCO3, 97 mg insulin,
53.5 mg holo-transferrin, 320 mg L-ascorbic acid, 70 mg
sodium selenite. Mix and make 10 mL aliquots into 15 mL
tubes. Store in designated box at 80 C.
4. 3 mM IWR-1-endo: 10 mg IWR, 8.1 mL DMSO. Make 50 μL
aliquots and store at 80 C.
5. 3.7 M sodium chloride (NaCl): 10.95 g NaCl, 50 mL cell
culture grade water. NaCl stock solution cannot be easily filter
sterilized.
6. 100 μM smoothened agonist (SAG): 1 mg SAG, 16 mL
DMSO. Make 100 μL aliquots and store at 80 C.
7. 400 mM Taurine: 500 mg taurine, 10 mL sterile cell culture
grade water. Make 1.25 mL aliquots and store at 80 C.
8. Matrigel-Growth Factor Reduced (GFR).
9. Accutase.
10. Hanks buffered saline solution (HBSS).
2.3 Medium 1. mTeSR1 medium.

Solutions 2. Neural induction medium (NIM): Dulbecco’s modified Eagle
medium (DMEM), 1% B27 minus vitamin A, 1 mM pyruvate,
iPSC-Derived Corneal Organoids 53
1 nonessential amino acids (NEAA), 1 Glutamax, 2 E6

supplement, 0.88 g/L NaCl. NaCl should raise the osmolarity
by +30 mOsm to ~330–340 mOsm. Sterilize using a 0.22 μm
filter.
3. Optic vesicle induction medium: NIM, 100 nM of the smooth-
ened agonist (SAG).
4. Long-term retina maturation medium (LTR): 3:1 mix of
DMEM:F12, 1% B27 supplement, 10% heat-inactivated fetal
bovine serum (FBS), 1 mM pyruvate, 1 NEAA, 1 Gluta-
max, 1 mM taurine. LTR is sterilized with a 0.22 μm sterile
filter. (For inclusion of 500 nM all-trans-retinoic acid,
see Note 1).
2.4 Tungsten Tungsten needles are made by embedding a 0.64 mm diameter

Needles for Neural tungsten wire into a hollow glass rod cut to size with a glass cutter
Vesicle Excision and mounted using epoxy resin. The tungsten wire is connected to
the cathode of a gel electrophoresis DC power supply; a carbon rod
is connected to the anode. Tips are immersed in a 0.5 M NaOH
solution and low voltage (30 V) applied. Bubbles forming at the tip
of the tungsten wire indicate active electrolytic sharpening. The tip
of the tungsten wire should be repeatedly immersed in an up/down
motion until the tip becomes very sharp. The first-time electrodes
are sharpened it can be helpful to bevel the edges with a sharpening
stone. Sharpening is recommended prior to each cutting session.
3 Methods
This method is permissive for generating retinal organoids, corneal

organoids, and hybrids. It is at the discretion of the investigator to
select for those organoids which meet their needs. A visual repre-
sentation of this method is provided in Fig. 1.
3.1 Stem Cell Culture 1. IMR-90.4 iPSCs should be maintained in mTeSR1 medium on
dishes coated with 1% (v/v) Matrigel-GFR™ at 37 C under
hypoxic conditions prior to reaggregation.
2. Replace medium every day as per standard cell culture
conditions.
3. Once colonies reach 70% of a field of view under 10 magnifi-
cation, treat cells with Accutase for 8–10 min.
4. Dissociate into a single cell suspension and quench Accutase
activity with mTeSR1 plus 5 μM blebbistatin (B) to improve
single cell survival.
5. Spin cells at 80 g for 5 min.
6. Resuspended in mTeSR1 + B and plated at 5000 cells per
35 mm well.
Fig. 1 Visual representation of corneal organoid methodology. IPS cells are taken through sequential stages of
differentiation from pluripotency, neural phenotype, optic induction, and finally corneal selection over 30 days
7. After 48 h, feed cells with mTeSR1 alone.

8. The entire protocol is carried out without antibiotics to mini-
mize cell stress.
3.2 Forced 1. On day 1, passage cells as described in Subheading 3.1.

Aggregation 2. Resuspend cells in mTeSR1 containing 5 μM blebbistatin and
count with a hemocytometer.
3. For a full 96-well plate, 3.0 105 cells are diluted in 5.0 mL of
media (60,000 cells/mL) and mixed by inversion 10 times.
4. Transfer 50 μL of cell suspension (corresponding to 3000 cells)
to a ChannelMate reservoir and add to each well of an ultralow
attachment U-bottom 96-well plate using a multichannel
pipette. This approach quickly dispenses cells and minimizes
variability.
5. After plating, place cells back in hypoxia overnight at 37 C to
facilitate cell survival and aggregate formation (see Note 2).
3.3 Neural Induction 1. On day 2 induce forced aggregates to a neuronal phenotype by

supplementing with an equal volume (50 μL) of NIM supple-
mented with 2% Matrigel and 6 μM IWR1-e.
2. Transfer organoids to normoxia on day 2 (5% CO2, 20% O2)
and maintain for the duration of the experiment.
3. On days 3 and 4, feed aggregates daily by addition of 50 μL of
NIM supplemented with 1% Matrigel and 3 μM IWR1-e.
4. On days 5–7, feed aggregates daily with a 50% media change
(100 μL) with NIM supplemented with 1% Matrigel and 3 μM
IWR1-e.
5. On day 9, feed aggregates with a 50% media change (100 μL)
with NIM supplemented without matrigel or 3 μM IWR1-e.
Media is changed every other day at this point.
6. On day 11, pool aggregates into 15 mL tubes and are allowed
to sink (see Notes 3 and 4).
7. Rinse the aggregates three times with Hanks buffered saline
solution (HBSS) to remove Matrigel, cell debris, and other
components of NIM.
8. Resuspend in 20 mL of optic vesicle medium in ultralow
attachment T-75 flasks or 15 mL of medium in untreated
deep dish 10 cm polystyrene petri dishes and keep in a standard
tissue culture incubator.
9. Feed vesicles every other day with >75% media exchange (see
Note 5).
3.4 Neural Vesicle 1. On days 11 and 13, feed organoids with NIM media +
Excision 100nM SAG. On days 11–15, visually identify and manually
excise vesicles from the rest of the organoids using a micro-
scope and a scissoring motion with fine-tipped tungsten nee-
dles (see Note 6).
2. Once isolated from the central mass, place the presumptive
neural vesicles back in optic vesicle medium containing
100 nM SAG in ultralow binding T-75 flasks or untreated
10 cm polystyrene petri dishes in a standard normoxic incuba-
tor at 37 C (see Note 7).
3.5 Organoid 1. On days 15–19, feed vesicles with LTR media + SAG every
Maturation 2 days.
and Selection 2. On day 21, feed vesicles with LTR + ATRA every 2–3 days for
the duration of the experiment (see Note 8). By day 31, corneal
organoids should be apparent by their translucent cystic
appearance. These structures differ from retinal organoids or
RPE organoids that have distinctive cup-shaped or darkly pig-
mented appearances, respectively (see Note 9).
3.6 Validation Corneal organoids can be differentiated from retinal progenitors by

their lack of pigmented cells, transparent appearance, and larger
cystic morphology. For molecular validation we recommend assay-
ing by immunofluorescence where good antibodies are available or
a transcript level assessment by RT-PCR:
1. PAX6—Ocular phenotype.
2. P63α—Limbal epithelium.
3. KERA—Stromal keratocyte.
4. KRT14—Basal epithelium.
5. COL8A1—Bowman’s membrane.
6. Six6—Retinal phenotype (should be absent as tested by
RT-PCR).
7. OCT4 and Nanog—Should be absent.
Additional information on assays and antibodies for these mar-
kers are provided in our previous work [15, 20].
4 Notes
1. ATRA is not added to LTR between day 15 and 20; it is added

fresh for each feeding after day 20.
2. It is strongly recommended to tap the sides of the 96-well dish
intermittently for up to 1 h after plating to ensure that solitary
organoids form.
3. To minimize damage to the aggregates we recommend using

large diameter 1000 μL tips for collection or 200 μL tips which
have had the lower ¼ removed to expand their smallest
diameter.
4. Failure to form well-organized vesicles from between days
6 and 12 indicates a failure of differentiation which can lead
to unreliable corneas. These aggregates should be discarded.
5. Organoids are visualized every day until day 10, to verify that
neural vesicles budding from the central mass have formed.
6. Detailed images of the budding neuronal vesicles can be found
in “Wahlin, K. J. et al. Photoreceptor Outer Segment-like
Structures in Long-Term 3D Retinas from Human Pluripotent
Stem Cells. Sci. Rep. 7, 766 (2017)”.
7. Too many organoids can adversely affect cell survival; thus, it
may be necessary to titrate different numbers of organoids for
long-term survival if other cell lines are used. A range from
20 to 50 vesicles per 10 cm dish is recommended.
8. Corneal organoids can be allowed to mature at the discretion of
the researcher. It is possible to mature these organoids
>120 days if so desired.
9. It is common for organoids to contain areas of dense pigmen-
ted cells and nonpigmented cells which correspond to pre-
sumptive RPE and corneal regions, respectively.
Acknowledgements
This work was supported by the NIH EY026104 (SC) and

R00EY024648 (KW).
References
1. Cotsarelis G, Cheng SZ, Dong G et al (1989) Mol Biol Transl Sci 134:43–59. https://doi.
Existence of slow-cycling limbal epithelial basal org/10.1016/bs.pmbts.2015.04.003
cells that can be preferentially stimulated to 5. Graw J (2010) Eye development. Curr Top
proliferate: implications on epithelial stem Dev Biol 90:343–386. https://doi.org/10.
cells. Cell 57:201–209 1016/S0070-2153(10)90010-0
2. Lavker RM, Tseng SCG, Sun T-T (2004) Cor- 6. Hassell JR, Birk DE (2010) The molecular
neal epithelial stem cells at the limbus: looking basis of corneal transparency. Exp Eye Res
at some old problems from a new angle. Exp 91:326–335. https://doi.org/10.1016/j.
Eye Res 78:433–446 exer.2010.06.021
3. Scott S-G, Jun AS, Chakravarti S (2011) 7. Carey BW, Markoulaki S, Hanna J et al (2009)
Sphere formation from corneal keratocytes Reprogramming of murine and human somatic
and phenotype specific markers. Exp Eye Res cells using a single polycistronic vector. Proc
93:898–905. https://doi.org/10.1016/j. Natl Acad Sci 106:157–162. https://doi.org/
exer.2011.10.004 10.1073/pnas.0811426106
4. Lwigale PY (2015) Corneal development: dif- 8. Eiraku M, Takata N, Ishibashi H et al (2011)
ferent cells from a common progenitor. Prog Self-organizing optic-cup morphogenesis in
three-dimensional culture. Nature 472:51–56. Biol 801:157–164. https://doi.org/10.1007/

https://doi.org/10.1038/nature09941 978-1-4614-3209-8_20
9. Nakano T, Ando S, Takata N et al (2012) Self- 15. Wahlin KJ, Maruotti JA, Sripathi SR et al
formation of optic cups and storable stratified (2017) Photoreceptor outer segment-like
neural retina from human ESCs. Cell Stem Cell structures in long-term 3D retinas from
10:771–785. https://doi.org/10.1016/j. human pluripotent stem cells. Sci Rep 7:766.
stem.2012.05.009 https://doi.org/10.1038/s41598-017-
10. Zhong X, Gutierrez C, Xue T et al (2014) 00774-9
Generation of three-dimensional retinal tissue 16. de la Torre RA MG, Nieto-Nicolau N,
with functional photoreceptors from human Morales-Pastor A, Casaroli-Marano RP
iPSCs. Nat Commun 5:4047. https://doi. (2017) Determination of the culture time
org/10.1038/ncomms5047 point to induce corneal epithelial differentia-
11. Vergara MN, Flores-Bellver M, Aparicio- tion in induced pluripotent stem cells. Trans-
Domingo S et al (2017) Three-dimensional plant Proc 49:2292–2295. https://doi.org/
automated reporter quantification (3D-ARQ) 10.1016/j.transproceed.2017.09.047
technology enables quantitative screening in 17. Joseph R, Srivastava OP, Pfister RR (2014)
retinal organoids. Development Generation of induced pluripotent stem cells
144:3698–3705. https://doi.org/10.1242/ from normal and keratoconus corneal fibro-
dev.146290 blasts using viral- and non-viral methods.
12. Eldred MK, Charlton-Perkins M, Muresan L, Invest Ophthalmol Vis Sci 55:523–523
Harris WA (2017) Self-organising aggregates 18. Zhao JJ, Afshari NA (2016) Generation of
of zebrafish retinal cells for investigating human corneal endothelial cells via in vitro
mechanisms of neural lamination. Develop- ocular lineage restriction of pluripotent stem
ment 144:1097–1106. https://doi.org/10. cells. Invest Ophthalmol Vis Sci
1242/dev.142760 57:6878–6884. https://doi.org/10.1167/
13. Maruotti J, Wahlin K, Gorrell D et al (2013) A iovs.16-20024
simple and scalable process for the differentia- 19. Hayashi R, Ishikawa Y, Sasamoto Y et al (2016)
tion of retinal pigment epithelium from human Co-ordinated ocular development from human
pluripotent stem cells. Stem Cells Transl Med iPS cells and recovery of corneal function.
2:341–354. https://doi.org/10.5966/sctm. Nature 531(7594):376–380. https://doi.
2012-0106 org/10.1038/nature17000
14. Wahlin KJ, Maruotti J, Zack DJ (2014) Mod- 20. Foster JW, Wahlin K, Adams SM et al (2017)
eling retinal dystrophies using patient-derived Cornea organoids from human induced plurip-
induced pluripotent stem cells. Adv Exp Med otent stem cells. Sci Rep 7:41286. https://doi.
org/10.1038/srep41286
Chapter 6
Gene Editing for Corneal Stromal Regeneration

Tara Moore, Connie Chao-Shern, Larry DeDionisio, Kathleen A. Christie,
and M. Andrew Nesbit
Abstract
CRISPR/Cas9 gene editing holds the promise of sequence-specific alteration of the genome to achieve
therapeutic benefit in the treated tissue. Cas9 is an RNA-guided nuclease in which the sequence of the RNA
can be altered to match the desired target. However, care must be taken in target choice and RNA guide
design to ensure both maximum on-target and minimum off-target activity. The cornea is an ideal tissue for
gene therapy due to its small surface area, accessibility, immune privilege, avascularity, and ease of visualiza-
tion. Herein, we describe the design, testing, and delivery of Cas9 and guide RNAs to target genes
expressed in the cornea.
Key words Gene editing, sgRNA expression construct, Dual luciferase assay, In vitro digest, Lym-
phoblastoid cell line, Nucleofection, Polymerase chain reaction (PCR), Corneal endothelial and
epithelial cell culture, In vivo imaging, Intracameral injection, Adeno-associated virus (AAV)
1 Introduction
The discovery, development, and application of Clustered Regu-

larly Interspaced Palindromic Repeats (CRISPR)/CRISPR asso-
ciated protein (Cas) systems has brought the promise of gene
therapy, tailored specifically for the patient, closer to realization.
CRISPR/Cas9 is a bacterial RNA-guided endonuclease that has
been modified for use in mammalian cells, consisting of a Cas9
nuclease that forms a complex with a simplified single guide RNA
(sgRNA), formed by the fusion of crRNA and tracrRNAs, which
guides the nuclease to its DNA target [1–3]. In addition to a match
between the RNA and DNA sequences, the Cas9 nuclease itself
recognizes and binds to a protospacer adjacent motif (PAM)
directly upstream of the guide RNA sequence (Fig. 1). Following
target recognition, the Cas9 nuclease makes a double strand break
(DSB) which triggers the DNA repair machinery of the cell, leading
to either error-prone nonhomologous end joining (NHEJ) or
precise homology directed repair (HDR) [4]. It is by harnessing
59
60 Tara Moore et al.
Fig. 1 S. pyogenes Cas9 (purple outline) can be directed to cut any sequence in the genome (DNA target in
grey), provided it is directly upstream of a protospacer adjacent motif known as PAM (pink box). This can be
achieved by altering the 20 nucleotide guide sequence, which is associated with a 82 nucleotide scaffold [8]
these different cellular responses that different forms of gene edit-

ing can be achieved. It is important to understand, however, that
DNA cleavage by Cas9 nuclease can occur in some cases where
there are mismatches between the guide sequence and the target
[5, 6]. This can lead to off-targeting elsewhere in the genome or, as
we have shown [7], failure to discriminate between mutant and
wild-type alleles. Selectivity between wild-type and mutant alleles
can be improved if the mutation is located within the PAM [7–9].
Successful application of targeted CRISPR/Cas9 gene editing
requires a stepwise approach by (1) careful selection of the target,
(2) testing the efficiency with which the Cas9/sgRNA nuclease
cleaves the selected target (on-target), (3) the specificity for the
on-target compared to off-target, and (4) consideration of the
method of delivery to the chosen tissue.
2 Materials
2.1 SpCas9 sgRNA 1. Sequence of target gene.

Design 2. Online design program (Benchling, CasOFFinder, etc.).
2.2 sgRNA Construct 1. S. pyogenes Cas9 vector plasmid or S. aureus Cas9 vector
Preparation plasmid.
2. BbsI restriction endonuclease.
3. Guide oligos containing the following template:
4. Top guide oligo—50 -CACCG - insert 20 nt of sgRNA – 30 .
5. Bottom guide oligo—50 -C - insert complement of 20 nt
CAAA – 30 .
Gene Editing for Corneal Stromal Regeneration 61
6. T4 DNA ligase.
7. LB Broth.
8. LB Agar.
9. Antibiotic—check plasmid map to see which is needed.
10. dH5α E. coli competent cells.
11. Plasmid DNA isolation kit.
2.3 Human 1. HEK AD293 cells.

Embryonic Kidney 2. DMEM low glucose.
Cells (HEK) AD293
3. Fetal bovine serum.
Maintenance
4. Trypsin/EDTA (0.25%).
5. Phosphate-buffered saline.
2.4 Dual Luciferase 1. DMEM low glucose.

Assay 2. Fetal bovine serum.
3. S. pyogenes Cas9 vector plasmid containing sgRNA of interest
pSpCas9(BB)-2A-Puro (PX459) V2.0.
4. psiTEST-LUC-target vector containing target sequence of
sgRNA.
5. Renilla luciferase reporter plasmid.
6. Lipofectamine 2000.
7. Optimem.
8. Dual-Luciferase® Reporter Assay System.
9. BMG Labtech, LUMIstar Optima plate reader (or equivalent).
2.5 In Vitro DNA 1. S. pyogenes EnGen Cas9 S.pyogenes and reaction buffer.
Cleavage Assay 2. Modified synthetic sgRNA.
3. Nuclease-free H2O.
4. Proteinase K.
5. DNA cleavage template containing 20 bp target site and adja-
cent PAM (The cleavage template can either be circular or
linearized plasmid, PCR products, or synthesized
oligonucleotides).
2.6 Generate 1. RPMI media.

Lymphocyte Cell 2. Fetal bovine serum.
Line (LCL)
3. Ficoll-Paque PLUS.
4. Epstein-Barr virus (Human gammaherpesvirus 4 (HHV-4),
ATCC® VR-1492).
5. Phytohemagglutinin.
2.7 Nucleofection 1. S. pyogenes EnGen Cas9 NLS.

of LCLs 2. Modified synthetic sgRNA.
with Ribonucleic
3. SF Cell line 4D-Nucleofector X kit.
Proteins (RNPs)
4. Lonza 4D nucleofector.
5. QIAamp DNA Mini Kit.
2.8 Polymerase 1. Dream Taq.

Chain Reaction (PCR) 2. Primers flanking region of interest.
3. DNA to be amplified.
4. Thermocycler.
2.9 Intrastromal 1. Capillary Glass, 1.0 mm outer diameter, 0.58 mm inner diam-
Injection of CRISPR/ eter, 10.16 cm (4 in.).
Cas9 2. DMZ universal puller (or equivalent).
3. Hamilton 701 RN 10 μL syringe without needle.
4. Hamilton RN Compression Fitting 1 mm.
5. Ketamine.
6. Xylazine.
7. dH2O for injections,
8. Tropicamide.
9. Phenylephrine.
10. Tooth (palate) bar only from model 923-B Mouse Gas Anes-
thesia Head Holder.
11. Gas mask.
12. Surgical microscope.
13. Fusidic gel.
2.10 IVIS In Vivo 1. Isoflurane.

Imaging 2. Isoflurane chamber.
of Fluorescence
3. Xenogen IVIS Lumina in vivo imager.
4. Custom mouse holder tube (Fig. 2).
5. 1.5–2% isoflurane.
3 Methods
3.1 sgRNA Design sgRNAs are designed using an online design program (Benchling,
CasOFFinder, etc.) and chosen on the basis of good on- and
off-target scores. Overhangs for the BbsI restriction site (shown
in bold) are added to the 20 bp sgRNA sequence of the target site.
Fig. 2 Custom-made mouse holder for use in the IVIS that allows eyes to be
placed at optimal angle to the camera
Top and bottom sgRNA oligos are ordered using the following
template:
Top guide oligo—50 -CACCG insert 20 nt of sgRNA - 30
Bottom guide oligo—50 -C insert complement of 20 nt CAAA - 30
3.2 CRISPR/Cas9 1. Digest pSpCas9(BB)-2A-Puro (PX459) V2.0 with BbsI and

Construct Preparation Buffer 2.1 for 2 hours at 37 C using the following reaction
mixture:
Volume
H2O Up to 50 μL
Buffer 2.1 (10) 5 μL
Plasmid DNA 1 μg
BbsI restriction enzyme 1 μL
2. Digestion mix is then electrophoresed on a 1% agarose gel

alongside a DNA ladder; linearized pSpCas9(BB)-2A-Puro
(PX459) V2.0 plasmid will run at 9 kb. Purify gel fragment
containing your cloning backbone.
3. Resuspend the top and bottom strands of oligos for each guide
oligo previously designed to a final concentration of 10 μM.
Prepare the following mixture for annealing the sgRNA oligos
(top and bottom strands):
Volume
Top Oligo (100 μM) 1 μL
Bottom Oligo (100 μM) 1 μL
H2O 7 μL
5 DNA Ligase Buffer 2 μL
4. Anneal the oligos in a thermocycler by using the following

parameters: 37 C for 30 min; 95 C for 5 min; ramp down
to 25 C at 5 C/min.
5. Dilute the oligos (2 μL annealed oligo: 98 μL water).
6. Set up a ligation reaction for each sgRNA, as described below.
Include a no-insert, pSpCas9(BB)-only negative control for
ligation.
Volume
Digested SpCas9 Plasmid 100 ng
Diluted annealed Oligos 2 μL
5 DNA Ligation Buffer 4 μL
H2O X μL
DNA Ligase Buffer 1 μL
Total Reaction Volume 20 μL
7. Incubate the ligation mixture for 1 hour at room temperature.

8. Add 4 μL ligation mixture to 45 μL of competent E. coli strain
e.g. DH5α cells in a prechilled tube.
(a) Leave on ice for 30 min.
(b) Heat shock for 45 s in a preheated water bath at 42 C.
(c) Leave on ice for 2 min.
(d) Add 150 μL of prewarmed LB broth to centrifuge tube.
(e) Incubate in shaking incubator at 37 C for 1 hour .
(f) Plate all 200 μL transformation mix onto an LB agar plate
containing 100 μg/mL ampicillin and place in 37 C
incubator overnight.
9. Inspect the plates for colony growth. Pick colonies with a sterile
tip and place into 20 μL LB broth in a 0.5 mL tube, select ~4
colonies per plate. Shake tip in LB to dislodge bacteria, remove,
and discard tip.
10. Vortex each tube briefly to disperse bacterial colony.
11. Perform colony PCR to determine successful ligation using the

reaction below:
Reagents 1 Reaction
5 Dream Taq Buffer 4 μL
Forward primer 10 μM 1 μL
(Cas9 BB seq Fwd–5-gggaaacgcctggtatcttt-30 )
Reverse primer 10 μM 1 μL
(Bottom guide oligo—custom for each sgRNA)
Colony broth 2 μL
H2O 11.92 μL
Dream Taq 0.08 μL
Total reaction volume 20 μL
12. Run PCR program:
Pre-PCR holding stage 95 C for 3 min

Cycling stage (35 cycles) 95 C for 15 s
60 C for 15 s
72 C for 30 s
Post-PCR holding stage 72 C for 5 min
13. After PCR, run the material on 1% agarose 1 TBE gel.

14. If the sgRNA has been inserted, there should be a product size
of 210 bp. The reverse primer used in the PCR reaction is the
bottom guide oligo; therefore, this customized for each reac-
tion. The forward primer primes off the pSpCas9(BB)-2A-
Puro (PX459) V2.0 backbone. Consequently, there should
only be a product in cases where the 20 bp sgRNA has been
inserted into the digested plasmid.
15. Inoculate 3 mL LB broth with 100 μg/mL ampicillin with
broth containing dispersed insert-positive colony in LB broth.
16. Incubate culture in a shaking incubator at 37 C overnight.
17. Isolate plasmid DNA from the cultures using a plasmid DNA
purification kit, following the manufacturer’s protocol.
3.3 HEK AD293 Cell 1. AD293 cells are cultured in 1 DMEM (containing 1 g/L
LINE MAINTENANCE glucose, 4.0 mM L-Glutamine, 1.0 mM Sodium Pyruvate) and
10% heat-inactivated fetal bovine serum (FBS) in an incubator
at 37 C with 5% CO2.
2. Subculture conditions: Split sub-confluent cultures (70–80%)

from 1:2 to 1:3, seeding at 1 106 cells/cm2 in a 75 cm2 flask,
using 1 trypsin.
3. Remove the growth medium by aspiration.
4. Wash cells once with 2 mL phosphate-buffered saline.
5. Trypsinize cells for 1–3 min in 1 trypsin/EDTA at 37 C.
6. Dilute the cells with growth medium (volume at least equal to
the volume of trypsin/EDTA solution added) to inactivate the
trypsin.
7. Transfer cell suspension to a conical tube and centrifuge at
1000 g for 5 min at room temp.
8. Aspirate the supernatant and resuspend cells in 10 mL growth
medium.
9. Seed cells at appropriate cell density. Place the cells in a 37 C
incubator at 5% CO2. Monitor cell density daily.
3.4 Preparation 1. Co-digest the psiTEST-LUC-target vector with NheI and

of psiTEST-LUC-target KpnI in Buffer 1.1 for 2 hours at 37 C. This plasmid will be
Vector used for both the dual-luciferase assay and the in vitro digest
reaction.
Volume
H2O Up to 50 μL
Buffer 2.1 (10) 5 μL
Plasmid DNA 1 μg
KpnI restriction enzyme 1 μL
NheI restriction enzyme 1 μL
Digestion mix is then electrophoresed on a 1% agarose gel

alongside a DNA ladder; desired band will run at 5223 bp and
excised band will run at 12 bp. Purify gel fragment containing
your cloning backbone.
2. Design a 50 bp template containing your 20 bp target sequence
and adjacent PAM including overhangs complementary to the
NheI and KpnI restriction sites.
3. Following steps 3–8 in Subheading 3.2 ligate the 50 bp
dsODN into the digested psiTEST-LUC-target vector.
4. Verify the 50 bp dsODN has been inserted into the digested
backbone using a colony PCR as previously described (Sub-
heading 3.2, steps 9–13). Correct insertion will result in a
PCR product of 641 bp; an empty backbone will result in a
PCR product of 591 bp.
5. This construct can now be used directly in the dual-luciferase

assay.
6. For the in vitro digest use the psiTEST-LUC-target vector
containing the 50 bp region encompassing your target site
and adjacent PAM for PCR. Prepare PCR product containing
the target sequence using the following primers:
(a) FWD: 50 – ACCCCAACATCTTCGACGCGGGC-30
(b) REV: 50 – TGCTGTCCTGCCCCACCCCA – 30
7. Purify PCR product and use in the in vitro cleavage reaction at
the specified concentration of 3 nM. The PCR product gener-
ated using the primers described in step 6 generates a template
of 587 bp which has a M.W of 356701.7 g/mol (see Note 6).
3.5 Dual Luciferase 1. Dual Luciferase assay workflow (Fig. 3).

Assay 2. The day prior to transfection seed AD293 cells in a volume of
100 μL in a 96-well plate at a density of 6.5 103 cells per well.
3. Cells are transfected using Lipofectamine 2000 according to
the manufacturers’ instructions.
4. For each well the reaction mix is as follows:
Reagents 1 reaction
Firefly luciferase reporter plasmid 20 ng
CRISPR/Cas9 expression construct 80 ng
Renilla luciferase plasmid 1 ng
Lipofectamine 2000 0.2 μL
Optimem 25 μL
DMEM with 10% FBS 25 μL
5. Prepare a mastermix for your experimental design.

6. Combine Lipofectamine 2000 and Optimem and incubate at
room temperature for 5 min.
7. Add Renilla plasmid.
8. Add Reporter construct plasmid.
9. Add sgRNA construct plasmid and incubate at room tempera-
ture for 5 min.
10. Add DMEM with 10% FBS.
11. Remove 50 μL culture medium from each well.
12. Add 50 μL transfection mixture to each well.
13. Incubate for 72 h FBS in an incubator at 37 C with 5% CO2.
14. Prepare the lysates using the Dual-Luciferase® Reporter Assay
System kit, following the manufacturer’s instructions
(see Note 1).
Fig. 3 Explanation of dual-luciferase assay. Diagram showing the different constructs that were transfected for
the dual luciferase assay. The Renilla luciferase plasmid (Red) is used to normalize the firefly luciferase read, as
it is not affected by the addition of sgGeneWT or sgGeneMut(ant). Firefly luciferase is shown in orange, while the
sgRNAs for GeneWT and GeneMut(ant) are shown in green and blue, respectively [8]
3.6 Post-transfection 1. Perform post-transfection luciferase readout on the LUMIstar

Luciferase Readout Optima microplate reader.
2. Aliquot the Dual-Luciferase® Reporter Assay System reagents
prior the experiment into 15 mL centrifuge tubes and store at
80 C.
3. Defrost the LAR and Stop&Glo reagents immediately before
reading the plate (see Note 2).
4. To prepare the Stop&Glo reagent, dilute 50 Stop and Glo
reagent (supplied with kit and kept at 20 C) with the S&G
buffer previously aliquoted to make a 1 solution.
5. In the LUMIstar Reagent Box, insert reagent injector 1 into
the LAR solution and reagent injector 2 into the S&G solution.
Ensure each tube is inserted properly (see Note 3).
6. Open the LUMIstar software.
7. Prime pumps (see Note 4).
8. Click test protocols and select the pre-made program (see
Note 5).
9. Edit the layout, based on the experiment.
10. Press OK and click the measurement icon (Traffic lights icon).
11. When the plate run is complete, open the LUMIstar analysis
software and select the plate run from the drop-down menu.
12. Save data as Excel spreadsheet for analyses.
13. After using the machine, wrap the remaining solutions in foil
and store for future use.
14. Remove injector needles from holes in the bottom of the
machine and place in a waste tube.
15. Place tubing 1 and 2 in a tube containing dH2O, click the
priming icon, and perform the first wash of machine using
dH2O.
16. Place tubing 1 and 2 in a tube containing isopropanol, click the
priming icon, and perform the second wash of machine using
isopropanol.
17. Finally, remove the tubing from isopropanol and prime pumps
using air, ensuring that the needles are still in waste tube.
3.7 In Vitro DNA 1. Prepare a cleavage template containing the target sgRNA
Cleavage Assay sequence and an adjacent PAM. The cleavage template can
either be circular or linearized plasmid, PCR products, or
synthesized oligonucleotides (Subheading 3.4).
2. Prepare the reaction at room temperature in the following
order:
Reagents Volume
H2O 20 μL
10 Cas9 Nuclease Reaction Buffer 3 μL
300 nM sgRNA 3 μL (30 nM)
1 μM Cas9 S. pyogenes 1 μL (30 nM)
Reaction volume 27 μL
3. Pre-incubate for 10 min at 25 C.

4. Add 3 μL 30 nM substrate DNA to the reaction (3 nM final).
5. Incubate at 37 C for 15 min.
6. Add 1 μL of Proteinase K to each sample, mix thoroughly, and
pulse-spin in a microfuge.
7. Incubate at room temperature for 10 min.
8. Run digested products on a 1% agarose 1 TBEgel.
9. The fraction of the PCR product (587 bp) that is digested
indicates the activity of the sgRNA.
10. An example of an in vitro cleavage reaction output is shown
below in Fig. 4.
3.8 Generate 1. Take 5 mL of freshly collected whole blood and place in a

Patient-Derived LCL sterile 50 mL conical tube.
2. Add an equal volume of RPMI medium containing 20% fetal
calf serum and mix by gentle inversion.
3. Place 6.25 mL of Ficoll-Paque PLUS in a separate sterile 50 mL
conical tube.
4. Very carefully add the 10 mL of blood/media mix to the Ficoll-
Paque. Hold the tube with the Ficoll-Paque at an angle of 45
and then using a sterile 3 mL aspirating pipette (Pastette)
gently run the blood down the side of the tube so that it
forms a separate layer above the Ficoll-Paque (see Note 7).
5. Spin the tube at 400 g for 20 min at room temperature using
slow acceleration and slow deceleration (brake off).
6. Remove the tube. The red blood cells will collect at the bottom
of the tube above which will be the Ficoll-Paque layer. The
lymphocytes should form a layer on top of the Ficoll-Paque
layer, while the top layer will be the medium.
7. Insert a clean sterile Pastette to just above the Ficoll-Paque
layer and draw off the lymphocytes, which are placed in a sterile
15 mL conical tube.
8. Centrifuge the lymphocytes at 60–100 g for 4 min.
9. Carefully aspirate and discard the media.
Fig. 4 Confirmation of the specificity achieved using a guide-specific system targeted to prevalent TGFBI
mutations. In vitro digestion of either wild-type or respective mutant TGFBI sequence via Cas9 protein
complexed with an sgRNA [8]
10. Wash the pellet of lymphocytes with RPMI media (20% FCS).
11. Centrifuge the lymphocytes at 60–100 g for 3 min.
12. Aspirate and discard the majority of the media.
13. Resuspend the pelleted lymphocytes in the residual media.
14. Rapidly thaw an aliquot of EBV at 37 C.
15. Add the thawed EBV suspension to the resuspended lympho-
cytes and mix gently.
16. Incubate for 1 hour at 37 C (infection period).
17. Add RPMI, 20% FCS media to EBV-treated lymphocytes to
give a total volume of 3 mL.
18. Add 40 μL of 1 mg/mL phytohemagglutinin and mix gently.
19. Aliquot 1.5 mL of the lymphocyte mixture to 2 of the middle
wells of a 24-well plate.
20. Add PBS to the surrounding wells of the plate to maintain
humidity.
21. Incubate for 24 hours at 37 C with 5% CO2 in tissue culture
incubator.
22. After 24 hours, the lymphoblastoid cells should be observed to
be aggregating.
23. When the media begins to turn yellow, replace with fresh media
and expand cells as appropriate into 6-well plates and small
flasks.
24. Once in flasks, the serum content of the media can be reduced
to 10%.
3.9 Nucleofection 1. RNPs are formed directly in the Lonza Nucleofector SF solu-
of LCLs with RNPs tion (SF Cell line 4D-Nucleofector X kit), and incubated for
10 min at room temperature.
2. Prepare the reagents according to the following order:
Reagent Volume
Synthego modified sgRNA (30 pmol) 2.66 μL
NEB EnGen SpCas9 (20 pmol) 2 μL
Lonza Nucleofector SF solution 20.34 μL
Total Volume 25 μL
3. Using a hemocytometer calculate the number of cells required

per transfection and total number of cells required for the
experiment.
4. Collect the total number of cells by centrifugation (300 g x
5 min) and resuspend in Nucleofector solution by gently pipet-
ting (5 μL/per reaction).
5. 5 μL of each cell solution was added to 25 μL of corresponding

preformed RNPs, mixed and transferred to the nucleofector
16-well strip.
6. The cells were electroporated using the 4D Nucleofector pro-
gram DN-100, 70 μL of pre-warmed media was added to each
well and allowed to recover at room temperature for 5 min.
7. The transfected cells were then transferred to a 24-well plate
containing 200 μL medium per well.
8. Incubate cells at 37 C for 48 hours.
9. 48 hours post nucleofection, extract gDNA from cells using the
QIAamp DNA Mini Kit following the manufacturer’s
instructions.
3.10 PCR 1. Design primers flanking the target site. Primers should gener-
for Tracking of Indels ate a ~700 bp product, the target site should be ~200 bp from
by Decomposition the sequencing start site.
(TIDE) 2. Perform a gradient PCR to determine the annealing
temperature.
3. Set up a PCR reaction as follows, include a non-edited control:
Reagents 1 Reaction
5 Dream Taq Buffer 4 μL
Forward primer 10 μM 1 μL
Reverse primer 10 μM 1 μL
DNA template Plasmid DNA ¼ 1 pg to 1 ng,
Human DNA ¼ 50–100 ng
H2O Make up 20 μL
Dream Taq 0.08 μL
4. Electrophorese PCR products on a 1% agarose gel alongside a

DNA ladder to confirm the PCR reaction has produced a
specific band at the expected size.
5. Purify the PCR product and send test and control PCR pro-
ducts for Sanger sequencing.
6. Upload the Sanger traces to the TIDE online program to
quantify indels in your test sample.
3.11 Intrastromal 1. Glass needles are prepared by pulling glass capillaries using
Injection of CRISPR/ program 17 on a DMZ Universal Puller.
Cas9 Constructs 2. Glass needles are fitted onto a Hamilton 10 μL syringe using a
compression fitting.
3. Anesthetize animals by an intra-peritoneal injection of keta-

mine and xylazine.
4. Position mouse on custom mouse tooth holder under the
surgical microscope. Rotate the head of the mouse to have a
bird’s-eye view of the cornea.
5. Dilate pupils by topically applying one drop of tropicamide and
phenylephrine on each eye.
6. 2 μg of CRISPR/Cas9 expression construct is back-filled into a
glass needle (1.0 mm) attached to a Hamilton 10 μL syringe.
7. The glass needle is used to make a track in the stroma just above
the periphery of the cornea, when the needle feels secure in the
stroma inject the CRISPR/Cas9 expression construct.
8. Fusidic gel was applied topically following injection as an anti-
biotic agent.
3.12 IVIS In Vivo 1. Anesthetize mice using 1.5–2% isoflurane in ~1.5 L/min flow
Imaging of oxygen (setting 2.5 on isoflurane control).
of Fluorescence 2. Mice are placed in a custom mouse holder tube (Fig. 2) and
in Mouse Eyes placed inside a Xenogen IVIS Lumina in vivo imager.
3. A custom program to detect fluorescence or luminescence is
selected and the signal is quantified.
4 Notes
1. Remove media from wells and wash with 100 μL PBS. Then
add 20 μL 1 Passive Lysis buffer to each well, shake gently for
15 min on a plate shaker. Finally insert into the LUMIstar
Optima plate reader.
2. Cover the 15 mL tubes completely in tin foil as reagents are
light sensitive. Put the reagents in the LUMIstar compartment
and close lid.
3. Care needs to be taken to prevent tube mix up.
4. Pump priming icon is on the top row of buttons in the LUMI-
star software.
5. Use Greiner 96 F-bottom black wall microplate with 36 inter-
vals, reading direction #. Volume for each injection is 36 μL.
6. Template ¼ 587 bp – M.W of 356701.7 g/mol, require 3 nM.
7. Avoid excessive mixing of the blood into the Ficoll-Paque layer.
8. Flasks must be coated with 10 μg/mL laminin and 10 mg/mL
chondroitin sulfate (acts like the Descemet’s membrane).
References
1. Jinek M, Chylinski K, Fonfara I, Hauer M, 31(9):827–832. https://doi.org/10.1038/nbt.
Doudna JA, Charpentier E (2012) A program- 2647
mable dual-rna-guided DNA endonuclease in 6. Fu Y, Foden JA, Khayter C, Maeder ML,
adaptive bacterial immunity. Science Reyon D, Joung JK, Sander JD (2013) High-
(New York, NY) 337(6096):816–821. https:// frequency off-target mutagenesis induced by
doi.org/10.1126/science.1225829 crispr-cas nucleases in human cells. Nat Biotech-
2. Mali P, Yang L, Esvelt KM, Aach J, Guell M, nol 31(9):822–826. https://doi.org/10.1038/
DiCarlo JE, Norville JE, Church GM (2013) nbt.2623
Rna-guided human genome engineering via 7. Moore CBT, Christie KA, Marshall J, Nesbit
cas9. Science (New York, NY) 339 MA (2018) Personalised genome editing - the
(6121):823–826. https://doi.org/10.1126/sci future for corneal dystrophies. Prog Retin Eye
ence.1232033 Res 65:147–165. https://doi.org/10.1016/j.
3. Cong L, Ran FA, Cox D, Lin S, Barretto R, preteyeres.2018.01.004
Habib N, Hsu PD, Wu X, Jiang W, Marraffini 8. Christie KA, Courtney DG, DeDionisio LA,
LA, Zhang F (2013) Multiplex genome engi- Shern CC, De Majumdar S, Mairs LC, Nesbit
neering using crispr/cas systems. Science MA, Moore CBT (2017) Towards personalised
(New York, NY) 339(6121):819–823. https:// allele-specific crispr gene editing to treat autoso-
doi.org/10.1126/science.1231143 mal dominant disorders. Sci Rep 7(1):16174.
4. Ran FA, Hsu PD, Wright J, Agarwala V, Scott https://doi.org/10.1038/s41598-017-16279-
DA, Zhang F (2013) Genome engineering using 4
the crispr-cas9 system. Nat Protoc 8 9. Courtney DG, Moore JE, Atkinson SD,
(11):2281–2308. https://doi.org/10.1038/ Maurizi E, Allen EH, Pedrioli DM, McLean
nprot.2013.143 WH, Nesbit MA, Moore CB (2016) Crispr/
5. Hsu PD, Scott DA, Weinstein JA, Ran FA, cas9 DNA cleavage at snp-derived pam enables
Konermann S, Agarwala V, Li Y, Fine EJ, both in vitro and in vivo krt12 mutation-specific
Wu X, Shalem O, Cradick TJ, Marraffini LA, targeting. Gene Ther 23(1):108–112. https://
Bao G, Zhang F (2013) DNA targeting specific- doi.org/10.1038/gt.2015.82
ity of rna-guided cas9 nucleases. Nat Biotechnol
Chapter 7
Preparation and Administration of Adeno-associated Virus

Vectors for Corneal Gene Delivery
Liujiang Song, Jacquelyn J. Bower, and Matthew L. Hirsch
Abstract
Gene delivery approaches using adeno-associated virus (AAV) vectors are currently the preferred method
for human gene therapy applications and have demonstrated success in clinical trials for a diverse set of
diseases including retinal blindness. To date, no clinical trials using AAV gene therapy in the anterior eye
have been initiated; however, corneal gene delivery appears to be an attractive approach for treating both
corneal and ocular surface diseases. Multiple preclinical studies by our lab and others have demonstrated
efficient AAV vector-mediated gene delivery to the cornea for immunomodulation, anti-vascularization,
and enzyme supplementation. Interestingly, the route of AAV vector administration and nuances such as
administered volume influence vector tropism and transduction efficiency. In this chapter, a detailed
protocol for AAV vector production and specific approaches for AAV-mediated gene transfer to the cornea
via subconjunctival and intrastromal injections are described.
Key words Adeno-associated virus (AAV), Cornea, Gene delivery, Subconjunctival injection, Intras-
tromal injection, Purification, Titering, Purity
1 Introduction
Adeno-associated virus (AAV) is a small single-stranded DNA virus

initially discovered in 1965 as a contaminant of an adenovirus
preparation [1, 2]. In the 1980s, it was reported that all AAV viral
coding sequences can be substituted with transgenic DNA and
packaged in the AAV capsid, thus establishing the field of AAV
gene therapy [3, 4]. To date, at least 12 naturally occurring ser-
otypes and hundreds of variants have been isolated via tissue mining
experiments in various species [5]. In general, AAV serotypes dem-
onstrate altered tropisms which utilize various cell surface entry
receptors or co-receptors [5–17]. Although the exact mechanism
of cellular entry, nuclear entry/trafficking, and virion uncoating
remains unknown, reports have demonstrated that desirable physi-
ological properties can be generated by random or rational capsid
77
78 Liujiang Song et al.
mutagenesis, including restricted tissue transduction or enhanced

whole body transduction [5, 18–22].
Over the past few decades, several AAV production systems
have been developed that yield high viral titers of transduction
competent AAV vectors for both preclinical and clinical applica-
tions. Producer cell lines include the mammalian HEK293 cell
system [4, 23] (adherent and serum free suspension), Vero cells
(Herpes Simplex Virus-based platform), and insect cell lines such as
Sf9 cells (Baculovirus-based system) [24–27]. Each packaging sys-
tem features scale-up options ranging from several dishes/flasks to
large-scale bioreactors (such as the iCellIS 500 m2 bioreactor or the
Abec 4000-L bioreactor) [28]. Ion exchange/affinity chromatog-
raphy is the most commonly used purification technique for high-
purity clinical-grade AAV manufacturing [25, 29], whereas gradi-
ent centrifugation (such as cesium chloride or iodixanol gradient) is
still the major method for general laboratory-scale AAV vector
purification [30, 31]. AAV titering methods include genome char-
acterization and quantification via quantitative PCR (qPCR), drop-
let digital PCR (ddPCR), Southern dot blot, and alkaline gel
electrophoresis. AAV capsids are characterized by Western dot
blots, ELISA, electron microscopy, and silver staining conveying
the purity of the preparation [32]. Among the different production
methods, a 30-year-old triple transfection protocol using adherent
HEK293 cells remains the most common and reproducible method
to produce small-scale AAV preparations for preclinical studies
[33]. A combination of iodixanol gradient centrifugation and ion
exchange chromatography utilizing prepacked columns is the most
widely used AAV purification strategy to ensure sufficient purity.
When used in combination, these methods consistently and effi-
ciently produce high yields of AAV vectors for preclinical use in
corneal gene delivery applications [34, 35].
The eye is a particularly attractive gene therapy target, due to its
unique anatomic accessibility and presumed immunological privi-
lege. In fact, AAV has taken center stage as the gene therapy of
choice for many ocular diseases; the most notable of these is Lux-
turna (voretigene neparvovec, STN:125610, Spark Therapeutics,
Inc.), which was the first commercially available gene therapy drug
approved by the FDA in the United States to treat a rare disease of
the posterior eye [36, 37]. Its success has encouraged the gene
therapy community to expand AAV applications to multiple ocular
diseases including those of the cornea, which is the anterior trans-
parent avascular tissue that acts as the major refractive surface of the
eye. It is naturally organized into 3 general layers: (1) the anterior
epithelia, (2) the relatively large central stroma region composed of
precisely aligned collagen lamellae interspaced primarily by quies-
cent keratocytes, and (3) a posterior single layer of generally non-
dividing endothelial cells. For corneal diseases, allogeneic cornea
transplantation is a commonly employed strategy for vision
Preparation and Administration of AAV for Corneal Gene Delivery 79
restoration. However, a lack of suitable transplantation-grade

donor corneas and immunologic rejection remain important gen-
eral obstacles for this treatment [38, 39]. Alternatively, several
in vivo animal studies have demonstrated efficient AAV gene deliv-
ery to the cornea for the correction of genetic or acquired corneal
diseases. For instance, AAV encoding an endostatin, decorin,
or angiostatin gene were reported to decrease or reverse corneal
fibrosis after injury or infection [35, 40–42] and prevent corneal
vascularization [35, 43–45]. Furthermore, AAV-mediated immu-
nomodulatory gene expression (such as HLA-G) may re-establish
tolerance in ocular surface immune-mediated diseases [35], and
delay/prevent rejection of allogeneic, and perhaps even xenoge-
neic, corneal transplants. Moreover, gene addition strategies may
prevent corneal opacity in children with hereditary lysosomal stor-
age diseases and have also been utilized for the treatment of Fuchs’
dystrophy [46, 47].
AAV serotype specificity and the route of administration play
key roles in AAV vector biodistribution and transgene expression
levels; thus, identification of the optimal serotype and route for
AAV delivery is a critical step for successful corneal gene therapy.
Currently, topical [40, 41, 45, 48–53], intrastromal [46, 54], sub-
conjunctival [34, 43, 44], and intracameral [55–59] delivery routes
have been utilized for AAV gene delivery to the cornea (Fig. 1).
While all of these administration routes are approved for clinical
use, some are not necessarily commonplace, and each delivery route
exhibits its own inherent potential complications. For example,
topical administration appears to be the most attractive route of
administration due to its minimal invasiveness. However, without
corneal epithelial cell removal, AAV transduction via topical eye
drops is minimal [48, 49, 60]. Corneal intrastromal and intracam-
eral injections of AAV vectors have been well characterized and
demonstrate impressive serotype-dependent transduction of
Fig. 1 Administration routes for corneal drug delivery. 1. Topical; 2. intrastromal;

3. intracameral; 4. subconjunctival
different anterior eye compartments [46, 54–59]. In addition, our

lab has recently demonstrated corneal transduction following a
simple and safe subconjunctival injection [34]. Although serotype
dependence has not been completely characterized for each route
of administration, the choice of the best serotype will be disease-
dependent, as different cells can be targeted even within an optimal
administration route.
This chapter provides a detailed description of essential meth-
ods for AAV production in adherent HEK293 cells by triple plas-
mid transfection using polyethylenimine (PEI), purification by
iodixanol density gradient ultracentrifugation and ion exchange
chromatography, titration by qPCR, and particle characterization
including genome integrity and vector purity for corneal gene
delivery (Fig. 2). Furthermore, protocols for AAV administration
via subconjunctival and intrastromal injections are also described.
Fig. 2 Diagram of the major steps in AAV preparation

2 Materials (See Note 1)
2.1 Vector 1. HEK293 cells (ATCC: CRL-1573).

Production 2. AAV production plasmids: pXX680, pXR2, pAAV-transgene
(These plasmids are commercially available at the University
of North Carolina at Chapel Hill Vector Core Facility or other
companies).
3. T-150 tissue culture dishes.
4. DMEM medium (with L-glutamine, high glucose, and sodium
pyruvate) supplemented with 10% heat-inactivated fetal bovine
serum (FBS), and 100 U/mL penicillin/100 μg/mL strepto-
mycin (for cell culture).
5. DMEM medium without FBS and antibiotics (for
transfection).
6. Phosphate-buffered saline (PBS).
7. PEI (linear, MW 25 kD). Dissolve in PBS at 1 mg/ml, adjust
the pH between 4~5 and filter sterilize.
8. 0.05% trypsin/EDTA.
9. 50 mL polystyrene conical tubes.
10. 250 mL centrifugation bottles.
11. Cell scrapers.
12. Serological pipettes.
13. 37 C water bath.
14. Dry ice/ethanol bath.
15. Vortex mixer.
16. Benzonase nuclease.
17. 4.8 M MgCl2.
2.2 AAV Purification 1. 60% Iodixanol: OptiPrep Density Gradient Medium.

2. 4.8 M MgCl2.
3. 2.5 M KCl.
4. 0.22 μm filtered Milli Q H2O.
5. 5 M NaCl.
6. 5 mL syringes with 21 G needles.
7. OptiSeal centrifuge tubes.
8. Tube rack.
9. Ultracentrifuge.
10. Syringe pump.
11. Prepacked strong ion-exchange columns.
12. Buffer A: 20 mM Tris, pH 9.0.

13. Buffer B: 20 mM Tris, pH 9.0, 2 M NaCl.
14. Buffer C: 20 mM Tris, pH 9.0, 200 mM NaCl.
15. Dialysis buffer: PBS with 50 g/L D-sorbitol, 210 mM NaCl.
16. Slide-A-Lyzer dialysis cassette.
2.3 AAV Titration 1. 0.5 M EDTA.

by qPCR 2. DNase digestion solution: 10 mM Tris (pH 7.5), 10 mM
MgCl2, 200 mM CaCl2, 100 μg/mL DNase I.
3. Proteinase solution: 100 μg/mL Proteinase K, 5 mM NaCl,
and 1% Sarkosyl.
4. Molecular-grade H2O.
5. 10 mM Tris, pH 8.0
6. Standard plasmid, such as a maxi preparation of pAAV-
transgene plasmid used to generate the AAV vector, or an
AAV reference standard stock from ATCC.
7. LightCycler® 480 SYBR Green I Master Mix.
8. Primers designed for the specific AAV vector, such as the
following primer sets are designed for an AAV2 vector with a
GFP transgene, a CMV promoter, and an SV40 polyA tail.
Target Primer ID Sequence (50 !30 )

GFP GFP-F AGCAGCACGACTTCTTCAAGTCC
GFP-R TGTAGTTGTACTCCAGCTTGTGCC
CMV promoter CMV-F CAAGTACGCCCCCTATTGAC
CMV-R AAGTCCCGTTGATTTTGGTG
SV40 polyA SV40 polyA-F AGCAATAGCATCACAAATTTCACAA
SV40 polyA-R CCAGACATGATAAGATACATTGATGAGTT
9. Gilson Pipetman.
10. Pipette tips.
11. qPCR 96-well plates and plate sealer sheets.
12. LightCycler 480 qPCR machine.

2.4 AAV Titering by 1. 100 mM NaOH.

Dot/Slot Blot 2. Benzonase nuclease.
3. ddH2O.
4. Church buffer: 1% (w/v) bovine serum albumin, 1 mM EDTA
(pH 8.0), 0.5 M phosphate buffer (0.5 M phosphate buffer is
134 g of Na2HPO4 · 7H2O, 4 mL of 85% H3PO4 (concen-
trated phosphoric acid), H2O to 1 L), 7% (w/v) SDS.
5. Bio-Rad thin absorbent filter papers or Whatman 3MM filter
papers.
6. Positively charged nylon transfer membrane.
7. Maxi preparation of pAAV-transgene plasmid used to generate
the AAV vector standard plasmid or AAV reference standard
stock from ATCC.
8. 96-Well dot-blot apparatus.
9. Stratagene Stratalinker 1800 UV Crosslinker.
10. Probe template (such as a restriction enzyme digested fragment
of the targeted vector).
11. Random Primer DNA Labeling Kit.
12. 20 SSC (saline-sodium citrate): 3 M NaCl, 0.3 M sodium
citrate, pH 7.0.
13. Low salt buffer: 0.1 SSC, 0.1% SDS.
14. High salt buffer: 2 SSC, 0.1% SDS.
15. Hybridization bottle.
16. X-ray film.
17. 8 10 in. autoradiography cassette.
2.5 scAAV Genome 1. 1 M NaOH.

Integrity by Alkaline 2. Benzonase nuclease.
Gel Electrophoresis
3. ddH2O.
4. 20 mM EDTA.
5. 1 kb DNA marker.
6. 10 alkaline gel buffer: Add 50 mL of 10 N NaOH and 20 mL
of 0.5 M EDTA (pH 8.0) to 800 mL of H2O and then adjust
the final volume to 1 L.
7. 1 TAE buffer: 40 mM Tris-acetate, 1 mM EDTA, pH 8.3.
8. 6 Alkaline Gel-loading Buffer: 300 mM NaOH, 6 mM
EDTA, 18% (w/v) Ficoll, 0.15% (w/v) bromocresol green,
0.25% (w/v) xylene cyanol.
9. General electrophoresis agarose powder.

12. SYBR™ Gold Nucleic Acid Gel Stain.
2.6 Determination 1. NuPAGE 4–12% Bis-Tris gels.

of AAV Purity 2. 2-Mercaptoethanol.
3. NuPAGE™ Sample Reducing Agent (10).
4. NuPAGE™ LDS 4 Loading buffer.
5. Silver staining kit.
6. Protein marker with broad range allowing for the determina-
tion of the AAV capsid protein sizes (87, 72, and 62 kDa).
7. 10 SDS-PAGE Running Buffer.
8. SDS-PAGE gel electrophoresis apparatus.
2.7 Corneal 1. Mouse model of choice.

Injections for Mouse 2. 1% Fluorescein solution, filter sterilized
Models
3. 0.5% Proparacaine hydrochloride eye drops
4. GenTeal ocular lubricant gel.
5. 1% Tropicamide
6. Oxygen (tank).
7. Isoflurane vaporizer and induction chamber set.
8. Mouse nose cone for anesthesia.
9. Standard Infuse/Withdraw Programmable Syringe pump.
10. Polyethylene tubing, I.D. 0.38 mm, O.D. 1.09 mm.
11. 10 μL Hamilton gastight removable needle syringe
12. Beveled 36 G needles.
13. Syringes with 27 G needle (O.D. about 0.41 mm).
14. Stereoscopic microscope.
15. Finely pointed precision forceps.
16. Heating pad.
17. Parafilm.
3 Methods
3.1 Vector 1. Culture HEK293 cells to a cell density of 80–90% confluence.

Production (see Note 2).
2. Calculate the amount of each plasmid needed for a 1:1:1 molar
ratio with a total mass of 62.4 μg/plate. For example, the
following volumes are calculated for ten T-150 plates for pack-
aging of AAV2-CMV-EGFP vector:
(a) The length of the pXX680, pXR2, and pAAV-CMV-

EGFP is 18,322 bp, 7462 bp, and 5848 bp, respectively.
Thus, the number of total base pairs will be:
18,322 + 7462 + 5848 ¼ 31,632 bp
(b) Calculate the μg/bp: 62.4 μg/31,632 bp ¼
0.0019769 μg/bp.
(c) Mass needed for each plasmid for 10 plates:
l pXX680: 0.0019769 μg/bp 18,322 bp 10 ¼
361.435508 μg
l pXR2: 0.0019769 μg/bp 7462 bp 10 ¼
147.516278 μg
l pAAV-CMV-EGFP: 0.0019769 μg/bp 5848 bp 10
¼ 115.609112 μg
(d) Then according to the plasmid concentration, determine
the volume needed for each plasmid. Assuming each plas-
mid is approximately 1 μg/μL, then a total volume of
624 μL containing all three plasmids will be used for the
transfection of 10 plates.
(e) If transfecting more/less plates, simply scale the reagents
up/down for the total number of desired plates (see Notes
3 and 4).
3. Add the calculated amounts of DMEM (without FBS or anti-
biotics) and plasmid DNA to a tube and mix well. Then add the
PEI to the DNA/DMEM solution, and vortex briefly to mix
(see Notes 5–8).
Regent Volume (mL) for 10 plates

PEI 1.25 mL
Plasmid 0.624 mL
DMEM 8.126 mL
Total 10 mL
4. Incubate the DMEM/DNA-PEI solution for 10 min at room

temperature to allow complex formation.
5. Carefully add 1 mL (per plate) of the DMEM/DNA-PEI
complexes dropwise to the cells.
6. Return the cells to the 37 C/5% CO2 incubator, incubate
4–8 h (or overnight). Cells will be more viable and give better
yields if the media is changed 4–8 h post transfection.
7. Scrape the dishes to harvest transfected cells at 72 h post
transfection (see Note 9).
8. Spin at 500 g and 4 C for 10 min. Discard the supernatant

(see Note 10).
9. Resuspend the cell pellet in ~5 mL of autoclaved MilliQ H2O.
10. Freeze/thaw by placing the resuspended pellet in a dry
ice/ethanol bath and subsequently thawing at 37 C for
10 min. Vortex the sample aggressively. Repeat the freeze/
thaw/vortex steps two additional times (the crude lysate may
be stored at 80 C for an extended period of time at this point
in the protocol).
11. Sonicate the sample on ice for 2 min.
12. Add 1 μL of 4.8 M MgCl2 (a final concentration of 1 mM), and
20 U of Benzonase (about 2 U of Benzonase for each T-150
plate harvested). Vortex and incubate at 37 C for 1 h.
13. Spin down at 2000 g at 4 C for 20 min and collect the
supernatant.
3.2 AAV Purification 1. Prepare iodixanol gradient as shown in the table below
(see Note 11):
Components 17% 25% 40% 60%

Iodixanol 60% stock 13.5 mL 21 mL 33.3 mL 50 mL
1 M Tris–HCl, pH 9.0 1.25 mL 1.25 mL 1.25 mL -
Autoclaved MilliQ H2O 25 mL 27.5 mL 15.3 mL -
5 M NaCl 10 mL - - -
1 M MgCl2 0.1 mL 0.1 mL 0.1 mL 0.1 mL
2.5 M KCl 0.1 mL 0.1 mL 0.1 mL 0.1 mL
2. In a biosafety cabinet hood, load the iodixanol gradient step-

wise with a glass Pasteur pipette and a pipette aid into a Quick-
Seal tube (29.9 mL capacity) in the following order, mark the
40% and 60% interface with a “+” (see Note 12):
(a) 6 mL of 17% iodixanol
(b) 6 mL of 25% iodixanol
(c) 7 mL of 40% iodixanol
(d) 6 mL of 60% iodixanol
(e) Crude lysate (at the top)
3. Spin in an ultracentrifuge with an appropriate rotor at
504,000 g (70,000 rpm in a Beckman Coulter Ti70 rotor,
for example) for 1 h at 18 C.
4. Collect the fraction at the 40–60% iodixanol interface by insert-
ing a 5 mL syringe with an 18 G needle (bevel up) between the
40% and 60% gradient fractions. Take ~ 4–5 mL of the fraction
(see Note 13).
5. Transfer each viral particle fraction obtained from an iodixanol

gradient/Quickseal tube into an individual 50 mL conical tube
and bring the final volume of all tubes to 12 mL with Buffer A.
6. Assemble the ion exchange apparatus by placing a syringe on
the pump and connecting the syringe to the prepacked column
using the appropriate tubes.
7. Load the following buffers to the syringe in succession and
collect the fractions eluted with Buffer C:
(a) 5 mL of 0.5 M NaOH (set speed at 1 mL/min)
(b) 5 mL of Buffer B (set speed at 1 mL/min)
(c) 10 mL of Buffer A (set speed at 1 mL/min)
(d) 12 mL of sample (diluted in buffer A) (set speed at
0.5 mL/min)
(e) 10 mL of Buffer A
(f) 6 mL of Buffer C
8. Collect the fraction eluted with Buffer C into multiple tubes,
the first tube 0.5 mL, the second tube 1 mL, the third tube
1 mL, then 0.5 mL for each tube until finish the collection
(see Notes 14).
3.3 Titration by qPCR 1. Add 10 μL of virus to 90 μL of the DNase I digestion solution.

2. Incubate for 1 h at 37 C to digest any DNA remaining outside
of the viral particles.
3. Add 6 μL of 0.5 M EDTA and vortex to inactivate the DNase I.
4. Add 120 μL of Proteinase K (ProK) solution to the DNase-
treated AAV samples to digest the capsid and mix.
5. Incubate at 55 C for 2 h to digest the viral particles and
release the AAV DNA.
6. Incubate the sample for 10 min at 95 C to inactivate the ProK.
7. Dilute the sample 100-fold in molecular-grade water or 10 mM
Tris (pH 8.0) for qPCR.
8. Prepare the standard plasmid by making serial 1:10 dilutions
with 10 mM Tris (pH 8.0) at concentrations ranging from
50 pg/μL to 0.05 fg/μL (see Notes 15 and 16).
9. Plan the sample layout for the 96-well qPCR plate. Negative
controls and plasmid standards should be assayed at least in
duplicate. To minimize the risk of cross-contamination, assign
the plasmid standard to the end of the plate only after adding all
of the samples (Fig. 3).
10. Determine the number of PCR reactions to be performed
including plasmid control dilutions, samples, negative con-
trols, and add additional 10% to the sample number in order
Fig. 3 qPCR plate layout template
to calculate the total number of reactions (N) required for the

master mix solution.
11. Dilute the primer to a concentration of 20 μM (see Note 17).
12. Set up the qPCR SYBR master mix reactions (N ¼ total num-
ber of reactions):
Reagent Volume (μL)

2 SYBR mix 5 μL N
Forward Primer 0.25 μL N
Reverse Primer 0.25 μL N
H2O 2.5 μL N
13. Load 8 μL of the master mix described above into each well (see
Note 18).
14. Pipette 2 μL of DNA (sample, standard, H2O, or vehicle
control) into each well.
15. Seal the plate with the adhesive plate seal.
16. Spin down the plate briefly.

17. Cycle in the qPCR thermocycler. Set up the cycling conditions
according to the primer sets used and the manufacturer’s
instructions. For example, the GFP primer set listed in the
materials uses the following cycling conditions:
Step 1 95 C for 10 min (1 cycle)

Step 2 95 C for 10 s
58 C for 10 s (40 cycles)
72 C for 10 s
Step 3 72 C for 10 s (1 cycle)
Step 4 Melting curve analysis
18. Calculate the viral titer (see Note 19).

(a) Perform the data analysis using the instrument’s software:
(1) input each sample name and the plasmid standard
mass (fg) in the “Sample Editor,” (2) choose the “Ab
Quant/2nd derivative max” and click “calculate,” (3) the
software will automatically generate the standard curve
and corresponding mass for each sample from their
respective Ct values.
(b) Determine the copy number in each well:
fg 6:022 1023
Copy number ¼
650 size of plasmidðbpÞ 1015
Average weight of a DNA base pair ¼ 650 Da.
Avogadro’s number ¼ 6.022 1023 molecules/mol.
(c) Determine the sample volume (vol) dilution factor
(DF) in steps 2–7:
starting volume þ DNAse mix þ EDTA þ Prok mix

DF ¼
starting volume
dilution in step 7
(d) Determine the virus titer:
For self-complementary AAV:
Copy number DF
Titer ðvg=mLÞ ¼ 103
The final sample vol added into each well in step 14
For single-stranded AAV:
Copy number DF 2
Titer ðvg=mLÞ ¼ 103
The final sample vol added into each well in step 14
3.4 Titration by 1. Add 40 μL of ddH2O to 10 μL of virus.

Dot Blot 2. Aliquot 100 ng plasmid and bring the volume to 50 μL.
3. Add 0.2 μL of Benzonase to both the virus and plasmid samples
to generate the functional Benzonase control sample. Incubate
at 37 C for 1 h.
4. Aliquot another 100 ng control plasmid and bring the volume
to 50 μL to generate the control plasmid sample.
5. Add 50 μL of 100 mM NaOH to each sample.
6. Incubate at 65 C for 30 min, then chill on ice for 5 min.
7. Cut the filter paper and nylon membrane to the size of the slot
blot apparatus.
8. Mark a corner to orient the blot using a pencil.
9. Rinse the membrane in ddH2O for 30 min at RT.
10. Soak the membrane in 10 SSC (diluted from 20 SSC) for
10 min at RT.
11. Rinse three sheets of filter paper in 10 SSC.
12. Set up the slot blot apparatus.
13. Place three sheets of filter paper on the bottom and place the
membrane on top of the filter paper. Connect to a vacuum
flask.
14. Add 100 μL of 10 SSC to the slot blot apparatus to check the
vacuum leak.
15. Add 100 μL of 20 SSC to each of the samples + 1 μL of 6
DNA loading dye to visualize sample loading.
16. Transfer all samples and plasmids into a 96-well plate. Add
100 μL of each sample onto the blot membrane using a multi-
channel pipette.
17. Add 100 μL of 10 SSC to each sample, mix.
18. Take 100 μL of the diluted sample to the next well on the
membrane.
19. Repeat steps 17 and 18 until you run out of wells at the end of
the slot blot apparatus.
20. Remove the membrane and insert it into the UV Crosslinker.
Use the following settings: “Energy,” “Optimal Crosslink,”
and “Start.” The UV crosslinker will automatically stop when
finished.
21. Place the membrane in a hybridization bottle with the side
containing the crosslinked DNA facing toward the center of
the bottle.
22. Add 15 mL of Church Buffer, tightly seal the bottle, and place
it in a rotating hybridization incubator at 68 C for at least 1 h.
23. Make a radioactively labeled probe following the manufac-

turer’s instructions included in the Random Primer DNA
Labeling Kit (see Note 20).
24. Add the probe to the Church buffer in the hybridization
bottle.
25. Hybridize at 68 C for overnight.
26. Wash the membrane with high salt buffer for 10 min at 68 C.
Repeat this step with a fresh buffer change.
27. Wash with low salt buffer for 30 min at 68 C.
28. Wrap the blot with plastic wrap, ensuring that no air bubbles
are trapped between the blot membrane and the wrap. Place it
into the film cassette and tape the membrane to keep it in
position. Scan with the radioactivity meter to roughly get a
sense about the signal strength of radioactivity.
29. Add a film over the membrane in the dark room. Fold and then
unfold a corner of the film to mark the orientation of the dot
blot membrane.
30. Leave the cassette tightly closed in a drawer to protect from
light and decide approximately how long to wait to develop the
film according to the signal strength obtained at step 28. The
wait times can range from several minutes to days. One film
exposed for about 30 min may be used to obtain a rough idea
of the signal strength and then a decision can be made as to
whether a longer or shorter wait time is necessary for a second
film exposure. Several exposures may be needed to obtain the
optimal exposure quality.
31. Scan the exposed film to assess each sample’s corresponding
mass compared to the plasmid standards.
32. Calculate the titer using the same formula/method in the
qPCR titering section.
3.5 Vector Genome 1. Aliquot 10 μL of purified virus, and add 10 μL of ddH2O.

Integrity by Alkaline 2. Add 0.2 μL of Benzonase and incubate at 37 C for 1 h.
Gel Analysis
3. Add 1 μL of 1 M NaOH.
4. Incubate at 65 C for 30 min, then chill on ice for 5 min.
5. Add 1.3 μL of 20 mM EDTA, 2.5 μL ddH2O, and 5 μL of 6
alkaline gel-loading buffer to the virus sample.
6. Add 1.5 μL of 1 M NaOH and 1.5 μL of 20 mM EDTA to 5 μL
of the DNA Marker and 22 μL of ddH2O.
7. Prepare the agarose solution by adding 2 g of agarose powder
to 180 mL of ddH2O, and boil it in the microwave until
agarose has dissolved and the solution is clear (approximately
2–3 min).
8. Cool the agarose solution to 55 C by placing the flask in 55 C

water bath.
9. Add 20 mL of 10 alkaline gel buffer.
10. Perform the electrophoresis using 1 Alkaline gel buffer
(Dilute the 10 alkaline buffer with ddH2O to generate a
1 working solution immediately before use) at 35 V
overnight.
11. Stain the gel using SYBR™ Gold Nucleic Acid Gel Stain in 1
TAE buffer (see Note 21).
3.6 Silver Staining 1. Aliquot 10 μL of virus.

2. Add 5%, (v/v) 2-Mercaptoethanol to the 4 loading buffer
(make sure that a fresh solution of 2-Mercaptoethanol is used).
3. Add 4 loading buffer mix and 1/10 volume of NuPAGE™
Sample Reducing Agent to the sample.
4. Boil the samples for 10 min.
5. Load the samples onto the SDS-PAGE gel.
6. Start the denaturing gel electrophoresis at 120 V for 1 h using
1 SDS-PAGE Running Gel Buffer.
7. Perform the silver staining following the manufacturer’s
instructions included in the kit.
3.7 Injection 1. Set up the injection stage as shown in Fig. 4a in a biosafety

cabinet. A simple injection stage consists of a stereo micro-
scope, a syringe pump, and the anesthesia system.
2. Cut the polyethylene tubing at a length as needed (usually
around 50 cm is convenient for operation).
3. Insert the 36 G needle shank to the polyethylene tubing.
4. Fill a syringe with sterile water, manually flush the tubing/
needle to ensure no leaks, clogs, or damage throughout the
tubing.
5. Flush 70% EtOH through the tubing to disinfect.
6. Use a syringe to evacuate with air and rinse with water.
7. Fill the tubing with sterile water using a syringe and pull the
syringe out slowly to create an excess drop of water on a
parafilm at the tubing opening; leave the opening in the water
drop (see Note 22).
8. Fill a Hamilton syringe and eject some water out to ensure
there is no air in the syringe. Place the filled Hamilton syringe
needle tip in the drop of water on the parafilm and insert it into
the water filled tubing.
9. Once the tubing is connected with the Hamilton syringe
(Fig. 4b), gently place the Hamilton syringe on the
microinjection pump.
Fig. 4 Injection stage setup. A simple injection stage consists of a stereo microscope, a syringe pump, and the
anesthesia system (a). A Hamilton syringe and a 36G needle connected with polyethylene tubing (b). An
anesthesia nose mask designed for a tight fit to the mouse’s face and allowing access to the eyes (c).
Introducing a small air bubble in the tubing to create a gap between the sterile water and the virus (d)
10. Eject the water out in the Hamilton syringe (the tubing is still
filled with water), pull back the Hamilton syringe, and intro-
duce a small air bubble in the tubing/needle.
Fig. 5 Assessment of the subconjunctival injection. Microscopic view of AAV

vector administration to the murine subconjunctival space. One percent
fluorescein is added directly to the 70 μL AAV vector preparation to allow
visualization of bleb formation during the procedure using an operating
microscope
11. Place the needle tip into AAV vector and withdraw virus. There
should be a visible air bubble remaining between the virus and
the water in the tubing (Fig. 4d, see Note 23).
12. Allow the tubing/needle to sit at room temperature for 10 min
with viral vector solution to allow the saturation of potential
binding of virus on the wall of the needle and/or tubing;
discard the virus.
13. Place needle tip in the viral prep and withdraw desired amount
of vector.
14. Set the syringe pump at a rate of 10 μL/min (for the subcon-
junctival injection) or 2 μL/min (for intrastromal injection)
(see Note 24).
15. In an anesthesia chamber, anesthetize the mouse with 2.5%
isoflurane and an oxygen flow rate of 1 L per min until the
mouse reaches a state of unconsciousness.
16. Transfer the mouse from the anesthesia chamber to the surgical
pad, and place a nose cone (Fig. 4c) over the mouse’s nose to
maintain anesthesia.
17. Apply 0.5% proparacaine hydrochloride eye drops topically to
the mouse’s eyes.
18. Use the forceps to expose the eyeball and immobilize the
conjunctiva.
19. For a subconjunctival injection, use the dominant hand to hold
the needle with the bevel facing upward. Insert the needle into
the conjunctiva (Fig. 5).
For an intrastromal injection, hold the needle horizontally
with the bevel facing upward. Penetrate the cornea at a distance
of approximately one-third of the corneal radius as measured
from the temporal limbus. Allow the needle tip to remain in the
stroma layer (see Notes 25–29).
20. Start the injection by using the footpad to control the pump
(see Note 30).
21. Hold the needle in place for at least 10 s before removing the
needle from the conjunctiva.
22. Put a drop of the GenTeal lubricant gel on the mouse’s eyes
and then place the mouse on a heating pad to recover.
4 Notes
1. Most of the kits, regents, or apparatuses listed in the materials

are commercially available from several different suppliers.
Choose cost-effective and consistent ones according to each
individual’s preferences.
2. Always keep good track of the passage number of the HEK293
cells and use low passage cells to avoid potential cell line pas-
sage effects.
3. The molar ratio of the three plasmids used for the triple trans-
fection protocol may vary among labs [30, 32, 61].
4. The total mass of plasmid used for transfection of one T-150
plate ranges from 28 μg to 120 μg, all of which are reported to
produce good AAV yields [32].
5. The plasmid:PEI ratio is approximately 1:2. If less or more total
mass of the plasmid/plate is used, the amount of PEI should
also be adjusted accordingly.
6. PEI with MW of 25,000 and PEI-Max with MW of 40,000
both have been reported to produce good yields of AAV [30].
7. It is recommended to do a triple transfection with a fluorescent
reporter gene in parallel as a transfection positive control when
packaging a non-reporter gene especially for a beginner.
8. AAV package capacity, remains unknown, however appears to
be limited to ~5 kb (for single-stranded AAV) [62, 63] and
~2.3 kb [64, 65] (for self-complementary AAV) in length,
respectively. Vector constructs that attempt to exceed this
length are most commonly packaged as sub-genomic frag-
ments [66]; however, successful capsid packaging of larger
fragments were demonstrated by several groups [67, 68].
9. After transfection, the media can be changed after 4–8 h to
increase cell viability and virus yield. The total incubation time
can vary anywhere from 2 to 6 days.
10. AAV will be present in the cells and in the medium in a
serotype-dependent fashion. For example, the majority of
AAV2 particles is reported to be inside the cells, whereas the

majority of AAV1 and AAV8 particles are present in the super-
natant [69]. The supernatant could also be collected by PEG
precipitation and then purified using the same protocol [70].
11. When preparing the 25% and 60% gradient fractions, it is
advisable to add a few drops of phenol red to help with visuali-
zation of the fractions at later steps.
12. Always prepare the gradient very carefully and avoid disturbing
the gradient. When handling multiple samples, a variable speed
peristaltic pump may be used to speed up loading of the
iodixanol gradient, and to minimize the loading differences
between tubes to ensure tubes are balanced and gradient layers
are undisturbed.
13. One critical step for the iodixanol purification is to aspirate the
correct layer from the 40%/60% gradient boundary after ultra-
centrifugation. For this purpose, gently and slowly collect the
40%/60% boundary, avoiding the use of too much force. When
the top part of the 40% fraction is reached during collection
with the syringe, turn the bevel down to minimize the risk of
collecting contaminants such as the empty capsid. Do not
collect the top of the 40% fraction which contains cellular
protein and empty particles.
14. The second and third tube is usually where the majority of the
viral particles are located. Dialysis can be performed at 4 C for
overnight to exchange the buffer. Do not leave the virus in
buffer C long term; its high pH value may lead to AAV vector
instability.
15. For the standard plasmid serial dilutions, it is highly recom-
mended that the plasmid DNA obtained from a Maxi-
preparation be diluted to 10 ng/μL initially in 10 mM Tris
(pH 8.0), aliquoted, and stored at 20 C, as this will ensure
consistency in the plasmid stock used to generate the standard
curve. Use one aliquot for each viral titration experiment and
avoid repeat usage of the DNA once it has been thawed.
16. AAV reference stock standard materials from ATCC (Cat#:
VR-1616 for AAV2 and VR-1816 for AAV8) [71] or previ-
ously titered AAV may also serve as qPCR standards. However,
to ensure consistency among batches, use the same reference
materials for each viral titration experiment and record the
standard Ct values each time.
17. Different primer sequences and position of the target amplicon
within the vector genome may influence titers. It is recom-
mended to use 3 sets of primers for genome quantification (50 ,
central, and 30 ) to reduce concerns of incomplete genome
packaging and use the same primer set for different batches of
the same vector to minimize concerns regarding primer biases.
18. To minimize pipetting errors, an electronic repeating pipettor

can be used to dispense the master mix into the wells.
19. Always check that the Ct value obtained for the control virus or
plasmid is within the previously obtained range. Always check
the PCR efficiency, melting curve, and the quality of the stan-
dard curve to ensure the quality of the qPCR assay.
20. Nonradioactive labeling kit is also available such as digoxigenin
DNA labeling if radioactive probe is not preferred.
21. SYBR Gold is sensitive enough to detect 1010 viral genomes. If
the viral genome titer is less than 109, skip the SYBR Gold
staining step and proceed to southern hybridization steps fol-
lowing standard protocol.
22. It is critical to minimize the amount of air present in the
injecting syringe/tube/needle to ensure accuracy of the injec-
tion volume.
23. As an optional step, mixing the virus with sodium fluorescein
(final conc. 0.01%) is helpful for visualizing the virus/air bub-
ble interface during injection and monitoring AAV distribution
and/or any leakage after injection (Figs. 5 and 6).
24. When setting up the syringe pump parameters, it is very impor-
tant to select the right syringe diameter and then set the speed
of injection.
25. For intrastromal injection (Fig. 5), 1~2 μL for mouse cornea is
well tolerated. One of the major difficulties of intrastromal
injections of a mouse cornea for beginners is endothelial
Fig. 6 Assessment of the intrastromal injection. Corneal optical coherence tomography (OCT) (left panel) and
fluorescence imaging (right panel) using a Micron IV (Phoenix Research Labs, Pleasanton, CA, USA) following a
1 μL of intrastromal injection of AAV vectors in a 0.1% fluorescein/PBS solution
perforation due to the small thickness of the cornea and the

size/design of commercially available needles.
26. A 1~2 μL intrastromal injection results in a high distribution
area in mice (Fig. 6) [72], which is visible as a temporary
corneal plaque. This is usually completely resolved within
24 h post-injection.
27. Subconjunctival injection has less of a restriction on the admi-
nistered volume (1~100 μL for a mouse). However, volume
differences may play a role in AAV biodistribution and trans-
duction. In addition, larger volume of injection was reported
to be used to create the conjunctival scarring model [68].
28. Upon subconjunctival injection, a bleb may appear (Fig. 5). It
is usually completely resolved a few hours post injection.
29. A Micro IV imaging system (Phoenix Research Labs) could be
used to monitor the resolution of the injected solution.
30. The movement of the air should be synchronized with the
movement of the Hamilton plunger. A delay indicates excess
air in the injecting system or loose connection between the
tubing/needle/syringe.
References
1. Atchison RW, Casto BC, Hammon WM Nakamura T, Watanabe M, Oshimi K, Daida

(1965) Adenovirus-associated defective virus H (2005) Hepatocyte growth factor receptor is
particles. Science 149(3685):754–756 a coreceptor for adeno-associated virus type
2. Atchison RW, Casto BC, Hammon WM 2 infection. J Virol 79(1):609–614. https://
(1966) Electron microscopy of adenovirus- doi.org/10.1128/JVI.79.1.609-614.2005
associated virus (aav) in cell cultures. Virology 8. Summerford C, Bartlett JS, Samulski RJ
29(2):353–357 (1999) Alphavbeta5 integrin: a co-receptor
3. Hermonat PL, Muzyczka N (1984) Use of for adeno-associated virus type 2 infection.
adeno-associated virus as a mammalian DNA Nat Med 5(1):78–82. https://doi.org/10.
cloning vector: transduction of neomycin resis- 1038/4768
tance into mammalian tissue culture cells. Proc 9. Akache B, Grimm D, Pandey K, Yant SR,
Natl Acad Sci U S A 81(20):6466–6470. Xu H, Kay MA (2006) The 37/67-kilodalton
https://doi.org/10.1073/pnas.81.20.6466 laminin receptor is a receptor for adeno-
4. Senapathy P, Carter BJ (1984) Molecular clon- associated virus serotypes 8, 2, 3, and 9. J
ing of adeno-associated virus variant genomes Virol 80(19):9831–9836. https://doi.org/
and generation of infectious virus by recombi- 10.1128/JVI.00878-06
nation in mammalian cells. J Biol Chem 259 10. Di Pasquale G, Davidson BL, Stein CS,
(7):4661–4666 Martins I, Scudiero D, Monks A, Chiorini JA
5. Asokan A, Schaffer DV, Samulski RJ (2012) (2003) Identification of pdgfr as a receptor for
The aav vector toolkit: poised at the clinical aav-5 transduction. Nat Med 9
crossroads. Mol Ther 20(4):699–708. (10):1306–1312. https://doi.org/10.1038/
https://doi.org/10.1038/mt.2011.287 nm929
6. Qing K, Mah C, Hansen J, Zhou S, Dwarki V, 11. Weller ML, Amornphimoltham P, Schmidt M,
Srivastava A (1999) Human fibroblast growth Wilson PA, Gutkind JS, Chiorini JA (2010)
factor receptor 1 is a co-receptor for infection Epidermal growth factor receptor is a
by adeno-associated virus 2. Nat Med 5 co-receptor for adeno-associated virus serotype
(1):71–77. https://doi.org/10.1038/4758 6. Nat Med 16(6):662–664. https://doi.org/
7. Kashiwakura Y, Tamayose K, Iwabuchi K, 10.1038/nm.2145
Hirai Y, Shimada T, Matsumoto K,
12. Bell CL, Vandenberghe LH, Bell P, Limberis DE, McCown TJ, Petermann KB, Sharpless
MP, Gao GP, Van Vliet K, Agbandje- NE, Samulski RJ (2008) Engineering and
McKenna M, Wilson JM (2011) The aav9 selection of shuffled aav genomes: a new strat-
receptor and its modification to improve egy for producing targeted biological nanopar-
in vivo lung gene transfer in mice. J Clin Invest ticles. Mol Ther 16(7):1252–1260. https://
121(6):2427–2435. https://doi.org/10. doi.org/10.1038/mt.2008.100
1172/JCI57367 22. Zhong L, Li B, Mah CS, Govindasamy L,
13. Shen S, Bryant KD, Brown SM, Randell SH, Agbandje-McKenna M, Cooper M, Herzog
Asokan A (2011) Terminal n-linked galactose is RW, Zolotukhin I, Warrington KH, Weigel-
the primary receptor for adeno-associated virus Van Aken KA, Hobbs JA, Zolotukhin S,
9. J Biol Chem 286(15):13532–13540. Muzyczka N, Srivastava A (2008) Next gener-
https://doi.org/10.1074/jbc.M110.210922 ation of adeno-associated virus 2 vectors: point
14. Wu Z, Miller E, Agbandje-McKenna M, mutations in tyrosines lead to high-efficiency
Samulski RJ (2006) Alpha2,3 and alpha2,6 transduction at lower doses. Proc Natl Acad Sci
n-linked sialic acids facilitate efficient binding U S A 105(22):7827–7832. https://doi.org/
and transduction by adeno-associated virus 10.1073/pnas.0802866105
types 1 and 6. J Virol 80(18):9093–9103. 23. Xiao X, Li J, Samulski RJ (1998) Production of
https://doi.org/10.1128/JVI.00895-06 high-titer recombinant adeno-associated virus
15. Pillay S, Meyer NL, Puschnik AS, Davulcu O, vectors in the absence of helper adenovirus. J
Diep J, Ishikawa Y, Jae LT, Wosen JE, Naga- Virol 72(3):2224–2232
mine CM, Chapman MS, Carette JE (2016) 24. Urabe M, Ding C, Kotin RM (2002) Insect
An essential receptor for adeno-associated cells as a factory to produce adeno-associated
virus infection. Nature 530(7588):108–112. virus type 2 vectors. Hum Gene Ther 13
https://doi.org/10.1038/nature16465 (16):1935–1943. https://doi.org/10.1089/
16. Summerford C, Johnson JS, Samulski RJ 10430340260355347
(2016) Aavr: a multi-serotype receptor for 25. Chahal PS, Aucoin MG, Kamen A (2007) Pri-
aav. Mol Ther 24(4):663–666. https://doi. mary recovery and chromatographic purifica-
org/10.1038/mt.2016.49 tion of adeno-associated virus type
17. Pillay S, Zou W, Cheng F, Puschnik AS, Meyer 2 produced by baculovirus/insect cell system.
NL, Ganaie SS, Deng X, Wosen JE, Davulcu O, J Virol Methods 139(1):61–70. https://doi.
Yan Z, Engelhardt JF, Brown KE, Chapman org/10.1016/j.jviromet.2006.09.011
MS, Qiu J, Carette JE (2017) Aav serotypes 26. Mietzsch M, Grasse S, Zurawski C, Weger S,
have distinctive interactions with domains of Bennett A, Agbandje-McKenna M,
the cellular receptor aavr. J Virol. https://doi. Muzyczka N, Zolotukhin S, Heilbronn R
org/10.1128/JVI.00391-17 (2014) Onebac: platform for scalable and
18. Sen D (2014) Improving clinical efficacy of high-titer production of adeno-associated
adeno associated vectors by rational capsid bio- virus serotype 1-12 vectors for gene therapy.
engineering. J Biomed Sci 21:103. https://doi. Hum Gene Ther 25(3):212–222. https://doi.
org/10.1186/s12929-014-0103-1 org/10.1089/hum.2013.184
19. Shi W, Arnold GS, Bartlett JS (2001) Inser- 27. Joshi PRH, Cervera L, Ahmed I, Kondratov O,
tional mutagenesis of the adeno-associated Zolotukhin S, Schrag J, Chahal PS, Kamen AA
virus type 2 (aav2) capsid gene and generation (2019) Achieving high-yield production of
of aav2 vectors targeted to alternative cell- functional aav5 gene delivery vectors via fed-
surface receptors. Hum Gene Ther 12 batch in an insect cell-one baculovirus system.
(14):1697–1711. https://doi.org/10.1089/ Mol Ther Methods Clin Dev 13:279–289.
104303401750476212 https://doi.org/10.1016/j.omtm.2019.02.
20. Bowles DE, McPhee SW, Li C, Gray SJ, 003
Samulski JJ, Camp AS, Li J, Wang B, Monahan 28. Powers AD, Piras BA, Clark RK, Lockey TD,
PE, Rabinowitz JE, Grieger JC, Meagher MM (2016) Development and opti-
Govindasamy L, Agbandje-McKenna M, mization of aav hfix particles by transient trans-
Xiao X, Samulski RJ (2012) Phase 1 gene ther- fection in an icellis(®) fixed-bed bioreactor.
apy for duchenne muscular dystrophy using a Hum Gene Ther Methods 27(3):112–121.
translational optimized aav vector. Mol Ther https://doi.org/10.1089/hgtb.2016.021
20(2):443–455. https://doi.org/10.1038/ 29. Gao G, Qu G, Burnham MS, Huang J,
mt.2011.237 Chirmule N, Joshi B, Yu QC, Marsh JA, Con-
21. Li W, Asokan A, Wu Z, Van Dyke T, ceicao CM, Wilson JM (2000) Purification of
DiPrimio N, Johnson JS, Govindaswamy L, recombinant adeno-associated virus vectors by
Agbandje-McKenna M, Leichtle S, Redmond column chromatography and its performance
in vivo. Hum Gene Ther 11(15):2079–2091. Drack A, Stone E, Wachtel K, Simonelli F,

https://doi.org/10.1089/ Leroy BP, Wright JF, High KA, Maguire AM
104303400750001390 (2017) Efficacy and safety of voretigene nepar-
30. Crosson SM, Dib P, Smith JK, Zolotukhin S vovec (aav2-hrpe65v2) in patients with rpe65-
(2018) Helper-free production of laboratory mediated inherited retinal dystrophy: a rando-
grade aav and purification by iodixanol density mised, controlled, open-label, phase 3 trial.
gradient centrifugation. Mol Ther Methods Lancet 390(10097):849–860. https://doi.
Clin Dev 10:1–7. https://doi.org/10.1016/j. org/10.1016/S0140-6736(17)31868-8
omtm.2018.05.001 38. Williams AM, Muir KW (2018) Awareness and
31. Strobel B, Miller FD, Rist W, Lamla T (2015) attitudes toward corneal donation: challenges
Comparative analysis of cesium chloride- and and opportunities. Clin Ophthalmol
iodixanol-based purification of recombinant 12:1049–1059. https://doi.org/10.2147/
adeno-associated viral vectors for preclinical OPTH.S142702
applications. Hum Gene Ther Methods 26 39. Gain P, Jullienne R, He Z, Aldossary M,
(4):147–157. https://doi.org/10.1089/ Acquart S, Cognasse F, Thuret G (2016)
hgtb.2015.051 Global survey of corneal transplantation and
32. Grieger JC, Choi VW, Samulski RJ (2006) eye banking. JAMA Ophthalmol 134
Production and characterization of adeno- (2):167–173. https://doi.org/10.1001/
associated viral vectors. Nat Protoc 1 jamaophthalmol.2015.4776
(3):1412–1428. https://doi.org/10.1038/ 40. Mohan RR, Tandon A, Sharma A, Cowden JW,
nprot.2006.207 Tovey JC (2011) Significant inhibition of cor-
33. Samulski RJ, Chang LS, Shenk T (1989) neal scarring in vivo with tissue-selective, tar-
Helper-free stocks of recombinant adeno- geted aav5 decorin gene therapy. Invest
associated viruses: normal integration does Ophthalmol Vis Sci 52(7):4833–4841.
not require viral gene expression. J Virol 63 https://doi.org/10.1167/iovs.11-7357
(9):3822–3828 41. Chaudhary K, Moore H, Tandon A, Gupta S,
34. Song L, Llanga T, Conatser LM, Zaric V, Gil- Khanna R, Mohan RR (2014) Nanotechnol-
ger BC, Hirsch ML (2018) Serotype survey of ogy and adeno-associated virus-based decorin
aav gene delivery via subconjunctival injection gene therapy ameliorates peritoneal fibrosis.
in mice. Gene Ther 25(6):402–414. https:// Am J Physiol Renal Physiol 307(7):
doi.org/10.1038/s41434-018-0035-6 F777–F782. https://doi.org/10.1152/
35. Hirsch ML, Conatser LM, Smith SM, Salmon ajprenal.00653.2013
JH, Wu J, Buglak NE, Davis R, Gilger BC 42. Sharma A, Ghosh A, Hansen ET, Newman JM,
(2017) Aav vector-meditated expression of Mohan RR (2010) Transduction efficiency of
hla-g reduces injury-induced corneal vasculari- aav 2/6, 2/8 and 2/9 vectors for delivering
zation, immune cell infiltration, and fibrosis. genes in human corneal fibroblasts. Brain Res
Sci Rep 7(1):17840. https://doi.org/10. Bull 81(2–3):273–278. https://doi.org/10.
1038/s41598-017-18002-9 1016/j.brainresbull.2009.07.005
36. Bennett J, Wellman J, Marshall KA, 43. Lai LJ, Xiao X, Wu JH (2007) Inhibition of
McCague S, Ashtari M, DiStefano-Pappas J, corneal neovascularization with endostatin
Elci OU, Chung DC, Sun J, Wright JF, Cross delivered by adeno-associated viral (aav) vector
DR, Aravand P, Cyckowski LL, Bennicelli JL, in a mouse corneal injury model. J Biomed Sci
Mingozzi F, Auricchio A, Pierce EA, 14(3):313–322. https://doi.org/10.1007/
Ruggiero J, Leroy BP, Simonelli F, High KA, s11373-007-9153-7
Maguire AM (2016) Safety and durability of 44. Cheng HC, Yeh SI, Tsao YP, Kuo PC (2007)
effect of contralateral-eye administration of Subconjunctival injection of recombinant
aav2 gene therapy in patients with childhood- aav-angiostatin ameliorates alkali burn induced
onset blindness caused by rpe65 mutations: a corneal angiogenesis. Mol Vis 13:2344–2352
follow-on phase 1 trial. Lancet 388 45. Mohan RR, Tovey JC, Sharma A, Schultz GS,
(10045):661–672. https://doi.org/10.1016/ Cowden JW, Tandon A (2011) Targeted dec-
S0140-6736(16)30371-3 orin gene therapy delivered with adeno-
37. Russell S, Bennett J, Wellman JA, Chung DC, associated virus effectively retards corneal neo-
Yu ZF, Tillman A, Wittes J, Pappas J, Elci O, vascularization in vivo. PLoS One 6(10):
McCague S, Cross D, Marshall KA, Walshire J, e26432. https://doi.org/10.1371/journal.
Kehoe TL, Reichert H, Davis M, Raffini L, pone.0026432
George LA, Hudson FP, Dingfield L, Zhu X, 46. Vance M, Llanga T, Bennett W, Woodard K,
Haller JA, Sohn EH, Mahajan VB, Pfeifer W, Murlidharan G, Chungfat N, Asokan A,
Weckmann M, Johnson C, Gewaily D, Gilger B, Kurtzberg J, Samulski RJ, Hirsch
ML (2016) Aav gene therapy for mps1- (6):e0128759. https://doi.org/10.1371/jour

associated corneal blindness. Sci Rep 6:22131. nal.pone.0128759
https://doi.org/10.1038/srep22131 57. Buie LK, Rasmussen CA, Porterfield EC, Ram-
47. Luca T, Givogri MI, Perani L, Galbiati F, golam VS, Choi VW, Markovic-Plese S,
Follenzi A, Naldini L, Bongarzone ER (2005) Samulski RJ, Kaufman PL, Borrás T (2010)
Axons mediate the distribution of arylsulfatase Self-complementary aav virus (scaav) safe and
a within the mouse hippocampus upon gene long-term gene transfer in the trabecular mesh-
delivery. Mol Ther 12(4):669–679. https:// work of living rats and monkeys. Invest
doi.org/10.1016/j.ymthe.2005.06.438 Ophthalmol Vis Sci 51(1):236–248. https://
48. Igarashi T, Miyake K, Suzuki N, Kato K, doi.org/10.1167/iovs.09-3847
Takahashi H, Ohara K, Shimada T (2002) 58. O’Callaghan J, Crosbie DE, Cassidy PS, Sher-
New strategy for in vivo transgene expression wood JM, Flügel-Koch C, Lütjen-Drecoll E,
in corneal epithelial progenitor cells. Curr Eye Humphries MM, Reina-Torres E, Wallace D,
Res 24(1):46–50 Kiang AS, Campbell M, Stamer WD, Overby
49. Mohan RR, Schultz GS, Hong JW, Wilson SE DR, O’Brien C, Tam LCS, Humphries P
(2003) Gene transfer into rabbit keratocytes (2017) Therapeutic potential of aav-mediated
using aav and lipid-mediated plasmid DNA mmp-3 secretion from corneal endothelium in
vectors with a lamellar flap for stromal access. treating glaucoma. Hum Mol Genet 26
Exp Eye Res 76(3):373–383 (7):1230–1246. https://doi.org/10.1093/
50. Mohan RR, Sharma A, Netto MV, Sinha S, hmg/ddx028
Wilson SE (2005) Gene therapy in the cornea. 59. Wang L, Xiao R, Andres-Mateos E, Vanden-
Prog Retin Eye Res 24(5):537–559. https:// berghe LH (2017) Single stranded adeno-
doi.org/10.1016/j.preteyeres.2005.04.001 associated virus achieves efficient gene transfer
51. Sharma A, Tovey JC, Ghosh A, Mohan RR to anterior segment in the mouse eye. PLoS
(2010) Aav serotype influences gene transfer One 12(8):e0182473. https://doi.org/10.
in corneal stroma in vivo. Exp Eye Res 91 1371/journal.pone.0182473
(3):440–448. https://doi.org/10.1016/j. 60. Rocha EM, Di Pasquale G, Riveros PP,
exer.2010.06.020 Quinn K, Handelman B, Chiorini JA (2011)
52. Mohan RR, Sharma A, Cebulko TC, Tandon A Transduction, tropism, and biodistribution of
(2010) Vector delivery technique affects gene aav vectors in the lacrimal gland. Invest
transfer in the cornea in vivo. Mol Vis Ophthalmol Vis Sci 52(13):9567–9572.
16:2494–2501 https://doi.org/10.1167/iovs.11-8171
53. Mohan RR, Sinha S, Tandon A, Gupta R, 61. Ling C, Lu Y, Cheng B, McGoogan KE, Gee
Tovey JC, Sharma A (2011) Efficacious and SW, Ma W, Li B, Aslanidi GV, Srivastava A
safe tissue-selective controlled gene therapy (2011) High-efficiency transduction of liver
approaches for the cornea. PLoS One 6(4): cancer cells by recombinant adeno-associated
e18771. https://doi.org/10.1371/journal. virus serotype 3 vectors. J Vis Exp (49).
pone.0018771 https://doi.org/10.3791/2538
54. Hippert C, Ibanes S, Serratrice N, Court F, 62. Dong JY, Fan PD, Frizzell RA (1996) Quanti-
Malecaze F, Kremer EJ, Kalatzis V (2012) Cor- tative analysis of the packaging capacity of
neal transduction by intra-stromal injection of recombinant adeno-associated virus. Hum
aav vectors in vivo in the mouse and ex vivo in Gene Ther 7(17):2101–2112. https://doi.
human explants. PLoS One 7(4):e35318. org/10.1089/hum.1996.7.17-2101
https://doi.org/10.1371/journal.pone. 63. Hermonat PL, Quirk JG, Bishop BM, Han L
0035318 (1997) The packaging capacity of adeno-
55. Tsai ML, Chen SL, Chou PI, Wen LY, Tsai RJ, associated virus (aav) and the potential for
Tsao YP (2002) Inducible adeno-associated wild-type-plus aav gene therapy vectors. FEBS
virus vector-delivered transgene expression in Lett 407(1):78–84
corneal endothelium. Invest Ophthalmol Vis 64. McCarty DM, Fu H, Monahan PE, Toulson
Sci 43(3):751–757 CE, Naik P, Samulski RJ (2003) Adeno-
56. Bogner B, Boye SL, Min SH, Peterson JJ, associated virus terminal repeat (tr) mutant
Ruan Q, Zhang Z, Reitsamer HA, Hauswirth generates self-complementary vectors to over-
WW, Boye SE (2015) Capsid mutated adeno- come the rate-limiting step to transduction
associated virus delivered to the anterior cham- in vivo. Gene Ther 10(26):2112–2118.
ber results in efficient transduction of trabecu- https://doi.org/10.1038/sj.gt.3302134
lar meshwork in mouse and rat. PLoS One 10 65. Wang Z, Ma HI, Li J, Sun L, Zhang J, Xiao X
(2003) Rapid and highly efficient transduction
by double-stranded adeno-associated virus
vectors in vitro and in vivo. Gene Ther 10 serotype 1 (aav1) and aav8 vectors, using dual
(26):2105–2111. https://doi.org/10.1038/ ion-exchange adsorptive membranes. Hum
sj.gt.3302133 Gene Ther 20(9):1013–1021. https://doi.
66. Hirsch ML, Agbandje-McKenna M, Samulski org/10.1089/hum.2009.006
RJ (2010) Little vector, big gene transduction: 70. Ayuso E, Mingozzi F, Montane J, Leon X,
fragmented genome reassembly of adeno- Anguela XM, Haurigot V, Edmonson SA,
associated virus. Mol Ther 18(1):6–8. Africa L, Zhou S, High KA, Bosch F, Wright
https://doi.org/10.1038/mt.2009.280 JF (2010) High aav vector purity results in
67. Grieger JC, Samulski RJ (2005) Packaging serotype- and tissue-independent enhance-
capacity of adeno-associated virus serotypes: ment of transduction efficiency. Gene Ther 17
impact of larger genomes on infectivity and (4):503–510. https://doi.org/10.1038/gt.
postentry steps. J Virol 79(15):9933–9944. 2009.157
https://doi.org/10.1128/JVI.79.15.9933- 71. Moullier P, Snyder RO (2008) International
9944.2005 efforts for recombinant adeno-associated viral
68. Dong B, Nakai H, Xiao W (2010) Characteri- vector reference standards. Mol Ther 16
zation of genome integrity for oversized (7):1185–1188. https://doi.org/10.1038/
recombinant aav vector. Mol Ther 18 mt.2008.125
(1):87–92. https://doi.org/10.1038/mt. 72. Matthaei M, Meng H, Bhutto I, Xu Q,
2009.258 Boelke E, Hanes J, Jun AS (2012) Systematic
69. Okada T, Nonaka-Sarukawa M, Uchibori R, assessment of microneedle injection into the
Kinoshita K, Hayashita-Kinoh H, Nitahara- mouse cornea. Eur J Med Res 17:19. https://
Kasahara Y, Takeda S, Ozawa K (2009) Scal- doi.org/10.1186/2047-783X-17-19
able purification of adeno-associated virus
Chapter 8
The Self-assembly Approach as a Tool for the Tissue

Engineering of a Bi-lamellar Human Cornea
Gaëtan Le-Bel, Pascale Desjardins, Camille Couture, Lucie Germain,
and Sylvain L. Guérin
Abstract
Tissue engineering is a flourishing field of regenerative medicine that allows the reconstruction of various
tissues of our body, including the cornea. In addition to addressing the growing need for organ transplants,
such tissue-engineered substitutes may also serve as good in vitro models for fundamental and preclinical
studies. Recent progress in the field of corneal tissue engineering has led to the development of new
technologies allowing the reconstruction of a human bi-lamellar cornea. One unique feature of this model
is the complete absence of exogenous material. Indeed, these human corneal equivalents are exclusively
composed of untransformed human corneal fibroblasts (hCFs) entangled in their own extracellular matrix,
as well as untransformed human corneal epithelial cells (hCECs), both of which isolated from donor
corneas. The reconstructed human bi-lamellar cornea thereby exhibits a well-organized stroma as well as
a well-differentiated epithelium. This chapter describes the methods used for the isolation and culture of
hCFs, the production and assembly of hCFs stromal sheets, the seeding of hCECs, and the maturation of
the tissue-engineered cornea.
Key words Cornea, Human corneal fibroblasts, Tissue engineering, Stroma, Epithelium
1 Introduction
The cornea, the transparent anterior section of the eye, is a particu-

larly attractive tissue for tissue engineering. Its composition is quite
simple: a stratified epithelium composed of 5–7 layers, a stroma
essentially made up of collagen, and an endothelium, a cell mono-
layer that plays a very critical function in corneal deturgescence
[1, 2]. To reconstruct this tissue is, however, much more complex
than it seems. Tissue-engineered corneas have been widely
improved over time in order to better mimic the characteristics of
the native cornea, such as a well-differentiated epithelium and a
fully organized stroma, which are both essential to ensure proper
Gaëtan Le-Bel and Pascale Desjardins contributed equally with all other contributors.
103
104 Gaëtan Le-Bel et al.
corneal transparency [3]. Many laboratories first used collagen gels

mixed with cells that were cultured from animals such as the pig
[4], the rabbit [5], and the bovine [6]. At that time, the use of
animal cells was more convenient due to their availability. However,
these reconstructed corneas were not suitable for grafting. Since
then, great efforts have been made to replace these animal cells by
human cell lines. To maintain long-term culture of reconstructed
corneas, immortalized cells were used at first [7, 8]. However, both
the fact that they were virally transformed and unable to faithfully
mimic the behavior of the native corneal cells restricted their use for
corneal reconstruction.
Today, due to refinements in tissue engineering, we created a
reconstructed cornea that is entirely made of untransformed human
cells and devoid of exogenous material such as collagen gels. In our
model, stromal and epithelial cells are cultured directly from post-
mortem corneas and preserved in appropriate culture conditions
over long time periods. Our model is also self-assembled, as sup-
plementation of the culture medium with ascorbic acid triggers the
synthesis of collagen by the stromal fibroblasts [9] and increases its
stability [10, 11]. This method allows the formation of a stroma
whose composition and organization are very similar to those seen
with the native cornea [12–14].
This chapter details the methods for producing a bi-lamellar
cornea composed of a stratified epithelium laying on a complex
stroma using the self-assembly approach. Particular attention is
given to the culture of human corneal cells that come directly
from postmortem corneas, the use of ascorbic acid, which is
required for the production of the self-assembled stromal matrix
devoid of any exogenous material, and the use of the air-liquid
interface to trigger the appropriate differentiation of the corneal
epithelium.
2 Materials
2.1 Cell Culture 1. Dulbecco’s modified Eagle’s medium (DMEM): Dilute

DMEM powder in apyrogenic ultrapure water. Add 3.07 g/L
2.1.1 Base Media
of NaHCO3 (final concentration: 36.5 mM). Adjust pH to 7.1.
Sterilize by filtration through a 0.22 μm low-binding dispos-
able filter and store in the dark at 4 C.
2. DME-Ham: Combine 3 parts DMEM with 1 part Ham’s
medium in apyrogenic ultrapure water. Add 3.07 g/L
NaHCO3 (final concentration: 36.5 mM), 24.3 mg/L of
0.18 mM adenine solubilized in 312.5 μL/L of 2 N HCl.
Adjust pH to 7.1. Sterilize by filtration through a 0.22 μm
low-binding disposable filter and store in the dark at 4 C.
Bi-lamellar Cornea Produced by Tissue-Engineering 105
2.1.2 Sera 1. Fetal calf serum.

2. Fetal clone II serum.
Thaw sera at 4 C or in cold water (see Note 1). Gently swirl it
to resuspend its components. Inactivate in a 56 C water bath for
30 min. To avoid repeated thawing and freezing cycles, distribute in
single-use aliquots. Store at 20 C or at 80 C for long-term
storage.
2.1.3 Additives 1. Insulin: 5 mg/mL in 5 mM HCl. Yields a 1000 stock solu-

tion. Can be stored at 4 C.
2. Epidermal growth factor (EGF): 200 μg/mL in 10 mM HCl.
Dilute (1:20) the solution with DME-Ham containing 10%
v/v fetal clone II serum. Yields a 1000 stock solution.
3. Hydrocortisone: 5 mg/mL in 95% ethanol. Dilute (1:25) the
solution with DME-Ham. Yields a 500 stock solution.
4. Penicillin G/Gentamicin: 50000 IU/mL Penicillin G and
12.5 mg/mL gentamicin in apyrogenic ultrapure water. Yields
a 500 stock solution.
5. Isoproterenol hydrochloride: Store at 4 C. Vials are at
0.2 mg/mL and are single use. Leftover isoproterenol should
not be stored. 1000 stock solution.
6. Fungizone: 0.25 mg/mL amphotericin B in apyrogenic ultra-
pure water. Yields a 500 stock solution.
For all additives, except isoproterenol, sterilize by filtration
through a 0.22 μm low-binding disposable filter. To avoid repeated
thawing and freezing cycles, distribute in single-use aliquots. Store
at 80 C.
2.1.4 Complete Media 1. Tissue transport medium (tDMEM): High (4.5 g/L) glucose
DMEM (containing sodium pyruvate and L-glutamine), 10%
(v/v) fetal calf serum, 0.2% (v/v) penicillin G/gentamicin,
0.2% (v/v) fungizone. Store in the dark at 20 C to 80 C
for up to 6 months or at 4 C for up to 10 days.
2. Complete corneal fibroblast culture medium (cfDMEM):
DMEM, 10% (v/v) fetal calf serum, 0.2% (v/v) penicillin
G/gentamicin. Store in the dark at 4 C for up to 10 days.
3. Complete corneal epithelial cell culture medium (ccDME-
Ham): DME-Ham, 5% (v/v) fetal clone II serum, 0.1% (v/v)
insulin (see Note 2), 1.06 mL/L isoproterenol, 0.1% (v/v)
epidermal growth factor, 0.2% hydrocortisone, and 0.2%
(v/v) penicillin G/gentamicin. Store in the dark at 4 C for
up to 10 days.
4. Cryopreservation medium for human corneal fibroblast: 10%

(v/v) dimethyl sulfoxide (DMSO), 90% (v/v) fetal calf serum.
Keep on ice or store at 4 C and use within 24 h (see Note 3).
5. Air-liquid corneal epithelial cell culture medium (alcDME-
Ham): DME-Ham, 5% (v/v) fetal clone II serum, 0.1% (v/v)
insulin (see Note 2), 1.06 mL/L isoproterenol, 0.1% (v/v)
epidermal growth factor, 0.2% hydrocortisone, and 0.2%
(v/v) penicillin G/gentamicin. Store in the dark at 4 C for
up to 10 days (see Note 4).
All frozen components can be thawed at 4 C (see Note 1).
2.1.5 Solutions 1. Phosphate-buffered saline (PBS): 127 mM NaCl, 2.7 mM KCl,

6.5 mM Na2HPO4, 1.5 mM KH2PO4 in apyrogenic ultrapure
water. Verify pH is between 7.35 and 7.45. Yields a 10 stock
solution. Store at room temperature.
2. PBS—Penicillin G/Gentamicin/Fungizone (PBS-P/G/F):
Dilute 10 PBS to 1 with apyrogenic ultrapure water. Steril-
ize by filtration through a 0.22 μm low-binding disposable
filter. Add Penicillin G/Gentamicin and Fungizone (dilute to
1). Store at 4 C.
3. 10 HEPES/KCl/NaCl: 0.1 M 4-(2-hydroxyethyl)-1-piper-
azineethanesulfonic acid (HEPES), 67 mM KCL, and 1.42 M
NaCl in apyrogenic ultrapure water. Adjust pH to 7.3. Yields a
10 stock solution. Store in the dark at 4 C (see Note 5).
4. HEPES/KCl/NaCl - CaCl2: Dilute 10 HEPES/KCl/NaCl
to 1 with apyrogenic ultrapure water. Add CaCl2 to 1mM
final. Adjust pH to 7.45. Store in the dark at 4 C.
5. Collagenase H: 0.125 U/mL collagenase H in cfDMEM. Ster-
ilize by filtration through a 0.22 μm low-binding disposable
filter. Preheat at 37 C and use immediately .
6. Dispase II: 2.5 mg/mL dispase II in HEPES/KCl/NaCl –
CaCl2. Verify that pH is at 7.4 with pH paper. Sterilize by
filtration through a 0.22 μm low-binding disposable filter.
Store at 4 C and use within the hour.
7. Trypsin/EDTA: 0.05% (w/v) trypsin, 0.01% (w/v) EDTA,
2.8 mM D-glucose in 1 PBS. Add 100,000 IU/L penicillin G,
25 mg/L active gentamicin, 0.00075% (v/v) pre-sterile filtered
0.1% phenol red-water solution. Adjust pH to 7.45. Sterilize by
filtration through a 0.22 μm low-binding disposable filter. To
avoid repeated thawing and freezing cycles, distribute in single-
use aliquots. Store at 20 C to 80 C.
8. L-Ascorbic acid for corneal fibroblasts: 10 mg/mL L-ascorbic
acid in cfDMEM. Sterilize by filtration through a 0.22 μm
low-binding disposable filter. Store at 4 C and use within
the day.
9. L-Ascorbic acid for corneal epithelial cell: 10 mg/mL L-ascor-

bic acid in ccDME-Ham. Sterilize by filtration through a
0.22 μm low-binding disposable filter. Store at 4 C and use
within the day.
2.1.6 Tissues and Cells 1. Human corneal fibroblasts (hCFs) are isolated, cultured, and
cryopreserved from surgically removed corneal stroma. Cor-
neas from human postmortem donors unsuitable for transplan-
tation (Banque nationale d’yeux du CHU de Québec, Québec,
Canada).
2. Human corneal epithelial cells (hCECs) are isolated, cultured,
and cryopreserved from corneal epithelium. Corneas from
human postmortem donors unsuitable for transplantation
(Banque nationale d’yeux du CHU de Québec, Québec,
Canada).
2.1.7 Labware 1. For volumes inferior to 100 mL: 0.22 μm low-binding dispos-
able filter. For volumes superior to 100 mL: filtration unit
mounted with a 47 mm diameter and 0.22 μm filter set.
2. Sterile containers.
3. 15 and 50 mL centrifuge tubes.
4. 35 40 mm and 100 15 mm cell culture Petri dishes.
5. Dissecting curved forceps.
6. Size 4 and 22 scalpel blades.
7. Trypsination unit, Celstir® 50 mL suspension culture flask.
8. Parafilm® M.
9. 25 or 75 cm2 tissue culture flasks.
10. Sterile cryogenic vials.
11. Freezing container.
12. Sterile 7 cm 7 cm gauze.
13. Dissecting curved scissor.
14. 8 mm diameter trephine.
15. Dissecting stereomicroscope
16. 6-Well tissue-culture plates.
17. Anchoring papers, made from Whatman® grade 3 qualitative
filter paper (see Note 6).
18. Plastic rings (see Note 7).
19. Sterile air-liquid stands (see Note 8).
20. Small ligating clips LIGACLIP® EXTRA.
21. Surgical single-clip applier LIGACLIP®.
22. 100 25 mm cell culture Petri dishes.
3 Methods
3.1 Cell culture Ethical approval and informed consent must be obtained for each
human tissue.
3.1.1 Tissue Sampling
and Transport 1. Shortly after donor’s death, eyes or only the corneas are
extracted and transferred in a sterile container filled with cold
(4 C) tDMEM by qualified staff.
2. Samples must be kept on ice and transported to a cell culture
facility without delay. Manipulations must be performed under
a sterile laminar flow hood cabinet.
3.1.2 Isolation of Human 1. hCFs are obtained from postmortem human donors cor-
Corneal Fibroblasts (hCFs) neas that are unsuitable for transplantation (see Note 9).
2. Wash the eye specimen in a 50 mL centrifuge tube containing
30 mL PBS-P/G/F. Agitate gently for 1–2 min. With sterile
curved forceps, transfer the eye specimen into another tube
filled with PBS-P/G/F. Repeat this step three times.
3. Place the eye specimen into a 100 mm Petri dish.
4. Surround the eye specimen with a folded sterile gauze. This
helps in holding the eye without having to touch it.
5. With the size 22 scalpel blade, make a small opening of 2–3 mm
in the sclera.
6. With curved scissors, cut out the cornea to obtain only the
limbus and the central cornea. Avoid cutting the sclera to
prevent contamination with conjonctival epithelial cells.
7. With two curved forceps, peel off the iris. Do this step while
holding the cornea in the air to avoid any damage to the
epithelium during the procedure.
8. Place the central cornea into a 35 mm tissue culture Petri dish.
The central cornea is obtained by separating the limbus from
the central cornea with an 8 mm diameter trephine (see Note
10). Hold the epithelium upward and add 5 mL of cold (4 C)
dispase II. Seal the Petri dish with parafilm.
9. Incubate overnight at 4 C.
10. With two curved forceps, mechanically detach the epithelium
from the stroma under a dissecting microscope.
11. With curved forceps, transfer the corneal stroma in a new
35 mm tissue culture Petri dish. With the size 22 scalpel
blade, cut the corneal stroma into small pieces. Add 2 mL of
collagenase H and keep cutting the corneal stroma until you
get small pieces of 1–2 mm2.
12. Collect the 2 mL of collagenase H containing the small pieces

of stroma and transfer it in a Celstir® suspension culture flask
containing 18 mL warm (37 C) collagenase H.
13. Incubate under agitation for 2–3 h at 37 C, until the small
pieces of corneal stroma have been digested by the enzyme (see
Note 11).
14. Stir well and collect the supernatant, put it in a 50 mL centri-
fuge tube, and add 20 mL of cfDMEM.
15. Centrifuge (300 g) the hCFs suspension for 10 min at room
temperature.
16. Remove the supernatant, resuspend with 20 mL of cfDMEM,
and centrifuge (300 g) the hCFs suspension for 10 min at
room temperature (see Note 12).
17. Remove the supernatant and resuspend hCFs in 5 mL of warm
(37 C) cfDMEM.
18. Seed hCFs into a culture flask and add cfDMEM. Total
medium volume should not exceed 7 mL/25 cm2.
19. When the hCFs reach 90% confluence, subculture or freeze
cells.
3.1.3 Culture 1. Place hCFs seeded culture flasks in an 8% CO2 and 100%
humidity atmosphere incubator at 37 C.
2. Change the culture medium three times a week, every
2–3 days. Remove the medium from the culture flask. Replace
it with warm (37 C) cfDMEM.
3. Monitor cell confluence (see Note 13) daily under a
microscope.
4. When cells reach 75–95% confluence, subculture (Fig. 1) or
cryopreserve them. Do not let cells reach 100% confluence.
3.1.4 Subculture 1. Remove the culture medium.

2. Depending on culture flask size, swiftly rinse cells with either
1 or 2 mL (for either 25 or 75 cm2 culture flasks, respectively)
trypsin/EDTA. Remove it.
3. Depending on culture flask size, add either 2 or 3 mL (for
either 25 or 75 cm2 culture flasks, respectively) of trypsin/
EDTA into the culture flask.
4. Incubate at 37 C until all cells are completely detached from
the flask (verify cell detachment under a microscope). Time for
complete detachment should be around 10 min. Do not incu-
bate for more than 10 min (see Note 14).
5. Depending on culture flask size, neutralize trypsin activity by
adding either 2 or 3 mL of cfDMEM.
Fig. 1 Growth and morphological characteristics of cultured human corneal

fibroblasts. Human corneal fibroblasts (hCFs) isolated from the central corneal
6. Vigorously pipette the cell suspension up and down at least ten

times to ensure suspension homogeneity.
7. Transfer the cell suspension into a 50 mL centrifuge tube.
8. Depending on culture flask size, thoroughly rinse the culture
flask with 2 or 4 mL cfDMEM. Transfer the suspension into
the 50 mL tube.
9. Use an automated cell counter or a hemocytometer to count
cells.
10. Use trypan blue staining and a hemocytometer to estimate cell
viability. Cell viability is expected to be greater than 95%.
11. Centrifuge (300 g) the cell suspension for 10 min at room
temperature.
12. Remove the supernatant and resuspend cells at the desired
concentration in cfDMEM.
13. Seed cells at no less than 7000 cells/cm2 (see Note 15) into a
culture flask. Seed directly into the culture medium cfDMEM.
3.1.5 Cryopreservation 1. Fill a freezing container with 100% isopropyl alcohol. Store at
4 C until cool.
2. Follow steps 1 through 11 from the previous section (Sub-
heading 3.1.4).
concentration (max. 1 107/mL) in hCF cryopreservation
medium. Put the tube on ice.
4. Aliquot in cryogenic vials on ice.
5. Put the cryogenic vials in the freezing container.
6. Store the container overnight at 80 C. Under these condi-
tions, cell temperature should drop 1 C/min.
7. Store cryogenic vials in liquid nitrogen for long-term storage.
3.1.6 Thawing 1. Put the cryogenic vial in a 37 C water bath. Do not let the cell
suspension thaw completely. A small ice pellet should remain.
2. Add 0.5–1 mL of cold (4 C) cfDMEM into the cryogenic vial.
3. As soon as the remaining ice has melted, transfer the content of
the cryogenic vial into a 50 mL centrifuge tube containing
8–10 mL of cold (4 C) cfDMEM.
Fig. 1 (continued) stroma of a postmortem human eye were isolated and seeded
in 75 cm2 tissue culture flasks. hCFs reached 25% confluence after 4 days in
culture (a), 40% confluence after 7 days in culture (b) and 95% confluence after
9 days in culture (c). Scale bar: 25 μm

temperature.
5. Remove the supernatant and resuspend cells in 10 mL of warm
(37 C) cfDMEM.
6. Use an automated cell counter or a hemocytometer to count
the cells.
7. Use trypan blue staining and a hemocytometer to estimate cell
viability. Cell viability is expected to be greater than 80%.
temperature.
concentration in cfDMEM (see Note 16).
10. Seed cells at no less than 7000 cells/cm2 (see Note 15) into a
culture flask. Seed directly into the culture medium cfDMEM.
3.2 All further manipulations must be performed under a sterile lami-

Tissue-Engineered nar flow hood cabinet.
Human Cornea
3.2.1 Production of hCF 1. In a 6-well plate, dispose an anchoring paper in each well and a
Sheets plastic ring on top of it.
2. Add ascorbic acid solution in cfDMEM to obtain a final con-
centration of 50 μg/mL in the medium.
3. Seed 105 hCFs per well (9.6 cm2) in the prepared 6-well
culture plate in 5 mL of cfDMEM containing 50 μg/mL
ascorbic acid. Seed directly into the culture medium (see Note
17).
4. Incubate in 8% CO2, 100% humidity atmosphere at 37 C for
40 days. Change culture medium three times a week.
3.2.2 Assembly of hCF 1. Completely remove culture medium from two wells at a time
Sheets for Stromal and remove temporarily the plastic ring.
Reconstruction 2. With a curved forceps, gently scrub the inside of the well all
around the anchoring paper and carefully detach the stromal
sheet from the bottom of the well.
3. Transfer one stromal sheet on top of the other.
4. With a sterile surgical single-clip applier, take one ligating clip
at a time.
5. Using a sterile curved forceps in one hand and the surgical
single-clip applier in the other, attach together the two stromal
sheets by placing the ligating clip around the two anchoring
papers.
6. Repeat steps 4 and 5 three times while disposing the ligating

clips evenly all around the anchoring papers.
7. Replace the plastic ring on top of the reconstructed stroma and
add 5 mL of cfDMEM containing 50 μg/mL ascorbic acid.
8. Incubate in 8% CO2, 100% humidity atmosphere at 37 C for
1 week to merge. Change culture medium every 2 or 3 days
with cfDMEM containing 50 μg/mL ascorbic acid.
3.2.3 Seeding of Human 1. One week after the assembly of fibroblast sheets, hCECs can be
Corneal Epithelial Cells seeded on the reconstructed stroma. hCECs between their
(hCECs) and Maturation second and fourth passages at the time of seeding are favored.
at the Air-Liquid Interface 2. Resuspend hCECs at 2 105 or 2.5 105 cells/mL in
ccDME-Ham.
3. Add 10 mg/mL ascorbic acid solution to ccDME-Ham to
obtain a final concentration of 50 μg/mL.
4. Remove culture medium of reconstructed stromas and replace
it with 5 mL of warm ccDME-Ham containing 50 μg/mL
ascorbic acid.
5. Seed 1 106 hCECs per reconstructed stroma drop by drop
everywhere within the plastic ring. Replace the 6-well plate in
the incubator carefully to limit the dispersion of hCECs within
the well (Fig. 2).
6. Incubate in 8% CO2, 100% humidity atmosphere at 37 C.
Change culture medium three times a week with ccDME-
Ham containing 50 μg/mL ascorbic acid.
7. After 1 week, remove culture medium and plastic ring in
each well.
8. Place an air-liquid stand in a 100 25 mm cell culture
Petri dish.
9. Using a curved forceps, carefully detach the reconstructed
cornea from the bottom of the well.
10. Place the reconstructed tissue on the air-liquid stand with the
hCECs on top (Fig. 2).
11. Add the right volume of 10 mg/mL ascorbic acid solution in
alcDME-Ham to obtain a final concentration of 50 μg/mL.
12. Depending on the air-liquid stand, add the right volume (here,
18 mL) of alcDME-Ham containing 50 μg/mL ascorbic acid.
Make sure the bottom of the reconstructed cornea is in contact
with culture medium while the top is kept dry (see Note 18).
13. Incubate in 8% CO2, 100% humidity atmosphere at 37 C and
keep at the air-liquid interface for 1 week (Fig. 3). Change
culture medium every 2 or 3 days with alcDME-Ham contain-
ing 50 μg/mL ascorbic acid.
Fig. 2 Detail of the procedure used for the production of a human tissue-engineered bi-lamellar cornea. (a)
After assembly and maturation of hCF sheets for stromal reconstruction, hCECs were seeded on the
reconstructed human corneal stromas. (b and c) An air-liquid stand produced by 3D printing in PLA and
covered with a nylon membrane. (d and e) The reconstructed tissue is deposited on the air-liquid stand with
hCECs on top and left in culture for 1 week in order to induce vertical stratification of the epithelium. (f)
Mature, human bi-lamellar cornea
4 Notes
1. Sera and additives can be thawed more rapidly at room temper-

ature or in a 37 C water bath. However, do not refreeze.
Rather, we recommend using immediately or dilute in culture
medium at working dilution for further utilization.
2. Serum must be added first, followed by insulin. Insulin must be
added with a new sterile plastic pipette.
Fig. 3 Characteristics of the human tissue-engineered bi-lamellar cornea (a and b) Macroscopic views
showing the transparency of the human bi-lamellar cornea. (c) Electron microscopic analysis of the tissue-
engineered corneal stroma. Organization of collagen fibers is very similar to that observed in the native corneal
stroma with collagen fibers perpendicular to each other. BM basement membrane; hCEC human corneal
epithelial cell; C collagen fibers; hCFs human corneal fibroblasts. (d) Histological view of the human
bi-lamellar cornea that shows the stratified epithelium made up of five to seven cell layers attached to the
corneal stroma containing many hCFs (sections were stained with Masson trichrome; cells are purple and
collagen is bluish). Scale bar: 20 μm
3. DMSO is a toxic oxidative agent at temperatures above 10 C.

Working with cells in contact with a solution containing
DMSO must be done quickly and on ice.
4. The composition of the alcDME-Ham is substantially the same
as for the ccDME-Ham, except that EGF is absent, allowing
hCECs differentiation rather than proliferation.
5. HEPES may undergo degradation when exposed to light and
might become toxic.
6. To make the anchoring paper, cut a ring with a 31.8 mm
external diameter and a 22.2 mm internal diameter in the
Whatman filter paper grade 3. Anchoring papers are sterilized
by autoclave.
7. Plastic rings are made of 50 mL collection tube. Caps are

removed and tubes are cut at approximately 1.5 cm from the
aperture of the tube. The 1.5 cm high plastic ring is notched at
its base to allow the culture medium to move freely within the
well. Plastic rings are used to prevent the anchoring paper (with
cells attach to it) to float at the surface of the culture medium.
Ingots or anchoring rings in stainless steel may be used as well.
Sterilization of the plastic rings is performed by gas sterilization
with ethylene oxide (Fig. 2).
8. Air-liquid stands are made by 3D printing. Dimensions are
56 mm external diameter, 32 mm internal diameter, and
6 mm height. Three legs of 4 mm high support the air-liquid
stands. Stands are made of polylactic acid (PLA) and are cov-
ered with a nylon membrane fixed to the plastic stands with
chloroform (Fig. 2). Since it deforms the plastic, gas steriliza-
tion is not suitable for those air-liquid stands. Instead, we
sterilize them by immersing the stands in 100% isopropyl alco-
hol, followed by ultraviolet irradiation under a sterile laminar
flow hood cabinet. Each side of the air-liquid stand is exposed
to ultraviolet irradiation for at least 30 min.
9. Protocols must be approved by the institution’s committee for
the protection of human subjects. It is preferable to use corneas
with as short as possible postmortem time to improve cultiva-
tion of hCFs and hCEC.
10. Corneal epithelial stem cells are found in the limbal basal
epithelium, the transition area separating the cornea and
underlying sclera. Discarding the limbal ring prevents the for-
mation of hCECs colonies caused by corneal epithelial stem
cells contaminating the corneal fibroblast cultures as hCECs
can efficiently proliferate in cfDMEM.
11. Collagenase H breaks the peptide bonds in collagen, destroy-
ing extracellular matrix structures. Collagenase treatment
should be stopped as soon as there are no more visible pieces
in suspension. Indeed, collagenase turns out to be toxic for
cells when incubated for a long time. It is therefore necessary to
check the suspension every 30 min.
12. Contrary to trypsin, collagenase H activity cannot be neutra-
lized with fetal calf serum contained in cfDMEM. For long-
term use, collagenase H is toxic for the cells. A double wash is
then required to completely remove remaining collagenase H
from the suspension.
13. Here, we define confluence as the approximate percentage of
the culture flask surface covered by hCFs. It is estimated under
a microscope by assessing how much of a given field of vision is
occupied by hCFs. Do not let more confluent areas differenti-
ate for the sake of obtaining a higher mean confluence.
14. For optimal efficiency, do not stack culture flasks on top of each
other in the incubator. Temperature is typically higher on the
flask surface directly in contact with the incubator shelf.
15. The seeding density is determined by the operator. hCF den-
sities are expected to double each day. Proliferation can vary
from one population to another and decreases as cells are
passaged. It is recommended to seed at no less than
7000 cells/cm2 with an uncharacterized population. 95% con-
fluence should be reached within 5–7 days for the first few
passages.
16. A double wash is preferable to completely remove DMSO from
the suspension.
17. hCFs between their fifth and seventh passages are favored for
generating sheets.
18. Try to avoid air bubbles beneath the reconstructed cornea to
maintain the air-liquid interface. If air bubble should form,
remove it with the vacuum.
Acknowledgements
The authors would like to thank current and former members of

the LOEX and CUO-Recherche laboratories who contributed to
develop and improve the foregoing protocols. This work was sup-
ported by the Canadian Institutes for Health Research (CIHR)
grant MOP-12087 and FDN-143213 (L.G.), the Fondation des
Pompiers du Québec pour les Grands Brûlés (FPQGB), the Fonds
de Recherche du Québec-Santé (FRQS), and the Réseau de thér-
apie cellulaire, tissulaire et génique du Québec -ThéCell (a thematic
network supported by the FRQS). The Banque d’yeux Nationale is
partly supported by the Réseau de Recherche en Santé de la Vision
from the FRQS. P.D. and C.C. were supported by studentships
from the FRQS. L.G. is the recipient of a Tier 1 Canadian Research
Chair on Stem Cells and Tissue Engineering and a Research Chair
on Tissue-Engineered Organs and Translational Medicine of the
Fondation de l’Université Laval.
References
1. Bonanno JA (2012) Molecular mechanisms 3. DelMonte DW, Kim T (2011) Anatomy and
underlying the corneal endothelial pump. Exp physiology of the cornea. J Cataract Refract
Eye Res 95(1):2–7. https://doi.org/10. Surg 37(3):588–598. https://doi.org/10.
1016/j.exer.2011.06.004 1016/j.jcrs.2010.12.037
2. Eghrari AO, Riazuddin SA, Gottsch JD (2015) 4. Schneider AI, Maier-Reif K, Graeve T (1999)
Overview of the cornea: Structure, function, Constructing an in vitro cornea from cultures
and development. Prog Mol Biol Transl Sci of the three specific corneal cell types. In Vitro
134:7–23. https://doi.org/10.1016/bs. Cell Dev Biol Anim 35(9):515–526. https://
pmbts.2015.04.001 doi.org/10.1007/s11626-999-0062-0
5. Zieske JD, Mason VS, Wasson ME, Meunier extracellular matrix accumulation in vitro. Bio-
SF, Nolte CJ, Fukai N, Olsen BR, Parenteau chem J 269:175–181
NL (1994) Basement membrane assembly and 11. Michel M, L’Heureux N, Pouliot R, Xu W,
differentiation of cultured corneal cells: Impor- Auger FA, Germain L (1999) Characterization
tance of culture environment and endothelial of a new tissue-engineered human skin equiva-
cell interaction. Exp Cell Res 214(2):621–633. lent with hair. In Vitro Cell Dev Biol Anim 35
https://doi.org/10.1006/excr.1994.1300 (6):318–326. https://doi.org/10.1007/
6. Parnigotto PP, Bassani V, Montesi F, Conconi s11626-999-0081-x
MT (1998) Bovine corneal stroma and epithe- 12. Couture C, Zaniolo K, Carrier P, Lake J,
lium reconstructed in vitro: Characterisation Patenaude J, Germain L, Guérin SL (2016)
and response to surfactants. Eye (Lond) 12 The tissue-engineered human cornea as a
(Pt 2):304–310. https://doi.org/10.1038/ model to study expression of matrix metallo-
eye.1998.70 proteinases during corneal wound healing.
7. Griffith M (1999) Functional human corneal Biomaterials 78:86–101. https://doi.org/10.
equivalents constructed from cell lines. Science 1016/j.biomaterials.2015.11.006
286(5447):2169–2172. https://doi.org/10. 13. Carrier P, Deschambeault A, Talbot MV, Gias-
1126/science.286.5447.2169 son CJ, Auger FA, Guérin SL, Germain L
8. Griffith M (2002) Artificial human cornea. (2008) Characterization of wound reepithelia-
Cornea 21(2):54–61. https://doi.org/10. lization using a new human tissue-engineered
1097/01.ico.0000263120.68768.f8 corneal wound healing model. Invest Ophthal-
9. Geesin JC, Darr D, Kaufman R, Murad S, Pin- mol Vis Sci 49(4):1376–1385. https://doi.
nell SR (1988) Ascorbic acid specifically org/10.1167/iovs.07-0904
increases type i and type iii procollagen messen- 14. Proulx S, d’Arc Uwamaliya J, Carrier P,
ger rna levels in human skin fibroblasts. J Invest Deschambeault A, Audet C, Giasson CJ,
Dermatol 90(4):420–424. https://doi.org/ Guérin SL, Auger FA, Germain L (2010)
10.1111/1523-1747.ep12460849 Reconstruction of a human cornea by the self-
10. Chan D, Lamande SR, Cole WG, Bateman JF assembly approach of tissue engineering using
(1990) Regulation of procollagen synthesis the three native cell types. Mol Vis
and processing during ascorbate-induced 16:2192–2201
Chapter 9
Formation of Corneal Stromal-Like Assemblies Using

Human Corneal Fibroblasts and Macromolecular Crowding
Mehmet Gürdal, Gülinnaz Ercan, and Dimitrios I. Zeugolis
Abstract
Tissue engineering by self-assembly allows for the formation of living tissue substitutes, using the cells’
innate capability to produce and deposit tissue-specific extracellular matrix. However, in order to develop
extracellular matrix-rich implantable devices, prolonged culture time is required in traditionally utilized
dilute ex vivo microenvironments. Macromolecular crowding, by imitating the in vivo tissue density,
dramatically accelerates biological processes, resulting in enhanced and accelerated extracellular matrix
deposition. Herein, we describe the ex vivo formation of corneal stromal-like assemblies using human
corneal fibroblasts and macromolecular crowding.
Key words Corneal stroma, Tissue engineering by self-assembly, Macromolecular crowding, Extra-
cellular matrix
1 Introduction
The cornea is a highly specialized transparent tissue located at the

anterior most surface of the eye. The cornea is composed of three
layers: an outer epithelial layer, a middle stromal layer, and an inner
endothelial layer. The corneal stromal layer is synthesized by cor-
neal keratocytes, makes up 90% of the corneal thickness, and com-
prises a heterodimeric complex of type I and type V collagen fibers,
which are arranged in bundles referred to as lamellae [1–5].
Tissue engineering by self-assembly approaches utilize the nat-
ural sophistication of the cells to create tissue-specific or native
supramolecular assemblies [6, 7]. Ex vivo, the formation of native
supramolecular assemblies is dependent on the rate of extracellular
matrix (ECM) synthesis and deposition, which, under the dilute
traditional cell culture conditions, is extremely slow. In vivo, cells
reside in the dense network of the ECM, where the enzymatic
processing of procollagen to collagen type I is rapid. In vitro, in
the dilute culture media, the conversion of water-soluble
119
120 Mehmet Gürdal et al.
procollagen to water-insoluble collagen type I is very slow; thus,

prolonged culture times are required to produce an implantable
device [8, 9].
Macromolecular crowding (MMC) is a biophysical phenome-
non that accelerates biological processes (such as the enzymatic
conversion of procollagen to collagen), resulting in amplified and
accelerated ECM deposition in vitro in permanently differentiated
and stem cell cultures [10–14]. Among the various macromolecular
crowders that have been used to-date, carrageenan, a highly sul-
fated polysaccharide, has been shown to induce the highest and
fastest ECM deposition in various cell types due to its polydispersity
and negative charge [15, 16].
Herein, we describe the ex vivo formation of corneal stromal-
like assemblies using human corneal fibroblasts and ascending con-
centrations of carrageenan in order to identify the optimal concen-
tration for maximum ECM deposition. The influence of
carrageenan in corneal fibroblasts cultures was assessed with cell
morphology, metabolic activity, viability and proliferation and
sodium dodecyl sulfate poly-acrylamide gel electrophoresis
(SDS-PAGE) and immunocytochemistry.
2 Materials
All cell culture materials should be handled following aseptic tech-

nique. Human primary cells should be handled using Biosafety
Level 2 practices and containment. Diligently follow all the waste
disposal regulations of your country/institution when disposing
any kind of waste material.
2.1 Cell Culture 1. Human corneal fibroblasts (HCFs), cryopreserved.

2. Fibroblast Medium Kit (see Note 1).
3. Poly(L-lysine) solution.
4. Double distilled water (ddH2O), sterile.
5. Hank’s Balanced Salt Solution (HBSS), sterile.
6. Penicillin-streptomycin solution (PS) 100: 10,000 units/ml
penicillin, 10 mg streptomycin/mL.
7. Maintaining growth medium: Fibroblast basal medium, 2%
fetal bovine serum (FBS), 1% fibroblast growth supplement
(FGS), 1% PS.
8. Dimethyl sulfoxide (DMSO), suitable for cell culture.
9. Dulbecco’s modified eagle medium: nutrient mixture F12
(DMEM/F12).
10. Newborn calf serum (NBCS), heat inactivated.
11. Human basic fibroblast growth factor (FGF-2).
Macromolecular Crowding in Human Corneal Fibroblast Culture 121
12. Growth medium: DMEM/F12, 10% NBCS, 5 ng/mL FGF-2,

1% PS.
13. 0.25% Trypsin-EDTA solution.
14. Carrageenan powder, suitable for gel preparation.
15. 100 mM L-ascorbic acid 2-phosphate solution: 100 mM L-
ascorbic acid 2-phosphate in ddH2O. When dissolved filter
with a syringe and a 0.2 μm syringe filter. Store it in frozen
aliquots at 20 C and protect from the light.
16. 0.2 μm surfactant-free cellulose acetate sterile syringe filters.
17. 75 cm2 tissue culture flask.
18. 175 cm2 tissue culture flask.
19. Poly(L-lysine)-coated 75 cm2 tissue culture treated flask:
0.1 mg/mL poly(L-lysine) solution is prepared and 1.5 mL of
solution is transferred to 75 cm2 flask (see Note 2). The flask is
gently rocked to evenly coat the surface. After 5 min, the excess
solution is removed and the surface is thoroughly rinsed with
sterile ddH2O and allowed to dry for several hours.
20. 24- and 48-well tissue culture treated plates.
21. MMC growth medium: Weigh the appropriated amount of
carrageenan powder in a 1.5 mL microcentrifuge tube in
order to prepare 1 mL of medium per well of 24-well tissue
culture plate or 0.5 mL of medium per well of 48-well tissue
culture plate to be treated, considering the concentration of
carrageenan. Always prepare 2 mL extra in order to compen-
sate for pipetting errors. In order to disinfect the carrageenan,
irradiate it with UV-C light for 15 min (see Note 3). Supple-
ment growth medium with 1 μL the 100 mM ascorbic acid
solution per mL of medium in order to reach a concentration
of 100 μM. This will be used as the control medium and will
also be used to prepare crowding MMC growth medium by
adding carrageenan. For the preparation of MMC growth
medium, recover the disinfected carrageenan from the
1.5 mL microcentrifuge tube by suspending it in 1 mL of
growth medium supplemented with ascorbic acid and transfer
the volume to a tube with the remaining medium. Repeat this
step at least twice in order to ensure the complete recovery of
the carrageenan from the tube. For non-crowded control wells,
keep aside the appropriated volume of growth medium supple-
mented with L-ascorbic acid 2-phosphate for this purpose. For
the solubilization of the carrageenan in the growth medium
supplemented with ascorbic acid, incubate the tube in a ther-
mostatic bath at 37 C for at least 30 min (see Note 4).
2.2 Collagen 1. HBSS.

Deposition Analysis 2. 0.5 M Acetic acid: 26.622 mL glacial acetic acid in 973.378 mL
ddH2O.
3. 100 μg/mL pepsin from porcine gastric mucosa: 1 mg pepsin
in 0.5 M acetic acid and dilute 1/10 (v/v) in HBSS. Use pepsin
solution in 30 min.
4. Phenol red solution: 10 mg of phenol red in 50 mL of ddH2O.
5. 1 N sodium hydroxide (NaOH): 4 g of NaOH in ddH2O, final
volume 100 mL (see Note 5).
6. 37% hydrochloric acid (HCl).
7. 1.875 M Tris–HCl, pH 8.8: 22.70 g Tris-base, 80 mL ddH2O,
2 mL 37% HCl. Leave overnight to equilibrate, adjust pH to
8.8 with a few drops concentrated HCl and make it up to
100 mL with ddH2O. Keep it at 4 C.
8. 1.25 M Tris–HCl, pH 6.8: 15.14 g Tris-base, 70 mL ddH2O,
7 mL 37% HCl. Leave overnight to equilibrate, adjust pH to
6.8 with a few drops concentrated HCl and make it up to
100 mL with ddH2O. Keep it at 4 C.
9. Sodium dodecyl sulfate (SDS) (see Note 6).
10. 5 sample buffer: 0.25 g SDS, 0.625 mL 1.25 M Tris–HCl,
pH 6.8, 2 mL ultrapure water. Leave overnight for the foam to
settle. Top up with glycerol to 5 mL (approximately 2.3 mL).
Add 2.5 mg bromophenol blue per 10 mL buffer. Also prepare
1 sample buffer diluted from 5 sample buffer with ddH2O.
11. 5 running buffer: 15.1 g Tris-base, 72 g glycine, 5 g SDS, 1 L
ddH2O. Store at 4 C. 1 running buffer is made to run the
gel from 5 running buffer by diluting in ddH2O.
12. 30% Acrylamide/Bis (37.5:1).
13. 10% (w/v) SDS: 10 g SDS in 90 ml ddH2O and adjust the
volume to 100 mL. This solution can be stored at room
temperature for up to 6 months.
14. 100 mg/mL Ammonium Persulfate (APS): 500 mg APS, 5 mL
ddH2O. Aliquot in microcentrifuge tubes and keep them at
20 C. The solution is active for a few months (see Note 7).
15. N,N,N0 ,N0 -Tetramethylethylenediamine (TEMED) (see Note
8).
16. Absolute ethanol.
17. 10% (v/v) Ethanol: Dilute 10 mL of absolute ethanol in 90 mL
of ddH2O. Mix thoroughly. Store at room temperature for up
to 1 year.
18. 70% (v/v) Ethanol: Add 30 mL of ddH2O to 70 mL of
absolute ethanol. Mix thoroughly. Store at room temperature
for up to 1 year.
19. 1 mg/mL collagen type I standard: 1 mg collagen type I, 1 mL

0.5 M acetic acid. Make further dilution in 0.5 M acetic acid, if
needed.
20. Hamilton syringe.
21. SDS PAGE gel making and electrophoresis apparatus.
22. SilverQuest™ Silver Staining Kit (see Note 9).
23. Fixing solution: Mix 40 mL absolute ethanol, 10 mL glacial
acetic acid and adjust volume to 100 mL ddH2O. Mix thor-
oughly. Prepare the fixing solution immediately before using.
24. Sensitizing solution: Mix 30 mL absolute ethanol, 10 mL
sensitizer and adjust volume to 100 mL with ddH2O. Mix
thoroughly. Prepare the sensitizing solution immediately
before using.
25. Staining solution: Mix 1 mL stainer and 99 mL ddH2O. Mix
thoroughly. Prepare the staining solution immediately before
using.
26. Developing solution: Mix 10 mL developer, 1 drop developer
enhancer and adjust volume to 100 mL with ddH2O. Mix
thoroughly. Prepare the developing solution immediately
before using.
27. 30% (v/v) Ethanol: Dilute 30 mL of absolute ethanol in 70 mL
of ddH2O and mix thoroughly. Store at room temperature for
up to 1 year.
2.3 Metabolic 1. HBSS.

Activity Assessment 2. AlamarBlue® solution.
3. 96-Well plate.
4. Microplate reader.
2.4 Cell Viability 1. DMSO.

Assessment 2. HBSS.
3. 4 mM Calcein AM in DMSO.
4. 2 mM Ethidium homodimer-1 in DMSO.
5. Inverted fluorescence microscope.
2.5 Cell Proliferation 1. HBSS.

Assessment 2. DNase-free water.
3. Quant-iT™ PicoGreen® dsDNA assay kit.
4. 96-Well plate.
5. Microplate reader.
2.6 Immuno- 1. Phosphate-buffered saline (1 PBS): 137 mM sodium chloride

fluorescent (NaCl), 2.7 mM potassium chloride (KCl), 10 mM phosphate
Characterization of buffer. Add 8 g of NaCl, 0.2 g of KCl, 1.44 g of Na2HPO4, and
ECM 0.24 g of KH2PO4 to 800 mL of ddH2O. Adjust the pH to 7.4
with HCl before adding ddH2O to 1 L.
2. Fixing solution: 4% (w/v) paraformaldehyde (PFA) in 1 PBS.
Weigh 0.4 g of PFA 95% powder and dissolve it by heating and
stirring in a glass beaker containing 10 mL of 1 PBS (see Note
10). Allow dissolution for 1 h (see Note 11). Once the PFA has
been completely dissolved, transfer the resulting fixation solu-
tion to a 50 mL centrifuge tube. Store at 4 C. Filter the
solution through a 0.2 μm sterile syringe filter before use.
3. Blocking solution: 3% bovine serum albumin (BSA) in 1 PBS.
Weigh 0.3 g of BSA and dissolve it in 10 mL of 1 PBS in a
15 mL centrifuge tube by vortexing. Keep it at 4 C for short-
term or 20 C for long-term storage.
4. Primary antibodies: rabbit polyclonal anti-collagen type I (Bos-
ter, UK; PA2140-2), rabbit polyclonal anti-collagen type IV
(Abcam, UK; ab6586), rabbit polyclonal anti-collagen type V
(Abcam, UK; ab7046), rabbit polyclonal anti-collagen type VI
(Abcam, UK; ab6588), rabbit polyclonal anti-fibronectin
(Abcam, UK; ab2413) (see Note 12). Prepare the dilutions of
the primary antibodies in 1 PBS as indicated: rabbit poly-
clonal anti-collagen type I (1:200), rabbit polyclonal anti-
collagen type IV (1:200), rabbit polyclonal anti-collagen type
V (1:200), rabbit polyclonal anti-collagen type VI (1:200),
rabbit polyclonal anti-fibronectin (1:200).
5. Secondary antibody: Goat anti-rabbit conjugated to Alexa-
Fluor®488 (Invitrogen, Thermo Fisher Scientific, Ireland;
A-32731) (see Note 13). Prepare the dilution of the secondary
antibody in 1 PBS as indicated: goat anti-rabbit Alexa-
Fluor®488 (1:400) for collagen types I, IV, V, VI and fibro-
nectin immunofluorescent staining. Protect it from exposure to
light (see Note 14).
6. 20 mM Hoechst 33342 Solution.
7. Olympus IX-81 inverted fluorescence microscope, equipped
with 10 objective, filters suitable for Hoechst 33342
(Ex/Em: 358/461 nm), AlexaFluor®488 (Ex/Em:
495/519 nm), and a connected camera.
3 Methods
For the preparation of corneal stromal-like assemblies rich in ECM,

routine cell culture practices should be applied.
3.1 Preparation of 1. Fill the appropriated number of 75 cm2 poly(L-lysine)-coated

Corneal Stromal-Like tissue culture flasks (T75) with 9 mL of maintaining growth
Assemblies Rich in medium. Equilibrate the flasks for 30 min in a cell culture
ECM incubator humidified at 37 C at 5% CO2 level.
2. Thaw a vial of cryopreserved HCFs, low passage preferred (see
Note 15) in a thermostatic bath at 37 C and seed the
pre-equilibrated tissue culture flasks at 5000 cells/cm2. Incu-
bate the cells overnight at 37 C and 5% CO2 under humidified
atmosphere and replace the medium with 9 mL of fresh main-
taining growth medium per T75 flask. Replace the maintaining
growth medium every 2–3 days.
3. Once the cell cultures have reached approximately 85% con-
fluency, discard the medium and wash the cell layers twice with
5 mL of HBSS per T75 flask. Discard the HBSS and add 1 mL
of trypsin-EDTA solution to each flask and incubate for 5 min
at 37 C. Assess cell detachment by microscopical examination,
and once all cells are detached from the flasks, add 2 mL of
maintaining growth medium, collect the cell suspension, and
centrifuge it for 5 min at 360 g. Discard the supernatant and
resuspend the cell pellet in 1 mL of growth medium per each
cultured T75. Determine the cell concentration with the aid of
a hematocytometer.
4. To plate the desired number of wells in 24 or 48-well tissue
culture treated plates, adjust the cell concentration in order to
seed 25,000 cells/cm2 and ensure that each well has 1 mL of
growth medium for 24-well plates or 0.5 mL of growth
medium for 48-well plates. Incubate the plates at 37 C and
5% CO2. In order to appreciate the effects of MMC on ECM
deposition, at least three wells are used per analysis or per
immunofluorescence marker to be analyzed, and three wells
without macromolecular crowder are used as control group.
Plate the necessary number of wells per immunofluorescence
assay in duplicate in order to have non-primary antibody con-
trols in triplicate per condition.
5. After 24 h of culture, examine the cells by microscopy and
ensure that they are healthy and properly attached to the cell
culture substrate before replacing the medium.
6. Replace the medium of the wells to be treated with MMC
medium with 1 mL (for 24-well plate) or 0.5 mL (for 48-well
plate) of the growth medium containing ascorbic acid and
carrageenan or only ascorbic acid for control wells. Incubate
the plates at 37 C and 5% CO2.
7. According to determined time points (in this study day 4, 7,
and 10), the plates can be examined by microscopy (Fig. 1) and
processed by SDS-PAGE, metabolic activity assessment, viabil-
ity assessment, proliferation assessment, and immunofluores-
cent staining.
Fig. 1 Morphology of HCFs in the absence and presence of increasing concentrations of macromolecular
crowder carrageenan (CR). Phase contrast microscopy revealed that HCFs maintained their spindle-shaped
morphology at each time point (day 4, 7, and 10) independently of the presence or absence of CR. However,
higher than 100 μg/mL CR concentrations caused cell losses
3.2 SDS-PAGE 1. At the end of each time point, aspirate the medium from the
wells of 24-well plate.
3.2.1 Pepsin Treatment
and Neutralization of 2. Wash the cell layer portion with HBSS.
Samples 3. Add 100 μg/mL pepsin solution to the cell layer (150 μL/well
of 24-well plate).
4. Incubate the plate at 37 C for 2 h with continuous shaking at
200 rpm using a Thermo Scientific, MaxQ 4000 Benchtop
Orbital Shaker, or equivalent.
5. After 2 h of incubation, scrape off the cell layer portion using
1 mL pipette tips and transfer to pre-labeled tubes. Cell layer
sample from 3 wells can be pooled in a 1.5 mL
microcentrifuge tube.
6. Add 22.5 μL of phenol red solution to 450 μL of cell layer
sample. The samples will turn into yellow color. Add 22.5 μL of
1 N NaOH to 450 μL of cell layer sample. Repeat this step until
the samples will turn into pink color (see Note 16).
7. Vortex briefly.
8. Store at 4 C for short-term storage or at 20 C for long-term
storage.
3.2.2 Preparation and 1. Prepare 250 μg/mL concentration collagen type I in 0.5 M
Running of SDS-PAGE acetic acid. Prepare the standard by adding 42 μL of ddH2O,
12 μL of 5 sample buffer, 2 μL of 1 N NaOH, and 4 μL of
standard (250 μg/mL) into a 1.5 mL microcentrifuge tube to
get 15:1 dilution.
2. Prepare cell layer samples by adding 24 μL of ddH2O, 12 μL of
5 sample buffers, and 24 μL of each cell layer sample into
Table 1
5% Separation Gel (1 mm thickness) for collagen for mini gel
Component 1 Gel (μL) 2 Gels (μL)

30% Acrylamide/Bis (37.5:1) 830 1660
1.875 M Tris–HCl pH 8.8 1000 2000
10% SDS 50 100
ddH2O 3073 6146
APS (100 mg/mL) 42 84
TEMED 5 10
Total 5000 10,000
Table 2
3% Stacking Gel (1 mm thickness) for collagen for mini gel
Component 1 Gel (μL) 2 Gels (μL)

30% Acrylamide/Bis (37.5:1) 200 400
1.875 M Tris–HCl pH 8.8 200 400
10% SDS 33 66
ddH2O 1547 5094
APS (100 mg/mL) 17 34
TEMED 3 6
Total 2000 10,000
1.5 mL microcentrifuge tubes to get 2.5:1 dilution (see

Note 17).
3. Vortex the samples and centrifuge them briefly. Store them at
4 C.
4. Prior to running SDS-PAGE, denature the samples and stan-
dard by heating at 95 C for 5 min.
5. Vortex and then centrifuge the samples briefly.
6. Take out of 20 C an aliquot of APS to thaw (put it back at
20 C at the end).
7. Clean glass plates with 70% ethanol and wipe dry with micro-
scope tissue papers.
8. Set the gel making apparatus ensuring that the glass plates fit
snugly to the platform (mini gel: 1 mm space using appropriate
spacers) (see Note 18).
9. Add the gel ingredients to make the 5% resolving gel according
to the Table 1. This can be done in a 15 mL centrifuge tube.
10. Make sure to add the APS and TEMED last, right before the
gels are to be poured.
11. Using a pipette, pour the prepared mixture carefully into the
space between the two glass plates to reach about 1 cm (mini
gel) from the bottom of the wells etched out by the comb
(keep the excess solution to check how quickly the gels will be
polymerized).
12. Overlay the gel with 10% ethanol to cut off oxygen in contact
with the gels.
13. Leave it aside for approximately 30 min (check with the excess
solution remained).
14. During the setting period, prepare the 3% stacking gel accord-
ing to Table 2 (Do not add the APS and TEMED until the gel
is ready for pouring).
15. A line at the ethanol-gel interface that initially had disappeared
will reappear after the 30 min period indicating that polymeri-
zation is complete.
16. Carefully aspirate the ethanol out of the glass plates using a
syringe and imbibe any traces using filter paper.
17. Now add the APS and TEMED to the stacking gel and care-
fully pour it on top of the polymerized resolving gel. Immedi-
ately insert the comb taking care to avoid trapping any air
bubbles (keep the excess solution to check how quickly the
gels will be polymerized).
18. Allow it to set for 10–15 min and, in the meantime, denature
samples and standard at 95 C as described above.
19. After the gels have been set (10–15 min, check it with the
excess solution), remove slowly the combs.
20. Assemble the electrophoresis apparatus; for small gel appara-
tus, fit the gel plates on the electrode bar and fit the set into the
inner chamber and clamp them.
21. Fill the upper/inner chamber with 1 running buffer.
22. Wash the wells by squirting buffer into the wells with a hypo-
dermal needle syringe to remove all air bubbles.
23. Load the standard and samples using a 50 μL Hamilton
syringe. Wash the syringe in between using the running buffer
in the chamber (at least 5-times). Load 10 μL per well of
15-well. Load 15 μL of 1 sample buffer per empty well of
15-well.
24. Put the upper chamber on the main chamber, close the lid, and
run the gel(s).
25. For the mini gel ! run at constant voltage: 50 V until the front
reaches the end of the stacking gel (30–40 min), then 120 V
until the front reaches the end of the separating gel (1 h).
26. Remove the glass using a Hoefer wonder wedge, cut the lower
right-hand corner, and release the gel slowly into ddH2O.
27. Proceed Silver staining.
28. Add 50 mL fixing solution onto the gel. Incubate the gel for
20 min at room temperature with slight shaking. After incuba-
tion, remove the fixing solution.
29. Wash the gel with 50 mL of 30% (v/v) ethanol for 10 min at
room temperature with slight shaking. After washing, remove
the solution.
30. Add 50 mL sensitizer solution onto the gel. Incubate the gel
for 10 min at room temperature with slight shaking. After
incubation, remove the sensitizer solution.
31. Wash the gel with 50 mL of 30% (v/v) ethanol for 10 min at
room temperature with slight shaking. After washing, remove
the solution.
32. Wash the gel with 50 mL of ddH2O for 10 min at room
temperature with slight shaking. After washing, remove the
solution.
33. Add 50 mL staining solution onto the gel. Incubate the gel for
15 min at room temperature with slight shaking. After incuba-
tion, remove the staining solution.
temperature with slight shaking. After washing, remove the
solution.
35. Add 50 mL developing solution onto the gel. Incubate the gel
for 4–8 min (see Note 19) at room temperature with slight
shaking.
36. Add 5 mL stopper solution directly to developing solution.
Incubate the gel for 10 min at room temperature with slight
shaking. After incubation, remove the developing solution.
temperature with slight shaking.
3.2.3 Protein Band 1. Scan the SDS-PAGE gels on Scanner with Active Transparency
Quantification Using Adapter (Fig. 2a).
ImageJ (Densitometric 2. Measure the band density of α1(I) and α2(I) bands with Ima-
Analysis) geJ software.
3. Open ImageJ.
4. Go to File ! Open ! (your image).
5. If the image looks too dark or too light go to
Image ! Adjust ! Brightness/contrast.
6. Save the image with an updated name.
7. Go to Analyze ! Set measurements ! Tick the following
boxes.
Fig. 2 Collagen I deposition at the cell layers in the absence and presence of increasing concentrations of
macromolecular crowder carrageenan (CR). SDS-PAGE (a) and complementary densitometric analysis (b)
revealed that in the presence of 50 μg/mL and 100 μg/mL CR, the highest amount of collagen I was deposited.
In addition, 50 μg/mL CR deposited the highest amount of collagen I at day 10. Data shown are mean
3. ∗P < 0.05, ∗∗P < 0.0001
8. Area, Mean Gray Value, Standard Deviation.

9. Select the rectangle tool, and draw a box around the lane.
10. Go to Analyze ! Measure.
11. A new box with all the results will appear.
12. Repeat steps 8 and 9 for all the bands of your interest.
13. Be consistent with the area of rectangle box.
14. Copy all results and paste into Microsoft Excel sheet.
15. Add the mean band intensity of α1(I) with α2(I) band.
16. Normalize the mean value of collagen type I standard. (For

example, if the mean band density is 500 for 250 μg/mL of
collagen type I standard, multiply all the mean value of sample
with 2).
17. Plot the bands intensity as μg/mL of sample and standard
bands (Fig. 2b).
3.3 Cellular 1. Prepare a 10% alamarBlue® solution in HBSS.

Metabolic Activity 2. Remove culture medium from the cells that are in 24-well plate
and wash with HBSS.
3. Add 500 μL of the diluted alamarBlue® solution to the cells
and a negative control of alamarBlue® at 10% alone.
4. To obtain the background absorbance, add HBSS to empty
wells.
5. Incubate for 3 h at 37 C, 5% CO2.
6. Transfer 100 μL of the alamarBlue® solution and of the nega-
tive control and background to a clear 96-well plate.
7. Measure the absorbance at 550 nm and at 595 nm using a
microplate reader.
8. Subtract the values of HBSS to the values of alamarBlue® alone
from both absorbances to obtain the absorbance of alamar-
Blue®. For 550 nm this value is called absorbance of the oxi-
dized form at lower wavelength (AOLW) and for 595 nm it is
called absorbance of the oxidized form at higher wavelength
(AOHW).
9. Calculate the correlation factor: Ro ¼ AOLW/AOHW.
10. To calculate the percentage of alamarBlue® reduced (AR) by
the cells, use the following: AR ¼ ALW (AHW Ro) 100
(Fig. 3a).
3.4 Cellular Viability 1. Prepare staining solution by diluting calcein AM to 4 μM and

ethidium homodimer-1 to 2 μM in HBSS.
2. To prepare a negative control, sample can be immersed in
DMSO to kill all cells before staining.
3. Remove culture medium from the cells and wash cells
with HBSS.
4. Add staining solution to cells (enough volume to cover
completely the sample).
5. Incubate at 37 C, 5% CO2 for 30 min.
6. Image under inverted fluorescence microscope: For Calcein
AM use FITC filter, for Ethidium homodimer-1: use Texas
Red filter (Fig. 3b).
Fig. 3 Cell metabolic activity (a), viability (b), and proliferation (c) of HCFs in the absence and presence of
50 μg/mL CR (concentration that induced the highest collagen I deposition at day 10, as per SDS-PAGE and
complementary densitometric analysis). Macromolecular crowding (50 μg/mL CR) did not affect cellular
metabolic activity (a), viability (b), and DNA content (c) at a given time point as revealed by alamarBlue assay,
Live/Dead assay, and Quant-iT PicoGreen dsDNA assay, respectively. Data shown are mean SD. The
N values for each group are 3
3.5 Cell Proliferation 1. Remove the media from the cells that are in 24-well plate and
gently rinse the cells with HBSS.
2. Add 250 μL of DNase-free water.
3. Freeze-thaw cells three times (freeze at 80 C for 15 min
minimum and thaw at room temperature until it is completely
defrosted).
4. Prepare a standard curve in a 96-well plate in accordance with
Table 3 using 1 TE buffer (initial solution at 20, dilute in
DNase-free water).
5. Make up 2 μg/mL DNA stock solution from 100 μg/mL
DNA standard and 50 ng/mL DNA stock solution from
2 μg/mL stock solution.
6. Transfer 100 μL of each sample in the 96-well plate.
7. Make up diluted PicoGreen® solution by adding 5.376 mL 1
TE and 27 μL concentrated PicoGreen® (enough for the stan-
dard curve in triplicate and 24 samples).
8. Add 100 μL of diluted PicoGreen® to each well.
9. Incubate at room temperature for 2–5 min in the dark.
10. Read the plate for fluorescence (excitation: 480 nm, emission:
520 nm).
11. Plot a graph concentration vs. the fluorescence values. Deter-
mine the concentration of DNA as a function of the standard
curve (Fig. 3c).
Table 3
Standard curve for PicoGreen
Final DNA
concentration 1 TE Volume of 2 μg/mL stock DNA Volume of 50 ng/mL stock DNA
(ng/mL) Volume (μL) solution (μL) solution (μL)
1000 0 100
500 50 50
100 90 10
50 95 5
25 0 100
10 60 40
5 80 20
0 100 0
3.6 Immuno- All the following steps must be performed at room temperature
fluorescent (unless otherwise stated).
Characterization of the
1. At the end of each time point, aspirate the cell culture media of
ECM the corresponding 48-well plate and wash each well three times
during 5 min with 500 μL of 1 sterile HBSS. Fix the samples
with 150 μL/well of the filtered fixation solution (4% PFA)
pre-cooled at 4 C for 15 min (see Note 20).
2. Remove the fixation solution and wash the wells three times
during 5 min with 250 μL of 1 PBS. Samples can be kept at
4 C in 1 PBS if immunofluorescent staining is not to be
performed right away.
3. Add 150 μL per well of BSA blocking solution for 30 min to
block unspecific binding sites for the primary antibody.
4. After the aforementioned 30 min, drain the blocking solution
and incubate the samples with 80 μL per well of the
corresponding primary antibody solution for overnight at
4 C. For negative control wells, add 80 μL of 1 PBS per
well (see Note 21).
5. Remove the primary antibody solution and wash the wells three
times during 5 min with 250 μL of 1 PBS (see Note 22).
Incubate with 80 μL per well of the secondary antibody solu-
tion for 1 h. In this case, all the samples including the negative
controls must be incubated with the secondary antibody solu-
tion (see Note 23). Protect the samples from light from
here on.
6. Remove the secondary antibody solution and wash three times

during 5 min with 250 μL of 1 PBS.
7. Remove the PBS and incubate the samples for 5 min with
80 μL per well of diluted Hoechst 33342 solution in order to
stain nuclei.
8. Remove the Hoechst 33342 solution and wash three times
during 5 min with 250 μL of 1 PBS.
9. Add 80 μL per well of 1 PBS in each well.
10. Analyze the samples by inverted fluorescence microscope
equipped with 10 objective, filters suitable for Hoechst
33342 (Ex/Em: 358/461 nm), AlexaFluor®488 (Ex/Em:
495/519 nm), and a connected camera.
3.6.1 Imaging of Perform the steps relative to the image acquisition process in dark,
Immunofluorescent in order to prevent the loss of fluorescence intensity from the
Staining fluorophores as a consequence of light excitation.
1. Once the software has been launched, position the plate on the
sample stage of the microscope and visualize a live preview from
the camera input; be sure that the light path selector of the
microscope is set on “camera” to view the image on screen.
2. It is pivotal to keep the same exposure time for all the samples
examined for the same antigen (see Note 24). Adjust the
exposure time for Hoechst 33342, and carefully adjust the
focus on the cell layer. Select the correct filter to excite the
fluorophore bound to the secondary antibody of interest, reg-
ulate the exposure time in order to avoid saturation, and take
note of the longest exposure time void of saturation. Repeat
these actions for all the samples, excluding the negative con-
trols, then set the lowest obtained exposure time value to
examine all of the samples and controls.
3. Position the plate on the sample stage exposing the first sample
to the light path, select the violet filter, and acquire a 10
magnified picture of the nuclei stained with Hoechst 33342.
Without moving the sample, turn the filter block turret to the
appropriated filter for the fluorophore bound to the secondary
antibody of interest, regulate the focus, and acquire the image.
Repeat this step five times in randomly selected fields of the well
for each replicate of each condition.
4. Using the emission filter for the secondary antibody of interest,
acquire five pictures from five randomly selected fields of the
non-primary antibody negative controls (see Note 25). Save all
the acquired pictures in TIFF format before proceeding with
the analysis of fluorescence intensity, which can be performed
with the ImageJ software.
5. To create a representative picture of the assay, open the desired

image in ImageJ software together with the corresponding
Hoechst 33342 field. For both of the pictures, open the
“Image” menu, select “Type” and click on “32-bit.” Once all
the images that need to be merged in a single picture have been
turned to 32-bit format, open the “Image” menu, select “Col-
our,” then “Merge channels”: match each picture with the
desired channel, tick the “create composite” box and click
“Ok” to obtain the composite picture (see Note 26). Once
the picture has been saved in the preferred format, it is possible
to add a scale bar using the dedicated ImageJ tool (see Note
27) (Fig. 4).
Fig. 4 Immunofluorescent labeling of different extracellular matrix proteins in HCFs for each time point in the
absence and presence of macromolecular crowding (50 μg/mL CR; concentration that induced the highest
collagen I deposition at day 10, as per SDS-PAGE and complementary densitometric analysis). Macromolecu-
lar crowding enhanced collagenous proteins (I, IV, V, VI), which are the main components of corneal stroma, in
the cell layer of HCFs at each time point
4 Notes
1. Fibroblast medium kit is a kit purchased from Innoprot (Spain)

that includes 500 mL fibroblast basal medium, 10 mL fetal
bovine serum (FBS), 5 mL fibroblast growth supplement
(FGS), and 5 mL penicillin/streptomycin solution.
2. Poly(L-lysine)-coated 175 cm2 flask: 3 mL 0.1 mg/mL poly(L-
lysine) solution is transferred to 175 cm2 flask. The flask is
gently rocked to evenly coat the surface. After 5 min, the excess
solution is removed and the surface is thoroughly rinsed with
sterile ddH2O and allowed to dry for several hours.
3. The efficacy of UV-C disinfection depends on many para-
meters, including the intensity and wavelength of the UV
radiation, the time of exposure, the distance from the source
of irradiation, the presence of particles that can protect the
microorganisms from UV, and a microorganism’s ability to
withstand UV during its exposure. We currently disinfect the
carrageenan with success inside of an open 1.5 mL microcen-
trifuge tube in vertical position with the UV lamps typically
included in the biological safety cabinets used for cell culture,
but more dedicated equipment can also be used with the same
efficacy if similar conditions are met.
4. Once the carrageenan is suspended into the medium, if the
tube is vortexed and observed through a light source, particles
in suspension can be appreciated. Once the carrageenan is
completely dissolved into the medium, no particles can be
appreciated following the same observation procedure. It is
very important to ensure the completely dissolution of the
carrageenan into the medium as non-dissolved particles can
deposit on the bottom of the plates and negatively affect the
cell viability.
5. Dissolving sodium hydroxide in water creates an exothermic
reaction, which produces large amounts of heat. Concentrated
sodium hydroxide may result in boiling; thus, the flask should
not be touched, and personal protective equipment (PPE) is
required.
6. SDS is a respiratory, skin, and eye irritant. It affects fertility. It is
harmful if swallowed, inhaled, or absorbed through the skin.
The use of PPE is recommended. Weigh SDS under a fume
hood to avoid eye and skin contact and inhalation.
7. APS is highly toxic. It causes respiratory, skin, and eye irrita-
tion. It is harmful if swallowed. The use of PPE is recom-
mended. Weigh APS under a fume hood to avoid eye and
skin contact and inhalation.
8. TEMED is highly flammable and it is harmful if swallowed or

inhaled. The use of PPE is recommended. Work with TEMED
under a fume hood.
9. SilverQuest™ Silver Staining Kit includes Sensitizer, Stainer,
Developer, Developer Enhancer, and Stopper solution.
10. Avoid temperatures over 60 C (optimal are between
55–57 C), which result in methanol formation and the reduc-
tion of the effective concentration of PFA. This can damage the
cytoskeleton of the cells and increase the auto-fluorescence of
the samples.
11. For safety reasons, make the fixation solution under a chemical
hood and cover the glass beaker, with aluminum foil for exam-
ple, to prevent the release of PFA toxic vapors and the evapo-
ration of the PFA. Likewise, all the steps that involve the use of
the fixation solution must be executed under a chemical hood
due to the PFA toxicity.
12. To perform the localization of several molecules in the same
sample by indirect immunofluorescent staining, it is required
for the primary antibodies against the antigens of interest to be
raised in different species in order to be able to be recognized
by different specific secondary antibodies (e.g., the primary
antibody anti-collagen type I has been raised in mouse and
can be combined with any of the other primary antibodies
raised in rabbit for their use in multicolor immunostaining).
13. For the selection of the secondary antibodies, several guide-
lines must be followed. First, the secondary antibody must be
specific for the target species corresponding to the primary
antibody (e.g., goat anti-mouse secondary antibodies for pri-
mary antibodies derived from mice and goat anti-rabbit sec-
ondary antibodies for primary antibodies derived from rabbit).
Moreover, to avoid cross-reactivity with the non-intended pri-
mary antibody targets in multicolor immunostaining and
increase specificity, secondary antibodies cross-adsorbed
against IgGs from the nontarget species and sera from different
species can be used (e.g., for the present study, we used goat
anti-mouse secondary antibody cross-adsorbed against human
serum and rabbit and goat IgGs, as well as goat anti-rabbit
secondary antibody cross-adsorbed against mouse and goat
IgGs). The selection of the fluorochromes conjugated with
the secondary antibodies and the appropriated filter set it is
crucial to have the minimal signal overlapping and maximal
specificity in case of detecting several molecules at the same
time in multicolor immunofluorescent stainings. For this, it is
very important to excite each fluorochrome with wavelengths
as close as possible to its maximal excitation, avoiding at the
same time to excite the other fluorochromes present in the
sample. Also, for collecting the fluorescence of each fluoro-

chrome specifically, the emission filter should match as close as
possible the maximal emission wavelength of the
corresponding fluorochrome, avoiding at the same time the
emission of the other fluorochromes present in the sample.
14. Protect the samples from the light to avoid the loss of fluores-
cent signal due to photobleaching. The photobleaching effect
(also termed as fading) causes a permanent photochemical
alteration, by cleaving covalent bonds in the fluorochromes.
These irreversible modifications make the fluorochrome unable
to emit fluorescence and thereby decrease the signal in the
samples.
15. The HCFs that can be acquired from Innoprot (Sapin) are
generally in early passages (p0 - p1). To thaw the cells, wear
protective gloves and face shield and take out the tube contain-
ing the frozen cells from liquid nitrogen cylinder. Thaw the
contents of cryotube by rubbing in palm or water bath at
37 C. Transfer the contents of the tube in a 15 mL conical
tube containing 10 mL of maintaining growth medium inside
the laminar air flow hood. Centrifuge the tube at 360 g for
5 min. Remove the supernatant and discard it. Add 1 mL of
pre-warmed (37 C) maintaining growth medium into the
same falcon tube and mix properly with gentle aspiration to
distribute the cells homogeneously. Count the cells using
hemocytometer. Transfer appropriate amount of the cell sus-
pension (5000 cells/cm2) to a poly(L-lysine)-coated T75 flask
containing 9 mL (20 mL for T175) of pre-warmed maintaining
growth medium. Label the flasks with name, date, cell type,
and passage no. The subculture routine included washing with
5 mL (10 mL for T175) of 1 HBSS, incubating with 1 mL
(2 mL for T175) of trypsin-EDTA solution for 5 min at 37 C,
cell recovery with 3 mL (6 mL for T175) of maintaining
growth medium per each T75 flask, centrifugation for 5 min
at 360 g, resuspension, counting, and plating at 5000 cells/
cm2 to new poly(L-lysine)-coated flask. Normally, the cells
reach confluency after a week of culture, and medium is
changed thrice per week, using 9 mL (20 mL for T175) per
T75. For cell cryopreservation, a cell density of 500,000 cells/
mL of 10% DMSO in growth medium is used. In this study,
passage 4 cells were used for MMC experiments.
16. The amount of NaOH solution used for neutralization is
sample-dependent (e.g., dependent upon the acidity of the
solution) and therefore may require optimization.
17. To make densitometric analyze, dilution factor of collagen type
I standard will be 15 and dilution factor of each sample will be
2.5.
18. Check for any leaks by pouring ddH2O into the gel-making
apparatus and then removing the ddH2O with filter paper
before making the gel.
19. After adding the developing solution onto the gel, closely
observe the bands as they form within 0 to 8 min. Once all
bands become completely visible, immediately add the stopper
solution to prevent a dark background.
20. Be extremely careful during fixation and washing steps to
prevent the detachment of the cell layer.
21. Alternatively, incubation with the primary antibody solutions
can be performed 90 min at room temperature.
22. When working with a high number of experimental groups or
samples, always proceed with the washes in a fractional way in
order to avoid drying the samples. Drying may cause a back-
ground increase due to unspecific binding of the primary anti-
body to the sample.
23. In order to demonstrate the nonspecific binding of the second-
ary antibodies to the sample, the non-primary antibody control
wells should be incubated with the corresponding secondary
antibody solution. For multicolor immunostaining, these wells
should be incubated with a solution containing the mixed
secondary antibodies as applied to the sample wells.
24. The selection of an optimal exposure time suitable for all the
examined samples, defined as the maximal exposure time at
which the sample with the highest fluorescence intensity does
not show signs of saturation, will allow for the comparison of
the collected data between the different samples. The acquisi-
tion of an excessively saturated image would result in an under-
estimation of fluorescence intensity.
25. Background fluorescence can result from several factors, such
as nonspecific binding of the secondary antibodies, or auto-
fluorescence from the culture plate, the cultured cells, and from
collagen itself. The quantification of non-primary antibody
controls fluorescence is therefore performed in order to esti-
mate the amount of this nonspecific fluorescence, which will
eventually be subtracted from the mean gray value measure-
ments of the corresponding samples.
26. It is possible to merge more than two pictures, or channels, in
this step, in order to appreciate the relative localization of the
examined antigens. To reduce the background interference
that can derive from nonspecific binding of the antibodies as
well as from autofluorescence of several biological molecules,
or even from the plastic substrate, the “Subtract background”
ImageJ tool can be useful. To use it, right after the images have
been turned to 32-bit format, open the “Process” menu, click
on “Subtract background,” select the same rolling ball radius

for every picture in the assay, and confirm with “OK,” then
proceed with the merging of the channels. Choose a larger
rolling ball radius to maximize the preservation of positive
pixels intensity, or a smaller one to maximize background
subtraction.
27. There are several options to add a scale bar with ImageJ. If the
pixel/length ratio is known, it is possible to open the “Ana-
lyze” menu, select “Set scale,” type “1” in the “Pixels” box and
type the corresponding distance in the “Known distance” box,
then specify the unit of length and confirm with “OK.” If the
pixel/length ratio is not known, it is possible to use the
“Straight” tool from ImageJ toolbar in order to trace a seg-
mented line covering a known distance in the picture, then
open the “Set scale” window and fill the “Known distance” and
“unit of length” boxes, then confirm with “OK.” A third
option provides that a second picture, with the same pixel/
length ratio of the former and already containing its scale bar is
open and selected in ImageJ: in such case, it is only needed to
tick the “Global” box in the “Set scale” window before clicking
“OK” to apply the same scale to all the open pictures. To finally
add the scale bar to the desired picture, all the listed options
provide that the “Analyze” menu is open, in order to select
“Tools > Scale bar,” personalize the scale bar and confirm with
“OK.”
Acknowledgements
This work has been supported from: Science Foundation Ireland,

Career Development Award Programme (grant agreement num-
ber: 15/CDA/3629) and Science Foundation Ireland and the
European Regional Development Fund (grant agreement number:
13/RC/2073). Mehmet Gürdal was supported by The Scientific
€
and Technological Research Council of Turkey (TUBİTAK), Sci-
ence Fellowships and Grant Programmes Department (BİDEB),
Programme of 2214-A Ph.D. Research Scholarship for Abroad.
The authors have no competing interests.
References
1. Fini ME, Stramer BM (2005) How the cornea distribution of collagen lamellae in the human
heals: cornea-specific repair mechanisms affect- cornea. Cornea 17(5):537–543
ing surgical outcomes. Cornea 24(8 Suppl): 3. Fini ME (1999) Keratocyte and fibroblast phe-
S2–S11 notypes in the repairing cornea. Prog Retin Eye
2. Radner W, Zehetmayer M, Aufreiter R, Mal- Res 18(4):529–551
linger R (1998) Interlacing and cross-angle 4. Hay ED (1980) Development of the vertebrate
cornea. Int Rev Cytol 63:263–322
5. Kumar P, Pandit A, Zeugolis D (2016) Prog- development of supramolecular corneal

ress in corneal stromal repair: from tissue grafts stromal-like assemblies from corneal fibroblasts
and biomaterials to modular supramolecular in the presence of macromolecular crowders.
tissue-like assemblies. Adv Mater 28 Tissue Eng Part C Methods 21(7):660–670.
(27):5381–5399. https://doi.org/10.1002/ https://doi.org/10.1089/ten.TEC.2014.
adma.201503986 0387
6. Guillame-Gentil O, Semenov O, Roca AS, 12. Graceffa V, Zeugolis D (2019) Carrageenan
Groth T, Zahn R, Voros J, Zenobi-Wong M enhances chondrogenesis and osteogenesis in
(2010) Engineering the extracellular environ- human bone marrow stem cell culture. Eur Cell
ment: strategies for building 2d and 3d cellular Mater 37:310–332. https://doi.org/10.
structures. Adv Mater 22(48):5443–5462. 22203/eCM.v037a19
https://doi.org/10.1002/adma.201001747 13. Graceffa V, Zeugolis D (2019) Macromolecu-
7. Peck M, Dusserre N, McAllister TN, L’Heur- lar crowding as a means to assess the effective-
eux N (2011) Tissue engineering by self- ness of chondrogenic media. J Tissue Eng
assembly. Mater Today 14(5):218–224. Regen Med 13(2):217–231. https://doi.org/
https://doi.org/10.1016/S1369-7021(11) 10.1002/term.2783
70117-1 14. Cigognini D, Gaspar D, Kumar P, Satyam A,
8. Canty EG, Kadler KE (2005) Procollagen traf- Alagesan S, Sanz-Nogues C, Griffin M,
ficking, processing and fibrillogenesis. J Cell Sci O’Brien T, Pandit A, Zeugolis DI (2016) Mac-
118(Pt 7):1341–1353. https://doi.org/10. romolecular crowding meets oxygen tension in
1242/jcs.01731 human mesenchymal stem cell culture - a step
9. Sorushanova A, Delgado L, Wu Z, Shologu N, closer to physiologically relevant in vitro organ-
Kshirsagar A, Raghunath R, Mullen A, ogenesis. Sci Rep 6:30746. https://doi.org/
Bayon Y, Pandit A, Raghunath M, Zeugolis D 10.1038/srep30746
(2019) The collagen suprafamily: from biosyn- 15. Satyam A, Kumar P, Fan X, Gorelov A,
thesis to advanced biomaterial development. Rochev Y, Joshi L, Peinado H, Lyden D,
Adv Mater 31(1):e1801651. https://doi.org/ Thomas B, Rodriguez B, Raghunath M,
10.1002/adma.201801651 Pandit A, Zeugolis D (2014) Macromolecular
10. Kumar P, Satyam A, Fan X, Collin E, Rochev Y, crowding meets tissue engineering by self-
Rodriguez BJ, Gorelov A, Dillon S, Joshi L, assembly: a paradigm shift in regenerative med-
Raghunath M, Pandit A, Zeugolis DI (2015) icine. Adv Mater 26(19):3024–3034. https://
Macromolecularly crowded in vitro microen- doi.org/10.1002/adma.201304428
vironments accelerate the production of extra- 16. Gaspar D, Fuller K, Zeugolis D (2019) Poly-
cellular matrix-rich supramolecular assemblies. dispersity and negative charge are key modula-
Sci Rep 5:8729. https://doi.org/10.1038/ tors of extracellular matrix deposition under
srep08729 macromolecular crowding conditions. Acta
11. Kumar P, Satyam A, Fan X, Rochev Y, Rodri- Biomater 88:197–210. https://doi.org/10.
guez BJ, Gorelov A, Joshi L, Raghunath M, 1016/j.actbio.2019.02.050
Pandit A, Zeugolis DI (2015) Accelerated
Chapter 10
Preparation of Dried Amniotic Membrane for Corneal Repair

Andrew Hopkinson, Emily R. Britchford, and Laura E. Sidney
Abstract
Amniotic membrane transplantation is an established therapeutic and biological adjunct for several clinical
situations, including treatment of diabetic foot ulcers and ocular surface disease. However, poorly standar-
dized and validated clinical preparation and storage procedures can render the final product highly variable
and an unpredictable biomaterial. We have therefore developed a novel, standardized method for proces-
sing and dry-preserving amniotic membrane, minimizing biochemical, compositional, and structure dam-
age to produce a potentially superior membrane suitable for clinical use. The intellectual property associated
with this methodology was patented by the University of Nottingham and licensed to NuVision® Biothera-
pies which formed the basis of the Tereo® manufacturing process which is used to manufacture Omnigen®.
Key words Amniotic membrane, Standardization, Preparation, Transplantation, Cornea, Repair
1 Introduction
Amniotic membrane (AM) is the innermost layer of the placental

membrane that surrounds the fetus during pregnancy. At birth, the
placenta is normally discarded as waste; however, following elective
caesarean sections, the AM can be harvested and processed into
an allograft transplant material for the repair of many soft-tissue
damage situations, including the ocular surface. The clinical objec-
tives of AM transplantation (AMT), particularly in corneal regener-
ation, are to: (1) support and replace damaged tissue, (2) protect
the defect from further degeneration from external factors, and
(3) promote effective recellularization [1]. This is achieved through
a combination of beneficial biological properties, including
immune-privilege (minimizing the risk of an immune response)
[2], ability to modulate and inhibit inflammation [3, 4], angiogen-
esis [3, 5, 6], fibrosis [7], tissue adhesions [7, 8], and cancer
[9–13], while preserving and nurturing the host’s stem cell popu-
lation [14, 15]. Combined, these biological properties facilitate
recellularization to accelerate wound healing [28]. The physical
143
144 Andrew Hopkinson et al.
nature of AM, as a biological bandage, also significantly improves

pain [16–19], and delivers antimicrobial benefits [20–22].
Since its first clinical use as a skin transplant in 1910 [23], AM
has been adopted in numerous soft-tissue transplantation modal-
ities, including restoration of hearing [24], grafting of the abdomi-
nal cavity [25, 26], and replacement of the vaginal [27, 28], urethral
[29], oral, and pharyngeal mucous membranes [30, 31]. Moreover,
AMT has been widely adopted in the treatment of ocular surface
disease, as well as more recently for the treatment of chronic wounds
[32]. Application of AM in the treatment of corneal disease has been
reported to accelerate wound healing [33], especially in infected
corneas by reducing infection-causing microorganisms [34]. AMT
has been shown to be superior to antibiotic treatment alone
[35, 36]. Although the mechanism of action of AM is still not fully
understood, it is broadly attributed to bioactive substances such as ,
inflammatory-mediators, proteoglycans, such as hyaluronic acid,
and other classifications of beneficial proteins that enhance the
function of structural constituents, working synergistically as a
complex matrix to support local cell survival, adhesion, prolifera-
tion, and migration [37].
An effective method for processing and preserving AM for
clinical application must therefore aim to preserve the integrity of
the tissue without compromising the structural and biochemical
composition, and therefore therapeutic potential of the
AM. Freshly collected AM is rarely used directly for clinical applica-
tion due to an increased risk of disease transmission, posttransplant
complications [38], and a very short shelf life. The use of fresh AM
is also not permitted by many regulatory authorities. Therefore,
commercial clinical AM products are subject to various processing
and preservation steps [37]. The conventional methods for preserv-
ing AM are: (1) cryopreservation (extra-cold freeze-preservation at
around 80 C) [39], (2) freeze-drying (lyophilization of 80 C
frozen tissue via sublimation) [40], and (3) heat (up to 50 C) [41]
or air-dehydration. Despite the use of cryoprotectants such as
glycerol or dimethyl sulfoxide (DMSO), cryopreservation
(wet-preservation) procedures have been shown to damage the
AM [42], which results in a significant decrease in the levels of
many bioactive substances, including important wound healing
factors, such as epidermal growth factors (EGF) and transforming
growth factors (TGF) [6, 42–46]. Cryopreservation also creates
major cold-chain storage and shipping, shelf life, and quality assur-
ance challenges. Furthermore, cryopreserved AM must be thawed
and washed to remove cryoprotectants before clinical use, which
further depletes beneficial proteins from the damaged tissue.
Though offering practicality, storage, and logistic advantages,
heat-dried AM followed by radiation-sterilization has been shown
to also reduce growth factor content [40], increased fragility, and
compromised efficacy. More recently, the use of freeze-dried AM
Dried Amniotic Membrane Preparation 145
has been proposed in combination with sterilization, and in the

presence of novel complex saccharide lyoprotectants [47]. Unfortu-
nately, this involves aforementioned deleterious freezing of the
tissue before drying. Further challenges of dry-preserving AM is
friability of tissues in the dry state, and subsequently effective rehy-
dration of the material to restore normal physical, biochemical, and
functional characteristics.
This chapter describes an alternative process for AM preserva-
tion that overcomes the issues of conventional wet- and
dry-preservation methods, to produce a standardized and high-
quality AM substrate, which is robust and malleable in the dehy-
drated state, and rapidly and effectively rehydrates when applied to
the wound in vivo. AM is delicately dehydrated using a unique
freeze- and heat-free low temperature vacuum evaporation (Patent
WO2014195699A3, licensed to NuVision Biotherapies Ltd) pro-
cess which preserves the integrity and biochemistry of AM, compa-
rable to fresh AM, as a devitalized stable stromal matrix.
During pregnancy and whilst in vivo, AM is considered sterile;
however, potential microbial “bioburden” contamination can be
introduced from skin and vaginal flora during delivery of the pla-
centa [48]. Though elective caesarean section reduces the risk of
bioburden, the AM manufacturing process must involve measures
to effectively decontaminate potential bioburden and allow for
long-term preservation of AM. Due to potential degradation by
terminal (radiation) sterilization techniques [49], AM can be
prepared aseptically, involving antibiotics, in a Good
Manufacturing Practice (GMP) environment in order to ensure
natural bioburden is eliminated and no further contamination is
introduced during processing. In this protocol, decontamination is
achieved through a combination of: (1) an aseptic procedure, (2) a
sequential washing process involving a potent antibiotic cocktail
decontamination step, and (3) a drying procedure. Recent research
at the University of Nottingham has demonstrated that this com-
bined process can be considered a sterilization process [50, 51].
This chapter describes the research preparation of dried AM
using a standardized laboratory manufacturing process. The final
product is produced by isolating the amniotic layer from the pla-
centa, chorion, and spongy layer (SL) prior to washing and drying.
The salient features of this technique are:
l Human Tissue Authority (HTA) approved protocol for clinical
transplantation of AM on human ocular surface (see Note 1).
l Minimized inter- and intra-donor variation between AM
preparations.
l Unique spongy layer removal process (Patent GB1441939,
licensed to NuVision Biotherapies Ltd).
l A xeno-free preparation system suitable for clinical, in vivo, and
in vitro research.
l Unique dehydration process.

l A dried AM matrix as an ideal tissue engineering substrate for
direct ex vivo co-culturing of various cell types.
2 Materials
2.1 Reagents 1. 0.9% Sodium chloride (NaCl): 9 g NaCl in 1 L H2O.

2. Phosphate buffered saline (PBS): 72 mg potassium phosphate
monobasic (KH2PO4), 397.5 mg sodium phosphate dibasic
(Na2HPO4), and 4500 mg NaCl in 0.5 L H2O, pH 7.4.
3. Antibiotic/antifungal concentrate 5 mL: gentamicin sulfate,
imipenem, nystatin dihydrate, polymyxin B sulfate, vancomycin
hydrochloride (broad-spectrum).
2.2 Disposable 1. 175 cm2 (T175) straight neck vent flask.

Instruments 2. Scissors Supersnip B/B sterile.
3. Adson forceps (Bayonet) non-toothed 18 cm.
4. Bio assay dish standard.
5. Syringes 50 mL.
6. Serological pipette 25 mL.
7. Pipette aid.
8. Sartorius syringe filters 0.22 μm.
9. Blades No. 22 sterile.
10. Disposable scalpels No. 22 sterile.
11. Surgical spongy spears.
12. Bold line surgical pen.
13. Parafilm® M roll.
2.3 Equipment 1. Class II laminar air flow hood (biological safety cabinet).
2. Rocker.
3. Incubator.
4. Freeze Dryer.
2.4 Additional 1. Gloves.

Materials 2. Sharps bin.
3. Marker pen.
4. Cleaning tissues.
5. Clinical waste bags.
6. Black general waste bags.
7. Cleaning reagents (70% IMS).
3 Methods
The preparation of all materials and methods should be performed

in a class II laminar air flow hood (safety cabinet) following Good
Laboratory Practice (GLP), unless otherwise stated for clinical use
(see Note 2). Remove all packaging inside the safety cabinet to
maintain sterility and cleanliness.
3.1 Tissue Collection Prepare the tissue collection flask (TCF) prior to tissue collection.
Flask
1. Transfer the following materials to the safety cabinet:
(a) 1 T175 flask
(b) 1 1 L NaCl
(c) Marker pen
2. Clearly label the TCF using the marker pen.
3. Mark the 250 mL line and decant 250 mL NaCl into the flask.
4. Pre-weight the Raffinose powder (D-(+)-Raffinose pentahy-
drate 99 + %) in a T175 flask (14.86 g). Label “RAFFINOSE”
using the marker pen (see Note 3).
3.2 Donor 1. Identify and consent potential donors undergoing elective

Consenting caesarean sections (donor informed consent is mandatory
and Amniotic before processing) (see Notes 4 and 5).
Membrane Tissue 2. Prepare a collection kit by placing the following items into a
Procurement collection box:
(a) 1 TCF (prepared in Subheading 3.1)
(b) 1 Scissors
(c) 1 18 cm forceps
3. Collect the AM tissue immediately after birth from the delivery
suite sluice room.
4. Use the forceps and scissors to search along the edges of the
isolated sac for areas where the AM and chorion have separated.
Separate the AM using general force (by hand) to peel/pull
the chorion away from the AM. Carefully cut the AM around
the base of the cord to release the tissue from the placenta (see
Note 6).
5. Use the forceps to place the isolated AM tissue in the TCF
(containing sterile saline) ready for transportation.
6. Dispose of the placental body.
7. Discard the scissors and forceps in the sharps bin (within
sluice room).
8. Transport the TCF to the laboratory within 2 h of birth.
3.3 Preparation 1. Ensure appropriate Personal Protective Equipment (PPE) is

for Manufacturing worn before entering the laboratory (see Note 7).
2. Collect 1 antibiotics from the 80 C freezer before entering
the laboratory. Thaw the vial at room temperature.
3. Keep the TCF outside the safety cabinet until Subheading 3.5,
step 5.
4. Switch on equipment—rocker, incubator, and safety cabinet.
5. Clean the working area in the safety cabinet with IMS (see
Note 8).
6. Set aside IMS, tissues, sharps bin, and marker pen for later use.
3.4 Preparation 1. Transfer the following materials to the safety cabinet:

of Washing Solutions (a) 3 T175 flasks
(b) 1 1 L NaCl
2. Label the T175 flasks “WASH 1,” “WASH 2,” and “WASH 3”
using the marker pen.
3. Mark the 250 mL line on each flask and decant 250 mL NaCl
into each flask.
3.5 Amniotic 1. Transfer the following materials to the safety cabinet:

Membrane Washing (a) 1 Square dish pack (containing 4 square tray pairs)
and Incubation
(b) 1 18 cm forceps
2. Open the square dish pack and take one square tray base to use
as an “instruments” tray. Place forceps on the instruments tray.
3. Take another tray base to use as a “working” tray.
4. Set aside the remaining square dishes for later use.
5. Transfer the TCF to the safety cabinet.
6. Use the forceps to remove the AM from the TCF and spread it
over the working tray for inspection (see Note 9).
7. Use the forceps to place the AM in the WASH 1 flask (rest the
cap upward on the cabinet surface).
8. Remove the WASH 1 flask from the safety cabinet and place on
the rocker for 20 min, speed ¼ 50 RPM.
(a) 1 0.5 L PBS
(b) 1 RAFFINOSE flask
10. Mark the 250 mL line on the flask and decant 250 mL PBS into
the flask. Retain the PBS bottle for later use.
11. Remove the RAFFINOSE flask from the safety cabinet and
place on the rocker for at least 25 min.
12. When the rocking time for the WASH 1 flask has finished,
transfer it back to the safety cabinet.
13. Keep the RAFFINOSE flask on the rocker.

14. Use forceps to transfer the AM from the WASH 1 to the WASH
2 flask (rest the flask caps upward on the cabinet surface).
the rocker for 20 min.
16. Discard the WASH 1 flask in the clinical waste bag.
(a) 2 T175 flask
(b) 2 50 mL syringe
(c) 2 0.22 μm filters
(d) 1 25 mL Serological pipette
(e) 1 Pipette aid
18. Place all materials on the instruments tray.
19. Label the flasks with “INCUBATION” and “INCUBATION
WASH” using the marker pen.
20. Mark the 250 mL line on INCUBATION WASH flask.
22. Transfer the RAFFINOSE flask to the safety cabinet (make sure
the raffinose is fully dissolved).
23. Use the forceps to transfer the AM from the WASH 2 to the
WASH 3 flask (rest the caps upward on the cabinet surface).
the rocker for 20 min.
25. Discard the WASH 2 flask in the clinical waste bag.
26. Use the 50 mL syringes and 0.22 μm filters to filter the raffi-
nose solution from the RAFFINOSE flask into the INCUBA-
TION flask.
(a) 1 Antibiotics
28. Add the antibiotics to the INCUBATION flask containing
PBS. Swirl the flask to mix.
29. Use a serological pipette to withdraw 25 mL solution contain-
ing antibiotics from the INCUBATION flask and add to the
INCUBATION WASH flask.
30. Add PBS to the INCUBATION WASH flask up to 250 mL
(retain the PBS bottle for later use).
31. Remove the INCUBATION flask from the safety cabinet and
place in the incubator (set to 37 C).
32. Set aside the INCUBATION WASH flask for later use.
33. Discard the RAFFINOSE flask, filters, syringes, and pipettes in
the clinical waste bag.
3.6 Spongy Layer 1. Transfer the following materials to the safety cabinet:
Removal (a) 1 No. 22 blade
(b) 2 Pack of spongy spears
(c) 1 Bold surgical pen
3. Take a new square tray base to use as working tray.
4. Use the forceps to remove the AM from the WASH 3 flask and
place it on the working tray (rest the flask cap upward on the
cabinet surface).
5. Use fresh, dry, and clean spongy spears to detect the SL.
6. Carefully dab a spongy spear onto the AM and lift it. The spear
will stick to the SL but not the epithelial side. If it does not
stick, it is most likely facing the epithelial side. Flip the AM and
repeat test with a new spongy spear to ensure your acceptation
is correct (see Note 10).
7. When the SL side has been determined, spread the AM with its
epithelial side (reverse to SL side) up. Confirm you have the
correct side with a new spongy spear.
8. Use the surgical pen to write: “UP” on the epithelial side. Take
a picture of the whole AM, showing the marking.
9. Flip the AM to the SL side and spread out as much as possible.
10. Add sufficient washing solution from the WASH 3 flask to
maintain AM and SL hydration.
11. Use the reverse side of the scalpel to firmly but carefully peel,
but not scrape, back the SL off the AM as a continuous layer
(from the direction of the front of the cabinet to the back).
Take care not to cut the underlying AM stroma. Remove
substantially all the SL from the AM (see Note 11).
12. Use new spongy spears to check substantially all the SL has
been removed from across the AM (the spear should not detect
any SL).
(a) INCUBATION flask
14. Use the forceps to place the AM in the INCUBATION flask
(rest the cap upward on the cabinet surface).
15. Transfer out the INCUBATION flask and place in the incuba-
tor (set to 37 C) for 2 h (wipe off any contaminating liquids
before placing on the rocker with IMS).
16. Transfer any removed SL to the WASH 3 flask.
17. Discard the WASH 3 flask and working tray in the clinical waste
bag. Discard sharps in the sharps bin.
18. Switch on dryer (20 min before the incubation ends).

(a) INCUBATION flask
(b) 1 18 cm forceps
20. Place the forceps on the instruments tray.
21. Use the forceps to transfer the AM from the INCUBATION to
the INCUBATION WASH flask.
22. Remove the INCUBATION WASH flask from the safety cabi-
net and place on the rocker for 15 min, speed ¼ 50 RPM.
23. Discard the INCUBATION flask in the clinical waste bag.
3.7 Amniotic 1. Transfer the following materials to the safety cabinet:

Membrane Drying (a) 1 Scalpel
(Patent PCT/GB2014/
(b) Parafilm
051722)
2. Take 2–3 (depending on AM size) square lids. Place one
pre-cut Parafilm in each tray lid.
(a) INCUBATION WASH flask
4. Prepare the dryer ready for drying (see Note 12).
5. Take a new square tray base to use as a working tray.
6. Use the forceps to transfer the AM from the INCUBATION
WASH flask to the tray. Observe the AM inside the INCUBA-
TION WASH flask, if it has coiled onto itself, uncoil it while in
the liquid before taking it to the tray.
7. Spread the AM out, epithelial side upward (use the UP mark to
identify).
8. If the AM is bigger than the tray, use a scalpel to cut it into
smaller pieces, and place each AM piece on a tray, epithelial
side up.
9. Transfer trays out of the safety cabinet.
10. Place AM trays in dryer. Run recipe (see Note 13).
(a) TCF
12. Discard the remaining PBS to the INCUBATION flask.
13. Take the TCF flask and transfer the collection liquid to the PBS
bottle for bioburden analysis.
14. Transfer out all remaining materials and close the sharps bin.
15. Clean inside the safety cabinet thoroughly with IMS.
16. Discard the TCF in the clinical waste bag.
17. Turn off equipment—rocker, incubator, and safety cabinet.
18. After drying cycle has finished, remove the dried AM (see
Note 14).
19. Transfer the dried AM to the safety cabinet. Cut or store AM as
necessary (see Note 15).
20. Turn off dryer.
4 Notes
1. For preparation of dried AM for clinical use, the laboratory

must conform to strict regulatory bodies, e.g., the HTA in
accordance with the Human Tissue Act 2004, the Human
Tissue (Quality and Safety for Human Applications) Regula-
tions 2007, the Quality and Safety of Organs Intended for
Transplantation Regulations 2012, or other EU or interna-
tional regulatory bodies if processing outside the
UK. Hepatitis B immunization is required for individuals
working with human tissue and blood.
2. For clinical use, the preparation of all materials and methods
should be performed in a Grade A positive pressure environ-
ment following GMP. Settle plates and continuous particle
counting (using a particle analyzer) need to be used during
processing to ensure that the isolator conforms with EU GMP
Annex 1 Microbial Contamination limits for Grade A environ-
ments. The environment needs to be monitored for relative
humidity and temperature.
3. Raffinose is a natural preservative sugar protectant used to
prepare the tissue for drying. As an alternative, the AM can be
processed with and without raffinose or in the presence of other
novel complex saccharide lyoprotectants such as trehalose.
4. Ensure appropriate training and permission has been granted
from approved hospitals prior to consent and procurement.
Research ethics or regulatory approved consenting policies
(for clinical use) must be in place. Consent should be taken
by trained personnel on the morning of the patient undergoing
elective caesarean surgery.
5. For clinical use, donor blood samples should be taken during
the surgical process for communicable disease testing including
serological testing for HIV/HIV 2 antibodies (HIV antigen/
antibody), hepatitis B HBsAg, hepatitis B Anti HBc, HCV
antibodies, syphilis (Treponema Pallidum), HTLV antibodies,
and molecular nucleic acid amplification test testing for HIV
RNA, HBV DNA, HCV RNA, and CMV DNA. Tests must be
sent to an accredited pathology laboratory for the screening of
the mandatory infectious diseases in accordance with the
requirements of Annex II Directive 2006/17/EC. Blood
tubes for testing (purple top EDTA and yellow top) should be
placed in the collection box prior to use.
6. The fetal sac (placental membrane) is a thin tissue composed of
two layers, the AM and the chorion. The chorion forms the
outermost layer. It is thicker in appearance, opaque and vascu-
lar, with a rough texture. The AM forms the innermost layer
and is smoother in appearance, transparent, and avascular.
7. PPE must be worn, including lab coats and gloves, at all times.
Ensure gloves are category II PPE registered. For clinical use,
ensure PPE aligns with appropriate regulatory bodies for pro-
cessing human tissue.
8. Ensure that the working area is frequently cleaned during
processing. For clinical use, GMP grade cleaning reagents
must be used.
9. AM is a biological variable tissue, which can present with
significant physical, color, and visual abnormalities, depending
on the mother health, well-being, and lifestyle. The position in
the womb can also influence the AM physical appearance. For
clinical use, all raw tissue needs to be assessed by completion of
an in-process raw tissue assessment form to assess whether the
tissue is acceptable for processing. Examine the fetal sac for any
discoloration, debris, or signs of damage.
10. The “sticky” physical property of the SL is exploited in the
theatre by ophthalmology surgeons to identify the SL side of
the cryopreserved AM, though not always visible on all clinical
grade cryopreserved AM (Fig. 1). Tseng et al. in 2007 [52]
advocates the use of Weckcel cellulose surgical spears to detect
the SL, and orientate the stromal surface of AmnioGraft®.
Similarly, the proprietary CryoTek® method is also employed
to manufacture PROKERA®, meaning both these products
contain variable amounts of SL with ultimately variable bio-
chemistry. Having such close contact with the ocular surface,
the SL may therefore be integral for the reported “properties”
of the AM. Consequently, this may be crucial in explaining the
variable clinical efficacy of different membranes observed in
clinical practice [53]. The spongy spear test has therefore
been developed as a nondestructive validation test to ensure
SL removal (Fig. 1). Spongy spears are used to dab the surface
of the processed AM. Due to the sticky nature of the SL, even
single fibers adhere to the spongy spears as it lifts from the
AM. The absence of any stickiness will indicate substantially all
the SL has been removed.
11. Removal of the SL is only effective after prolonged washing
steps (Fig. 2). The excessive swelling enables the easy removal
of substantially all the SL, almost intact. The reverse edge of
the scalpel blade is used to break the SL, which is then
Fig. 1 Amnion SL removal. Image example of (a) spear testing of amnion before SL removal, and (b) following
SL removal still detecting small amount of SL at the edge
Fig. 2 Diagram illustrating complete removal of SL. (a) Standard blood contaminated AM, (b) blood removal by
continued washing without mechanical intervention (allowing SL layer to swell), (c) removal of swollen intact
SL, and (d) image of SL removed
delicately peeled off the AM. This process is performed across

the whole membrane, separating the spongy layer in its
entirety.
12. Drying parameters have been optimized to ensure effective
drying using an advantage Pro Laboratory benchtop freeze
dryer. The recipe controls the temperature and vacuum at
which the AM is exposed for a specific time in order to del-
icately dehydrate the tissue. Settings will need to be optimized
and validated for different dryers.
13. Turn off the shelf and condenser from pre-freeze. Shelf tem-
peratures should be set between 3.5 C and 6.0 C and con-
denser should be set between 60 C and 90 C. These can
be altered by turning the shelf and condenser buttons on
and off.
14. Verify the dryer receipt is complete. Defrost shelves, condenser,

and release vacuum. Wait until the system reaches atmospheric
pressure (approximately 1070 mBar) and shelves reach 30 C
before taking the trays out of the dryer.
15. Use new consumables for each AM cutting procedure. Identify
holes and tears in the tissue and do not include them in the
product. Use cutters or scalpel to cut AM into desired shapes.
Use forceps (grip lightly to ensure no damage is caused) to
carefully lift cut pieces from the tray. If any edges still appear to
be connected, use a scalpel blade to gently cut the connections.
Package product as required.
References
1. Meller D, Pauklin M, Thomasen H, X. [pii] 10.1002/(SICI)1097-4652(199906)
Westekemper H, Steuhl K-P (2011) Amniotic 179:3<325::AID-JCP10>3.0.CO;2-X
membrane transplantation in the human eye. 8. Meller D, Pires RT, Mack RJ, Figueiredo F,
Dtsch Arztebl Int 108(14):243–248. https:// Heiligenhaus A, Park WC, Prabhasawat P,
doi.org/10.3238/arztebl.2011.0243 John T, McLeod SD, Steuhl KP, Tseng SC
2. Dua HS, Gomes JA, King AJ, Maharajan VS (2000) Amniotic membrane transplantation
(2004) The amniotic membrane in ophthal- for acute chemical or thermal burns. Ophthal-
mology. Surv Ophthalmol 49(1):51–77 mology 107(5):980–989. discussion 990.
3. Hao Y, Ma DH, Hwang DG, Kim WS, Zhang S0161-6420(00)00024-5 [pii]
F (2000) Identification of antiangiogenic and 9. Kamiya K, Wang M, Uchida S, Amano S,
antiinflammatory proteins in human amniotic Oshika T, Sakuragawa N, Hori J (2005) Topi-
membrane. Cornea 19(3):348–352 cal application of culture supernatant from
4. Ogawa Y, He H, Mukai S, Imada T, human amniotic epithelial cells suppresses
Nakamura S, Su CW, Mahabole M, Tseng SC, inflammatory reactions in cornea. Exp Eye
Tsubota K (2017) Heavy chain-hyaluronan/ Res 80(5):671–679. https://doi.org/10.
pentraxin 3 from amniotic membrane sup- 1016/j.exer.2004.11.018
presses inflammation and scarring in murine 10. Niknejad H, Yazdanpanah G (2014) Antican-
lacrimal gland and conjunctiva of chronic cer effects of human amniotic membrane and
graft-versus-host disease. Sci Rep 7:42195. its epithelial cells. Med Hypotheses 82
https://doi.org/10.1038/srep42195 (4):488–489. https://doi.org/10.1016/j.
5. Bennett JP, Matthews R, Faulk WP (1980) mehy.2014.01.034
Treatment of chronic ulceration of the legs 11. Niknejad H, Yazdanpanah G, Ahmadiani A
with human amnion. Lancet 1 (2016) Induction of apoptosis, stimulation of
(8179):1153–1156 cell-cycle arrest and inhibition of angiogenesis
6. Hopkinson A, McIntosh RS, make human amnion-derived cells promising
Shanmuganathan V, Tighe PJ, Dua HS sources for cell therapy of cancer. Cell Tissue
(2006) Proteomic analysis of amniotic mem- Res 363(3):599–608. https://doi.org/10.
brane prepared for human transplantation: 1007/s00441-016-2364-3
characterization of proteins and clinical impli- 12. Niknejad H, Yazdanpanah G, Mirmasoumi M,
cations. J Proteome Res 5(9):2226–2235. Abolghasemi H, Peirovi H, Ahmadiani A
https://doi.org/10.1021/pr050425q (2013) Inhibition of hsp90 could be possible
7. Tseng SC, Li DQ, Ma X (1999) Suppression of mechanism for anti-cancer property of amni-
transforming growth factor-beta isoforms, otic membrane. Med Hypotheses 81
tgf-beta receptor type ii, and myofibroblast (5):862–865. https://doi.org/10.1016/j.
differentiation in cultured human corneal and mehy.2013.08.018
limbal fibroblasts by amniotic membrane 13. Magatti M, Munari S, Vertua E, Parolini O
matrix. J Cell Physiol 179(3):325–335. (2012) Amniotic membrane-derived cells
https://doi.org/10.1002/(SICI)1097-4652( inhibit proliferation of cancer cell lines by
199906)179:3<325::AID-JCP10>3.0.CO;2- inducing cell cycle arrest. J Cell Mol Med 16
(9):2208–2218. https://doi.org/10.1111/j. 23. Davis J (1910) Skin transplantation with a

1582-4934.2012.01531.x review of 550 cases at the johns hopkins hospi-
14. Prabhasawat P, Tesavibul N, tal. Johns Hopkins Med J 15:307
Prakairungthong N, Booranapong W (2007) 24. Schrimpf WJ (1954) Repair of tympanic mem-
Efficacy of amniotic membrane patching for brane perforations with human amniotic mem-
acute chemical and thermal ocular burns. J brane; report of fifty-three cases. Ann Otol
Med Assoc Thail 90(2):319–326 Rhinol Laryngol 63(1):101–115
15. Tejwani S, Kolari RS, Sangwan VS, Rao GN 25. Trelford JD, Trelford-Sauder M (1979) The
(2007) Role of amniotic membrane graft for amnion in surgery, past and present. Am J
ocular chemical and thermal injuries. Cornea Obstet Gynecol 134(7):833–845. doi:0002-
26(1):21–26. https://doi.org/10.1097/ICO. 9378(79)90957-8 [pii]
0b013e31802b4201. 00003226-200701000- 26. Matthews RN (1981) Human tissue response
00005 [pii] to amnion allograft. Lancet 2(8260–61):1428
16. Chawla B, Sharma N, Tandon R, Kalaivani M, 27. Ashworth MF, Morton KE, Dewhurst J, Lil-
Titiyal JS, Vajpayee RB (2010) Comparative ford RJ, Bates RG (1986) Vaginoplasty using
evaluation of phototherapeutic keratectomy amnion. Obstet Gynecol 67(3):443–446
and amniotic membrane transplantation for 28. Cox LW (1987) Constructive and reconstruc-
management of symptomatic chronic bullous tive gynaecology: uterine and vaginal grafts.
keratopathy. Cornea 29(9):976–979. https:// Aust N Z J Surg 57(5):284–286
doi.org/10.1097/ICO.0b013e3181ca369a
29. Brandt FT, Albuquerque CD, Lorenzato FR
17. Tamhane A, Vajpayee RB, Biswas NR, Pandey (2000) Female urethral reconstruction with
RM, Sharma N, Titiyal JS, Tandon R (2005) amnion grafts. Int J Surg Investig 1
Evaluation of amniotic membrane transplanta- (5):409–414
tion as an adjunct to medical therapy as com-
pared with medical therapy alone in acute 30. Lai DR, Chen HR, Lin LM, Huang YL, Tsai
ocular burns. Ophthalmology 112 CC (1995) Clinical evaluation of different
(11):1963–1969. https://doi.org/10.1016/j. treatment methods for oral submucous fibro-
ophtha.2005.05.022 sis. A 10-year experience with 150 cases. J Oral
Pathol Med 24(9):402–406
18. Sharma N, Singh D, Maharana PK, Kriplani A,
Velpandian T, Pandey RM, Vajpayee RB 31. Goto Y, Noguchi Y, Nomura A, Sakamoto T,
(2016) Comparison of amniotic membrane Ishii Y, Bitoh S, Picton C, Fujita Y,
transplantation and umbilical cord serum in Watanabe T, Hasegawa S, Uchida Y (1999) In
acute ocular chemical burns: a randomized vitro reconstitution of the tracheal epithelium.
controlled trial. Am J Ophthalmol Am J Respir Cell Mol Biol 20(2):312–318
168:157–163. https://doi.org/10.1016/j. 32. Haugh AM, Witt JG, Hauch A, Darden M,
ajo.2016.05.010 Parker G, Ellsworth WA, Buell JF (2017)
19. Kheirkhah A, Tabatabaei A, Zavareh MK, Amnion membrane in diabetic foot wounds: a
Khodabandeh A, Mohammadpour M, Raju meta-analysis. Plast Reconstr Surg Glob Open
VK (2012) A controlled study of amniotic 5(4):e1302–e1302. https://doi.org/10.
membrane transplantation for acute pseudo- 1097/GOX.0000000000001302
monas keratitis. Can J Ophthalmol 47 33. Guo Q, Hao J, Yang Q, Guan L, Ouyang S,
(3):305–311. https://doi.org/10.1016/j. Wang J (2011) A comparison of the effective-
jcjo.2012.03.014 ness between amniotic membrane homogenate
20. Jensen O, Gluud B (1985) Bacterial growth in and transplanted amniotic membrane in heal-
the conjunctival sac and the local defense of the ing corneal damage in a rabbit model. Acta
outer eye. Acta Ophthalmol 63(S173):80–82 Ophthalmol 89(4):e315–e319. https://doi.
org/10.1111/j.1755-3768.2010.02097.x
21. Sotozono CMDPD, Inagaki KMD, Fujita
AMD, Koizumi NMDPD, Sano YMDPD, Ina- 34. Dua HS (1999) Amniotic membrane trans-
tomi TMDPD, Kinoshita SMDPD (2002) plantation. Br J Ophthalmol 83(6):748–752
Methicillin-resistant staphylococcus aureus 35. Tabatabaei SA, Soleimani M, Behrouz MJ,
and methicillin-resistant staphylococcus epi- Torkashvand A, Anvari P, Yaseri M (2017) A
dermidis infections in the cornea. Cornea 21 randomized clinical trial to evaluate the useful-
(2):S94–S101 ness of amniotic membrane transplantation in
22. Goodman DF, Gottsch JD (1988) Methicillin- bacterial keratitis healing. Ocul Surf 15
resistant staphylococcus epidermidis keratitis (2):218–226
treated with vancomycin. Arch Ophthalmol 36. Chen H-C, Tan H-Y, Hsiao C-H, Huang
106(11):1570–1571. https://doi.org/10. SC-M, Lin K-K, Ma DH-K (2006) Amniotic
1001/archopht.1988.01060140738046 membrane transplantation for persistent
corneal ulcers and perforations in acute fungal membrane in ophthalmic surgery. In: Essen-
keratitis. Cornea 25(5):564–572 tials in ophthalmology cornea and external eye
37. Dua HS, Rahman I, Miri A (2010) Variations disease. Springer, New York, pp 29–41
in amniotic membrane: relevance for clinical 46. Allen CL, Clare G, Stewart EA, Branch MJ,
applications. Br J Ophthalmol 94(8):963–964 McIntosh OD, Dadhwal M, Dua HS, Hopkin-
38. Khokhar SMD, Sharma NMD, Kumar HMD, son A (2013) Augmented dried versus cryopre-
Soni AMD (2001) Infection after use of non- served amniotic membrane as an ocular surface
preserved human amniotic membrane for the dressing. PLoS One 8(10):e78441. https://
reconstruction of the ocular surface. Cornea 20 doi.org/10.1371/journal.pone.0078441
(7):773–774 47. Nakamura T, Sekiyama E, Takaoka M, Bentley
39. Cheng AM, Zhao D, Chen R, Yin HY, Tighe S, AJ, Yokoi N, Fullwood NJ, Kinoshita S (2008)
Sheha H, Casas V, Tseng SC (2016) Acceler- The use of trehalose-treated freeze-dried amni-
ated restoration of ocular surface health in dry otic membrane for ocular surface reconstruc-
eye disease by self-retained cryopreserved tion. Biomaterials 29(27):3729–3737.
amniotic membrane. Ocul Surf 14(1):56–63. https://doi.org/10.1016/j.biomaterials.
https://doi.org/10.1016/j.jtos.2015.07.003 2008.05.023. S0142-9612(08)00378-5 [pii]
40. Paolin A, Trojan D, Leonardi A, Mellone S, 48. Aghayan HR, Goodarzi P, Baradaran-Rafii A,
Volpe A, Orlandi A, Cogliati E (2016) Cyto- Larijani B, Moradabadi L, Rahim F, Arjmand B
kine expression and ultrastructural alterations (2013) Bacterial contamination of amniotic
in fresh-frozen, freeze-dried and gamma- membrane in a tissue bank from iran. Cell Tis-
irradiated human amniotic membranes. Cell sue Bank 14(3):401–406. https://doi.org/10.
Tissue Bank 17(3):399–406. https://doi.org/ 1007/s10561-012-9345-x
10.1007/s10561-016-9553-x 49. Marsit NM, Sidney LE, Branch MJ, Wilson SL,
41. Koob TJ, Rennert R, Zabek N, Massee M, Lim Hopkinson A (2017) Terminal sterilization:
JJ, Temenoff JS, Li WW, Gurtner G (2013) conventional methods versus emerging cold
Biological properties of dehydrated human atmospheric pressure plasma technology for
amnion/chorion composite graft: implications non-viable biological tissues. Plasma Process
for chronic wound healing. Int Wound J 10 Polym 14(7):e1600134. https://doi.org/10.
(5):493–500. https://doi.org/10.1111/iwj. 1002/ppap.201600134
12140 50. Marsit N (2019) Characterisation of vacuum
42. Hopkinson A, McIntosh RS, Tighe PJ, James dried amniotic membrane and validation of
DK, Dua HS (2006) Amniotic membrane for the processing protocol. PhD, The University
ocular surface reconstruction: donor variations of Nottingham Recently passed
and the effect of handling on tgf-beta content. 51. Marsit NM, Sidney LE, Britchford ER, McIn-
Invest Ophthalmol Vis Sci 47(10):4316–4322. tosh OD, Allen CL, Ashraf W, Bayston R, Hop-
https://doi.org/10.1167/iovs.05-1415 kinson A (2019) Validation and assessment of
43. Gicquel JJ, Dua HS, Brodie A, Mohammed I, an antibiotic decontamination manufacturing
Suleman H, Lazutina E, James DK, Hopkinson protocol for vacuum-dried human amniotic
A (2009) Epidermal growth factor variations in membrane. Sci Rep 9(1):12854. https://doi.
amniotic membrane used for ex vivo tissue org/10.1038/s41598-019-49314-7
constructs. Tissue Eng Part A 15 52. Tseng SC, Elizondo A, Cases V (2007) Amni-
(8):1919–1927. https://doi.org/10.1089/ otic membrane suturing techniques. In: Macsai
ten.tea.2008.0432 MS (ed) Ophthalmic microsurgical suturing
44. Hopkinson A, McIntosh RS, Layfield R, techniques. Springer, Berlin
Keyte J, Dua HS, Tighe PJ (2005) Optimised 53. Maharajan VS, Shanmuganathan V, Currie A,
two-dimensional electrophoresis procedures Hopkinson A, Powell-Richards A, Dua HS
for the protein characterisation of structural (2007) Amniotic membrane transplantation
tissues. Proteomics 5(7):1967–1979. https:// for ocular surface reconstruction: indications
doi.org/10.1002/pmic.200401073 and outcomes. Clin Exp Ophthalmol 35
45. Dua HS, Maharajan VS, Hopkinson A (2005) (2):140–147. https://doi.org/10.1111/j.
Controversies and limitations of amniotic 1442-9071.2006.01408.x
Chapter 11
Fabrication of Corneal Extracellular Matrix-Derived

Hydrogels
Mark Ahearne and Julia Fernández-Pérez
Abstract
Hydrogels derived from corneal extracellular matrix (ECM) represent a promising biomaterial for corneal
repair and regeneration. To fabricate these hydrogels, first corneas need to be decellularized using repeated
freeze-thaw cycles and nucleases to remove all nuclear and cellular components. The remaining corneal
ECM is lyophilized to remove all water and milled into a fine powder. The ECM powder is weighed and
dissolved in pepsin solution at a concentration of 20 mg/mL. Hydrogels are formed by neutralizing the pH
of the solution and maintaining it at 37 C until fibrillogenesis has occurred. Corneal stromal cells may be
suspended throughout the hydrogel solution prior to gelation to generate a corneal stromal substitute.
Key words Hydrogel, Scaffold, Cornea, Stroma, Collagen, Extracellular matrix, Keratocyte
1 Introduction
The global shortage of donor corneal tissues suitable for transplan-

tation [1] has led many researchers to investigate alternative treat-
ment modalities. One solution is to use tissue engineering to
manufacture corneal tissue in vitro by combining either corneal-
derived cells or stem cells with a suitable three-dimensional bioma-
terial scaffold. One of the main challenges in developing scaffolds
that are suitable for corneal stromal regeneration or the replace-
ment of damaged corneal tissue is how to accurately replicate the
composition of the real tissue. While the stroma primarily consists
of collagen type 1, the presence of other collagens, proteoglycans,
and cytokines in the extracellular matrix (ECM) is vital to regulat-
ing the phenotype of corneal stromal keratocytes [2–5] and main-
taining the corneas structure and transparency [6–8]. While
decellularized corneal scaffolds are capable of retaining most of
these ECM components of the cornea, the main difficulty asso-
ciated with these types of scaffolds is how to successfully repopulate
the matrix with new healthy cells after decellularization [9, 10].
159
160 Mark Ahearne and Julia Fernández-Pérez
Fig. 1 Overview of steps involved in fabricating corneal ECM hydrogels: (a) removal of cornea from eye;
(b) dissection of cornea; (c) suspended for freeze-thaw decellularization; (d) ECM after freeze-drying; (e) ECM
after cryomilling; (f) suspended in pepsin; (g) final hydrogel solutions; (h) final hydrogel
The recent development of corneal ECM-derived hydrogels

represents a novel approach for generating biocompatible scaffolds
that can retain most of the native tissues ECM components [11, 12].
These hydrogels form when solubilized collagen from ECM under-
goes polymerization under specific temperature and pH conditions
to form a water swollen collagen network that still contains other
ECM components [13]. Unlike decellularized corneal scaffolds,
cells can be easily suspended throughout the hydrogel prior to
gelation. This allows for the generation of a three-dimensional
corneal stromal construct embedded with viable cells. In addition,
corneal ECM hydrogel solutions have previously been used as a
bioink for the 3D bioprinting of corneal constructs [14, 15].
This chapter will explore the process of fabricating corneal
ECM-derived hydrogels. While there are several variations in how
to fabricate these hydrogels, they all tend to follow the same basic
steps (Fig. 1). First, the corneal tissue needs to be decellularized to
remove any cells and cellular components that could potentially
illicit a negative immune response if present (see Note 1). Next, the
tissue needs to be dehydrated and broken down into fine particles
to allow the ECM to be more easily digested. The ECM particles
are then digested in an acidic solution. Finally, by neutralizing the
pH of the ECM digest, the collagen undergoes fibrillogenesis
resulting in a stable hydrogel being formed. Since gelation of the
hydrogel takes 20–30 min, cells can be mixed into the neutralized
ECM solution prior to gelation occurring. Further crosslinking
may then be applied to improve the stability and mechanical prop-
erties of the hydrogels [16].
Corneal ECM Hydrogels 161
2 Materials
Where possible, fresh porcine corneas should be obtained directly

from a local slaughterhouse or abattoir (see Notes 2 and 3). Ultra-
pure water is water that has been purified by double deionizing and
filtering to 18 MΩ-cm resistivity at room temperature.
2.1 Lab Equipment 1. Freeze dryer.

2. Cryomill—a “SPEX Freezer/Mill Cryogenic Grinder” was
used in this study.
3. Cryomill polycarbonate tube, stainless steel end plugs, and
stainless steel impactor.
4. Type II biological safety cabinet.
5. CO2, temperate, and humidity controlled incubator.
6. Oven.
7. 80 C freezer.
8. Tube rotator.
9. Electronic weighing scale.
10. 12-Well plates.
11. 50 mL syringes.
12. 10 mL syringes.
13. 0.2 μm sterile filters.
14. 50 mm diameter Petri dishes.
15. 5 mL sterile tubes.
16. 1.5 mL sterile centrifuge tube.
17. 8 mm diameter biopsy punches.
18. Gilson pipettes (P1000 and P200) and sterile pipette tips.
19. Scalpel.
20. Spatula.
21. Tape.
22. Sharp scissors.
23. Sterile fine tipped forceps.
2.2 Preparation of 1. Reconstituting buffer: 5 mM CaCl2 in ultrapure water. Sterile

Solutions filter solution using a syringe and 0.2 μm sterile filter.
2. 10 mM MgCl2 buffer solution: 203.3 mg of magnesium chlo-
ride hexahydrate in 100 mL of ultrapure water. Adjust the pH
to 7.5 using 5 M sodium hydroxide solution (NaOH) and
sterile filter solution using a syringe and 0.2 μm sterile filter.
3. 400 U/mL deoxyribonuclease I (DNAse): Dissolve 10 mg of
400 U/mg DNAse from bovine pancreas in 10 mL of recon-
stituting buffer. Aliquot and store at 20 C for up to 1 year.
4. 900 U/mL Ribonuclease (RNAse): Dissolve 50 mg of

90 U/mg RNAse A from bovine pancreas in 5 mL sterile-
autoclaved ultrapure water. For different initial U/mg adjust
the mass accordingly. Aliquot and store at 20 C for up to
6 months.
5. DNAse/RNAse solution: Add 75 μL of DNAse solution and
33 μL of RNAse solution to 3 mL MgCl2 buffer solution.
6. 1 mg/mL pepsin solution: Add 20 mg of pepsin to 20 mL
0.1 M HCl.
7. 2% iodine solution.
8. Phosphate-buffered saline (PBS).
2.3 Hydrogel 1. 1 N NaOH.

Fabrication 2. 10 PBS.
3. Whatman Grade 1 filter paper.
4. PTFE inserts (see Note 4).
3 Methods
3.1 Decellularization While there are numerous methods that have been developed to
decellularize cornea [17], the method outlined here has been
shown to be the most suitable for removing cells while still allowing
a hydrogel to be formed [18].
1. Under sterile conditions in a biological safety cabinet, rinse
enucleated porcine eyes with 2% iodine solution followed by
washes with PBS.
2. Using a sterile scalpel and forceps, remove the cornea from the
eye and place the cornea into a Petri dish. Remove any
non-corneal tissue from the cornea such as iris or conjunctiva
and chop the remaining cornea into several pieces
(2–3 mm3 each).
3. Place cornea pieces into a 5 mL tube with 4 mL sterile deio-
nized water.
4. Freeze tube at 80 C for at least 4 h.
5. Allow water to fully thaw to room temperature without using
the water bath and then replace the water with fresh ultrapure
water.
6. Repeat steps 4 and 5 so the corneas undergo 5 freeze-thaw
cycles in total.
7. Place each cornea in 3 mL of DNAse/RNAse solution and
place in an oven under gentle rotation at 37 C for 1 h.
8. Remove solution and wash each cornea using ultrapure water

three times for 24 h under rotation to remove any residual
DNAse or RNAse (see Note 5).
3.2 Lyophilization 1. Place the corneas in a Petri dish and submerge with sterile
ultrapure water.
2. Place the dish in a freeze dryer (see Note 6) and program to
reduce temperature to 30 C at a rate of 1 C/min from
room temperature and hold at this temperature for 1 h.
3. Increase temperature to 10 C at a rate of 1 C/min. When
the temperature reaches 10 C, switch on the vacuum to
generate a pressure of 200 mbar inside the freeze dryer cham-
ber and leave for 18 h.
4. Switch off vacuum and remove dish from freeze dryer.
5. Check that the tissue has been completely dried (see Note 7).
3.3 Cryomilling 1. Sterilize cryomills polycarbonate tube, stainless steel end plugs,
and stainless steel impactor before use (see Note 8).
2. Place lyophilized corneas in a polycarbonate tube with the
impactor, seal both ends with the end plugs, and place tube
into the cryomill (see Note 9).
3. Carefully fill tank with liquid nitrogen (see Note 10).
4. Close the cryomill chamber and run three times a 1-min cycle
at a rate of 5 CPS with a 1-min break between each cycle (see
Note 11).
5. Open and remove contents into a pre-weighed 5 mL tube in a
flow hood. Calculate the weight of corneal ECM powder by
re-weighing the tube (see Note 12).
3.4 Hydrogel 1. Add 1 mg/mL pepsin solution to ECM powder to give a final
Formation concentration of 20 mg/mL ECM in solution.
2. Place in a rotator and gently rotate for at least 24 h at room
temperature. Check that ECM powder has fully dissolved
before continuing.
3. To make filter paper rings first use a 12 mm diameter circular
disk to draw several circles on a sheet of filter paper. Cut around
the circles with a sharp scissors to make 12 mm diameter filter
paper disks. Next, use an 8 mm biopsy punch to cut holes in the
center of the filter paper disks. Filter paper rings can be auto-
claved to sterilize.
4. Place a PTFE disk into each well of a 12-well plate and place a
filter paper ring on each disk.
5. To make 15 mg/mL hydrogels, cool a 1.5 mL tube on ice and
add 42.7 μL of sterile water, 50 μL of 10 PBS, and 32.3 μL of
1 N NaOH and mix by pipetting up and down. Add 375 μL to

tube to give a total volume of 500 μL and gentle mix by
rotating the tube (see Notes 13–15).
6. Pipette 150 μL of the hydrogel solution into each well so to
cover the filter paper ring and hole (see Notes 16 and 17).
7. Place in an incubator at 37 C for 1 h and then submerge in
pre-warmed culture media.
8. Separate the hydrogel and filter paper from the PTFE disk and
remove the disk from the well (see Note 18).
4 Notes
1. It is important that all cells and most of the cellular DNA be

removed from the ECM to reduce the potential for negative
immune responses in vivo. To confirm the success of decellu-
larization it is believed that the ECM should contain less than
50 ng of double-stranded DNA per mg of ECM dry weight
and any DNA fragments be less than 200 base pairs [19]. Cells
or cell nuclei should also not be visible after histological sec-
tioning and staining.
2. The quality of the eyes that are being used should be inspected
before proceeding. Corneas that show signs of swelling, cloud-
iness, or physical damage should not be used. Ideally, the eyes
should be removed shortly after the animal has been sacrificed.
Some slaughterhouses or abattoirs will burn or chemically treat
the pigs after death to remove hair, and if the eyes are still in
place when this is happening, this will compromise the quality
of the organs.
3. This protocol has been designed for use with porcine eyes but it
should be possible to apply this protocol to eyes from different
species. Some modifications to the decellularization process
may have to be made and there is no guarantee that the out-
come will be the same.
4. PTFE is less adhesive to proteins than cell culture plastic and
allows the hydrogels to detach more easily. PTFE disks can be
easily manufactured by purchasing a 0.5 mm to 1 mm thick
sheet of PTFE and cutting it into circular disks using sharp
scalpel. Thinner sheets may also be used but these tend to bend
and can lead to difficulties when casting the hydrogel. Thicker
sheets are more difficult to cut. PTFE can be sterilized by
autoclaving.
5. Here we used a freeze-thaw process combined with nucleases
to decellularize the tissue. We have tested several alternative
decellularization procedures and found that this approach
offered good retention of ECM components while still
allowing the removal cell and cellular components when com-

pared to some alternative decellularization treatments. Other
decellularization techniques could potentially work but would
need to be tested.
6. When using a Petri dish in the freeze dryer, it is important to
tape the lid onto the dish; otherwise it is likely to be sucked off
once a vacuum is applied. Other containers, such as tubes, may
also be used as alternative to a Petri dish; however it is impor-
tant that there is a hole or space to allow water vapor to escape
from the container. Make sure that the water is completely
frozen before applying the vacuum.
7. After freeze-drying is complete, there should be no water or ice
left in the Petri dish. If there is still water present, a longer
vacuum cycle may be required and less water used in future.
Alternatively, if the lid on the Petri dish or container housing
the corneas is closed too tightly, this will prevent the water
from escaping.
8. The cryomill polycarbonate tube, stainless steel end plugs, and
impactor need to be sterilized before use. Particular care needs
to be taken with the tube as it is likely to warp if autoclaved and
undergo cracking if subjected to ethanol for prolonged periods
of time. UV exposure in a cell culture flow hood for 30 min
allows the tube to become sterile. Ethylene oxide may also be
used to sterile the tube without damaging it.
9. We place between 10 and 20 freeze-dried corneas in the tube
for cryomilling at the same time. Fewer corneas than this can
lead to a low volume of ECM powder being obtained since the
powder sticks to the side of the tube and grinder due to static
energy. Higher numbers of corneas than this and the tube can
become too full, preventing the impactor from successfully
grinding the tissue into a powder.
10. It is recommended that several precautions are taken when
using liquid nitrogen for cryomilling. Personnel protective
equipment, especially a facemask and nitrogen handling gloves,
should be worn as the nitrogen will probably spit and bubble
when it first encounters the room temperate cryomill chamber.
Adequate space and ventilation is also necessary to prevent the
accumulation of nitrogen gas in the room. Ideally, an oxygen
sensor should be present to detect if the nitrogen has displaced
too much oxygen within the room. The nitrogen chamber in
the cryomill should only be closed when the nitrogen has
stopped bubbling. Care should also be taken when opening
the chamber and removing the tube after the cryomilling cycles
have been complete as liquid nitrogen is still present. The
remaining nitrogen should be allowed to evaporate rather
than trying to pour into a separate container.
11. The cryomill used in this description is very loud, so it is

recommended that the user wear ear protection and warns
any users nearby when the machine is going to start.
12. Some of the ECM will stick to the sides and ends of the tube
and the grinder after cryomilling making it more difficult to
extract. A sterile spatula is the best way to collect this ECM
powder and moving it to a tube; however the user can expect
some powder to be lost. The powder can be centrifuged once
transferred to the tube.
13. Cells may be mixed into the hydrogel prior to gelation. This is
best done by reducing the amount of dH2O and suspending
the cells in culture media of an equivalent volume to that
reduced. The cell suspension should be added after the other
reagents have been mixed together; otherwise the pH of the
hydrogel solution may not be evenly neutralized and result in
cell death where the pH is too high or low. The cell suspension
can be mixed by gentle inverting or stirring using a pipette tip.
Pipetting up and down into a tip is not recommended as this
results in the formation of bubbles. However for hydrogels
with high concentrations (over 15 mg/mL) pipetting may be
required to mix.
14. 10x DMEM (Dulbecco’s modified eagles medium) can be used
as an alternative to 10 PBS when forming the hydrogels. The
advantage of 10x DMEM is that it allows the user to visually
determine if the solution has been evenly mixed due to its
color. The color can also be used to indicate if the pH is
approximately between 7 and 7.5. The drawback with using
10 DMEM is that the hydrogel is now red in color rather
than being fully transparent. However this color can be
removed after gelation by washing with PBS.
15. Try to avoid generating any bubbles while pipetting the hydro-
gel solution into the filter paper rings. If bubbles are present,
they can be removed by either bursting with a needle tip or by
pushing to the edge of the hydrogel using the pipette tip.
16. The shape and volume of the hydrogel can be changed depend-
ing on the user requirements. The use of filter paper rings when
casting the hydrogels does have a number of advantages includ-
ing it makes it easier for the hydrogels to be picked up and it
limits contraction of the hydrogel by cells in the horizontal
plane [20]. The size of the rings can also be varied if necessary.
If using larger rings (e.g., 20 mm diameter), it is better to use
two rings placed on top of each other rather than one.
17. Whatman grade 1 filter paper is effective at allowing hydrogel
adhesion without absorbing too much liquid. We have test
other grades of filter paper and found this to be the most
practical to use.
18. The hydrogel can be detached from the PTFE disks by gently
pushing it from the edge while submerged in solution, e.g.,
culture media. If the hydrogel is still sticking to the disk, lift the
disk and use a scalpel to separate.
Acknowledgements
The research is supported by funding from the European Research

Council (ERC) under the European Union’s Horizon 2020
research and innovation program (grant agreement no. 637460)
and Science Foundation Ireland and Marie-Curie Action
COFUND (grant no. 11/SIRG/B2104).
References
1. Gain P, Jullienne R, He Z, Aldossary M, Biophys Rev 1(2):83–93. https://doi.org/10.
Acquart S, Cognasse F, Thuret G (2016) 1007/s12551-009-0011-x
Global survey of corneal transplantation and 9. Lynch AP, Wilson SL, Ahearne M (2016) Dex-
eye banking. JAMA Ophthalmol 134 tran preserves native corneal structure during
(2):167–173. https://doi.org/10.1001/ decellularization. Tissue Eng Part C Methods
jamaophthalmol.2015.4776 22(6):561–572. https://doi.org/10.1089/
2. Fernandez-Perez J, Ahearne M (2019) Influ- ten.TEC.2016.0017
ence of biochemical cues in human corneal 10. Wilson SL, Sidney LE, Dunphy SE, Dua HS,
stromal cell phenotype. Curr Eye Res 44 Hopkinson A (2016) Corneal decellulariza-
(2):135–146. https://doi.org/10.1080/ tion: a method of recycling unsuitable donor
02713683.2018.1536216 tissue for clinical translation? Curr Eye Res 41
3. Petroll WM, Miron-Mendoza M (2015) (6):769–782. https://doi.org/10.3109/
Mechanical interactions and crosstalk between 02713683.2015.1062114
corneal keratocytes and the extracellular 11. Ahearne M, Lynch AP (2015) Early observa-
matrix. Exp Eye Res 133:49–57. https://doi. tion of extracellular matrix-derived hydrogels
org/10.1016/j.exer.2014.09.003 for corneal stroma regeneration. Tissue Eng
4. Jester JV, Ho-Chang J (2003) Modulation of Part C Methods 21(10):1059–1069. https://
cultured corneal keratocyte phenotype by doi.org/10.1089/ten.TEC.2015.0008
growth factors/cytokines control in vitro con- 12. Lu Y, Yao QK, Feng B, Yan CX, Zhu MY, Chen
tractility and extracellular matrix contraction. JZ, Fu W, Fu Y (2015) Characterization of a
Exp Eye Res 77(5):581–592 hydrogel derived from decellularized corneal
5. Lynch AP, O’Sullivan F, Ahearne M (2016) extracellular matrix. J Biomater Tiss Eng 5
The effect of growth factor supplementation (12):951–960. https://doi.org/10.1166/jbt.
on corneal stromal cell phenotype in vitro 2015.1410
using a serum-free media. Exp Eye Res 13. Saldin LT, Cramer MC, Velankar SS, White LJ,
151:26–37. https://doi.org/10.1016/j.exer. Badylak SF (2017) Extracellular matrix hydro-
2016.07.015 gels from decellularized tissues: structure and
6. Hassell JR, Birk DE (2010) The molecular function. Acta Biomater 49:1–15. https://doi.
basis of corneal transparency. Exp Eye Res 91 org/10.1016/j.actbio.2016.11.068
(3):326–335. https://doi.org/10.1016/j. 14. Kim H, Park MN, Kim J, Jang J, Kim HK, Cho
exer.2010.06.021 DW (2019) Characterization of cornea-specific
7. Massoudi D, Malecaze F, Galiacy SD (2016) bioink: high transparency, improved in vivo
Collagens and proteoglycans of the cornea: safety. J Tissue Eng 10:2041731418823382.
importance in transparency and visual disorders. https://doi.org/10.1177/
Cell Tissue Res 363(2):337–349. https://doi. 2041731418823382
org/10.1007/s00441-015-2233-5 15. Kim H, Jang J, Park J, Lee KP, Lee S, Lee DM,
8. Meek KM (2009) Corneal collagen-its role in Kim KH, Kim HK, Cho DW (2019) Shear-
maintaining corneal shape and transparency. induced alignment of collagen fibrils using 3d
cell printing for corneal stroma tissue engineer- extracellular matrix derived hydrogels. Sci Rep
ing. Biofabrication 11(3). https://doi.org/10. 9(1):14933. https://doi.org/10.1038/
1088/1758-5090/ab1a8b s41598-019-49575-2
16. Ahearne M, Coyle A (2016) Application of 19. Crapo PM, Gilbert TW, Badylak SF (2011) An
uva-riboflavin crosslinking to enhance the overview of tissue and whole organ decellular-
mechanical properties of extracellular matrix ization processes. Biomaterials 32
derived hydrogels. J Mech Behav Biomed (12):3233–3243. https://doi.org/10.1016/j.
Mater 54:259–267. https://doi.org/10. biomaterials.2011.01.057
1016/j.jmbbm.2015.09.035 20. Ahearne M, Liu KK, El Haj AJ, Then KY,
17. Fernandez-Perez J, Ahearne M (2019) Decel- Rauz S, Yang Y (2010) Online monitoring of
lularization and recellularization of cornea: the mechanical behavior of collagen hydrogels:
progress towards a donor alternative. Methods influence of corneal fibroblasts on elastic mod-
171:86–96. https://doi.org/10.1016/j. ulus. Tissue Eng Part C Methods 16
ymeth.2019.05.009 (2):319–327. https://doi.org/10.1089/ten.
18. Fernandez-Perez J, Ahearne M (2019) The TEC.2008.0650
impact of decellularization methods on
Chapter 12
Synthesis and Application of Collagens for Assembling

a Corneal Implant
Elle Edin, Fiona Simpson, and May Griffith
Abstract
Recombinant or artificial designer collagens have developed to a point where they are viable candidates for
replacing extracted animal collagens in regenerative medicine applications. Biomimetic corneas made have
shown promise as replacements for human donor corneas, and have previously been fabricated from several
different collagens or collagen-like peptides (CLPs). Prokaryotic expression systems allow for cheap, rapid,
gram scale production of collagens/CLPs. Here, we describe a procedure for production of collagen-like
peptides for the manufacture of a biomimetic cornea.
Key words Biomimetic, Artificial, Cornea, Artificial collagen, Collagen, Corneal regeneration,
Transgenic
1 Introduction
1.1 Collagen Collagen is the most abundant protein present in the extracellular
and Collagen-Like matrix that surrounds the cells of various tissues and organs in the
Peptides mammalian body, including the cornea [1]. The defining feature of
collagen is its unique supercoiled triple-helix structure [2, 3]. Fibril-
lar collagens, in particular, are robust structural macromolecules
that contain cell-interactive domains. Hence, they have excellent
properties for creating regenerative, cell-free scaffolds for corneal
repair as seen in early clinical evaluation (Fig. 1) [4].
Most commercially available collagen is extracted from animal
sources and purified using different methods, resulting in hetero-
geneity of size and helicity [5]. Recombinantly produced
human collagens and short collagen mimetic peptides (CMPs) or
collagen-like peptides (CLPs) developed as alternatives to animal
collagens have the benefit of low heterogeneity. Also, unlike xeno-
geneic collagens [6], there is little/no risk of allergy to xenogeneic
protein or zoonotic disease transfer.
Collagen was initially considered a protein that is unique to
multicellular animals, as hydroxyproline residues within collagen
169
170 Elle Edin et al.
Fig. 1 Before and after photos of patients who had been grafted with recombi-
nant human collagen-based implants to treat ulcers and scarring due to infection
or burns. These patients showed stable integration of the implants and regen-
erated neo-corneas after an average of 2 years post-surgery. Modified from
Fig. 2, Islam et al. [4]
have been considered the main determinants for structural stability

[7]. However, CLPs have since been identified in prokaryotes, such
as bacteria [8, 9]. These proteins have been isolated from biofilm,
and they have been shown to also have triple helical structures and
similar thermal stability to mammalian collagens [10]. As such,
researchers have been able to design new CLPs that are based on
bacterial collagen sequences and analyzed the structure-mechanical
property relationships between these fibrils [8, 11, 12].
Collagens are chains of G-X-Y amino acid motifs, where the G
amino acid (Gaa) is glycine [12]. The amino acid at the X (Xaa)
position is frequently proline, and in animal collagens, the Y amino
acid (Yaa) is often hydroxyproline. Important features that are
known to affect collagen assembly, stability, as well as the related
melting temperature include the Grand Average of Hydropathicity
(GRAVY) score, hydroxyproline spacing, and the frequency of the
six amino acid sequence Xaa1Yaa1Gaa1Xaa2Yaa2Gaa2 where the
Yaa1 position hosts a lysine and the Xaa2 is occupied by a negatively
charged residue (glutamic, or aspartic acid) [12–15].
In longer collagen peptides (>50 amino acids), assembly
regions are often necessary for collagen fibril formation
[16, 17]. In small CLPs that have high inter-strand interactions,
assembly regions are not needed. Therefore, when selecting or
designing a CLP, the experimenter needs to consider the availability
of functional groups that can be used to stabilize the collagen helix,
as well as stabilizing inter-fibrillar interactions.
1.2 CLP Production Solid state synthesis is the method of choice for shorter CLPs
(<40–50 amino acids). However, longer peptides have been pro-
duced using a combination of solid phase peptide synthesis, poly-
merization, and self-assembly [18]. The final products of such a
Synthesis and Application of Collagens for Assembling a Corneal Implant 171
combination strategy are triple helical nanofibers of 10–20 nm.

The Hartgerink group used an N-terminal cysteine and
C-terminal thioester to achieve selective head to tail polymerization
of peptides under aqueous conditions, without the need for pro-
tecting groups [18]. However, this method is not cost-effective for
producing longer polypeptides.
Recombinant DNA technology is more efficient and cost-
effective for production of full-length recombinant human col-
lagens. Full-length human and recombinant human collagens
have been produced in a range of transgenic species ranging from
yeast (Pichia pastoris for types 1 and III human collagen [19]) to
human fibroblasts [20], silkworms [21], and plants (tobacco for
type I human collagen) [22]. CLPs (in this case, gelatins) have also
been recombinantly produced, e.g., in silkworms [23].
1.3 CLP Protocols The protocols that we provide cover the production of recombi-
nantly produced CLP to fabrication of corneal shaped and sized
implants (Fig. 2). This protocol covers in particular CLPs that are
based on bacterial sequences or synthetic sequences that lack
hydroxyproline residues, as these are significantly less demanding
to produce. If a sequence is reliant on hydroxyprolines for stable
fibril formation, a system designed particularly for allowing this
type of post-translational modification must be used. For shorter
sequences, however, solid state synthesis is recommended and
hydroxyprolines can be incorporated more easily.
We have not provided any specific CLP sequence but instead
have provided a general protocol that can be used in its entirety or
in part for fabrication of implants, using a CLP that the reader has
access to. The specific cloning protocol given here is optimized for
CLPs that are between 20 and 50 kDa, with either an assembly
initiation region or which can fold in the absence of such a region.
For convenience, we have named the generic CLP we are preparing
“exColA”.
Fig. 2 Collagen-like peptide-based corneal implant manufactured according to

Subheading 3.6 of this chapter
For preparation of hydrogel implants, various crosslinkers can

be used. Water-soluble carbodiimides such as 1-ethyl-3-
(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) and
its co-reactant, N-hydroxysuccinimide (NHS) can be used to stabi-
lize hydrogels. A protocol for making hydrogels for implantation
into mini-pig corneal models based on the CLP sequence from
O’Leary et al. [24] and EDC-NHS crosslinking can be found in
Islam et al. [25]. In Islam et al., the CLP was conjugated to an
8-arm PEG prior to crosslinking. Here, however, our exColA
peptide is crosslinked without prior conjugation to a polymeric
backbone. In Samarawickrama et al. [26], 4-(4,6-dimethoxy-
1,3,5-triazin-2-yl)-4-methyl-morpholinium chloride (DMTMM)
was used as a crosslinker. Unlike EDC-based crosslinking,
DMTMM does not require accurate pH control or pH shift during
the reaction to be effective [27].
2 Materials
Please note that examples of equipment given are those we have

used. They can be substituted with equipment or reagents from
other manufacturers or suppliers.
2.1 Synthetic DNA 1. pUC57-exColA: Commercially available DNA cloning vector

with recombinant CLP sequence, exColA. The exColA
sequence should be codon optimized. This vector contains a
Tobacco Etch Virus (TEV) endoprotease cleavage site after
His-tag (see Notes 1 and 2). As mentioned above, the name
“exColA” is simply a placeholder for the CLP selected by the
reader.
2.2 Preparation 1. E. coli cloning strain: 5-α cold shock competent E. coli.
of Cloning System 2. E. coli expression strain: ClearColi® BL21(DE3) Electrocom-
and Expression petent Cells (see Note 3).
System
3. Expression plasmid: pColdIII.
4. Ampicillin stock solution: 3 g of ampicillin in 30 mL distilled
water. The ampicillin solution is decanted into a 60 mL syringe
and filtered through a 0.22 μm syringe filter. The sterile filtered
solution is divided into 3 mL aliquots and stored at 20 C.
5. LB Miller broth: 100 g of powdered LB broth is weighed and
placed in a 6 L glass flask. The flask is filled to 4 L with distilled
water. The flask is covered with aluminum foil and secured with
autoclaving tape. The bottle is autoclaved for 20 min at 121 C
and then allowed to cool. When the broth has cooled to below
45 C, 4 mL of ampicillin stock is added to make a final
concentration of 100 μg/mL.
6. LB Miller agar plates: 5 g of LB broth is added to a 1 L bottle.

Two g of agar is added and the volume is made up to 200 mL
by adding distilled water. The powder is dissolved by heating in
a microwave oven at high intensity until boiling. The solution is
boiled by microwaving at ~30% intensity for an additional
5 min. When flask has cooled to below 55 C, 4 mL of ampicil-
lin stock is added to make a final concentration of 100 μg/mL.
The solution is poured into 35 mm petri dishes (see Note 4).
7. Culture tubes: 100 mm size, translucent microbiology poly-
propylene tubes.
8. Shaking incubator: A closed, temperature-controlled, shaking
incubator capable of maintaining 250 RPM and 37 C.
9. Incubator: A closed, temperature-controlled, incubator capa-
ble of maintaining 37 C.
10. Agar gel electrophoresis system: Horizontal electrophoresis
system with combs to make 30 μL wells.
11. 1% Agar gels prepared according to external protocol [28].
12. Tris, Acetic acid, EDTA (TAE) buffer: Ready-made TAE buffer
concentrate is diluted 1:50 in water.
13. Loading buffer.
14. Invitrogen™ SYBR™ Safe™ DNA Gel Stain.
15. Blue light table: Light source capable of exciting at ~470 nm.
16. Gel imaging box: Imaging system with light source capable of
exciting at ~470 nm.
17. Freezing solution: 10 mL glycerol solution and 10 mL water.
18. Miniprep kits.
19. Vacuum manifold.
20. Centrifuge.
2.3 Expression 1. Culture flasks.

of CLP 2. Shaking incubator.
3. Magnetic stirrer.
4. Thermometer.
5. LB Miller broth (see Subheading 2.2, item 5).
6. 1 M Isopropyl β-D-1-thiogalactopyranoside (IPTG): 2.38 g
IPTG in 10 mL of water. The solution is filtered using a
0.22 μM syringe filter and stored at 20 C.
7. Lysis buffer: 690 mM NaCl, 13.5 mM KCl, 50 mM Na2HPO4,
9 mM KH2PO4, 20 mM imidazole, 5 M urea, 5 g/L Triton
X100, 10% v/v glycerol.
8. 1 mM phenylmethanesulfonyl fluoride (PMSF) in ethanol.
9. Probe sonicator.
10. Ultracentrifuge.
11. SDS-PAGE system.
12. SDS-PAGE Precast Protein gels (10%). Can be substituted
with polyacrylamide gels cast by the reader for protein
electrophoresis.
13. SDS-PAGE running buffer: 10 Tris/Glycine/SDS, diluted
1:10 in water.
14. Laemmli buffer.
15. Coomassie solution.
21. Gel imaging box: Imaging system with white light
illumination.
2.4 FPLC Purification 1. FPLC system.

of CLP 2. FPLC columns.
3. Loading/washing buffer: PBS powder, pH 7.4, 5 M urea,
20 mM imidazole to make 1 L.
4. Elution buffer: PBS powder, pH 7.4, 5 M urea, 500 mM
imidazole to make 1 L.
5. 20% ethanol: Prepared from sterile water and 100% ethanol.
Filter with 0.20 μm bottle filters.
6. TEV reaction buffer: 500 mM NaCl, 20 mM Tris, pH 7.5 in
water.
7. Bradford protein assay kit. Other protein assays may be
substituted.
2.5 Preparation 1. Dialysis tubing (see Note 5).

of Lyophilized CLP 2. Dialysis buffer container (see Note 6).
3. Urea in water: 4 M, 3 M, 2 M, and 1 M solutions of urea in
ultrapure water is prepared and sterile filtered using a 0.2 μm
bottle filter.
4. Liquid nitrogen.
5. Lyophilizer.
2.6 Preparation The apparatus used for molding is identical to that used for making
of Corneal-Shaped collagen-based implants and can be found in Islam et al. [29].
Implants
1. Pre-weighed 4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methyl-
morpholinium chloride (DMTMM) powder aliquots of
50–100 mg. Store at 20 C.
2. T-junction connector and fittings.
3. Glass syringes: 2 mL, Luer lock.
4. Rubber septum: 2 mm thickness.
5. Biopsy punch: 4 mm diameter.
6. 500 μL, 100 μL and 50 μL Hamilton syringes.

7. Needle (22 s/51/2)L for 500 μL syringe.
8. Needle (22 s/51/2)S for 100 μL and 50 μL syringes.
9. Wide glass dish.
10. Metal jigs: stainless steel jigs machined to tightly hold PTFE
corneal molds.
11. Polypropylene corneal implant molds: 500 μm thick, 10 mm
diameter, x0 curvature (custom-made).
12. Allen key.
13. Hydrated chamber (a pipette tip box can be used).
3 Methods
3.1 Preparation 1. Transform 5-α cold shock competent E. coli with pUC57-
of Cloning System exColA according to the bacterial strain supplier protocol.
2. Pick at least 10 colonies by touching the colony with a 10 μm
sterile pipette tip and place the pipette tip in 6 mL of Miller
media in a bacterial culture tube.
3. Culture the 5-α E. coli clones for 24 h at 37 C at 250 RPM;
cultures can be taken to the next step after the optical density
(OD) reaches 0.6 or higher.
4. Freeze stocks of 5-α pUC57-exColA. 500 μL of bacterial cul-
ture is mixed with 500 μL of freezing solution in a 2 mL
cryotube and placed in 80 C.
5. Clone exColA from pUC57 to pColdIII according to any
conventional cloning protocol [30, 31].
6. Transform 5-α E. coli with pColdIII-exColA according to the
protocol supplier with the bacterial strain.
7. Pick at least 10 colonies by touching the colony with a 10 μm
sterile pipette tip and place the pipette tip in 6 mL of Miller
media within a bacterial culture tube.
8. Culture the 5-α E. coli pColdIII-exColA clones for 24 h at
37 C at 250 RPM; cultures can be taken to the next step
after OD reaches 0.6 or higher.
9. Freeze stocks of 5-α pUC57-exColA. 500 μL of bacterial cul-
ture is mixed with 500 μL of freezing solution in a 2 mL
cryotube and placed in 80 C.
10. Perform Miniprep plasmid isolation according to manufacturer
protocol on the 10 clones.
11. Digest the 10 clones with restriction enzymes, or run analytical
PCR. Use a different set of enzymes than was used for the
cloning in step 5.
12. Run agarose gels according to external protocol (see Notes 7

and 8) [28].
13. Image and validate the presence of expected bands using a gel
imaging system.
14. Discard any clones that do not cut or amplify in expected
manner.
3.2 Preparation 1. Transform ClearColi with pColdIII-exColA according to sup-

of CLP Expression plier protocol.
System 2. Pick at least 10 colonies by touching each colony with a 10 μm
sterile pipette tip and then placing the pipette tip in 6 mL of
Miller media in a bacterial culture tube.
3. Culture the ClearColi clones for 24–48 h at 37 C at 250 RPM;
cultures can be taken to the next step after OD reaches 0.6 or
higher.
4. Freeze stocks of ClearColi pColdIII-exColA. To freeze, 500 μL
of bacterial culture is mixed with 500 μL of freezing solution in
a 2 mL cryotube and placed in a 80 C freezer.
5. Perform a diagnostic restriction digestion according to external
protocol [32].
6. Run agarose gels [28].
7. Image and validate the presence of expected bands.
8. Discard any clone that does not show the appropriate restric-
tion digested bands or amplifies in expected manner.
9. Optional: perform sequencing of plasmid.
3.3 Expression 1. ClearColi pColdIII-exColA is scraped on 10 μm pipette tip and

of CLP placed in 40 mL of culture media.
2. Culture is maintained at 37 C until OD > 0.5.
3. 40 mL of bacterial culture is added to 3 L of media in baffled
culture flask.
4. Culture is maintained at 37 C at 250 RPM until an OD of 1 is
reached (see Note 9).
5. IPTG is added to culture to a final concentration of 1 mM.
6. Culture is brought to 5–8 C by placing in ice bath on a stir-
plate at 250 RPM. Monitor temperature with an analog
thermometer.
7. Culture is placed at 5–16 C at 250 RPM for 24 h.
8. Culture is divided into 1 L centrifugation flasks and pelleted at
4 C, 4000 rcf.
9. Supernatant is discarded.
10. Pellet is dissolved in lysis buffer: 5 the pellet volume of buffer
is used.
11. Bacteria is lysed with sonication: 60% amplitude; 1 s on, 200 ms

off; 4 2 min with 10 min of cooling on ice between each cycle.
12. Solution is centrifuged at 15 kRCF at 4 C for 30 min.
13. Supernatant is decanted and combined in a sterile glass flask.
14. Pellets are resuspended in 5 mL of lysis buffer with fresh PMSF.
15. Suspension is sonicated using the same parameters as above.
16. Perform SDS-PAGE of the supernatant from step 13 and
solution from step 14 according to external protocol [33].
17. If either of solutions from step 13 or 14 is devoid of target
protein, that fraction is discarded.
18. Pool fractions that contain significant amounts of exColA.
3.4 FPLC Purification 1. The HisPrep™ FF 16/10 column containing Ni-loaded

of CLP sepharose is attached to FPLC system (see Note 10).
2. System cleaning is performed according to FPLC system man-
ufacturer handbook.
3. Column is equilibrated with 5 column volumes (CVs) of run-
ning buffer (see Note 11).
4. Sample from step 3.3.18 is loaded by direct injection at 3 mL/
min flow (see Note 12).
5. Wash the sample flow path with 1 CV of washing buffer.
6. Column with bound sample is washed with 10 CVs of washing
buffer (see Note 13).
7. A linear gradient from washing buffer to elusion buffer is
performed over 10 CVs. Fraction collection is performed
over the whole span of the gradient, and one additional
CV. Fraction volume is set to 20 mL (see Note 14).
8. FPLC system and column is washed according to manufacturer
instructions.
9. Chromatogram is used to identify which fractions contain
eluted protein. The peak is expected in the range of
100–400 mM of imidazole. Peak base width is expected to be
no wider than 30 mL (see Note 15).
10. Dialysis tubing should be cut to the correct length for the
sample volume. The length can be calculated using the
formula:
Tubing Length ¼ ðð4 þ ððð376:992 Sample VolumeÞ=Tubing Flat WidthÞ=
Tubing Flat WidthÞ 100Þ þ 0:5Þ=100
It is recommended to add 5–10 cm of tubing to this
measurement to have space to attach the dialysis tubing clamps.
11. The dialysis tubing should be pre-equilibrated in the buffer to

remove the residual preservatives from the tubing before use.
12. Clamp the bottom end of the dialysis tube by folding over the
end of the tubing and attaching the weighted clamp over the
folded edge.
13. Carefully pipette the solution into the tubing.
14. Open the tubing fully at the top end to allow air into the tubing
above the sample. Fold the top end of the tubing over and
clamp at the top ensuring that the air bubble remains between
the clamp and the sample. The air is critical for the buoyancy of
the sample during dialysis.
15. Fill the buffer container 80% with buffer and an appropriately
sized stir bar. Add the sample and fill to 100% of the volume to
prevent spillover.
16. Dialysis is performed by stepwise dialysis against urea solution.
The sample is dialyzed against a total of 50 dialysis volumes of
urea solution for a total of 12H per buffer step. 4 M, 3 M, 2 M,
and 1 M is used and finally the solution is dialyzed against
100 dialysis volumes of TEV reaction buffer (see Notes 16
and 17).
17. The protein content is quantified using a Bradford assay
according to the manufacturer instructions.
18. The sample is digested with TEV protease. 0.25 mg of TEV
protease is used per mg of protein.
19. FPLC column is equilibrated with 5 CV of running buffer.
20. The sample is loaded on the column. Due to the lack of His-tag
after digestion, only His-rich bacterial proteins should bind to
the column while the protein of interest should run straight
through. The flow-through is run to outlet and collected in a
sterile glass bottle. Do not run to waste (see Note 18).
21. Wash buffer is run for 1 CV and collected in the same vessel as
sample load flow through.
22. Bound protein is eluted using 3 CV of elution buffer. Flow-
through goes to waste (see Note 19).
3.5 Preparation The protein flow-through should be dialyzed again from FPLC
of Lyophilized CLP buffer until it is in ultrapure water.
1. Dialyzed exColA is transferred into 50 mL liquid nitrogen-safe
tubes.
2. Sample tubes are frozen in liquid nitrogen for 10 min (see Note
20).
3. Sample tubes are opened slightly to allow air flow and placed in
lyophilizer flasks.
4. The lyophilizer system is closed, and cycle is started.
3.6 Preparation 1. Remove the plunger from a sterile 10 mL syringe and wipe the
of Corneal-Shaped interior of the barrel with a particle-free wipe to remove the
Implants syringe’s coating as it can interfere with hydrogel formation.
Cap the syringe using a rubber cap held in place with parafilm.
2. Weigh the empty syringe and record the weight. Carefully
transfer the lyophilized exColA into the syringe and weigh
the assembly to determine the exColA mass.
3. Add ddH2O for a final concentration of 20% w/w. Cap the top
of the syringe with a rubber stopper and parafilm. Centrifuge
for 1 min at 200 rcf to ensure protein and water are in contact
with one another. Dissolving can be expedited by cycles of
heating to 37 C for 30 min followed by cooling on ice. Store
at 4 C (see Note 21).
4. Centrifuge dissolved exColA at 1000 RCF at 4 C for 1 h;
repeat until solution is free of visible bubbles.
5. Transfer 0.7 g of exColA solution from plastic syringe to a glass
syringe using a 2 mm inner diameter PTFE tube to connect the
two syringes. Ensure that no bubbles are produced during the
transfer.
6. Prepare water bath in large glass beaker using ultrapure water
(dd-water).
7. Fill syringe mixing system with dd-water and violently expel
any trapped air bubbles into a water bath. Eject all water from
the attached syringe and keep the mixing system submerged.
Attach the glass syringe containing collagen to the empty Luer
adapter on the mixing system, take care not to introduce bub-
bles. Place assembled mixing system on ice (see Note 22).
8. Dissolve DMTMM to 20% w/w in H2O. Sterile filter through
a 0.2 μm syringe filter.
9. Inject dissolved DMTMM through the septum of the mixing
system; use a volume equivalent to 0.7 times the molar amount
of primary amines in exColA (see Note 23).
10. Mix the solution by alternating pressing the two plungers of
the mixing system; pass the solution through the central t-piece
40 times to ensure sufficient homogeneity.
11. Eject 150 μL of exColA/DMTMM solution to each cornea
mold. Assemble molds and jigs, expelling any surplus
CLP/DMTMM solution around the edges of the molds.
Place in a hydrated chamber overnight at room temperature.
12. Open the jigs and place the mold assemblies in PBS overnight
at 4 C.
13. Carefully pry the jigs open and incubate the open molds over-
night in PBS at 4 C.
14. Gently lift the corneal implants out of the molds once fully
hydrated. Wash in PBS at 4 C for 7 days, changing the buffer daily.
4 Notes
1. When an appropriate CLP has been designed or selected,

codon optimization for the desired expression system must be
performed. This service can be performed by commercial enti-
ties such as GenScript. We suggest incorporation of an
N-terminal 6x His tag and TEV endoprotease cutting site.
The design should take into account the nucleases that will be
used for future enzymatic cloning, ensuring that these are
avoided within the coding sequence.
2. Cleavage sites other than the TEV site can be used. TEV
endoprotease was chosen due to high specificity and due to
the rarity of the motif.
3. Endonuclease deficient bacterial strains are required for long-
term stability of cloning strains. This protocol uses a cold shock
competent bacterial strain. If different competency bacteria is
used, the reader should follow the protocols supplied together
with that cloning strain for transformation. We use ClearColi®,
an E. coli strain that was modified to have diminished or
non-existent activation of LPS response in mammalian cells.
A distinction should be made between endotoxin pathway
activation and lack of immunogenicity. LPS is not the only
bacterial constituent that can trigger inflammation or
rejection [34].
4. Microwaving will sterilize the solution sufficiently for no spon-
taneous growth to occur for several weeks on properly stored
agar plates, even in the absence of antibiotics.
5. Dialysis tubing made from nitrocellulose can generally be used.
Mw cutoff needs to be based on target protein Mw. Dialysis
tubing should be purchased with a pore size that is at least
5 kDa smaller than a single subunit of your protein of interest.
The dialysis tubing width should be chosen so that a standard
batch results in tubing that is the correct length for your
dialysis chamber. If necessary, use two shorter lengths of dialy-
sis tubing for the sample to fit the beaker, so it floats free and
unencumbered within the chamber.
6. The dialysis buffer container should be sufficiently large to hold
a minimum of 50 the sample volume. A large graduated
cylinder may be ideal as it allows for longer lengths of dialysis
tubing for large samples than a large beaker.
7. Optional: at this point sequencing can be used in place of
restriction digestion to ensure that no mutations have been
introduced into the sequence.
8. We recommend use of SYBR Safe in place of toxic and muta-
genic ethidium bromide.
9. Oxygenation is of critical importance during the initial growth

of the bacterial colony. The culture conditions outlined here
assumes a large aerated space in the shaking incubator, or a
ventilated/actively oxygenated incubator.
10. Binding capacity is heavily dependent upon the geometry and
surface chemistry of the electrophoresis matrix. If a different
matrix is used, the reader must reference manufacturer instruc-
tions and adjust bed volume to facilitate a sufficient protein
binding capacity.
11. Depending on the predisposition of the protein used, a higher
ionic strength might be needed for wash/bind buffer and
elution buffer. If the protein is noted to form insoluble parti-
cles when exposed to the wash/bind buffer, 5 PBS can be
used instead of 1.
12. Ensure that the column manufacturer-specified max delta pres-
sure is not exceeded in this step. If column becomes visibly
compressed or the delta pressure exceeds manufacturer recom-
mendations, a lower pump speed should be used.
13. If this is a routine run, the flow-through in this step can be run
to waste. If this is an early optimization run, the sample should
be collected by running it to an outlet valve with a clean
collection vessel.
14. Average peak base width using HisPrep™ FF 16/10 columns
is 1 CV, which is why 20 mL fractions are used.
15. If yields are poor the imidazole absorption at 280 nm can make
resolving the exact range of target elusion difficult; in these
cases, a “dry run” with all the same parameters but without
protein in the loading buffer can supply a baseline that can be
deducted from the absorbance chromatogram.
16. The first dialysis stage should be no more than 3 h before the
buffer is changed.
17. The stepwise dialysis is necessary to avoid protein falling out of
solution.
18. This flow-through contains your target protein. Only bacterial
His rich proteins will bind the column.
19. This step is a cleaning step, removing His rich bacterial pro-
teins, TEV protease, and digested His-tags from the column.
20. Samples should be frozen and lyophilized according to the
instructions provided by the lyophilizer manufacturer. These
instructions are based on protocols for most research
lyophilizers.
21. If the protein cannot be kept soluble in water at room temper-
ature, other buffers can be used. Note that many collagens do
not lyophilize well in phosphate-based buffers; PB and PBS
should be avoided in this step.
22. Depending on the size and inter-strand interaction strength,

the viscosity of different CLPs at any given concentration will
vary. A mechanical syringe mixer can facilitate mixing of solu-
tions that are not possible to safely mix by hand.
23. DMTMM has a MW of 276.72 Da. The volume of exColA is
0.7 mL. The concentration of exColA is 0.2 g/mL. The con-
centration of DMTMM is 0.2 g/mL. To calculate the volume
to inject use the formula:
V DMTMM ¼ ð0:7 ð0:2=MW exColA Þ ðNumber of primary amines per exColAÞ MW DMTMM Þ=0:2
References
1. Montes GS, Junqueira LC (1982) Biology of 9. Peng YY, Yoshizumi A, Danon SJ, Glattauer V,
collagen. Rev Can Biol Exp 41(2):143–156 Prokopenko O, Mirochnitchenko O, Yu Z,
2. Brodsky B, Persikov AV (2005) Molecular Inouye M, Werkmeister JA, Brodsky B, Ram-
structure of the collagen triple helix. In: shaw JA (2010) A streptococcus pyogenes
Advances in protein chemistry, vol 70. Aca- derived collagen-like protein as a
demic Press, Cambridge, pp 301–339 non-cytotoxic and non-immunogenic cross-
3. Kadler KE, Holmes DF, Trotter JA, Chapman linkable biomaterial. Biomaterials 31
JA (1996) Collagen fibril formation. Biochem J (10):2755–2761. https://doi.org/10.1016/j.
316(1):1. https://doi.org/10.1042/ biomaterials.2009.12.040
bj3160001 10. An B, Kaplan DL, Brodsky B (2014) Engi-
4. Islam MM, Buznyk O, Reddy JC, neered recombinant bacterial collagen as an
Pasyechnikova N, Alarcon EI, Hayes S, alternative collagen-based biomaterial for tis-
Lewis P, Fagerholm P, He C, Iakymenko S, sue engineering. Front Chem 2(40). https://
Liu W, Meek KM, Sangwan VS, Griffith M doi.org/10.3389/fchem.2014.00040
(2018) Biomaterials-enabled cornea regenera- 11. Cosgriff-Hernandez E, Hahn MS, Russell B,
tion in patients at high risk for rejection of Wilems T, Munoz-Pinto D, Browning MB,
donor tissue transplantation. NPJ Regen Med Rivera J, Hook M (2010) Bioactive hydrogels
3:2. https://doi.org/10.1038/s41536-017- based on designer collagens. Acta Biomater 6
0038-8 (10):3969–3977. https://doi.org/10.1016/j.
5. Zeugolis DI, Paul RG, Attenburrow G (2008) actbio.2010.05.002
Factors influencing the properties of reconsti- 12. Xu C, Yu Z, Inouye M, Brodsky B, Miroch-
tuted collagen fibers prior to self-assembly: ani- nitchenko O (2010) Expanding the family of
mal species and collagen extraction method. J collagen proteins: recombinant bacterial col-
Biomed Mater Res A 86A(4):892–904. lagens of varying composition form triple-
https://doi.org/10.1002/jbm.a.31694 helices of similar stability. Biomacromolecules
6. Lynn AK, Yannas IV, Bonfield W (2004) Anti- 11(2):348–356. https://doi.org/10.1021/
genicity and immunogenicity of collagen. J bm900894b
Biomed Mater Res B Appl Biomater 71B 13. Shoulders MD, Raines RT (2009) Collagen
(2):343–354. https://doi.org/10.1002/jbm. structure and stability. Annu Rev Biochem
b.30096 78:929–958. https://doi.org/10.1146/
7. Jenkins CL, Bretscher LE, Guzei IA, Raines RT annurev.biochem.77.032207.120833
(2003) Effect of 3-hydroxyproline residues on 14. Fallas JA, Gauba V, Hartgerink JD (2009)
collagen stability. J Am Chem Soc 125 Solution structure of an abc collagen heterotri-
(21):6422–6427. https://doi.org/10.1021/ mer reveals a single-register helix stabilized by
ja034015j electrostatic interactions. J Biol Chem 284
8. Yu Z, An B, Ramshaw JAM, Brodsky B (2014) (39):26851–26859. https://doi.org/10.
Bacterial collagen-like proteins that form 1074/jbc.M109.014753
triple-helical structures. J Struct Biol 186 15. Nassa M, Anand P, Jain A, Chhabra A,
(3):451–461. https://doi.org/10.1016/j.jsb. Jaiswal A, Malhotra U, Rani V (2012) Analysis
2014.01.003 of human collagen sequences. Bioinformation
8(1):26–33. https://doi.org/10.6026/ 25. Islam MM, Ravichandran R, Olsen D, Ljungg-

97320630008026 ren MK, Fagerholm P, Lee CJ, Griffith M, Pho-
16. Boudko SP, Engel J, B€achinger HP (2012) The pase J (2016) Self-assembled collagen-like-
crucial role of trimerization domains in colla- peptide implants as alternatives to human
gen folding. Int J Biochem Cell Biol 44 donor corneal transplantation. RSC Adv 6
(1):21–32. https://doi.org/10.1016/j.biocel. (61):55745–55749. https://doi.org/10.
2011.09.009 1039/C6RA08895C
17. Sundaramoorthy M, Meiyappan M, Todd P, 26. Samarawickrama C, Samanta A, Liszka A,
Hudson BG (2002) Crystal structure of nc1 Fagerholm P, Buznyk O, Griffith M, Allan B
domains. Structural basis for type iv collagen (2018) Collagen-based fillers as alternatives to
assembly in basement membranes. J Biol Chem cyanoacrylate glue for the sealing of large cor-
277(34):31142–31153. https://doi.org/10. neal perforations. Cornea 37(5):609–616.
1074/jbc.M201740200 https://doi.org/10.1097/ico.
18. Paramonov SE, Gauba V, Hartgerink JD 0000000000001459
(2005) Synthesis of collagen-like peptide poly- 27. D’Este M, Eglin D, Alini M (2014) A system-
mers by native chemical ligation. Macromole- atic analysis of dmtmm vs edc/nhs for ligation
cules 38(18):7555–7561. https://doi.org/10. of amines to hyaluronan in water. Carbohydr
1021/ma0514065 Polym 108:239–246. https://doi.org/10.
19. Olsen D, Yang C, Bodo M, Chang R, Leigh S, 1016/j.carbpol.2014.02.070
Baez J, Carmichael D, Perala M, Hamalainen 28. Voytas D (2000) Agarose gel electrophoresis.
ER, Jarvinen M, Polarek J (2003) Recombi- Curr Protoc Mol Biol 51(1):2.5A.1–2.5A.9.
nant collagen and gelatin for drug delivery. https://doi.org/10.1002/0471142727.
Adv Drug Deliv Rev 55(12):1547–1567 mb0205as51
20. Bagley D (2019) Declaration of source material 29. Islam MM, Griffith M, Merrett K (2013) Fab-
- vitricol. https://www.advancedbiomatrix. rication of a human recombinant collagen-
com/wp-content/uploads/2011/09/Declara based corneal substitute using carbodiimide
tion-of-Source-Material-9-3-15.pdf. Accessed chemistry. Methods Mol Biol 1014:157–164.
10 June 2019 https://doi.org/10.1007/978-1-62703-432-
21. Tomita M, Munetsuna H, Sato T, Adachi T, 6_10
Hino R, Hayashi M, Shimizu K, Nakamura N, 30. Plasmid cloning by restriction enzyme digest.
Tamura T, Yoshizato K (2003) Transgenic silk- addgene. https://www.addgene.org/
worms produce recombinant human type iii protocols/subcloning/. Accessed 10 June
procollagen in cocoons. Nat Biotechnol 21 2019
(1):52–56. https://doi.org/10.1038/nbt771 31. Struhl K (1991) Subcloning of DNA frag-
22. Shoseyov O, Posen Y, Grynspan F (2014) ments. Curr Protoc Mol Biol 13
Human collagen produced in plants: more (1):3.16.11–13.16.12. https://doi.org/10.
than just another molecule. Bioengineered 5 1002/0471142727.mb0316s13
(1):49–52. https://doi.org/10.4161/bioe. 32. Diagnostic restriction digest (2018) addgene.
26002 https://www.addgene.org/protocols/diagnos
23. Adachi T, Wang X, Murata T, Obara M, tic-digest/. Accessed 10 June 2019
Akutsu H, Machida M, Umezawa A, Tomita 33. Gallagher SR (2006) One-dimensional sds gel
M (2010) Production of a non-triple helical electrophoresis of proteins. Curr Protoc Mol
collagen alpha chain in transgenic silkworms Biol 75(1):10.12.11–10.12A.37. https://doi.
and its evaluation as a gelatin substitute for org/10.1002/0471142727.mb1002as75
cell culture. Biotechnol Bioeng 106 34. Mamat U, Wilke K, Bramhill D, Schromm AB,
(6):860–870. https://doi.org/10.1002/bit. Lindner B, Kohl TA, Corchero JL,
22752 Villaverde A, Schaffer L, Head SR,
24. O’Leary LER, Fallas JA, Bakota EL, Kang MK, Souvignier C, Meredith TC, Woodard RW
Hartgerink JD (2011) Multi-hierarchical self- (2015) Detoxifying escherichia coli for
assembly of a collagen mimetic peptide from endotoxin-free production of recombinant pro-
triple helix to nanofibre and hydrogel. Nat teins. Microb Cell Factories 14(1):57. https://
Chem 3:821. https://doi.org/10.1038/ doi.org/10.1186/s12934-015-0241-5
nchem.1123. https://www.nature.com/arti
cles/nchem.1123#supplementary-information
Chapter 13
Development and Validation of a 3D In Vitro Model to Study

the Chemotactic Behavior of Corneal Stromal Fibroblasts
Evrim Ceren Kabak, Julia Fernández-Pérez, and Mark Ahearne
Abstract
Chemotaxis plays a pivotal role in crucial biological phenomena including immune response, cancer
metastasis, and wound healing. Although many chemotaxis assays have been developed to better under-
stand these multicomplex biological mechanisms, most of them have serious limitations mainly due to the
poor representation of native three-dimensional (3D) microenvironment. Here, we describe a method to
develop and validate a novel 3D in vitro chemotaxis model to study the migration of corneal fibroblasts
through a stromal equivalent. A hydrogel was used that contained gelatin microspheres loaded with
platelet-derived growth factor-BB (PDGF-BB) in the inner section and corneal fibroblasts in the outer
section. The cell migration toward the chemical stimuli over time can be monitored via confocal micros-
copy. The development of this in vitro model can be used for both qualitative and quantitative examinations
of chemotaxis.
Key words Chemotaxis assays, Chemoattractant, Corneal wound healing, Tissue engineering, 3D
microenvironment, In vitro model, Corneal stromal fibroblasts, Growth factor delivery system, Plastic
compression
1 Introduction
Chemotaxis involves the movement of cells in response to chemical

stimuli. This phenomenon is involved in many physiological and
pathological processes including immune response, wound healing,
and cancer metastasis by allowing immune cells to reach infection
sites, facilitating tissue repair by supporting the migration of cells to
close a wound and enabling the cancer cells to relocate at different
sites in the body, respectively [1, 2]. The ability to study chemotaxis
is important in the development of novel therapeutics and treat-
ments. During corneal wound healing cells respond to chemical
signals produced by other cells resulting in chemotaxis [3, 4].
Corneal stromal cells residing in the stromal layer of cornea become
activated and migrate toward the wound site to remodel and repair
the damages tissue [5–7].
185
186 Evrim Ceren Kabak et al.
In order to study the chemotactic behavior of cells, many assays

have been developed. One of the widely used assays is a Boyden
chamber, which is based on the quantification of a cell movement
from one compartment to another through a porous filter mem-
brane toward the chemoattractant gradient [8]. Despite its robust
structure and ability to screen large numbers of cells, it has serious
limitations mainly due to poor representation of the 3D microen-
vironment and inability to maintain a well-defined chemoattractant
gradient [9]. Although microfluidic-based chemotaxis assays are
able to generate a well-defined chemoattractant gradient with an
accurate quantitative analysis capacity, they are unsuitable to be
routinely used in biological and biomedical labs, due to the expense
and complexity of the system, mainly during the setup and data
analysis processes [10].
In order to overcome the limitations in the current models, we
developed a method to fabricate a novel in vitro 3D chemotaxis
model based on utilizing tissue engineering and biomaterial fabri-
cation techniques, such as fabrication of a growth factor delivery
microspheres [11] and plastic compression [12]. The chemotaxis
platform developed by the given protocol below will provide a wide
range of applicability to a different type of cells and chemotactic
agents with distinct concentrations. Moreover, the versatile feature
of the model allows a variety of analyses including real-time imag-
ing, immunocytochemical and biochemical analyses to be per-
formed on the model.
2 Materials
2.1 Fabrication and 1. Deionized (DI) water.

Characterization of 2. Olive oil.
Gelatin Microspheres
3. Gelatin from bovine skin, type B.
2.1.1 Medium and 4. Acetone.
Solution
5. Tween 80.
6. Glutaraldehyde.
7. 3.73 mg/mL Glycine solution.
2.1.2 Plastic and Glass 1. Pasteur pipettes (3 mL).

2. Serological pipettes (2 mL, 5 mL, 10 mL, 25 mL, 50 mL).
3. 50 μm cell microsieve.
4. 70 μm cell strainer.
5. Syringe.
6. 18G needles.
7. 2-cm-long flea magnet.
Development and Validation of a 3D In Vitro Model to Study the Chemotactic. . . 187
8. Round-bottom flasks.
9. Rubber or cork bases.
10. Centrifuge tubes (5 mL, 15 mL, 30 mL, 50 mL).
11. Glass and plastic Petri dishes.
12. Glass and plastic beakers (50 mL, 100 mL, 250 mL, 500 mL).
2.1.3 Equipment 1. Pipettor.

2. Magnetic stirrer.
3. Laboratory ice machine.
4. Freeze dryer.
5. Vacuum oven.
6. Epi-fluorescence microscope.
7. Fume hood.
2.2 Cell Culture 1. Culture medium for cell growth: Low glucose Dulbecco’s
Modified Eagle Medium (DMEM), 10% (v/v) fetal bovine
2.2.1 Medium and
serum (FBS), 100 U/mL penicillin, 100 mg/mL streptomy-
Solution
cin, 250 ng/mL amphotericin B.
2. Culture medium for chemotaxis studies: DMEM/F12 (1:1),
50 μg/mL ascorbic acid, 1 insulin-transferrin-selenium solu-
tion (ITS).
3. Sterile phosphate buffered saline (PBS), pH: 7.4.
4. Trypsin-EDTA solution (0.25%).
5. Ethanol 70% (v/v) in water.
6. Trypan blue.
2.2.2 Plastic and Glass 1. Microcentrifuge tubes (0.5 mL, 1 mL, 1.5 mL).
2. Cryogenic vials (1 mL, 2 mL).
3. T75 and T175 angled neck cell culture flasks.
4. Micropipette tips (1–10 μL, 2–20 μL, 20–200 μL,
100–1000 μL).
5. Hemocytometer.
2.2.3 Equipment 1. Micropipettes (1–10 μL, 2–20 μL, 20–200 μL, 100–1000 μL).
2. Water bath.
3. Cell culture CO2 incubator.
4. Inverted phase contrast microscope.
5. Class II biosafety cabinet.
6. Refrigerated centrifuge.
7. Refrigerator (+4 C).
8. Deep freezer (20 C).

9. Ultralow temperature freezer (80 C).
10. Nitrogen tank.
2.3 Development of 1. Corneal fibroblasts.

3D Chemotaxis Model: 2. High concentration rat tail collagen type I in acetic acid.
The Outer and Inner
3. 1 N Sterile NaOH solution.
Ring Gel Model
4. 10 PBS.
2.3.1 Cells, Medium, and
5. Sterile deionized water (dH2O).
Solution
6. Trypsin-EDTA (0.25%).
7. 1 PBS, pH: 7.4
8. 100 μg/mL recombinant human platelet-derived growth
factor-BB (PDGF-BB).
2.3.2 Plastic and Glass 1. 24-Well plates.

2. RAFT™ 3D cell culture system.
3. Aid kit box.
2.3.3 Equipment 1. Tweezers.

2. Scissors.
3. Mini and microcentrifuge.
2.4 Validation of 3D 1. 2 μM Calcein acetoxymethyl (Calcein AM).

Chemotaxis Model 2. 4 μM Ethidium homodimer-1 (EthD-1).
2.4.1 Cell Viability 3. Confocal laser scanning microscope (CLSM).
Analysis 4. ImageJ software.
2.4.2 Enzyme-Linked 1. ABTS ELISA Buffer Kit.

Immunosorbent Assay 2. Human PDGF-BB Mini ABTS ELISA Development Kit.
3. Disposable reagent reservoir.
4. 8-Channel micropipette (10–100 μL, 30–300 μL).
2.4.3 Immuno- 1. Fixation solution: 4% paraformaldehyde (PFA) in PBS.

cytochemistry 2. F-Actin staining solution: Tetramethylrhodamine B isothiocy-
anate (TRITC)-conjugated phalloidin in PBS with a ratio of
1:2000.
3. DAPI nuclear staining solution: 40 ,6-diamidino-2-phenylin-
dole (DAPI) in PBS with a ratio of 1:2000.
2.4.4 Live Cell Imaging Live cell imaging solution: Reconstitute one 50 μg vial of CellTrace
Oregon Green in 20 μL DMSO.
3 Methods
3.1 Fabrication of Gelatin microspheres are fabricated by using water-in-oil emulsion

Gelatin Microspheres technique, crosslinked, freeze-dried, and sterilized by dehydrother-
mal (DHT) treatment (Fig. 1a).
1. Pour 100 mL olive oil into a round-bottom flask (labeled flask
1) with 2-cm-long flea magnet.
2. Place the flask on the rubber base in a large beaker surrounded
with warm water and allow the oil to heat up to 42 C (see Note
1) on the magnetic stirrer.
3. Weigh out 2 g of gelatin and pre-warm DI water in the 30 mL
tube while stirring on the magnetic stirrer (see Note 2).
4. Once the gelatin is fully dissolved, add the gelatin solution to
the olive oil dropwise approximately 1 cm from the center of
the flask.
5. Turn off the heat and continue to stir for 10 min.
6. Remove the warm water surrounding the flask and place the
flask into the beaker with enough ice to surround it, while
continuing to stir for 30 min.
7. Add 40 mL prechilled acetone on ice to the flask and continue
to add ice if it starts to melt.
8. Pour the contents of the round-bottom flask through a 50 μm
cell microsieve in a funnel into bottle marked waste.
9. Wash all the olive oil through the sieve using chilled acetone
into a waste bottle (see Note 3).
10. Once the microspheres (MS) have been fully washed, leave
them in a glass Petri dish for 5 min to be completely dried off
(see Note 4).
11. Prepare 0.1% (v/v) solution containing Tween 80 and glutar-
aldehyde in 30 mL tube surrounded with ice. Add the dry MS
into the solution and stir for 1 h.
12. Pour the solution through a new 50 μm cell microsieve into a
waste bottle and wash with approximately 100 mL DI water.
13. Add the MS into a prechilled glycine solution, prepared by
dissolving glycine in DI water to a concentration of
3.73 mg/mL in a round-bottom flask and stir the solution
on ice for 1 h.
14. Pour the solution through a new 50 μm cell microsieve into a
waste bottle and wash with approximately 100 mL DI water.
15. Store the MS at 4 C overnight in 40 mL glycine solution.
16. Pour through a new 50 μm cell microsieve and repeat steps 13
and 14 twice.
Fig. 1 Illustration of work scheme for the development of the 3D chemotaxis model. (a) Fabrication and
characterization of gelatin microspheres. (b) Fabrication and characterization of outer and inner ring gel model
17. Place the MS into a Petri dish with approximately 10 mL DI

water (see Note 5).
18. Freeze-dry the samples by initially freezing at 30 C for 1 h
and then freeze-drying under 200 mbar vacuum at 1 C
overnight (see Note 6).
19. Aliquot 6 mg freeze-dried MS into 1.5 mL labeled centrifuge
tubes.
20. Place tube into a vacuum oven and set the temperature and
pressure to 110 C and 50 mbar, respectively, and leave for
24 h.
21. Once the treatment procedure is complete, leave the samples
to cool.
22. Collect and store the samples in a cool, dry place (preferably at
4 C).
3.2 Fabrication of 1. Calculate the required volume of collagen using the stock
Collagen Hydrogels collagen concentration (SCC) and the required collagen con-
centration (RCC) (see Note 7).
2. Determine the volume of each reagent required for preparing
outer and inner collagen hydrogels based on the given
formulas:
Total volume
Volume of 10 PBS ¼
10
Total volume RCC
Volume of stock collagen ¼
SCC
Volume of NaOH ¼ Volume of stock collagen 0:023
Volume of dH2 O ¼ Total volume 10 PBS collagen
NaOH
3. Keep all reagents on ice to prevent the early polymerization.
3.3 Fabrication of 1. Remove the culture medium in the cell culture flask containing
Cell-Laden Hydrogel: corneal fibroblasts and wash cells with PBS.
Outer Gel 2. Add trypsin-EDTA to the flask and incubate for 5 min at 37 C
to detach the cells from the surface.
3. Add medium to halt the trypsinization and collect the cell
suspension into a tube.
4. Centrifuge the tube for 5 min to obtain a pellet of cells.
5. Remove the supernatant from the tube and resuspend in 2 mL
of medium.
6. Mix 10 μL of the cell suspension and 10 μL of trypan blue and
count the cells in each 1 mm2 quadrant of a hemocytometer.
Calculate the number of cells in the solution (see Notes 8).
7. Suspend cells in 10 PBS prior to the hydrogel to being

formed with a concentration of 15,000 cells per 90 μL
10 PBS.
8. Mix the 10 PBS and cells with the other collagen reagents
(NaOH, dH2O, and stock collagen).
3.4 Fabrication of 1. Add 15 μL PDGF-BB with 100 μg/mL stock concentration to

Microsphere the 1.5 mL tube with 6 mg microspheres for a final concentra-
Embedded Hydrogel: tion of 300 ng per 300 μL gel and store the loaded micro-
Inner Gel spheres at 4 C overnight.
2. Wash the microspheres with 1 mL PBS to remove unbound
PDGF-BB, centrifuge for 5 min, and remove the supernatant.
3. Once the loading step is completed, suspend the microspheres
in 150 μL, 10 PBS (see Note 9).
4. Add 10 PBS and microspheres into the other collagen
reagents.
3.5 3D Chemotaxis The outer and inner ring gel model is composed of two parts: outer
Model ring gel with the cells and inner ring gel with the microspheres
(Fig. 1b). The preparation of the model includes several steps that
should be carried out in a sterile container that allows the inverted
micropipette tips placed in a 24-well plate to fit when closed
(Fig. 2).
1. Place an inverted P1000 micro-pipette tip in the inner part of a
well of 24-well plate using sterile tweezers (Fig. 2a).
2. Add 900 μL cell-laden collagen solution into the region
between the micro-pipette tip and the well (Fig. 2b) and
place the aid box in the incubator at 37 C for 15 min.
3. Remove the micro-pipette tip gently from the well (Fig. 2c).
4. Pipette 300 μL microspheres embedded collagen solution
directly into the inner area of the well (Fig. 2d) and incubate
at 37 C for another 15 min to allow polymerization of the
inner collagen gel.
5. Apply a RAFT absorber to the top of the gel to plastically
compress the gels (see Note 10).
6. Add the culture medium to each well to initiate chemotaxis
studies.
3.6 Validation of 1. Mix 4 μm EthD-1 and 2 μm Calcein with PBS in a 15 mL

Model falcon tube and vortex for 3 min.
3.6.1 Cell Viability 2. Wash the gels with PBS and then add approximately 1 mL
Analysis freshly prepared Live/Dead solution to each well.
3. Incubate the samples for 1 h at 37 C.
Fig. 2 Fabrication steps of the outer and inner ring gel model. (a) Placement of an inverted micropipette tip. (b)
Addition of the collagen solution for the outer gel. (c) Removal of micropipette tip after the gel polymerization.
(d) Addition of collagen solution for the inner gel and (e) placement of RAFT™ absorber on the top of the gel
model
4. Wash the gels with PBS three times to remove unbound

reagents and examine using a CLSM.
5. Cell viability can be quantified by counting the number of
green (live) and red (dead) cells present.
3.6.2 ELISA 1. Once the gelatin microspheres are loaded with various concen-
tration of PDGF-BB, place the microspheres in culture
medium and maintain them in the incubator, at 37 C (see
Note 11).
2. Change the medium every 3–4 days, collect and store the
aspirated medium at 20 C.
3. Quantify the PDGF-BB released into the medium by utilizing a
suitable ELISA Kit (e.g., ELISA Kit including ABTS ELISA
Buffer Kit and Human PDGF-BB Mini ABTS ELISA Devel-
opment Kit).
3.6.3 Immunocyto- 1. Fix the prepared outer and inner ring gel model samples with
chemistry 4% PFA for 15 min at room temperature and wash the samples
with PBS to remove the PFA residuals. Keep the samples at
4 C.
2. Mix the TRITC-conjugated phalloidin and DAPI dyes with
PBS in a ratio of 1:2000.
Fig. 3 Confocal microscopy Z-stack image of the overall outer and inner ring gel
model. Phalloidin and DAPI staining display F-actin and nucleus of cells,
respectively, in the outer gel. Microspheres are embedded in the inner gel.
The green dashed line indicates the interface of the two gels (magnification:
10, scale bar: 500 μm)
3. Stain the samples with the two dyes for 45 min at room tem-
perature (see Note 12).
4. Examine the samples and acquire Z-stack images of the samples
using LSCM (Fig. 3).
3.6.4 Live Cell Imaging 1. Trypsinize the desired concentration of corneal fibroblasts and
collect the cells from the cell culture flask to the falcon tube (see
Note 13).
2. Prepare 2.5 mg/mL stock solution of CellTrace by reconstitut-
ing the content of one vial (50 μg) of CellTracein 20 μL of
DMSO in the meantime.
3. Add 1 μL of freshly made stock solution in DMSO to each mL
of cell suspension for a final working concentration of 2.5 μg/
mL, cover the tube with an aluminum foil, and incubate for
20 min at 37 C.
4. Centrifuge the suspension for 5 min, remove the supernatant,
and resuspend the pellet in the fresh medium for chemotaxis
studies.
5. Mix the cell suspension with 10 PBS to prepare the cell-laden
collagen hydrogels.
6. Once the outer and inner gels are prepared, examine the sam-
ples by imaging them using LSCM.
4 Notes
1. In order to reach the heat of 42 C, we recommend setting the

temperature control of magnetic stirrer to 100 C and moni-
toring the temperature of the oil. Once it reaches 42 C, switch
off the heat.
2. The gelatin solution should be stirred until all the gelatin
particles have been dissolved homogeneously. We recommend
trying to avoid producing bubbles while stirring. The tube can
be placed under warm water tap if it starts to get cell too
quickly. Once the gelatin is dissolved uniformly in water, the
large beaker with olive oil should be warmed again using a
magnetic stirrer.
3. Washing the olive oil residuals off the microspheres on the
50 μm cell microsieve (no.1) requires approximately
700–800 mL acetone.
4. Microspheres should turn a lighter color when they are dry.
5. Prior to freeze-drying, the size uniformity of the microspheres
should be inspected. A microscopy can be used to record
images of the microspheres placed in DI water and their size
distribution can then be quantified using appropriate software,
e.g., ImageJ.
6. In the case that the freeze-dryer is unavailable at the time that
the MS are ready, they can be stored in dH2O at 4 C. The
ramp speed should be set at 1 C per minute in the freeze-
drying procedure.
7. We recommend the required RCC of collagen hydrogels
should be 3.5 mg/mL and the SCC should be checked on
the collagen bottle.
8. In order to calculate the cell number, the following equation
should be utilized:

Total cell number ¼ Cell count 1 104 Volume
Dillution factor
9. In order for inner gel of the chemotaxis model to be differ-
entiated from the outer gel, one of the reagents of the collagen
solution—10 PBS—can be mixed with phenol before the
inner gel preparation.
10. Once you place the RAFT™ absorber on the top of the che-
motaxis models, we suggest leaving the 24-well plate for
15 min in the class II biosafety cabinet. During that time, a
controlled amount of liquid from hydrogels will be taken up by
the absorbers. Consequently, the concentration of hydrogels
will be increased, and the thickness of them will reach approxi-
mately 100 μm.
11. ELISA should be conducted for two purposes: (1) to obtain

the cumulative release curve of PDGF-BB with various con-
centrations of the growth factor, (2) to quantify the release of
PDGF-BBs into medium directly from collagen hydrogel.
12. We suggest using TRITC-conjugated phalloidin and DAPI
dyes to visualize F-actin and nucleus of cells, respectively.
13. The staining of cells with CellTrace should be conducted prior
to the preparation of cell-laden collagen hydrogels so that
when the cells are embedded in the gels, no additional staining
steps are required.
Acknowledgement
This research is supported by European Research Council starting

grant [EYEREGEN-637460].
References
1. Kolsch V, Charest PG, Firtel RA (2008) The controlled-rate freeze. Invest Ophthalmol Vis
regulation of cell motility and chemotaxis by Sci 26(9):1210–1215
phospholipid signaling. J Cell Sci 121 8. Cano PM, Vargas A, Lavoie JP (2016) A real-
(Pt 5):551–559. https://doi.org/10.1242/ time assay for neutrophil chemotaxis. BioTech-
jcs.023333 niques 60(5):245–251. https://doi.org/10.
2. Rees PA, Greaves NS, Baguneid M, Bayat A 2144/000114416
(2015) Chemokines in wound healing and as 9. Whitehead BC, Bezuidenhout D, Chokoza C,
potential therapeutic targets for reducing cuta- Davies NH, Goetsch KP (2016) Cast tube
neous scarring. Adv Wound Care 4 assay: a 3-d in vitro assay for visualization and
(11):687–703. https://doi.org/10.1089/ quantification of horizontal chemotaxis and
wound.2014.0568 cellular invasion. BioTechniques 61(2):66–72.
3. Jester JV, Petroll WM, Cavanagh HD (1999) https://doi.org/10.2144/000114442
Corneal stromal wound healing in refractive 10. Li J, Lin F (2011) Microfluidic devices for
surgery: the role of myofibroblasts. Prog studying chemotaxis and electrotaxis. Trends
Retin Eye Res 18(3):311–356 Cell Biol 21(8):489–497. https://doi.org/10.
4. Fernandez-Perez J, Ahearne M (2019) Influ- 1016/j.tcb.2011.05.002
ence of biochemical cues in human corneal 11. Ahearne M, Buckley CT, Kelly DJ (2011) A
stromal cell phenotype. Curr Eye Res 44 growth factor delivery system for chondrogenic
(2):135–146. https://doi.org/10.1080/ induction of infrapatellar fat pad-derived stem
02713683.2018.1536216 cells in fibrin hydrogels. Biotechnol Appl Bio-
5. Matsuda H, Smelser GK (1973) Electron chem 58(5):345–352. https://doi.org/10.
microscopy of corneal wound healing. Exp 1002/bab.45
Eye Res 16(6):427–442. https://doi.org/10. 12. Cheema U, Brown RA (2013) Rapid fabrica-
1016/0014-4835(73)90100-0 tion of living tissue models by collagen plastic
6. Kratz-Owens KL, Hageman GS, Schanzlin DJ compression: Understanding three-
(1992) An in-vivo technique for monitoring dimensional cell matrix repair in vitro. Adv
keratocyte migration following lamellar kerato- Wound Care (New Rochelle) 2(4):176–184.
plasty. Refract Corneal Surg 8(3):230–234 https://doi.org/10.1089/wound.2012.0392
7. Lee TJ, Wan WL, Kash RL, Kratz KL, Schan-
zlin DJ (1985) Keratocyte survival following a
Chapter 14
Femtosecond Laser-Assisted Surgery for Implantation

of Bioengineered Corneal Stroma to Promote Corneal
Regeneration
Neil Lagali and Mehrdad Rafat
Abstract
The femtosecond laser has achieved widespread use in ophthalmology owing to its ability to deliver focused
high energy that is rapidly dissipated and thereby does not damage surrounding tissue outside the precise
focal region. Extremely accurate and smooth cuts can be made by the laser, enabling a range of applications
in anterior segment surgery. Minimally invasive corneal surgical procedures can be performed using the
femtosecond laser, and here we describe the application of such procedures to improve implantation of
bioengineered materials into the cornea. Bioengineered corneal tissue, including the collagenous corneal
stroma, promises to provide a virtually unlimited supply of biocompatible tissue for treating multiple causes
of corneal blindness globally, thereby circumventing problems of donor tissue shortages and access to tissue
banking infrastructure. Optimal implantation of bioengineered materials, however, is required, in order to
facilitate postoperative wound healing for the maintenance of corneal transparency and avoidance of
postoperative complications such as scarring, inflammation, and neovascularization. Moreover, the avoid-
ance of a detrimental physiological physiological wound healing response is critical for facilitating the
corneal stromal regeneration enabled by the bioengineered stroma. Without proper implantation, the tissue
response will favor inflammation and pathologic processes instead of quiescent keratocyte migration and
new collagen production. Here we describe several procedures for optimized biomaterial implantation into
the corneal stroma, that facilitate rapid wound healing and regenerative restoration of corneal transparency
without the use of human donor tissue. A step-by-step methodology is provided for the use of the
femtosecond laser and associated techniques, to enable seamless integration of bioengineered materials
into the corneal stroma.
Key words Femtosecond laser, Cornea, Biomaterial, Corneal transplantation, Corneal blindness,
Artificial cornea
1 Introduction
Bioengineered corneas, either the full cornea, or epithelial, stromal,

and endothelial parts separately or in combination, have been the
subject of intense research due to the worldwide shortage of suit-
able human donor tissue for transplantation [1]. Bioengineered or
biosynthetic corneas, unlike keratoprostheses, are implanted
197
198 Neil Lagali and Mehrdad Rafat
scaffolds designed to support endogenous corneal regeneration

and repopulation of the implanted region with host cells and nerves
[2]. The implantation technique itself, however, can represent an
additional challenge for healing and regeneration, as the surgical
technique may trigger inflammation and neovascular response. For
example, in preclinical studies implanting a biomaterial in the alkali-
injured rabbit cornea [3], or in the rat cornea [4], the surgical
sutures themselves trigger inflammation and neovascularization.
Likewise, in a phase I human clinical trial conducted in Sweden,
recombinant human collagen III (RHCIII) corneal substitutes
implanted as deep anterior lamellar grafts were secured with over-
lying sutures [5]. The sutures delayed wound healing and caused
inflammation and stromal melting in some patients [6].
There is therefore a need to develop less invasive implantation
techniques for bioengineered tissue in the eye that minimize the
surgical trauma and facilitate faster healing of the surgical wound.
Using less invasive techniques, regeneration instead of
inflammation-driven wound healing would be favored within the
recipient tissue. Here, we focus on bioengineering of the corneal
stroma, and the means by which stromal replacements can be
surgically implanted into the cornea in a minimally invasive manner.
Intra-stromal surgery has been facilitated by the advent of the
femtosecond surgical ophthalmic laser [7], enabling precise inci-
sions to be made within the cornea, at a selectable depth. The
femtosecond laser technique is based on the principle of long-
wavelength, deep infrared light with a very short pulse duration
(in the hundreds of femtoseconds regime), and high repetition rate
(in kHz range). The light is focused to a very small spot size of a few
μm in diameter, forming a localized plasma in the tissue that creates
a cavitation bubble of carbon dioxide and water [8], which sepa-
rates the layers (lamellae) of the surrounding corneal stromal tissue,
as the bubble expands. By this method, a very smooth interface
between cut surfaces can be obtained. The femtosecond laser is
nowadays widely used in ophthalmic surgery, for laser in situ kerat-
omileusis (Femtosecond Laser-Assisted LASIK), implantation of
intracorneal ring segments, small incision lenticule extraction
(SMILE) procedures, penetrating/lamellar keratoplasty, and cata-
ract surgery [8]. We recently introduced the use of the femtosecond
laser for the intrastromal implantation (femtosecond laser-assisted
intrastromal keratoplasty; FLISK) of bioengineered corneal substi-
tutes [9] in order to avoid the need for surgical sutures. This
method can reduce suture-induced astigmatism in the transplant
and may avoid an excessive epithelial and stromal wound healing
response [6], thereby facilitating stromal regeneration following
biomaterial implantation [9]. The minimally invasive FLISK proce-
dure resulted in rapid wound healing due to retention of the
recipient epithelium and Bowman’s layer [10], without major dis-
ruption to the corneal subbasal nerve plexus and associated
Laser-Assisted Biomaterial Implantation 199
postoperative complications. The procedure avoids inflammation

and achieves a 100% implant retention rate without affecting the
corneal transparency and structure. A similar procedure is starting
to be adopted in a clinical setting for intra-stromal implantation of
donor refractive lenticules [11].
Here we detail the method of femtosecond laser surgical
implantation of laboratory-made biomaterials in the cornea. Such
bioengineered stromal materials could avoid the need for use of
human donor tissue and associated immune responses. This chap-
ter describes preclinical implantation in animal models; however,
the steps may be applicable (with appropriate modifications) to
human implantation.
2 Materials
1. Femtosecond laser. Here, we describe the use of the IntraLase

iFs 150 kHz 5th Generation femtosecond laser manufactured
by Abbot Medical Optics. Equivalent ophthalmic surgical fem-
tosecond lasers are available from other manufacturers; how-
ever, the procedures may require some adaptation from the
methods described here, in order to achieve the same effect.
Femtosecond laser placement in a temperature and humidity
controlled environment is required to ensure optimal operating
conditions for the laser. Although not strictly required for
animal studies, a clean room/sterile environment is recom-
mended to avoid the possible contamination by airborne par-
ticles that could become trapped within the implant and/or
cornea during the surgical procedures.
2. Single-use disposable patient interface module for the femto-
second laser (Fig. 1). This consists of a sterile “applanation
cone” with flat glass plate that is fitted into the laser unit. The
laser beam travels through the glass plate, which applanates the
cornea. The interface module also contains a sterile plastic ring
with suction that is designed to be placed on the cornea and
under the eyelids and is designed to hold the glass ring and
applanation cone in place during surgery (see Note 1).
3. Holder for eye globes. It may be desirable to practice the
proposed laser procedures ex vivo in whole porcine or bovine
eye globes, prior to performing the procedures in vivo. In this
case a suitable holder for eye globes is required, which can
stabilize the eye tissue and provide adequate outward pressure
to simulate the intraocular pressure. An apparatus used for this
purpose is shown in Fig. 2 below.
4. Eyelid speculum (blepharostat), for maintaining open eyelids
during the procedure.
Fig. 1 Patient interface module. This disposable single-use module consists of a suction ring connected to a
syringe to achieve vacuum suction on the eye, and an applanation cone. The right image is a detail of the
cone, where the glass plate is marked. The glass applanates the cornea by downward pressure, and also
transmits the laser beam
Fig. 2 Eye globe holder. The ex vivo eye is placed into the holder and the syringe is used to hold the eye in
place and ensure sufficient outward pressure from the eye globe
5. Topical anesthetic eye drops, to be instilled prior to blepharo-

stat placement, prior to the laser procedure, and immediately
postoperatively.
6. Surgical forceps for removing cut disc of tissue, handling and
inserting of biomaterial.
7. Blunt or flattened surgical tool for dissecting tissue or lifting a
flap of tissue after laser cutting.
8. Corneal button punch-type trephine for cutting the biomate-
rial into a circular button of a predetermined diameter.
9. Surgical marking pen for marking the center of the pupil prior
to laser alignment.
10. Postoperative analgesics, steroid eye drops, antibiotic eye

ointment.
11. Anterior segment optical coherence tomography (OCT)
equipment, to examine the cornea preoperatively and
postoperatively.
3 Methods
3.1 Preparation 1. Start the laser and allow a warmup period after a cold start,
of Laser usually 30–60 min.
2. After warm-up, the procedural details must be entered into the
laser system, including identification of patient and eye and
selection of type of procedure. It is important to specify the
desired diameter of the operation zone in the cornea, taking
into account the available diameter of biomaterial and the
diameter of the cornea in the animal model used. Some femto-
second laser systems may have a minimum cutoff value for the
diameter of the cutting region, such as 7 mm. If a smaller
diameter is required (due to a smaller eye, or for testing
implantation in only a specific zone of the cornea), then the
laser may need to be used in a “research mode,” which may
allow a more flexible choice of cutting diameter. Check the
laser manual and contact the laser supplier for more details.
3.2 Choice of Laser 1. Determine the laser procedures to be used for stromal implan-
Procedure(s) tation of bioengineered materials. Several options are available,
depending on the specific application to be tested. The femto-
second laser provides the flexibility to combine various types of
basic cuts to develop different surgical procedures:
(a) A curved, arc-shaped cut of selectable arc length and
depth is available to provide a limited-size incision for
tissue insertion and excision
(b) A full 360 circular cut of selectable depth and location
(anterior or posterior) is available for anterior, posterior,
or mid-stromal procedures
(c) A flap-cut of selectable flap depth, arc length, and hinge
position is available for the creation of an anterior corneal
flap
(d) A full circular lamellar cut of selectable depth and diameter
is available to separate different layers of the stromal tissue
2. If required, combine the above cuts for implantation of bioma-
terials of different sizes by different methods. As an example, in
Fig. 3, the procedure of intra-stromal keratoplasty [9] is illu-
strated. Here, an implant of 3 mm diameter and 150 μm thick-
ness is to be implanted in a rabbit cornea of thickness 370 μm.
Fig. 3 Femtosecond laser-enabled intra-stromal keratoplasty (FLISK), implemented in a rabbit cornea. A small
3 mm diameter biomaterial implant is inserted into an intra-stromal pocket created by three laser cuts
(2 lamellar, 1 circular side cut). The removal of native tissue and insertion of the biomaterial is achieved by
means of an access cut that is angled at 45 and provides access to the implant region. Note that no sutures
are required to maintain the implant in place in the stroma
Fig. 4 Hybrid flap—anterior lamellar keratoplasty procedure for intra-stromal implantation of larger diameter
and thicker implants, for replacing the majority of the corneal stroma (in cases of dystrophies or scars for
example)
The procedure involves cutting and removing a disc of native

tissue from the cornea and inserting a biomaterial with similar
dimensions into the created pocket. The laser cuts used are two
circular lamellar cuts (top and bottom interfaces), a 360 circu-
lar side cut, and a 60–90 arc-shaped side access cut, used to
extract the native tissue and insert the biomaterial. The access
cut is the only region where the epithelium is cut. Outside this
access cut, the epithelium remains intact.
3. Determine if a hybrid intra-stromal procedure is required, to
facilitate placement of larger and deeper stromal implants. An
example of this is given in Fig. 4, where a LASIK flap is
combined with an anterior lamellar keratoplasty procedure.
The laser cuts used are a flap cut, a circular side cut, and a
lamellar cut. To remove the native tissue, the flap is “peeled
back” (held in place only by the hinge) and the underlying
lamellar disc is removed. The biomaterial implant is then placed
into the lamellar bed, and the flap is replaced and fixed in place
using typically four superficial interrupted sutures around the
circumference of the flap.
3.3 Laser 1. Design the laser cutting procedure. When the laser is pro-
Programming grammed to perform a series of cuts in succession, it must be
kept in mind that the most posterior cuts will be made first,
followed by the more anterior cuts. This procedure of deepest-
to-shallowest cuts is important, as one must avoid focusing the
laser beam through a plane of the cornea that has already been
cut. This is because the interface and refractive index changes in
the cut plane can disrupt the laser beam. For a similar reason,
cutting through dense scar tissue or other opacities should be
avoided as these can disperse the laser beam and result in
unpredictable results.
2. If possible, program multiple laser procedures ahead of time,
with appropriate parameter values entered and saved into the
system prior to starting the procedure (see Note 2).
3. Prior to programming the laser, it is recommended to first
perform OCT imaging of the eye to be cut, to determine the
pachymetric thickness of the cornea within the proposed cut-
ting zone (see Note 3). Once the appropriate dimensions are
known, the parameters for the laser cuts can be entered. For
lamellar cuts, the desired lamellar depth can be entered. For
deep lamellar cuts, the depth should be OCT guided (see Note
3). In general, more anterior lamellar cuts can be associated
with a greater wound healing-fibroblast response and asso-
ciated corneal haze, compared to more posterior cuts.
4. The diameter of the lamellar cut should be specified, and
should exceed the desired diameter by 50 μm on each side
(100 μm total). For example, if a 7 mm diameter lamellar cut
is required, then specify a diameter of 7.1 mm. This is done to
ensure that the subsequent circular side cut will intersect the
lamellar cut, to aid in ease of tissue removal.
5. Specify the laser energy for the lamellar cut (0.6 is used in the
iFs 150 system, but will need to first be tested with ex vivo
tissue on other systems, to determine optimum energy for
smooth separation of tissue). A raster pattern of the laser spot
is chosen for cutting. Next, the laser spot separation is chosen
to be 3 μm.
6. Specify the 360 anterior side cut. The diameter should be
exactly the desired diameter (7 mm in the above example).
The depth of the side cut is specified by two parameters, the
posterior depth (at which the laser cutting starts) and the
“depth in glass” at which the laser cutting stops. “Depth in
glass” refers to the surface of the applanation cone glass plate in
Fig. 1, which is in contact with the cornea. The stop depth of
the side cut is referenced from this glass plane. If the anterior
side cut should go through the entire epithelium (for instance
in standard lamellar keratoplasty), then a “depth in glass” value
of 50 μm should be used (which means the laser will actually
cut 50 μm into the glass, ensuring that the full epithelium is
also cut). Otherwise, for purely intra-stromal procedures, a
negative value for “depth in glass” should be used, for example,
100 μm if the side cut should end 100 μm below the epithelial
surface. As in the case of the lamellar cut, the depth of the
anterior side cut should exceed the desired depth by 20 μm on
each side (top and bottom), to ensure overlap with the lamellar
cut. So if a stromal disc of 200 μm thickness is to be removed
(and equivalent disc of biomaterial to be inserted), then the
posterior depth should be 20 μm deeper than the posterior
lamellar cut depth, and 20 μm shallower than the anterior
lamellar cut depth. A higher energy (2.0) is used for the ante-
rior side cut, and the full 360 arc length should be specified to
obtain a circular/cylindrical cut. Usually for an applanated
cornea, the angle of the side cut should be 90 (vertical cut).
Laser spot separation is again 3 μm.
7. Specify the flap cut. A flap diameter should be chosen that is
larger than the biomaterial to be implanted, typically 1 mm
larger. This facilitates suturing of the flap without contact to
the implanted biomaterial, and a wound healing response that
is kept away from the biomaterial. Possible downgrowth of
epithelium under the flap is also minimized when the flap size
is larger than the implant size. Typical flap depth is 50–100 μm
depending on the corneal thickness. The hinge can be placed at
any location (nasal, temporal, superior, or inferior).
8. Specify the arc-shaped access cut. For this cut, an arc length of
60–90 can be used, to facilitate tissue excision and biomaterial
insertion; however, typically 60–70 is recommended if a
suture is to be avoided. With larger openings, one or two
sutures need to be placed to ensure the biomaterial does not
extrude through the opening. The arc length is also dependent
on how pliable and elastic the biomaterial is; with biomaterials
that can easily be folded or rolled, a smaller opening can be
used, while more brittle materials will risk tearing if folded and
thus require larger openings. The access cut angle is typically
45 , extending from the intra-stromal pocket and through the
epithelium (see Note 4).
3.4 Preparation 1. Using the corneal trephine punch for the desired diameter of
of Biomaterial implant, carefully cut the biomaterial button manually (see
for Implantation Note 5).
2. After trephination, if the biomaterial will not be directly
implanted into the recipient cornea, it is important to keep
the biomaterial immersed in PBS liquid prior to insertion into
the cornea, to ensure it does not dry out and maintains its shape
and optimal hydration state.
3.5 Laser Cut, Tissue 1. Mark the pupil center of the eye to be cut with a surgical
Excision, marking pen.
and Biomaterial 2. Apply several drops of topical anesthetic to the eye, and using a
Insertion sponge tip, absorb the excess fluid.
3. Place the eye under the laser and align with the applanation
cone (Fig. 5). Guidance LEDs or other laser features (such as a
camera/video screen) can be used to assist in manual align-
ment. Once aligned, proceed to dock the applanation cone to
the cornea, to achieve applanation on the eye. Make sure that
the glass part of the cone is still centered after applanation, or
else make small manual lateral adjustments with the joystick to
center the cone on the pupil.
Fig. 5 Applanation cone after insertion into the femtosecond laser. The laser
head is moved downward to make contact with the eye during the applanation
procedure
Fig. 6 Real-time monitoring of laser procedures using the video monitor. The left image indicates the posterior
lamellar cut forming the lower interface of the intrastromal pocket, while the right image was taken after
completion of the posterior lamellar and circular side cuts, and indicates the progress of the anterior lamellar
cut
4. Start the laser cutting program using the software interface.

When signaled by the software, press down the foot pedal and
hold, to perform the programmed cutting sequence. The cut-
ting can usually be followed in real time on the laser video
monitor (Fig. 6). Once the sequence has completed, release the
foot pedal.
5. Next, without moving the applanation cone, initiate any
subsequent cutting procedures through the software interface,
and complete these procedures (see Note 6).
6. Once the final laser procedure has been completed, raise the
applanation cone from the eye, to return the eye to a normal
curved shape. Remove and dispose of the applanation cone.
7. The operated eye should be brought under a surgical micro-
scope located in the vicinity of the femtosecond laser. Using the
surgical microscope and a blunt spatula-type surgical instru-
ment, the corneal tissue should be separated gently by drawing
the spatula across the cut lamellar surfaces. For a flap, the
spatula can be used to open the flap and expose the underlying
tissue. For an intra-stromal pocket, the spatula is used to enter
the pocket via the access cut, and verify separation of the upper
and lower lamellar interfaces.
8. Once the separation of stromal tissue has been achieved, for
intra-stromal procedures surgical forceps are used. The forceps
are inserted into the pocket via the arc-shaped angled access
cut, and the forceps are placed such that the stromal tissue
within the pocket is sandwiched between the forceps.
9. With a gentle pulling motion, the native stromal tissue button

is pulled out of the pocket and taken out of the cornea through
the access cut. The cut disc of native stromal tissue should be
excised in one piece. If, however, there is resistance or there are
residual attachments of the button to the surrounding stromal
tissue, use the blunt spatula to gently separate these attachment
points. Do not forcefully pull out the stromal button. The
optimal laser spot size, separation, and energy will ensure a
smooth separation of the stromal button from the surrounding
tissue.
10. In the reverse manner from the previous step, place the forceps
around the biomaterial button such that the biomaterial is
sandwiched between the forceps. The entire diameter of the
biomaterial should be placed between the forceps, with the
biomaterial ending at the tip of the forceps.
11. Next, carefully insert the forceps through the access cut and
into the stromal pocket. Proceed forwards into the pocket with
the forceps until the forceps reach the circular side cut of the
pocket opposite the access cut. Then, slowly release pressure on
the forceps and gently draw out the forceps from the pocket via
the access cut (see Note 7). An example of the tissue excision
and biomaterial insertion procedure in the rabbit eye is given in
Fig. 7.
12. For lamellar procedures such as the flap—lamellar hybrid
implantation, the blunt spatula is used to lift the flap to expose
the underlying stromal tissue. Next, sharp-tipped forceps are
used to gently separate the button from the underlying stroma.
The forceps are used to lift the button from the residual corneal
bed, leaving exposed a thin layer of posterior stroma, with
underlying Descemet’s membrane, and endothelium. The bio-
material is gently placed on this stromal bed, flattened with a
blunt spatula, and the flap is then drawn over the biomaterial,
and flattened to remove wrinkles. Surgical sutures are then
used to anchor the flap to the surrounding anterior stroma,
without contact with the biomaterial. Typically 10-0 nylon
sutures are used, and knots are buried in the stroma. Examples
of suturing patterns are shown in Fig. 8 below.
3.6 Final Surgical 1. Following biomaterial implantation and suturing, OCT imag-
Procedures ing (see Note 8) should be performed to verify the correct
positioning of the biomaterial within the recipient stroma
(Fig. 9). At this point, the native corneal stroma and biomate-
rial may appear swollen, yielding an appearance of a very thick
cornea. This is a temporary intraoperative phenomenon and
will subside within the first postoperative days.
Fig. 7 Procedure for excision of native tissue cut by the femtosecond laser, followed by implantation of the
bioengineered stromal implant. (a) Surgical forceps are used to gently pull the native stromal tissue out of the
intra-stromal pocket. (b) Native tissue (white arrow) seen after removal from the femtosecond pocket (black
arrow, points to access cut region). (c) Surgical forceps are used to grip the biomaterial, with the forceps
holding the biomaterial across the entire diameter, in a sandwiched configuration. (d) The rabbit eye showing
the stromal pocket (asterisk), native disc of stromal tissue (white arrow), and bioengineered stroma (black
arrow) of identical thickness and diameter (3 mm). (e) Insertion of the bioengineered implant into the stromal
pocket via the access cut. The forceps are drawn into the entire length of the pocket until the circular side cut
opposite the access cut is reached
Fig. 8 Suturing pattern after biomaterial implantation in porcine eyes. (a) An anterior lamellar implant, where
the biomaterial is held in place postoperatively by the use of overlying sutures. (b) An intra-stromal (FLISK)
implant where two surgical sutures (arrow) are used to close the access cut region
Fig. 9 Optical coherence tomography (OCT) images taken after biomaterial implantation in the porcine cornea.
(a) Native porcine cornea. (b) Bioengineered stromal tissue (arrow) immediately after implantation intra-
stromally using the FLISK procedure. Note the postoperative swelling. (c) Bioengineered stromal implant (long
arrow) as an anterior lamellar graft. Note the overlying sutures (short arrows)
2. Following OCT examination, animals should be given topical

ophthalmic antibiotic ointment and topical steroid eye drops.
If desired, a bandage contact lens can be placed over the
operated cornea (Fig. 10).
Fig. 10 OCT images of bandage contact lens (arrows) placed onto operated corneas, immediately postopera-
tively. The bandage lens helps protect the operated eye and facilitates smooth eyelid movement over the
implanted eyes in the postoperative period. The bandage lens facilitates epithelial healing and allows topical
eye drops to penetrate into the cornea
3.7 Postoperative 1. Postoperatively (and possibly even preoperatively—consult

Care and Follow-up with a veterinarian) animals should be given systemic analgesics
daily for at least the first few postoperative days.
2. Topical corticosteroids (e.g., dexamethasone) should be given
in the immediate postoperative period and should be main-
tained at least twice daily for a period determined by the specific
laser procedure and the general appearance of the eye postop-
eratively. For intra-stromal procedures where sutures are not
used, the course of topical steroids may last for about 1 week,
whereas anterior lamellar and flap procedures may require cor-
ticosteroids for 1 month after sutures are removed, typically
3–6 months postoperatively (see Note 9).
3. Minimize the use of general anesthesia during subsequent
longitudinal follow-up to preserve the well-being of the ani-
mals. Utilize restraining procedures for animals while awake,
where possible, to obtain ocular data.
4 Notes
1. Different laser systems will have differing methods of corneal

applanation and stabilization/alignment of the cornea to the
laser beam. It is suggested to follow the laser manufacturer’s
instructions to prepare the cornea prior to laser cutting. It is
also possible and may be desirable to modify this procedure for
animal experiments. For example, in large animals (rabbits and
pigs), omitting the suction ring may be advantageous, as
obtaining a vacuum may not be feasible in eyes of different
sizes and anatomies. Omitting the suction ring is not recom-
mended, however, in humans.
2. It may be advantageous to create several “hypothetical”
patients and save the laser parameters for these procedures
ahead of time. For example, if one will perform the hybrid
flap-lamellar keratoplasty procedure in Fig. 4, the laser config-
urations for the deep lamellar procedure and for the flap pro-
cedure should be saved separately. Then, during surgery, once
the deep lamellar procedure is completed, the flap procedure
can be quickly started while the eye is still under applanation.
Saving the laser parameters for the procedures ahead of time
will minimize the time between procedures. There is a risk that
the eye moves slightly between subsequent procedures if left
too long. Such movement, even of a few micrometers, will
bring the cuts out of alignment, resulting in suboptimal results.
3. OCT is essential for determining the corneal thickness prior to
laser surgery. This will minimize the risk of corneal perforation
with deeper stromal cuts, and result in biomaterial implantation
in the desired region of the cornea. With animal corneas, it is
important to keep in mind that these also have a variability in
corneal thickness. For example, rabbit corneas of the same litter
can vary from 320 μm to 420 μm in thickness, depending on
the age and individual. For cases where opacities are to be
removed, OCT will indicate the exact depth and extent of the
opacity, which will aid in planning the laser procedures and
parameters.
4. It is important to determine the orientation of the laser system
relative to the programming interface. In some cases, the sur-
geon will have a view of the eye that is upside-down relative to
the programming interface. It is important to know whether
the surgeon is left- or right-handed, and which hand is pre-
ferred for holding the forceps during tissue excision and bio-
material insertion through the access cut. This will determine
the exact location of the access cut around the circular
implanted region.
5. Depending on the size of the biomaterial button and the

hydration properties, it may be desirable to use a diameter of
biomaterial exceeding the diameter of the stromal pocket. This
could partially compensate for de-swelling of biomaterials post-
implantation and/or relaxation of the cornea after creation of
the femtosecond pocket, resulting in a larger diameter pocket
than initially intended. The diameter of the biomaterial can
typically be increased by 250–500 μm relative to the stromal
pocket diameter.
6. Depending on the animal model, some adjustments may be
required. For example, rabbit eyes are typically small and
require smaller implant sizes and/or omission of the suction
ring. Porcine eyes and the anatomic bony plate around the eye
may prevent proper alignment of the laser. These features
should be checked prior to performing the laser procedures.
It may, however, be unavoidable that the applanation to the
animal eye is not possible to center on the center of the pupil.
In such cases, the laser cuts will be de-centered and may enter
the limbal region. This will trigger an inflammation and neo-
vascularization postoperatively (see Fig. 11).
7. It is important that biomaterial insertion is performed using an
operating surgical microscope. Under the microscope, it may
be possible to see areas of unevenness and folding or wrinkling
of the biomaterial. In this case, after biomaterial insertion, use a
blunt spatula to flatten the biomaterial within the pocket.
8. An anterior segment OCT unit that is portable and can be
configured to examine eyes on an operating table is ideal.
Figure 12 below shows examination of a porcine eye using
the Optovue iVue OCT.
9. One should aim to perform regular postoperative assessment of
implanted eyes, ideally without general anesthesia. Addition-
ally, procedures should be established to handle animals regu-
larly, to facilitate postoperative eye drops to be given at least
Fig. 11 Effect of de-centration of biomaterial implant in the porcine eye. (a) The bony plate surrounding the
porcine eye (arrows) may prevent proper alignment of the applanation cone to the pupil center. (b)
De-centered cut region (arrow) close to the limbus on one side of the cornea. (c) The side nearest the limbus
has limited vessel invasion (arrows) and the inflammation and vascular leakage results in an opaque implant
Fig. 12 Examination of the implanted eye postoperatively using OCT with the
intubated animal lying on the operating table. An OCT that can be placed both
vertically and horizontally facilitates the examination
twice daily without use of anesthesia. Starting at 1 month

postoperative, general anesthesia can be used to perform
more extensive eye examinations (e.g., OCT, slit lamp photos)
that typically require a greater degree of compliance than is
possible while awake. Eye drops, fluorescein staining, and pho-
tography, however, can be performed in rabbits while they are
awake.
Acknowledgements
This work was supported by a Research Grant (Grant No. 667400)

from the European Commission, under the Horizon2020 project
“ARREST BLINDNESS.”
References
1. Gain P, Jullienne R, He Z, Aldossary M, 3. Hackett JM, Lagali N, Merrett K,
Acquart S, Cognasse F, Thuret G (2016) Edelhauser H, Sun Y, Gan L, Griffith M, Fager-
Global survey of corneal transplantation and holm P (2011) Biosynthetic corneal implants
eye banking. JAMA Ophthalmol 134:167–173 for replacement of pathologic corneal tissue:
2. Lagali N, Fagerholm P, Griffith M (2011) Bio- performance in a controlled rabbit alkali burn
synthetic corneas: prospects for supplementing model. Invest Ophthalmol Vis Sci 52:651–657
the human donor cornea supply. Expert Rev 4. van Essen TH, Lin CC, Hussain AK, Maas S,
Med Dev 8:127–130 Lai HJ, Linnartz H, van den Berg TJ, Salvatori
DC, Luyten GP, Jager MJ (2013) A fish
scale–derived collagen matrix as artificial cor- 9. Koulikovska M, Rafat M, Petrovski G, Veréb Z,

nea in rats: properties and potential. Invest Akhtar S, Fagerholm P, Lagali N (2015)
Ophthalmol Vis Sci 54:3224–3233 Enhanced regeneration of corneal tissue via a
5. Fagerholm P, Lagali N, Carlsson D, Merrett K, bioengineered collagen construct implanted by
Griffith M (2009) Corneal regeneration fol- a nondisruptive surgical technique. Tissue Eng
lowing implantation of a biomimetic tissue- Part A 21:1116–1130
engineered substitute. Clin Transl Sci 10. Lagali N, Germundsson J, Fagerholm P (2009)
2:162–164 The role of Bowman’s layer in corneal regener-
6. Fagerholm P, Lagali NS, Merrett K, Jackson ation after phototherapeutic keratectomy: a
WB, Munger R, Liu Y, Polarek JW, prospective study using in vivo confocal
Söderqvist M, Griffith M (2010) A biosyn- microscopy. Invest Ophthalmol Vis Sci
thetic alternative to human donor tissue for 50:4192–4198
inducing corneal regeneration: 24 month 11. Moshirfar M, Shah TJ, Masud M, Fanning T,
follow-up of a Phase I clinical study. Sci Transl Linn SH, Ronquillo Y, Hoopes PCS (2018) A
Med 2:46ra61 modified small-incision lenticule intrastromal
7. Ozulken K, Cabot F, Yoo SH (2013) Applica- keratoplasty (sLIKE) for the correction of
tions of femtosecond lasers in ophthalmic sur- high hyperopia: a description of a new surgical
gery. Expert Rev Med Dev 10:115–124 technique and comparison to lenticule intras-
8. Roszkowska A, Urso M, Signorino A, Aragona tromal keratoplasty (LIKE). Med Hypothesis
P (2018) Use of the femtosecond lasers in Discov Innov Ophthalmol 7:48–56
ophthalmology. EPJ Web of Conferences
167:05004
Chapter 15
The Use of Animal Models to Assess Engineered Corneal

Tissue
Robert Thomas Brady and Peter W. Madden
Abstract
Tissue-engineered corneal constructs offer the potential of readily available corneal substitutes for trans-
plantation. As with all medical devices and implants, these constructs require rigorous safety assessments,
combined with well-described analyses of the implant’s physical and biological characteristics. Although the
constructs are developed in vitro, such studies are currently unable to fully emulate the complex bio-
mechanical and biochemical conditions within living tissue, as well as the interplay between this environ-
ment and immunological factors. For these reasons, animal models remain essential to characterize such
interactions. They form a stage where corneal implants can be tested for utility and survival in a living
location to assess their ability to provide vision and avoid adverse event. Here, we examine the surgical
considerations of animal models and we describe how the rabbit can be used for this purpose. This animal
has been the routine model for ophthalmological studies and we set out methods to implant corneal
constructs with this species.
Key words Cornea, Animal surgery, Rabbit, Engineered tissue, Implant
1 Introduction
It is difficult to mimic the complex environment of the living body

in order to assess the ultimate value and utility of tissue-engineered
corneal constructs. In vitro models, such as those involving tissue
or organ culture, are an initial vital source of information with
which to devise and optimize the most appropriate corneal con-
struct. Such methods should always be used before contemplating
implantation into any living animal or human. Nevertheless, animal
surgery still remains the best method to imitate the complex envi-
ronment of biophysiological processes, biomechanical forces, and
immunological challenges that the implant will encounter. With the
cornea, this is the most accepted proof of value to justify clinical
human trials.
The expected scientific benefit of any study must sufficiently
justify the use of animals. Scientific justification has to form the
215
216 Robert Thomas Brady and Peter W. Madden
basis of plans to use animals and the scientific value must vindicate
the harm that all animal studies incur. The seminal work of Russell
and Burch [1], where animal use is Replaced, Reduced, and Refined,
the so-called 3 R’s, should always be paramount, not only leading
up to but also following animal surgery. If the required scientific
outcome can be met without using live animals, then this path
should be followed.
Here we present an overview of factors that need to be consid-
ered prior to commencing an animal study. We highlight that all
animal studies are an ethical balance of harm to benefit and that the
selection of which model to use relies upon the scientific outcome
to be achieved. The rabbit has been the conventional ophthalmic
model and we describe the surgical stages of using this species to
assess the survival of an engineered corneal construct.
2 Preliminary Considerations
All animal research must be legal and comply with ethical require-
ments. Regulation and guidance may be available in your local
institution, or through your governmental animal welfare regulat-
ing body. There are broad guideline documents available for such
studies in Europe [2] and in the USA [3]. One, or more, of these
guides should always be followed in the planning for any animal
study and the 3R’s rigorously applied. There are also readily avail-
able organizational guidance documents, such as the ARVO state-
ment on the use of animals in research [4].
Although the requirements for every study needs to be individ-
ually produced, they all have core aspects: (1) the animals should be
sourced from a reputable breeder; (2) there must be a veterinarian
physically available as necessary, having oversight of the animal’s
health and housing; (3) there must be appropriate husbandry facil-
ities and trained staff managing them; and (4) the surgery is to be
performed by trained staff, working in adequate surgical facilities. A
pilot study with a small number of animals can also often be
gainfully used to optimize the procedures in order to minimize
animal harm and maximize the scientific outcome.
2.1 Animal Model In the selection of an animal model, a review is required indicating
Selection which species have provided the best, most applicable data in the
past and which will be the most useful for current and future
research. The final choice should be at the lowest point on the
phylogenetic tree to still allow the most scientifically accurate and
interpretable results to be achieved, while requiring the fewest
number of animals.
Which animal model to be used depends upon the scientific
requirement of the study. Corneal transplants have been performed
with many animals, but there are advantages and disadvantages of
Corneal Surgical Animal Models 217
any model. Simians, with their closeness to humans, are the ulti-
mate model before clinical trials, but the high intelligence of these
species result in complex and expensive management, in addition to
significant ethical boundaries. It is routine therefore to begin with a
species lower on the phylogenetic tree, and historically the labora-
tory rabbit has been the routine animal to launch ophthalmic
studies [5]. Here, we will further examine this value and explain
how surgery can be performed with this species. An additional
animal model after the rabbit can also be advantageous to gather
further scientific value, such as using the rat [6] or mouse [7],
where genetic factors can be more precisely varied to examine
immunological events, or the pig [8], which has a cornea that is
mechanically closer to that of a human.
2.2 Planning for Prior to an animal study commencing, there is a requirement to

Animal Studies review the suitability of the model and also to assess the confound-
ing aspects of the particular model. The benefits and weaknesses of
the rabbit as a model are shown in Table 1. Weaknesses need to be
weighed against the particular scientific outcome that is being
investigated. The breed of rabbit is also an important determinant,
e.g., some breeds that grow significantly can make long-term
handling and housing more complex.
As a part of refining an animal study to cause the least harm,
having the shortest experimental period with the least invasive
procedures will clearly be advantageous. Many studies, however,
will have set requirements to achieve the desired outcome. One
mitigant that may be available is to perform implantation on only
one eye of the animal. This allows the animal to retain many normal
behaviors if the other eye is left untouched. Importantly, this also
has the additional benefit of acting as an internal control. Rabbits
are nonetheless troubled if one eye is not optimal; their eye posi-
tion, one each side of the head, allows two very wide fields of view,
but it also means there is a relatively narrow bilateral vision field.
Sight loss in the operated eye may be ameliorated to an extent, if the
implant can be placed off the visual axis, or by having an implant
smaller than the pupil.
All procedures should not only be documented with standard
operating procedures of what is expected to occur, but there also
needs to be procedures for adverse events that may occur. Impor-
tantly, this includes procedures to remove or add sutures, or per-
form lid reconstruction surgery. Consultation with a veterinarian is
essential in all animal experiment planning, with specific regard to
important determinants such as anesthesia, possible complications
or harms of the surgery, and analgesic adjuncts. Seeking the view of
a veterinary ophthalmologist who works with the specific species is
of value; human ophthalmic surgeons may not be fully aware of the
nuances of animal ocular physiology and behavior.
Table 1
Issues of rabbit corneal animal model of human transplants
Issue Effect
Historically, the model used for corneal studies Vast accumulated data on the anatomy and
physiology of the rabbit eye and its similarity
with that of the human
Similar size of the eye to humans Conventional human surgical instruments can
generally be used
Historically, the rabbit has been used for the Vast information on the response of the eye to drug
infamous Draize test and chemicals on eye irritation [9]
Relatively large eye size to body weight Smaller size makes for easy husbandry
Easy to breed and economical compared with larger Economy allows scientific outcome to be achieved
animals at a lower cost
Rabbits are docile Makes for easy handling and examination
Difference in anterior chamber shape compared to Can make deep corneal constructs difficult to use
a human, including shallow central depth [10]
A nictitating membrane is present This effectively acts as a third eyelid and so the
mechanobiology of the corneal surface differs
from humans. The membrane is also delicate and
vascularized, and its damage can be problematic
All rabbit corneal endothelial cells retain the ability This may make it unsuitable as a model of human
to divide even during adult life [11] penetrating keratoplasty as host endothelial cells
may more readily repopulate a construct
Limited range of genotypes This makes genetic studies difficult
No vomit reflex Fasting can be avoided and aspirating vomit is not
an issue
Breath-holding with anaesthetic gases Requires injectable anaesthesia for induction and
careful monitoring
Difficult to intubate and anaesthesia balance can be Requires involved anaesthesia
difficult
Younger rabbits demonstrate a more active Such age-related differences can model paediatric
postoperative inflammatory response and there humans and so this can be used to advantage.
can be marked production of fibrin in the Alternatively, it may adversely interfere with a
anterior chamber [12] corneal implant
Lack of corneal rigidity Constructs may be subject to differing mechanical
forces than in humans
3 Equipment
3.1 Animal The anesthetic equipment will vary with the type of anesthesia, but
Anesthesia and oxygen and a mechanism to administer it will be required in every
Physiological case. If gaseous anesthesia is used, then there must also be a
Monitoring scavenger unit to capture noxious waste gas, avoiding a risk of
exposure to the surgeon and assistants. Intubation is generally the
preferred method with many animals; however it can be problem-
atic in rabbits because it is difficult to directly visualize the tracheal
opening well, even with a laryngoscope, and so there has been
much debate with regard to the best method [13]. We therefore
use anesthesia by mask, but in the absence of intubation equip-
ment, it can be more demanding to achieve stable anesthesia and
provide assisted respirations where there may be drug-induced
bradycardia or breath-holding episodes.
Heart rate and oxygen saturation are required monitoring best
provided by a pulse oximeter: many pulse meters designed for
human use will not read heart rates above 250 beats per minute
and are therefore unsuitable for use with the rabbit. Meters that
also display waveform are advantageous to demonstrate heartbeats.
Hypothermia can result from a lack of movement, or interrup-
tion of thermoregulatory mechanisms such as cutaneous vasodila-
tion induced by anesthesia, a cold surgery table, or environment.
For this reason, core body temperature should be measured and
this can readily be achieved by the use of a rectal thermometer.
Breathing needs to be monitored. This can be done visually,
counting the breaths against time. If surgical drapes cover the
animal, then a breathing monitor that displays physical chest move-
ment is easier.
3.2 Eye Examination l Operating microscope: A binocular microscope with appropriate

lighting is essential for any corneal surgery. Foot controls can
maintain sterility, or focus handles can be decontaminated with
70% v/v alcohol, or covered with sterile aluminum foil.
l Tonometer: Significant changes from preoperative intraocular
pressure can indicate eye perforation or be a sign of rejection.
We advise that a contact tonometer is used, as non-contact
methods may not measure at specific sites and air-puff methods
may give noise that is disconcerting to rabbits.
l Slit lamp: The cornea can best be examined with a handheld slit
lamp for opacities or other pathologies.
l Corneal thickness: A range of equipment can review corneal
thickness and ideally implant depth: optical coherence tomogra-
phy (OCT), confocal microscopy, or an ultrasonic pachymeter.
While there are published values of corneal thickness of varying
rabbit species, outliers exist. Surgical planning to detect an
unusually thin cornea by one of these methods may therefore
avoid perforation of the anterior chamber. As such, OCT or

confocal microscopy give not only thickness, but also a structural
evaluation and are to be preferred. With any equipment that is
calibrated for the normally hydrated human cornea there will
need to be calculated adjustment to provide true linear measure-
ment of animal tissue.
3.3 Consumables The desirable aim is to only use consumables that are certified for
and Medicaments human clinical use. If this is not possible, then animal-certified
items are acceptable without review. If only research grade items
are available, then these should be reviewed as part of the project
risk assessment planning. Medicament sources may need to be
sourced by a different rank, with animal-certified, human-certified,
and then research-certified, depending upon the regulatory control
in place.
4 Operative Procedure in Rabbits
Here we focus on anterior lamellar implants, as these involve many

of the techniques used in all the procedures.
4.1 Selection Criteria Standard operating procedures need to include assessment by a

and Acclimatization supervising veterinarian of individual animals, as part of their
admission to the surgical program. All animals should be reviewed
against the animal selection criteria, such as sex, species, size, and
weight and any required preexisting health issues for the particular
scientific study. They should also have been given sufficient time to
acclimatize to the facility conditions before any surgical procedures
are undertaken. This is regularly taken to be 7 days, as a minimum.
Prior to surgery, a general health review is required, ideally by a
qualified veterinarian, plus a microscopic examination of the eye,
ideally by an ophthalmologist. Thus, there is the starting point of
an animal that is prepared for surgery, with good general and eye
health, or with specific morbidities within the study selection
criteria.
4.2 Preparation of No animal should be prepared for surgery unless the implant
Operating Theatre construct and the necessary staff, including animal welfare staff,
and equipment are available. Surgical planning should start with
having all surfaces prepared prior to procedure. Floors should be
wet-mopped to reduce dust and operating benches disinfected. If
immunocompromised animals are used, additional requirements
will be indicated. It is advisable to have a full dry-run of the
operating procedure, with the surgeon and all the ancillary workers
present. The team can simulate how the animal will be moved
through the procedure and where instruments and consumables
are sourced.
4.3 Health Regardless of whether the animal has previously appeared to be

Examination Including suitable for surgery, on the day of the procedure its health should
Preparatory Eye be checked for any change. Changes in weight, temperature, or
Examination heart rate could indicate an underlying infection. Both eyes should
be examined microscopically prior to anesthesia as corneal lesions
are common [14]. Intraocular pressure should also be measured as
this base level is valuable in monitoring eye response following
construct implantation.
4.4 Anesthesia and Preoperative and operative procedures need to be out of sight,
Administration of sound, and smell from other animals in a quiet area. Rabbits have
Preoperative Drugs a strong digestive tract cardiac sphincter [15], which precludes true
vomiting and so there is no requirement for preoperative fasting,
other than until one to two hours before surgery if intubation may
be used, to ensure that the mouth does not contain ingested
material.
The anesthesia of rabbits can be difficult. Here we describe a
technique that initially uses injectable anesthesia that is then sup-
plemented with gaseous anesthesia to extend operating time. An
approach of this type permits the long-term anesthesia that may be
required to adequately position a construct, especially in the
learning phase of surgery with a new device.
Having weighed the animal, administer 0.01–0.03 mg/kg of
buprenorphine subcutaneously and shave the marginal ear veins of
both ears. Lignocaine 3% gel is also applied to these veins to lessen
the sensation of needle insertion. The animal is placed in a holding
cage alone for one hour without food or water. Buprenorphine can
increase the duration of anesthesia and may provide beneficial
analgesia for the first hours after surgery [6]. An ear vein is cleaned
with an alcohol wipe and a butterfly catheter inserted. Its position
and patency is confirmed by use of sterile injectable saline. If the
patency is not adequate, then the other ear is used and direct light
pressure applied to the needle site of the first until bleeding stops.
Once a catheter is functional, it is taped into position on the ear
using gauze padding and micropore tape.
Xylazine at 3.0 mg/kg and ketamine, 10 mg/kg, mixed well in
preprepared 1 ml Luer lock syringes are then administered using an
injection port of the catheter, into which a butterfly needle has been
inserted. Using a tubed system like this allows movement of the
syringe to be decoupled from the catheter to ensure it is not
dislodged during movement of the syringe, or the animal. Multiple
syringes should be prepared, sufficient for the total anesthetic
period. Similarly, 3 mL syringes of sterile 0.9% NaCl should also
be readied. This saline should be used periodically throughout the
anesthesia to flush the anesthetic line, and not only check that the
catheter is patent, but also ensure that the small volume of anes-
thetic is fully dispensed from the tubing. If the surgery is greater
than one hour, then 5 mL of 0.9% NaCl should be given every hour
peritoneally, as maintenance fluids.
As soon as the animal is anesthetized with the injectable agents,
it is put on the heated pad, and supplied with pure oxygen by face
mask at a rate of 5 L/min. A rectal thermometer is put in place and
a pulse oximeter applied over a prominent vein of the shaved ear.
The pulse oximeter may be covered with aluminum foil to prevent
the surgical lights interfering with the oximeter light sensor.
Testing of anesthetic depth using a toe pinch and corneal reflex
should be performed every 10 min and an additional 0.05–0.1 mL
bolus anesthetic given if there is any reaction. Rabbits do have a
very pronounced breath-holding response to gaseous agents
[16]. As soon as anesthesia is confirmed, a very small concentration,
<0.5% isoflurane, is administered in combination with oxygen, so
that the animal becomes accustomed to the smell. Habituated in
this way, the full level of dosage can then be applied at a 5% level a
few minutes later and this will normally avoid breath-holding.
Bolus anesthesia of xylazine and ketamine can result in tempo-
rary respiratory suppression, but this rapidly resolves. It can be
made more dramatic in that the xylazine can also cause vasocon-
striction of the ear veins which will change the reading of the pulse
oximeter, which may alarm. Monitoring of vital signs should be
continuous but this temporary suppression requires no specific
intervention or reversal.
Under anesthesia, the nictitating membrane will fall back into a
relaxed position, exposing the cornea and ocular surface. If at any
stage during the operation the anesthesia becomes light, the nicti-
tating membrane may move laterally to encroach upon the cornea.
An intravenous bolus of 0.05–0.1 mL anesthetic will result in the
prompt return of the membrane to its resting position by the
medial canthus.
4.5 Check of Due to their larger size and coats, rabbits are not as susceptible to
Physiological Vital heat loss as smaller animals. However, intraoperative heat loss and
Signs hypothermia can result in prolonged recovery and should be
avoided. A heating pad should be placed under the animal for the
duration of anesthesia and recovery. Core body temperature of
38.5–40 C also requires monitoring. Monitoring heat gain is just
as important as heat loss to ensure that the temperature does not
rise enough to result in heat stroke.
Vital signs should be monitored throughout the procedure.
The rabbit heart rate is typically between 150 and 300 bpm. Bra-
dycardia can occur with bolus doses of anesthetic. In contrast,
tachycardia can result from distress, pain, or hyperthermia. Simi-
larly, the respiratory rate should be in the order of 30–60/min. This
can drop with breath-holding episodes occurring with bolus doses
of anesthetic agents. Oxygen saturations should be maintained
above 95%.
At this point, a surgical drape can be applied to the animal’s

body. This may be a full coverage drape if intubation is used or an
elevated drape if the animal requires direct observation during the
procedure. Operating instruments and sutures should be prepared
in a separate sterile tray, as with any other surgery.
4.6 Eye Preparation The eyelashes may need to be trimmed for better access, but it is
preferable to retain them if possible, as they will assist recovery,
reducing foreign body entry and protectively closing the eye when
touched.
Once the animal is in position and the anesthetist gives approval
to proceed, the eye should be prepared by instilling a topical
anesthetic of 0.4% w/v oxybuprocaine hydrochloride. If it is neces-
sary to enlarge the corneal pupil to allow a better view or better
surgical depth, then a cycloplegic such as atropine sulfate 1% eye
drop can be instilled. The common drug preservative benzalko-
nium chloride can be an irritant and single-use products that avoid
this and other preservatives are to be preferred [17].
Topical povidone iodine solution can then be instilled to
decontaminate the eye. The preferred solution for the ocular sur-
face is a 1% w/v concentration without preservative [18], as higher
concentrations may cause cellular damage.
While anesthetized, the eye should be closely examined to
assess for any small conjunctival or corneal abnormalities. A healthy
corneal reflex should be apparent in addition to the absence of any
signs of ocular inflammation. Microscopic examination is advised as
subtle hypopyon may be present. If any abnormalities are found,
then the animal may not be a suitable candidate for the surgery.
The surgical field should next be prepared by the use of povi-
done iodine or chlorhexidine applied on sterile gauze around the
eyelids, working outward in a circular manner without retracing. A
keyhole sterile drape can then be used to give a sterile field up to
the eye.
To gain access to the eye, management of the nictitating mem-
brane as well as the eyelids must be controlled. With sufficient
anesthesia, the nictitating membrane should be in a position close
to its medial canthus origin. This method is to be preferred as
membrane removal can change the tear film [19] and stay suturing
during the operation can lead to hemorrhage and swelling. Fur-
thermore, injury or irritation to the nictitating membrane can result
in excessive blinking and this may negatively affect the sited
implant. The eyelids proper may need a speculum to completely
expose the ocular surface. Depending on the size of the animal,
even human pediatric speculums may be too big for this purpose.
Oversized speculums, used for even short periods of time, may
result in significant stretching of the eyelids. This should be avoided
as the lids may not return to their original size and there may be
subsequent eyelid ectropion/entropion. Any irritation from
disfigured lids can also recruit the nictitating membrane to sweep

more often across the corneal surface and further aggravate implant
location. An alternative to a speculum are retraction sutures,
although these also pose a risk. A loop of 6/0 braided silk through
the lid margin is held in place by clamping with an arterial forceps
onto the sterile drape. Care must be taken to limit tension across
the suture, with the risk of lid injury increasing with operating time
length.
During surgery, the eye that is not being operated on needs to
be protected from damage. The proptosis of the rabbit eye increases
drying risk, which can lead to irritation and ultimately ulceration.
Adhesive tape to hold the eye closed is not recommended, given the
fur-bearing skin. Instead, unless this interferes with the scientific
outcomes of the project, a viscous eye lubricant, e.g., a carbomer,
can be instilled and the eye manually closed during general anes-
thesia. This combination will protect the eye and prevent it from
drying.
4.7 Fine Examination Fine examination of the eye should follow routine veterinary and
of Eye human clinical practice, such that there can be a determination of
whether the eye is normal. Unless there is a specific need to assess a
particular cell layer, this can be accomplished by examination by slit
lamp, intraocular tonometry and fluorescein staining for epithelial
defects. The outcome is to ensure there is no active infection or
disease that is not part of the scientific requirement.
4.8 Eye Irrigation For cell-seeded implants, a period of dryness could result in cell
death. Therefore, it is preferable that a surgical assistant is tasked to
regularly irrigate the ocular surface, allowing the surgeon to focus
on operating and complete the procedure more speedily. However,
there are species differences in stromal swelling rate, and following
de-epithelization of the rabbit cornea, its swelling can be variable at
different depths [20] and it may swell more than in humans which
can make positioning the implant more difficult. As such, during
the operation, hydration should be limited to just sufficient to
maintain cell survival and only after implantation a sufficiently wet
ocular surface ensured to prevent postoperative complications; a
dry eye will result in more pain, inflammation, and animal irritation.
4.9 Stabilization of A stable eye is needed for surgery. In addition, eye position can
the Eye with Sutures fluctuate with the depth of anesthesia. While an intubated animal is
the most stable option, long procedures or those utilizing bolus
doses are more variable. As such, eye position may vary. As the
anesthetic concentration lessens, the eye will usually adduct and will
meet the extending nictitating membrane. Additional doses of
agents may stabilize the eye; however, a conjunctival stay suture
can be used to give additional stability to the eye.
4.10 Epithelial The epithelium can add an extra complexity to thickness calcula-
Removal tions. It is advised that the epithelial layer be debrided by abrasion
before any trephine is applied. This will give a more accurate
approximation of the trephine depth achieved. Additionally, epi-
thelium removed from the host edge is less likely to allow epithelial
cells to contaminate between implant and host, preventing stromal
integration.
4.11 Depth Cut Circular trephines are used to remove host tissue in which to place a
transplant and similar methods are used with an implant. In
humans, 6–8 mm diameters are typical, balancing a size large
enough to replace central vision, but small enough to limit involve-
ment with the vascularized limbus to minimize rejection risk. In
animal studies, it is important to select a trephine that is suitable for
the animal’s size, and many adult New Zealand White rabbit studies
employ 6 mm diameter, as we do. We also use a 6.5 mm diameter
implant, as the 0.5 mm oversize ensures close apposition between
host and implant to ensure a good seal.
It is strongly suggested to complete a series of pretrial trephine
cuttings upon cadaveric samples. Trephines are optimized for
human use and we have found that those with a narrow suction
area best suit rabbit use. Wider suction areas can deform the more
flexible rabbit cornea and make the depth of cut irregular, possibly
leading to corneal perforation.
During the operation, normal saline is used to prevent corneal
drying. If an adequate vacuum cannot be maintained, then repla-
cing this with a viscous eye lubricant can be advantageous.
4.12 Removal of Host Having trephined to the desired depth, the stroma may be dissected
Tissue and Check of away. The bubble technique for generating a lamellar cleavage plane
Depth does not work as effectively with rabbit tissue compared to human
tissue and hydro-dissection with saline can also increase stromal
hydration and impair visualization. Preferred is the use of a blunted
crescent-shaped blade to find a cleavage plane and avoid perfora-
tion. Toothed forceps can be used to assist in determining this
plane. To ensure the correct final defect depth and dimensions, a
specifically sized shim is valuable. A shim is particularly useful to
determine if depth is sufficient or if the construct may sit proud,
thereby increasing opportunity of failure.
4.13 Preparation of Once incised, corneal tissue can hydrate and expand significantly. It
Surgical Bed is therefore important to promptly complete the dissection of the
host tissue while also avoiding excessive irrigation. Although the
blunted crescent blade is usually sufficient to have generated a clear
dissection plane, an Alger brush may be considered to smooth the
implant bed. Moreover, edges need to be smooth for a close
apposition of implant and host. If having found that the implant
sits proud in the defect in spite of an accurate dissection and the
minimal thickness of recipient tissue remains, then it is possible to

undermine the stroma at the deep margin of the defect along the
plane of the base. This will allow the surface edge to be approxi-
mated anteriorly and be sutured in line with the implant to ensure a
surface as smooth as possible.
4.14 Interrupted or The implant needs to be fixed in position. This is predominantly

Continuous Sutures done using sutures and suture technique can vary. A single contin-
uous or interrupted suture can be used to secure the implant. A
single suture has the benefit of achieving even tension across the
implant edges, reduced astigmatism, and a more rapid surgical
completion time. However, if a continuous suture fails, the con-
struct will be lost. An alternative is interrupted sutures, typically
16 in total, which are sited with 4 initial cardinal sutures as shown in
Fig. 1. This technique requires skill to maintain centration and limit
displacement of the implant. It can be a more time-consuming
technique; however unlike the continuous type, the failure of a
knot or development of a suture abscess can be easily addressed
while still maintaining implant integrity. A third method combines
both 4–8 cardinal sutures with an additional continuous suture
around the entire implant.
For most implant work interrupted sutures provide the most
adaptable technique that also allows for revision. We use a nonab-
sorbable suture of 10/0 monofilament nylon to maintain strength
for an extended period. After all the sutures are placed and the
surgeon believes the implant is well placed, the suture tails should
be trimmed to 2 mm and the knots buried.
Fig. 1 A corneal construct after surgery. A 6.5 mm diameter decellularized pig

corneal stroma, initially 300 μm deep before being populated with keratocytes
added in vitro, has been sewn into a 300 μm deep, 6.0 mm diameter, anterior
depression bed. Sixteen interrupted nylon sutures with buried knots locate the
implant
4.15 Surgical Checks Having completed the procedure, ensure not only that there are no
suture ends protruding, but also that sutures have not become
loose upon burying. Uncorrected, such issues may elicit a foreign
body sensation resulting in irritation, excessive blinking, and
recruitment of the nictitating membrane. Moreover, loose sutures
may act as a nidus for mucus and debris to collect, possibly resulting
in infection. If interrupted sutures are loose, a replacement should
be placed adjacently and the original removed.
Anterior cornea removal should not interfere with anterior
chamber integrity, but penetrating corneal surgery can result in
leaking of the aqueous humor. Leaks can be identified by Seidel’s
test [21], applying fluorescein dye to the ocular surface and use of a
cobalt blue light to reveal leaks. Once identified, a leak can be
addressed through additional sutures or replacing loose sutures
around the leak. A leak which is not managed appropriately will
result in hypotony and can lead to endophthalmitis from microor-
ganism entry.
Remove the eye stabilization suture and then speculum or
retraction sutures. The lids should resume a good anatomical
approximation with the ocular surface.
4.16 Anesthesia Most agents routinely used in rabbit anesthesia have a rapid half-life
Reversion and reversal of anesthesia is not normally required. Some medica-
tions such as benzodiazepines and opioids have reversal agents
available and in this case it should always be ready for use; individual
rabbits may respond atypically and may have difficulty in recovering
otherwise. With the use of xylazine and ketamine facilitated anes-
thesia, the animal may be given oxygen alone, and if it does not
appear to be rousing after 5 min, then 0.5 mg/kg atipamezole
hydrochloride may be given.
5 Postoperative
5.1 Removal to Once all checks are completed, the rabbit may be transferred to the
Recovery Site with recovery area. For a few hours it should be held in isolation for very
Monitoring close observation. Solid food, hay, and water should be in place in
the recovery area from the start. Early eating and drinking is advan-
tageous to recovery, avoiding problems with gastrointestinal stasis.
Once the animal is behaving normally, ideally with demonstration
of eating, drinking, urination, and defecation, it should be returned
to its social housing environment with other animals. There should
be heightened vigilance of behavior at this time and a veterinarian
should be consulted regarding any atypical behavior. If there are
concerns, some cases may need an additional time of separation
using a visually and olfactory transparent material such as a chicken
wire barrier. Sufficient space and safety refuges should be available
in animal housing. All the rabbits, not just the surgical one, should
be observed following removal or return from surgery as social
dynamics change in the colony.
5.2 Postoperative All eye surgery that involves epithelial loss will result in a foreign
Medication body sensation. Such surgery can also result in a temporary dys-
function of the ocular lubricating mechanism and postoperative
dryness. There should be a significant effort made to reduce any
discomfort from these sequelae. Firstly, analgesia is required such as
0.01–0.03 mg/kg buprenorphine subcutaneously, twice a day, for
the first three days. Secondly, methods to reduce the foreign body
sensation which is enhanced by blinking are required. Tarsorrhaphy
has been used, but this disadvantages the animal to one eye and
changes the conditions of the implant [22]. External eye shields or
bandaging will not be tolerated by rabbits in a social setting, and
bandage contact lenses that are routine following human surgery
are problematic in rabbits with a nictitating membrane. For these
reasons we use an ocular lubricant which are viscous agents that act
as a barrier between the compromised ocular surface and the eyelid.
They should be used for as long as it takes to restore an epithelial
cover. Thirdly, because the epithelial barrier will have been dis-
turbed by the surgery, then infection may traverse into the cornea
and so an ocular antibiotic is used until the barriers has been
restored. We use chloramphenicol 0.5% eye drops. Fourthly, if the
construct contains any cell or cell antigen other than autologous,
then a steroid to reduce inflammation is required. Here 0.1% w/v
betamethasone sodium phosphate is employed, initially at least
three times a day. Topical medication should be followed by
5 min delay before the next for best effect. However, these agents
may often also be available as ointments rather than liquid drops
and using ointment is to be preferred, allowing shortened delay
between applications and maintaining concentration of the active
agent for longer.
5.3 Postoperative The general health of the animal is paramount to assess the health
Reviews of the eye. Useful parameters to assess health and nutritional status
are scoring of locomotion, grimace scale [23], body condition,
feeding/drinking, and socialization. There should be regular
weighing. It is important to identify poor feeding quickly, with
isolation and review of fecal output for any animal at risk. A veteri-
narian should promptly investigate any animal of concern.
Full examinations may require sedation of the animal and a
schedule agreed with the veterinarian as part of project planning.
However, if pain is well managed, the docile nature of rabbits
should allow most examinations without sedation. If temporary
pain relief is required, then topical application of a 0.4% w/v
oxybuprocaine hydrochloride drop can be given, although this
does hinder reepithelization and should not be routinely used.
All scoring, regardless of those measurements required by the

scientific outcome specifics, should initially include assessment of
suture integrity, presence or absence of infection, and epithelializa-
tion. The type and extent of postoperative checks are determined by
both the implant and the type of surgery carried out. Corneal
transparency will need to be scored as will implant apposition
with the eye.
5.4 Revision There should be planning for revision treatment. Lid laxity can
Treatments develop from retraction tension during surgery, particularly with
long surgical duration, or eye swelling. This needs to be addressed
promptly with a wedge excision to remove excess lid tissue and
restore their anatomical position.
Nonabsorbable nylon sutures can remain in place unless they
become dislodged. Inevitably some sutures will become unburied
or loose and can cause irritation and foreign body response. If the
integrity of the implant is compromised, re-suturing under general
anesthesia is required, but if their mechanical support is no longer
necessary, individual sutures can normally be removed with local
anesthesia using 0.01–0.03 mg/kg buprenorphine subcutaneously
and topical 0.4% w/v oxybuprocaine hydrochloride.
6 Conclusions
The rabbit eye allows human clinical-sized constructs to be used

with conventional instruments. It remains as an important model to
assess the usability and viability of corneal implants.
References
1. Russell WMS, Burch RL (1959) The principles 15(3):267–291. https://doi.org/10.2174/

of humane experimental technique. Methuen, 1871527315666151110124957
London 6. Murphy KL, Roughan JV, Baxter MG, Fleck-
2. Council directive 2010/63/eu on the protec- nell PA (2010) Anaesthesia with a combination
tion of animals used for scientific purposes of ketamine and medetomidine in the rabbit:
(2010) vol L276. European Parliament effect of premedication with buprenorphine.
3. Council NR (2011) Guide for the care and use Vet Anaesth Analg 37(3):222–229. https://
of laboratory animals, 8th edn. The National doi.org/10.1111/j.1467-2995.2009.
Academies Press, Washington, DC 00525.x. VAA525 [pii]
4. Arvo statement for the use of animals in oph- 7. She SC, Steahly LP, Moticka EJ (1990) A
thalmology and vision research (2016) The method for performing full-thickness, ortho-
Association for Research in Vision and Oph- topic, penetrating keratoplasty in the mouse.
thalmology. https://www.arvo.org/About/ Ophthalmic Surg 21(11):781–785
policies/statement-for-the-use-of-animals-in- 8. Brunette I, Rosolen SG, Carrier M,
ophthalmic-and-vision-research/ Abderrahman M, Nada O, Germain L, Proulx
5. Zernii EY, Baksheeva VE, Iomdina EN, Aver- S (2011) Comparison of the pig and feline
ina OA, Permyakov SE, Philippov PP, Zamyat- models for full thickness corneal transplanta-
nin AA, Senin II (2016) Rabbit models of tion. Vet Ophthalmol 14(6):365–377. https://
ocular diseases: new relevance for classical doi.org/10.1111/j.1463-5224.2011.
approaches. CNS Neurol Disord Drug Targets 00886.x
9. York M, Steiling W (1998) A critical review of 17. Aguayo Bonniard A, Yeung JY, Chan CC, Birt
the assessment of eye irritation potential using CM (2016) Ocular surface toxicity from glau-
the Draize rabbit eye test. J Appl Toxicol 18 coma topical medications and associated pre-
(4):233–240. https://doi.org/10.1002/(sici) servatives such as benzalkonium chloride (bak).
1099-1263(199807/08)18:4<233::aid- Expert Opin Drug Metab Toxicol 12
jat496>3.0.co;2-y (11):1279–1289. https://doi.org/10.1080/
10. Thomasy SM, Eaton JS, Timberlake MJ, Miller 17425255.2016.1209481
PE, Matsumoto S, Murphy CJ (2016) Species 18. Lee S, Khun D, Kumarasinghe GL, De Zoysa
differences in the geometry of the anterior seg- GH, Sarojini V, Vellara HR, Rupenthal ID,
ment differentially affect anterior chamber cell Thakur SS (2019) Ex vivo evaluation of the
scoring systems in laboratory animals. J Ocul stability, safety and antibacterial efficacy of an
Pharmacol Ther 32(1):28–37. https://doi. extemporaneous povidone-iodine preparation
org/10.1089/jop.2015.0071 for ophthalmic applications. Clin Exp Optom.
11. Olsen EG, Davanger M (1984) The healing of https://doi.org/10.1111/cxo.12899
rabbit corneal endothelium. Acta Ophthalmol 19. Larry AM, Nemiroff B, John LU, Hf E, Char-
(Copenh) 62(5):796–807. https://doi.org/ lesworth LD (1987) Role of the rabbit nictitat-
10.1111/j.1755-3768.1984.tb05808.x ing membrane in ocular irritancy testing. J
12. Gwon A (2008) The rabbit in cataract/IOL Toxicol Cutan Ocul Toxicol 6(1):43–56.
surgery. In: Tsonis PA (ed) Animal models in https://doi.org/10.3109/
eye research. Academic Press, London, pp 15569528709052164
184–204. https://doi.org/10.1016/B978-0- 20. Wilson G, O’Leary DJ, Vaughan W (1984)
12-374169-1.00013-8 Differential swelling in compartments of the
13. Wenger S, Müllhaupt D, Ohlerth S, Prasse S, corneal stroma. Invest Ophthalmol Vis Sci 25
Klein K, da Silva VB, Mosing M (2017) Exper- (9):1105–1108
imental evaluation of four airway devices in 21. Cain W Jr, Sinskey RM (1981) Detection of
anaesthetized new zealand white rabbits. Vet anterior chamber leakage with seidel’s test.
Anaesth Analg 44(3):529–537. https://doi. Arch Ophthalmol 99(11):2013. https://doi.
org/10.1016/j.vaa.2016.05.009 org/10.1001/archopht.1981.
14. Holve DL, Mundwiler KE, Pritt SL (2011) 03930020889015
Incidence of spontaneous ocular lesions in lab- 22. Koch JM, Refojo MF, Leong FL, Kenyon KR
oratory rabbits. Comp Med 61(5):436–440 (1989) Keratopathy of the rabbit cornea fol-
15. Botha GS (1958) Histological observations on lowing complete eyelid closure. Acta Ophthal-
the gastro-esophageal junction in the rabbit. J mol Suppl 192:108–114
Anat 92(3):441–446 23. Keating SC, Thomas AA, Flecknell PA, Leach
16. Flecknell PA, Roughan JV, Hedenqvist P MC (2012) Evaluation of emla cream for pre-
(1999) Induction of anaesthesia with sevoflur- venting pain during tattooing of rabbits:
ane and isoflurane in the rabbit. Lab Anim 33 changes in physiological, behavioural and facial
(1):41–46. https://doi.org/10.1258/ expression responses. PLoS One 7(9):e44437.
002367799780578516 https://doi.org/10.1371/journal.pone.
0044437. PONE-D-12-17063 [pii]
Chapter 16
X-Ray Diffraction Imaging of Corneal Ultrastructure

Keith M. Meek, Andrew J. Quantock, Sally Hayes, and James Bell
Abstract
X-ray scattering enables the structure of collagen-rich tissues, such as the cornea, to be examined at both the
molecular and fibrillar level. The high-intensity X-rays available at synchrotron radiation sources, coupled
with minimal sample preparation requirements, facilitates the rapid generation of high-quality X-ray
scattering data from corneal tissue at a close-to-physiological state of hydration. Analysis of resulting
X-ray scatter patterns allows one to quantify numerous structural parameters relating to the average
diameter, lateral arrangement and alignment of collagen fibrils within the cornea, as well as the axial and
lateral arrangements of collagen molecules within the fibrils. Here we describe the typical experimental
setup and considerations involved in the collection of X-ray scattering data from corneal tissue.
Key words X-ray, Cornea, Synchrotron, Collagen hydrogels
1 Introduction
The corneal stroma has a well-defined, tissue-specific ultrastructure

[1]. The size, spacing, and spatial arrangement of the collagen
fibrils confer strength and transparency to the cornea. In particular,
the collagen fibrillar organization allows the cornea to retain its
shape during the intraocular pressure pulse, when blinking and
during eye movement. These structural and biomechanical proper-
ties need to be considered when designing a successful artificial
corneal replacement.
X-ray diffraction, which in the case of noncrystalline samples is
now generally referred to as X-ray scattering, is a noninvasive
technique that requires minimal sample preparation. It allows the
tissue to be examined in a close to physiological hydration state
with no requirement for chemical fixation and/or sectioning, such
as is necessary for examination using light or electron microscopy.
The technique involves passing a beam of high-intensity, focused
X-rays through a specimen whose molecular assemblies contain
repeating (periodic) units. The data generated from each specimen
represents an average of every molecule or molecular assembly
231
232 Keith M. Meek et al.
(such as a fibril in the case of collagen in the corneal stroma) within

the path of the X-ray beam; this can number many billions. Analysis
of the subsequent distribution of X-ray scatter provides structural
information from the molecular scale upward. Constructive inter-
ference between scattered X-rays leads to the formation of so-called
X-ray reflections, and the angle of scatter of these is defined by
Bragg’s law:
nλ ¼ 2d sin θ ð1Þ
where n is the order of the reflection, λ is the wavelength of the
X-rays, θ is half the angle of scatter, and d is the repeat spacing
within the sample under investigation. Bragg’s law indicates an
inverse relationship between d and θ for a fixed wavelength, so
larger spacings lead to lower scatter angles and vice versa. The
dimensionality within an X-ray scatter pattern is therefore referred
to as “reciprocal space” whereas distances within the sample itself
are said to be in “real space.” The X-ray scatter is usually divided
into small-angle scattering (SAXS) and wide-angle scattering
(WAXS) components, depending on the sizes of the structures
being investigated. SAXS covers the angular range up to about 1
while WAXS typically covers the 5–60 range. In the cornea, SAXS
is used to measure the axial periodicity and the electron density
distribution along the collagen fibrils (from so-called meridional
reflections), the average diameter of the fibrils as a population, and
the center-to-center lateral spacing of the fibrils (from equatorial
reflections) [2]. WAXS, on the other hand, is used to measure the
axial amino acid residue repeat along the collagen molecules within
the fibrils (from meridional reflections), as well as the lateral inter-
molecular collagen spacing (from equatorial reflections) [3]. SAXS
and WAXS also allow one to quantify the orientation distribution of
the collagen fibrils and molecules, respectively [3–7]. SAXS and
WAXS patterns are produced together, and it is the setup of the
experimental system, predominantly based on the placement of the
X-ray detector, which dictates whether a SAXS or WAXS analysis is
conducted (on some setups, SAXS and WAXS data can be collected
concurrently). Figure 1 shows schematically how X-ray patterns
from a cornea are generated, where the X-ray beam is passed either
along the corneal axis or edgeways through strips of cornea. Such
X-ray scatter patterns have been obtained in the laboratory using
conventional X-ray generators, but these systems can take many
hours to generate a single X-ray scatter pattern, whereas the much
greater X-ray intensity and high-performance imaging detectors
available at synchrotron sources allow multiple, high-resolution
patterns to be obtained in a matter of seconds. Suitable facilities
are now available at many synchrotron X-ray sources worldwide (see
Note 1). This article will focus on synchrotron-based methods, but
most of the details and concepts apply to laboratory-based X-ray
X-Ray Diffraction from Cornea 233
Fig. 1 Formation of X-ray scatter patterns (SAXS and WAXS), with the X-ray beam passed parallel to the optical
axis and edge-on through strips of cornea. In the parallel case the presence of collagen in all azimuthal
directions leads to a pattern consisting of concentric equatorial and meridional circles (example schematic
pattern on the left). In the edge-on case (example schematic pattern on the right), collagen is orientated such
that equatorial reflections (arising from lateral dimensions and packing of the collagen) occur in the horizontal
direction and meridional reflections (arising from axial periodicities along the collagen molecules and fibrils)
occur in orthogonal directions to the predominant alignment of fibrils (reproduced from ref. 2 with permission
of the copyright holder)
sources as well. Two types of experiment are possible, static and

time resolved. Static studies involve studying the sample at rest,
whereas time resolved studies take full advantage of the high inten-
sity of synchrotron light to examine dynamic processes, such as
those that occur when mechanical forces are applied.
2 Materials
No specific materials or solutions are necessary, save those required

to construct the specimen cells and calibrate the system. If a sample
is too thin, or too hydrated, other reagents are sometimes used to
enhance the scattering.
2.1 Specimens Freshly excised (e.g., corneas) or prepared (e.g., hydrogels) speci-
mens. Where this is not possible, they may need to be stored and
transported to the synchrotron in a solution in which they will not
swell (see Note 2), such as balanced salt solution or proprietary
storage solutions such as Optisol-GS. If absolutely necessary, speci-
mens can be fixed (4% paraformaldehyde works well) or snap-
frozen in liquid nitrogen-cooled isopentane and then thawed at
the synchrotron just before examination [8] (see Note 3).
2.2 Specimen Cells Specimen cells need to be designed according to the required
experiment (Fig. 2). The design may depend on the mounting
arrangement available in the path of the X-ray beam, which can
vary depending on the specific synchrotron facility being used. In
all cases, the cell should be made from a rigid material to support
the specimen and have a window that allows the uninterrupted
Fig. 2 Examples of specimen cells and a cell holder for stretching, compressing, and inflating corneal tissue
(left). Not all corneas can be fully enclosed within a sealed cell and the image on the right shows a prototype
gas inflation setup in situ at the beamline.
passage of X-rays through the sample. If it is necessary to minimize

hydration changes, the window should be constructed of a material
whose X-ray scatter does not interfere with the reflections being
studied. Mylar or mica are suitable materials for use in the construc-
tion of specimen cells for cornea experiments as their scatter does
not overlap with SAXS or WAXS patterns from collagen. The
specimen cell should be of a suitable size and have an attachment
mechanism (usually a separate cell holder) for it to be compatible
with the space and setup available on the beamline; it is essential to
ascertain this prior to conducting the experiments. All synchrotrons
have beamline managers who advise how to mount the sample,
prior to the experimental run. Modern cells can be readily 3-D
printed.
2.3 Calibration For quantitative measurements, the X-ray scatter patterns need to
Materials be calibrated. It is very difficult to measure accurately the exact
distance from the specimen to the detector, which is necessary to
calculate the scattering angle 2θ (Fig. 3) and hence the unknown
spacing d (from Eq. 1), which in the case of experiments with
corneas relates to collagen molecular and fibrillar structures and
arrangements. To overcome this, an X-ray exposure from a material
with a known spacing (d) is recorded at the start of the experimen-
tal run (see Note 4). Diffraction gratings are routinely used for
calibration by the beamline managers at the start of every run, but
it is advisable to carry out an independent calibration. For SAXS
(d values roughly in the range 300 nm to 1 nm), silver behenate
(d ¼ 5.838 nm) is used. In the past, rat tail tendons were typically
used, but are not now recommended as the known spacing (i.e., the
axial D-periodic repeat along a collagen fibril axis) can change if the
tendon starts to dry. For WAXS (d values roughly in the range 1 nm
to 0.1 nm), calcite powder (d ¼ 0.305 nm) is suitable.
Fig. 3 Schematic showing the geometry of the X-ray scatter pattern in two
dimensions and the positioning of a lead beam stop between the sample and the
detector to suppress intense un-scattered X-rays and prevent damage to the
detector. The diffracted X-rays appear on the detector at a distance, y, from the
center of the pattern. The distance, x, is not easily measured, so is determined
by using a calibrant of known periodicity, d, then applying Bragg’s law (see
Note 4)
2.4 Other Materials 1. A balance, if not available at the synchrotron, should be taken
That May Be Required in order to measure the weight of the specimen before and after
at the Synchrotron data collection, to monitor changes in tissue hydration.
2. Mylar®. This is a polyester film that is available in rolls for X-ray
use. It is needed to construct/repair X-ray transparent win-
dows in the specimen cells. Cyanoacrylate glue is used to stick
the windows to the cells.
3. Commercially available food wrap film (e.g., Cling Film or
Saranwrap). This is useful for wrapping specimens before they
are mounted in the specimen cell. This helps to minimize any
dehydration during handling, and during the X-ray exposure.
3 Methods
3.1 How to Get Synchrotron sources offer facilities for a number of different types
Synchrotron Access of experiment utilizing intense light of different wavelengths,
including X-rays. The different experiments are carried out on
so-called “beamlines.”
1. Go to the synchrotron website and identify the beamline that
offers small-/wide-angle X-ray scattering, SAXS and/or WAXS
(or contact the Synchrotron Facility User Office for this
information).
2. Contact the beamline manager to discuss the experiment and
its feasibility.
3. Follow the online procedure for applying for beam time to
carry out the experiment (see Note 5).
3.2 Setting Up The positioning of the detector with respect to the specimen will be
at the Synchrotron done by beamline staff either before your arrival or at the start of
the experiment (see Note 6). The beamline staff will then help to set
up the experiment and to focus the X-ray beam onto your specimen
cell (Fig. 4) (see Note 7). They will also instruct on the software
available for remotely moving the specimen in the beam and for
collecting the X-ray scatter patterns. For time-resolved studies
involving the synchronization of data acquisition with a separate
piece of apparatus (usually supplied by the user), the apparatus
needs to be integrated into the synchrotron’s systems (see Note
8). For safety reasons, the X-ray camera is housed within a special
room (typically referred to as a hutch), which will be sealed closed
when the camera shutters are open and the X-ray beam
on. Synchrotron X-ray radiation is very intense and is fatal within
minutes, so safety training is provided at every visit to demonstrate
how to operate the fail-safe systems in the hutch to prevent acci-
dental exposure. This training also covers the proper use of the
delicate (and expensive) beamline apparatus.
Fig. 4 The general experimental setup. For simple static experiments, where no
mechanical forces are being applied, the corneas are wrapped in Cling Film
(a) which produces weak and diffuse background scatter that can be removed
during image analysis but maintains specimen hydration during X-ray exposure.
They are usually mounted in a sealed cell with X-ray transparent Mylar windows
(b). The cell is then positioned into a cell holder in the path of the X-ray beam
(c) on a table that allows movement to accurately center the beam on the
appropriate part of the specimen. The X-rays emerge from the evacuated tube
(shown on the right of panel c) and the scattered X-rays are recorded on a
detector (seen on the far left of panel c). The distance from the specimen to the
detector can vary from centimeters (for WAXS) to several meters as shown here
(for SAXS). In the latter case they pass through an evacuated tube before
reaching the detector, in order to reduce the scattering effects of air
3.3 Data Collection 1. Check any additional equipment that you may be adding to the
beamline prior to data acquisition. These checks are to ensure
that the controllers and scripts for the apparatus cause it to
behave as it should, and that its use will not disrupt or damage
other parts of the beamline.
2. Calibrate the beamline by exposing a calibrant material for a

few seconds (see Note 9). Make sure that the calibrant, which is
a tissue or chemical with a known periodicity of appropriate
dimensions that relates to the experimental specimen, is posi-
tioned in the same position axially along the path of the X-ray
beam as where the experimental sample will later be placed. For
this reason, it is best to use identical specimen cells and cell
holders for the calibrant and for the specimens under
investigation.
3. For collection of static data, weigh the specimen, wrap
itand place it within the specimen cell. All specimen handling
must be carried out in the appropriate laboratory, and only
sealed samples taken to the beamline. All synchrotrons have
laboratories close to the beamlines.
4. Mount the specimen cell in the X-ray beam (Fig. 4).
5. Locate the position of the sample with respect to the beam.
This can be done in numerous ways. If an in-line microscope is
available, this may be used to either locate and record the
motor coordinates that correspond to the edges of the speci-
men (necessary for a raster scan) or to align the beam with a
desired position on the specimen ready for data collection
(Fig. 5a). In some cases, it may be difficult to see a transparent
sample wrapped in transparent film, so the use of a fine marker
pen to highlight the region of interest is recommended
(Fig. 5b).
6. If an in-line microscope is not available, a piece of X-ray sensi-
tive paper will be provided at the synchrotron. This can be
adhered to the front of an empty specimen cell and a brief
X-ray exposure will provide a mark of the beam on the paper
(Fig. 5c). As the mark on the paper corresponds to known
motor coordinates, it can be used as an aid for the consistent
positioning of samples within the cell or for making adjust-
ments to the position of the sample holder to ensure that the
Fig. 5 Location of sample with respect to X-ray beam can be achieved in various ways. (a) An in-line
microscope is used to locate the edges of a corneal strip (highlighted in red). (b) A fine tip marker pen is used
to highlight the region of interest. (c) X-ray sensitive paper is used to identify the position of the beam
sample lines up with the beam. Alternatively, perform a diode

scan across the sample and cell. The coordinates of the sample
edges can be accurately measured from the plot of absorbance
(absorbance increases across the sample and drops off sharply at
either side of the sample).
7. For raster scans, it is a good idea to do a pilot scan with a large
raster increment, to check that satisfactory images are being
recorded across the area of interest; then a final scan can be
carried out with the desired increment (see Note 10).
8. Leave, then seal off the hutch following the safety protocols.
9. The X-rays are controlled from outside the hutch using
in-house software for the use of which training will be
provided. The software allows one to move the motorized
platform on which the specimen holder is mounted and to
open the shutter to allow X-rays to pass through the specimen
(at all times the hutch remains sealed). For time-resolved stud-
ies, the data acquisition script will control the whole apparatus
during the experiment. When conducting an experiment, make
sure any critical variables are shown on-screen so that in the
case of a malfunction or failure, the experiment can be quickly
terminated and later restarted after corrective action.
10. Determine a suitable X-ray exposure for the sample under
investigation. This can be done on a separate test sample that
is similar to the ones to be examined. Record a series of X-ray
scatter patterns each using a different exposure time, making
sure to displace the specimen slightly between each to avoid
exposing the same region twice. The length of the exposure
will vary depending on the sample and the X-ray source. Start
with a 0.5 s exposure and gradually increase the exposure time
in 0.1–0.5 s increments up until the point when no further
gains, in terms of X-ray scattering intensity, can be achieved.
Beyond this point, longer exposures will start to cause excessive
specimen damage and the quality of the pattern will reduce.
During this optimization period, great care should be taken to
avoid exceeding the maximum intensity limits of the detector
and causing irreversible damage to the beamline equipment.
Choose the minimum exposure time needed to get an accept-
able scatter pattern.
11. Collect the required X-ray scatter patterns from the specimens
to be investigated, remembering to log all relevant details of
each specimen for future identification and reference purposes.
3.4 Data Analysis X-ray scatter images produced at synchrotrons are typically
acquired on Dectris units, which output data as HDF files (hierar-
chical data format), appearing as FILENAME.h5. These files can be
read and analyzed using DAWN (data analysis workbench), a
versatile, open-source software package that is designed and devel-

oped by synchrotron scientists. Alternatively, the local contact for
the beamline can convert the HDF files into a more common image
format (.TIF, usually), allowing use of an alternative analysis pack-
age, such as SAXS4COLL, which has been designed by our group
specifically for the analysis of corneal X-ray scattering data
[9]. Alongside the image files, variables describing the experiment
(such as date and time, motor positions, camera length, diode light
intensities, and any parameters for synchronized user equipment)
will be saved in a nexus file (FILENAME.nxs), which can also be
opened in DAWN or converted to text by beamline staff. While the
HDF/nexus file formats are standards most synchrotron beamlines
have now adopted, some beamlines still use the CBF file format
(crystallographic binary file), which amalgamates the image and
parametric data. CBF files can also be opened and analyzed in
DAWN, and beamline staff can convert them to more common
image formats if necessary. The ESRF synchrotron uses a proprie-
tary EDF format, which can be opened and converted in its own
EDF Explorer software. Whether using DAWN or an alternative
package to analyze the data, the list below outlines the essential
steps. Note that several data reduction steps are omitted here,
which are briefly outlined in Note 11.
1. Use the scatter pattern from the calibrant to find the center
point and determine the real spacing for a reference point on
the image, from which the real spacing of any feature can then
be extrapolated. This calibration will be invariant throughout
the experiment if the sample position and beamline optics were
not changed. If, for instance, silver behenate was used, it will
produce circular diffraction rings, and a circle can thus be fitted
to one of these reflections to find the center point, and the
radius of the circle will correspond to a real spacing of 5:838n nm,
where n is the order of the X-ray reflection. Use of the inner-
most ring (first order reflection), which corresponds to a real
spacing of 5.838 nm, is recommended due to its relatively high
intensity which facilitates fitting.
2. Calibrated images should then be converted into a polar coor-
dinate system (r,φ) for ease of analysis, and the scatter
integrated over all values of φ to provide radial intensity data.
3. Fill in lost data from detector module edges. There are regions
of synchrotron detectors that do not contain pixels (see Fig. 6a,
b). These regions can be mostly filled using reflection about the
center, assuming the sample under investigation exhibits two-
fold symmetry (collagen fibrils and molecules exhibit this sym-
metry). As long as the beam is not aligned with the center of a
detector module, reflection can remove most of this dead
space. The remaining dead space can then either be ignored
Fig. 6 Representative SAXS (a) and WAXS (b) images from a human cornea, with
corresponding radial intensity plots (c, d) below. The black circles at the center
show the beam stop which blocks the main X-ray beam which passes straight
through the cornea, and the black lines mark the edges of detector modules
where there are no pixels. Labeled peaks arising from the X-ray reflections
correspond to the following structural features: IF—collagen interfibrillar Bragg
spacing (~52 nm in humans); M3 and M5—third and fifth meridional reflections
from the collagen axial periodicity (~65 nm in humans); B—the square of the
fibril (cylinder) transform takes the form of a Bessel function squared, and
provides a measure of fibril diameter (~33 nm in humans); IM—intermolecular
Bragg spacing (1.6 nm in humans). From [7], by license (CC BY https://
creativecommons.org/licenses/by/4.0/)
(note that azimuthal distributions may then be incomplete), or

interpolation can be used to fill them.
4. Remove background scatter. Background scatter can arise from
components in the tissue other than the fibrillar collagen as well
as from Cling Film, etc. The nature of the background subtrac-
tion will depend on the package used. DAWN now has an
accurate Porod scatter removal tool, along with specific feature
removal functionality (for removing features associated with
the specimen cell and beamline). A simple way for removing
background is to plot the radial intensity data as a function of
radius on log axes (see Fig. 6c, d). Given that Porod scatter
follows a power law, this background will appear as a straight
line on a log plot, and thus is easy to fit. With the Porod
background removed, the main features of the data should be
clear.
5. Equatorial reflections associated with fibril spacing and diame-
ter (which appear on a plot of radial intensity as a Gaussian and
Bessel function, respectively—see Fig. 6a, c) overlap and need

to be separated. The “lobes” of the Bessel function can be fitted
(taking care not to let meridional reflections skew this fit),
giving the square of what is commonly referred to as the
cylinder transform [2]. The fit will take the form:

2J1 ðkr f Þ 2
F 2 ðkr f Þ ¼ ð2Þ
kr f
where J1 is a first order Bessel function of the first kind, k is
the wavevector, and rf is the fibril diameter. The radial intensity
can then be divided by the cylinder transform to yield the
interference function, from which the fibril spacing can be
extracted.
6. Determining peak positions and widths can be accomplished in
a variety of ways, and it is advisable to check by eye the good-
ness of fit for any fitting algorithms before accepting their
results. The simplest way to fit a peak is to take the position
of the highest value, and the distance between the points
closest to 50% of the highest value. While this avoids fitting a
function to the data, the fit is limited by the resolution (features
such as meridional peaks can be as little as 5 pixels wide). If the
peak is symmetrical, then a Gaussian can usually be fitted
accurately, from which peak position and width can be precisely
determined. If the peak is asymmetrical or has a shoulder, then
a multiple Gaussian fit can be used, but take care to constrain
each Gaussian to the approximate radial position of each fea-
ture to avoid errors in fitting. Lastly, where a peak is irregular in
shape and a multiple Gaussian fit is not suitable, a cubic spline
can be used.
7. With an approach to peak fitting finalized, the azimuthal distri-
bution of the feature can be obtained by applying the fit to the
radial intensity distribution for each value of φ. Note that
angle-dependent radial intensity distributions will by much
noisier than those integrated over all angles, and peak fitting
will be sensitive to this noise, as well as any artefacts on the
images.
4 Notes
1. At the time of writing, suitable synchrotron facilities that we

have used to study the cornea are available at the Diamond
Light Source (UK), ESRF (France), Soleil (France), EMBL
(Germany), APS (USA), SPring-8 (Japan), ALBA (Spain),
and the National Synchrotron Light Source at the Brookhaven
National Laboratory (USA).
2. Swelling or abnormally high hydration of a cornea is a signifi-

cant problem as it will change the values of many of the struc-
tural parameters being measured [10].
3. The intensity of the scattered X-rays will depend on several
factors, but in general, the thinner the specimen, the less
collagen and hence the weaker the scatter patterns. Water is a
strong X-ray scatterer, so too much water, even in thicker
specimens, will lead to weak scattering. Conversely, if the spec-
imen is too thick, X-ray absorption can reduce the scatter
intensity. As a rough guide, natural corneal tissue should not
be at a hydration much above 90%, or thicker than about
1–2 mm. If scattering from thin specimens is too weak, it can
be increased by either layering or rolling up the sample to
increase the thickness [11, 12] or, as a last resort, by adding
heavy metal stains to increase the scattering from the constitu-
ent proteins. In the latter case, to stain collagen, soak the
sample for about 20 min in a solution of 2% phosphotungstic
acid pH 1.7 followed by a 5-min wash in distilled water
[13]. Some biological hydrogels have a very high water-
content and it may be necessary to reduce this until the protein
concentration is increased enough to produce a detectable
X-ray scatter pattern.
4. From Bragg’s law, nλ ¼ 2d sinθ, we can calculate θ (half the
scattering angle) because we know n (the order of the meridio-
nal reflection we choose), λ (the wavelength of the X-rays (this
information will be provided by the beamline staff at the syn-
chrotron), and d (the spacing of the calibrant structure giving
rise to the diffraction pattern). If we imagine a right-angle
triangle in which a particular order is formed by X-rays scat-
tered through an angle, 2θ, subtended with the “straight-
through” beam direction (Fig. 3), then the specimen-to-detec-
tor distance (x) can be calculated from a simple tangent rela-
tionship as ½ tan1(y/x), where y is the distance from the
center of the X-ray scatter pattern to the chosen X-ray reflec-
tion. Knowing the specimen-to-detector distance, x, allows the
converse calculation to be done, whereby measuring y for a
periodic structure of unknown dimensions, d, that gives rise to
an X-ray reflection allows us to compute θ and thus d. This
approach is commonly utilized for the analysis of data from
WAXS experiments (where calcite is the calibrant). The situa-
tion is simplified somewhat for SAXS experiments because we
can use the “small-angle” approximation owing to the fact that
the base and the hypotenuse of the right-angle triangle seen in
Fig. 3 are essentially the same. This is because the specimen-to-
detector distance, at several meters, dwarfs that of the distance
from the center of the diffraction pattern to the chosen X-ray
reflection, which typically is on the cm scale. Thus, an unknown
scattering structure, d, can be calculated from the calibrant d-

value (i.e., 5.838 nm) multiplied by ycalibrant/ycornea, the ratio of
the distances of the X-ray reflection from the center of the
diffraction pattern for silver behenate and for cornea.
5. Some industrial users have a different mode of access to syn-
chrotron beam time. Academic users apply as they would for a
research grant. The application is reviewed and, if successful,
beam time is awarded. Synchrotrons operate 24 h a day, so a
team of people will be needed to utilize any time awarded
efficiently. Deadlines for applications are advertised on the
synchrotron websites. Some, but not all, beamtime awards
provide funding for travel to the synchrotron and on-site
accommodation for the duration of the experimental run.
6. For SAXS experiments the specimen-to-detector distance is
large, typically being in the region of 3 m on beamline 40XU
in the SPring8 facility in Japan [14], 2.5 m on beamline X12B
at the National Synchrotron Light Source at the Brookhaven
National Laboratory, USA [15], and 2–10 m on beamline
I22 at the Diamond synchrotron facility in Oxfordshire
[7, 16]. Once the specimen-to-detector distance has been set
for a particular experimental run (which typically would take
place over a few days, examining multiple specimens), it is
generally left unchanged for the whole of the run, and we
recommend that experiments are designed to avoid altering
the specimen-to-detector distance during the experiments. If
this is unavoidable, a significant amount of beamtime that
could more effectively have been used collecting data will be
lost because shifting the detector can take many hours, not least
because the diffracted X-rays pass through an evacuated tube
on their way to the detector and any significant shifts in the
detector’s position will require this to be reconfigured and
pumped down to vacuum once more.
7. Before attending the synchrotron session, decisions will need
to be made and discussed with beamline staff regarding the
wavelength of the X-rays that is available, the distance between
the specimen and the detector, and the size (in cross-section) of
the X-ray beam. Most facilities these days provide the option of
a standard beam width of ~60–400 μm or a microfocus beam
(dimensions from about 0.1 μm to about 20 μm). Microfocus
beams are useful for examining small samples, or for high-
resolution raster scans, but there is a trade-off: the greater the
beam size, the higher the signal-to-noise ratio, the smaller the
beam size, the higher the spatial resolution; so sometimes, the
largest non-microfocus beams are preferable.
8. To synchronize a piece of apparatus with the beamline’s sys-

tems, an EPICS (Experimental Physics and Industrial Control
System) controller needs to be written. This allows the appara-
tus to be controlled from a synchrotron computer. For many
applications, the beamline may already have suitable apparatus,
or your local contact could advise on a suitable brand/model
they already have an EPICS controller for. Once an EPICS
controller is in place, it can be incorporated into the scripts
that control data acquisition (typically written in the Java or
Python programming languages). The writing of EPICS con-
trollers and adaptation of data acquisition scripts can take
several hours or days for a dedicated synchrotron technician,
so it is advisable to discuss the requirements of an experiment as
far in advance as possible.
9. The specimen-to-detector distance is fundamental for the cal-
culation of collagen dimensions in the cornea using SAXS/
WAXS, but rather than measure this distance directly it is
calculated by obtaining X-ray scatter patterns from materials
with structures of known dimensions, such as silver behenate or
powdered calcite. The calibrant is placed in a specimen cell,
ordinarily between two sheets of Mylar, and a short X-ray
exposure taken for several seconds. In each case this gives rise
to a characteristic X-ray scatter patterns with reflections that
arise from the 5.838 nm or 0.305 nm periodicities for SAXS or
WAXS, respectively. The pattern from silver behenate or cal-
cite is a series of concentric rings. Historically, rat tail tendon
has been used as a calibrant; however this has fallen out of use
due to the scattering angle being more variable than for other
calibrants.
10. The beamline manager or technical assistant will be able to
provide details about the size in cross section of the incident
focused X-ray beam. As they penetrate the specimen, the
intense X-rays damage the specimen. This is not a problem as
the patterns are recorded before the damage occurs, but it
would be a problem if successive exposures overlap. Therefore,
do not choose a raster increment smaller than the beam dimen-
sions as successive exposures will overlap. The increment cho-
sen for the pilot run needs to be large because the X-rays will
damage the points of the specimen they penetrate, which will
then not give good scatter patterns on the final raster scan.
11. While a good approximation of feature sizes can be made with
few data reduction steps when peaks are large and clear, in
order to maximize the accuracy and precision of the measure-
ments it is important to be careful with these initial steps.
Compensating for beam flux variations, background radiation,

variations in pixel solid angle with radius, etc. will help you
obtain the best possible results from your sample. For more
information, see ref. [17].
Acknowledgements
Our corneal X-ray program has been supported by many organiza-

tions, but principally by the UK Medical Research Council
(Grants G0001033; G0600755MR; MR/K000837/1 and MR/
S037829/1).
References
1. Meek KM, Hayes S (2018) Corneal stroma: 8. Fullwood NJ, Meek KM (1994) Ultrastruc-
collagen ultrastructure and orientation in tural, time-resolved study of freezing in the
health and disease. In: Roberts C, Dupps WJ, corneal stroma. J Mol Biol 236(3):749–758.
Downs JC (eds) Biomechanics of the eye. Kug- https://doi.org/10.1006/jmbi.1994.1187
ler, Amsterdam, pp 15–30 9. Abass A, Bell JS, Spang MT, Hayes S, Meek
2. Meek KM, Quantock AJ (2001) The use of KM, Boote C (2017) Saxs4coll: an integrated
x-ray scattering techniques to determine cor- software tool for analysing fibrous collagen-
neal ultrastructure. Prog Retin Eye Res 20 based tissues. J Appl Crystallogr
(1):95–137. https://doi.org/10.1016/ 50:1235–1240. https://doi.org/10.1107/
S1350-9462(00)00016-1 S1600576717007877
3. Meek KM, Boote C (2009) The use of x-ray 10. Meek KM, Fullwood NJ, Cooke PH, Elliott
scattering techniques to quantify the orienta- GF, Maurice DM, Quantock AJ, Wall RS,
tion and distribution of collagen in the corneal Worthington CR (1991) Synchrotron x-ray-
stroma. Prog Retin Eye Res 28(5):369–392. diffraction studies of the cornea, with implica-
https://doi.org/10.1016/j.preteyeres.2009. tions for stromal hydration. Biophys J 60
06.005 (2):467–474. https://doi.org/10.1016/
4. Daxer A, Fratzl P (1997) Collagen fibril orien- S0006-3495(91)82073-2
tation in the human corneal stroma and its 11. Connon CJ, Nakamura T, Hopkinson A,
implication in keratoconus. Invest Ophthalmol Quantock A, Yagi N, Doutch J, Meek KM
Vis Sci 38(1):121–129 (2007) The biomechanics of amnion rupture:
5. Boote C, Dennis S, Huang YF, Quantock AJ, an x-ray diffraction study. PLoS One 2(11):
Meek KM (2005) Lamellar orientation in e1147. https://doi.org/10.1371/journal.
human cornea in relation to mechanical prop- pone.0001147
erties. J Struct Biol 149(1):1–6. https://doi. 12. Worthington CR, Inouye H (1985) X-ray-dif-
org/10.1016/j.jsb.2004.08.009 fraction study of the cornea. Int J Biol Macro-
6. Abass A, Hayes S, White N, Sorensen T, Meek mol 7(1):2–8. https://doi.org/10.1016/
KM (2015) Transverse depth-dependent 0141-8130(85)90057-1
changes in corneal collagen lamellar orienta- 13. Hayes S, Lewis P, Islam MM, Doutch J,
tion and distribution. J R Soc Interface 12 Sorensen T, White T, Griffith M, Meek KM
(104):20140717. https://doi.org/10.1098/ (2015) The structural and optical properties
rsif.2014.0717 of type iii human collagen biosynthetic corneal
7. Bell JS, Hayes S, Whitford C, Sanchez- substitutes. Acta Biomater 25:121–130.
Weatherby J, Shebanova O, Vergari C, Winlove https://doi.org/10.1016/j.actbio.2015.07.
CP, Terrill N, Sorensen T, Elsheikh A, Meek 009
KM (2018) The hierarchical response of 14. Quantock AJ, Boote C, Young RD, Hayes S,
human corneal collagen to load. Acta Biomater Tanioka H, Kawasaki S, Ohta N, Iida T, Yagi N,
65:216–225. https://doi.org/10.1016/j. Kinoshita S, Meek KM (2007) Small-angle
actbio.2017.11.015 fibre diffraction studies of corneal matrix struc-
ture: a depth-profiled investigation of the
human eye-bank cornea. J Appl Crystallogr 40: Shebanova O, Meek KM (2017) The structural
S335–S340. https://doi.org/10.1107/ response of the cornea to changes in stromal
S0021889807005523 hydration. J R Soc Interface 14
15. Quantock AJ, Kinoshita S, Capel MS, Schan- (131):20170062. https://doi.org/10.1098/
zlin DJ (1998) A synchrotron x-ray diffraction rsif.2017.0062
study of developing chick corneas. Biophys J 17. Pauw BR, Smith AJ, Snow T, Terrill NJ, Thu-
74(2):995–998. https://doi.org/10.1016/ nemann AF (2017) The modular small-angle
S0006-3495(98)74023-8 x-ray scattering data correction sequence. J
16. Hayes S, White T, Boote C, Kamma-Lorger Appl Crystallogr 50:1800–1811. https://doi.
CS, Bell J, Sorenson T, Terrill N, org/10.1107/S1600576717015096
INDEX
A E
Adeno-associated virus vectors.................................77–98 Endothelial cells .......................................................17–26,
Air-liquid interface ...................................... 104, 113, 117 51, 52, 78
Amniotic membrane (AM) ............................36, 143–155 Epithelial cells.................................................... 13, 29–36,
Animal models..................................................... 199, 201, 47, 51, 52, 79, 104–107, 113, 115, 225
212, 215–229 Extracellular matrix (ECM).........................................1, 8,
31, 51, 116, 119, 120, 124, 125, 133–135,
C 159–167, 169
Cell cultures....................................................... 2–4, 8, 11,
F
14, 30, 31, 34–36, 39–48, 52, 53, 81, 104–113,
119–121, 124, 125, 133, 136, 164, 165, 187, Femtosecond lasers .............................................. 197–213
188, 191, 194 Freeze drying............................................... 144, 191, 195
Cell self-assembly .......................................................... 119
Cell sheets............................................................... 35, 113 G
Chemotaxis.......................................................... 185–188, Gene editing ..............................................................59–75
190, 192, 194, 195
Gene therapy ...................................................... 59, 77–79
Clustered regularly interspaced palindromic repeats
(CRISPR)..................................................... 59, 60, H
62–65, 67, 73, 74
Collagen.............................................................1, 2, 4, 13, Hydrogels ............................................................ 159–167,
25, 30, 51, 78, 103, 104, 115, 116, 119, 120, 172, 179, 191, 192, 194–196, 234, 243
122–124, 126, 127, 130–132, 135, 138, 139,
I
159, 160, 169–182, 188, 191–196, 198,
231–233, 235, 240, 241, 243, 245 Immunocytochemical staining ..................................... 188
Collagen-like peptides (CLPs) ........................... 169–174,
176–178, 180 K
Cornea ...............................................................1, 2, 4, 13,
Keratocytes ............................................................ 1, 2, 51,
17, 18, 20, 21, 25, 29, 30, 34, 42, 43, 51–57, 74,
52, 56, 78, 119, 159, 226
78, 79, 94, 97, 98, 103–117, 119, 159–165, 169,
Keratoplasty ......................................................... 2, 4, 198,
179, 185, 197–202, 204–207, 209–212, 215,
201, 202, 204, 211
217, 219, 220, 222, 224, 225, 227, 228,
231–235, 237, 240, 241, 243–245 L
Corneal fibroblasts .............................................. 105–108,
110, 115, 116, 119–140, 188, 191, 194 Limbal stem cells.................................................. 2, 39–47
Corneal repair...............................................143–155, 169
M
D Macromolecular crowding (MMC) .................... 119–140
Decellularization ................................................. 159, 160,
162, 164, 165
https://doi.org/10.1007/978-1-0716-0599-8, © Springer Science+Business Media, LLC, part of Springer Nature 2020
249
CORNEAL REGENERATION: METHODS AND PROTOCOLS
250 Index
O Stromal cells .................................................................. 185
Surgery...............................................................5, 30, 152,
Organoids ..................................................................51–57 197–213, 215–217, 219–221, 223, 224, 226–229
P V
Pluripotent stem cells................................................51–57 Viral vectors..................................................................... 94
S X
Scaffolds................................................................. 60, 159, X-ray diffraction ................................................... 231–246
160, 169, 197

(Methods in Molecular Biology 2145) Mark Ahearne - Corneal Regeneration - Methods and Protocols-Springer US - Humana (2020)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Methods in Molecular Biology 2145) Mark Ahearne - Corneal Regeneration - Methods and Protocols-Springer US - Humana (2020)

Uploaded by

Copyright:

Available Formats

Methods in

Molecular Biology 2145

Mark Ahearne Editor

For further volumes:

Methods and Protocols

ISSN 1064-3745 ISSN 1940-6029 (electronic)

© Springer Science+Business Media, LLC, part of Springer Nature 2020

Dublin, Ireland Mark Ahearne

1 Isolation and Culture of Corneal Stromal Stem Cells . . . . . . . . . . . . . . . . . . . . . . . . 1

14 Femtosecond Laser-Assisted Surgery for Implantation

MARK AHEARNE • Department of Mechanical and Manufacturing Engineering, School of

ELLE EDIN • Maisonneuve-Rosemont Hospital Research Centre, Montréal, QC, Canada;

MEHRDAD RAFAT • Department of Biomedical Engineering, Linköping University,

Isolation and Culture of Corneal Stromal Stem Cells

Cells isolated from the corneal stroma are known to develop a

2.2 Solutions 1. 5% Betadine povidone-iodine ophthalmic solution.

All procedures detailed here are in accordance with the directives of

removed using forceps, and then the corneal button is placed in

Fig. 4 Pieces of stroma in a 12-well plate with culturing medium

growth of adherent elongated or spindle-shaped cells from

3.3 Isolation Steps 1, 2, 3, and 6 are performed in a laminar flow hood.

Fig. 6 Image showing minced corneal tissue in a plastic Petri dish

Fig. 8 Image shows straining of the digest of stromal tissue

5. Observe the cell pellet and aspirate the supernatant carefully,

3.4 Sub-culturing Steps 2, 4, and 6 are performed in a laminar flow hood.

Fig. 10 Counting CSSCs in a Bürker chamber

4. Collect cell suspension using the pipette controller into a ster-

Fig. 11 Refreshing the culture medium

Fig. 12 Freezing CSSC aliquots

1. DMEM with low glucose content is the most commonly used

Fig. 13 Epithelial contamination in a culture with explanted stromal tissue

coating was omitted. Alternatively, a coating solution, contain-

The work presented was supported by the South-Eastern Norway

1. Forrester J, Dick AD, McMenamin PG, 49:17–45. https://doi.org/10.1016/j.pre

In Vitro Expansion of Corneal Endothelial Cells

Key words Cornea, Corneal endothelium, Cell culture, Cell expansion

The corneal endothelium consists of a monolayer of corneal endo-

to cover the damaged zone and reform the endothelial barrier

2.1 Labware 1. Filtration unit: The filtration unit is mounted with a 47 mm

2.2.2 Base Media 1. DMEM: Dulbecco’s Modified Eagle Minimal Essential

2. Penicillin/Streptomycin: Aliquot penicillin/streptomycin

2.3 Donor Tissues 1. Human globes or corneas unsuitable for transplantation

3.2 Cell Expansion Cell morphology (fibroblast-like or endothelial morphology;

3.2.2 CECs Passages Prior to cell passages, trypsin/EDTA must be warmed in a 37 C

1. Remove medium from confluent or nearly confluent CECs

Total cell counted

1. Avoid freeze-thaw cycle that could damage serum or additives.

complete expansion medium to seed them on FNC-coated

Procurement of eyes for research was possible thanks to a partner-

Primary Culture of Cornea-Limbal Epithelial Cells In Vitro

The cornea epithelium accounts for approximately 10% of the total

into cornea epithelia cells and has a greater proliferative potential

even when tissue is sourced from accredited tissue banks there is

2.1 Preparation 1. Recombinant epidermal growth factor (rhEGF) stock solution:

fitted with a 0.22 μm filter to sterilize the solution. Add 200 μg

2.3 Other solutions 1. PHEM Buffer: 60 mM 1,4-piperazinediethanesulfonic acid

11.902 g of EGTA, 0.203 g of MgCl2. Add 400 mL of UHP

2.4 Immuno- 1. Methanol.

3.1 Option A: 1. Culture NIH 3T3 fibroblast cells in DMEM supplemented

3.2 Option B: 1. Thaw overnight at 4 C cyro-preserved human amniotic mem-

EDTA solution. Cover the dish and place in a 37 C incubator

8. The following day refeed explant cultures with growth media.

1. Avoid using media with phenol red during the irradiation

1. The shipping and culture media can be stored at 4 C for