an injectable gelatin-based conjugate incorporating egf

231
An Injectable Gelatin-Based Conjugate Incorporating EGF Promotes Tissue Repair and Functional Recovery After Spinal Cord Injury in a Rat Model By Adhvait M. Shah BS, Biomedical Engineering Tufts University, 2012 SUBMITTED TO THE HARVARD-MIT PROGRAM IN HEALTH SCIENCES AND TECHNOLOGY IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN MEDICAL ENGINEERING AND MEDICAL PHYSICS AT MASSACHUSETTS INSTITUTE OF TECHNOLOGY FEBRUARY 2019 © 2019 Adhvait M. Shah. All rights reserved. The author hereby grants to MIT permission to reproduce and distribute publicly paper and electronic copies of this thesis document in whole or in part in any medium now known or hereafter created. Signature of Author: Harvard MIT Program in Health Sciences & Technology December 21, 2018 Certified by: Myron Spector, PhD Professor of Orthopedic Surgery (Biomaterials), HMS Department of Orthopedic Surgery, BWH Director, Tissue Engineering/Regenerative Medicine Laboratory, VABHS Mechanical engineering and HST, MIT Thesis Supervisor Accepted by: Emery N. Brown, MD, PhD Director, Harvard-MIT Program in Health Sciences and Technology Professor of Computational Neuroscience and Health Sciences and Technology

Upload: others

Post on 25-Oct-2021

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: An Injectable Gelatin-Based Conjugate Incorporating EGF

An Injectable Gelatin-Based Conjugate Incorporating EGF Promotes Tissue

Repair and Functional Recovery After Spinal Cord Injury in a Rat Model

By

Adhvait M. Shah

BS, Biomedical Engineering

Tufts University, 2012

SUBMITTED TO THE HARVARD-MIT PROGRAM IN HEALTH SCIENCES AND TECHNOLOGY IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE

DEGREE OF

DOCTOR OF PHILOSOPHY IN MEDICAL ENGINEERING AND MEDICAL PHYSICS AT

MASSACHUSETTS INSTITUTE OF TECHNOLOGY

FEBRUARY 2019

© 2019 Adhvait M. Shah. All rights reserved.

The author hereby grants to MIT permission to reproduce and distribute publicly paper and electronic copies of this thesis document in whole or in part in any medium now

known or hereafter created.

Signature of Author:

Harvard MIT Program in Health Sciences & Technology December 21, 2018

Certified by:

Myron Spector, PhD

Professor of Orthopedic Surgery (Biomaterials), HMS

Department of Orthopedic Surgery, BWH

Director, Tissue Engineering/Regenerative Medicine Laboratory, VABHS

Mechanical engineering and HST, MIT

Thesis Supervisor

Accepted by:

Emery N. Brown, MD, PhD

Director, Harvard-MIT Program in Health Sciences and Technology

Professor of Computational Neuroscience and Health Sciences and Technology

Page 2: An Injectable Gelatin-Based Conjugate Incorporating EGF

2

An Injectable Gelatin-Based Conjugate Incorporating EGF Promotes Tissue

Repair and Functional Recovery After Spinal Cord Injury in a Rat Model

By

Adhvait M. Shah

Submitted to the Harvard-MIT program in Health Sciences and Technology on December 21st, 2018 in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in

Medical Engineering and Medical Physics

ABSTRACT

Spinal cord injury (SCI) is a devastating condition drastically reducing the quality of life that affects

about 300,000 patients in the USA. As a result of the injury, sensory perception and motor

functions are lost. Current treatments do not address the root cause - degeneration and loss of

neural tissue. The overall goal of this pre-clinical work was to evaluate a novel gelatin-based

conjugate (gelatin-hydroxyphenyl propionic acid; Gtn-HPA) capable of undergoing covalent

cross-linking in vivo after being injected as a liquid. Gtn-HPA incorporating epidermal growth

factor (EGF) and/or stromal cell-derived factor - 1α (SDF-1α) was evaluated for promoting tissue

healing and functional recovery using a standardized 2-mm hemi-resection SCI rat model, four

weeks after injection.

Injection of Gtn-HPA/EGF immediately after the surgical excision injury significantly improved

motor functional recovery, compared to gel alone and non-treated controls. Bladder function was

also improved in Gtn-HPA/EGF-treated animals. Functional improvement correlated with the

amount of spared tissue. The volume of gel in the defects was quantified by a newly developed

MRI-based method employing T1-weighted inversion recovery to unambiguously image Gtn-HPA

in the injury site in a non-destructive manner. Histological analysis showed the presence of

multiple islands of Gtn-HPA in the injury site after four weeks. There was a significantly greater

number of cells migrating into the Gtn-HPA/EGF, compared to the gel alone, and these cells

displayed neural progenitor cell markers: nestin, vimentin, and Musashi. The cells infiltrating Gtn-

HPA were negative for glial fibrillary acidic protein (GFAP), a marker for astrocytes. Injection of

the gel reduced the reactive astrocytic presence at the border outlining the injury site indicating

the reduction of glial scar. There was no notable inflammatory response to the Gtn-HPA gel,

reflected in the number of CD68-positive cells, including macrophages. Of note was the

demonstration by immunohistochemistry that the Gtn-HPA remaining at 4 weeks post-injection

contained EGF. MMP2 was found to be playing a role in in vivo degradation of the Gtn-HPA gel.

Additional behavioral and histological results were acquired injecting Gtn-HPA/EGF in 2-mm

complete resection SCI rat model. Collectively, the findings signaled that injury sites injected with

Gtn-HPA/EGF had greater potential for regeneration.

In summary, this work commends an injectable, covalently cross-linkable formulation of Gtn-HPA

incorporating EGF for further investigation in promoting functional recovery and potential

regeneration for treatment of SCI and thereby improve the quality of life of SCI patients.

Page 3: An Injectable Gelatin-Based Conjugate Incorporating EGF

3

Committee Members:

Thesis Supervisor:

Myron Spector, PhD

Professor of Orthopedic Surgery (Biomaterials), HMS

Department of Orthopedic Surgery, BWH

Director, Tissue Engineering/Regenerative Medicine Laboratory, VABHS

Mechanical Engineering and HST, MIT

Thesis committee Chair:

Michael J. Cima, PhD David H. Koch Professor of Engineering

Professor of Materials Science and Engineering, MIT

Thesis Readers:

Zhigang He, PhD, BM

Professor of Neurology and Ophthalmology, HMS

F. M. Kirby Neurobiology Center, BCH

Srinivasan Mukundan, Jr., MD, PhD

Department of Radiology, BWH

Associate Professor of Radiology, HMS

Page 4: An Injectable Gelatin-Based Conjugate Incorporating EGF

4

Acknowledgements

I am enormously grateful to my thesis advisor Dr. Myron Spector. I was inspired

during my first course with him at MIT. He has always supported me for my professional

and personal growth throughout my doctoral work. He has always given me ‘in-person’

time whenever I needed it. I was able to accomplish the work done in this project as a

result of various discussions with him. His suggestions have brought us closer to our goal.

He has been very accommodating of my other commitments as well. He has been my

guiding light in difficult days. Overall, he has groomed me into an able research scientist

and critical thinker. I could not have asked for a better mentor, teacher and motivator. He

is a legend and role model in my view.

I would like to thank my colleagues in the Tissue engineering/regenerative

medicine laboratory at Veterans Affairs Medical Center at Jamaica Plains. I thank Dr. Hu-

ping Hsu for his time and assistance with the animal surgeries. His excellent and

experienced hands created consistent and standardized defect in our rat model. He has

been an excellent mentor and teacher in the surgical suite. I would like to thank my current

lab members Dr. Wanting Niu, Christopher Love, and Dr. Yunfei Ma for their help in

various aspects of the project. Specifically, I wish to thank Wanting for her help with

animal care and in vitro experiments, Christopher Love for discussions and help with

mechanical testing, and Yunfei for his help with imaging of slides and ELISA. I also wish

to express my gratitude to previous lab members Dr. Daniel Macaya, Dr. Ravi Rajappa

and Dr. Teck Chuan Lim for laying the foundation of my work and teaching me various

skills that were crucial in successful completion of the doctoral work. I truly felt that our

lab operated as a family with celebrations and support in each other’s work. I thank Dr.

Spector for giving me an opportunity to work in stimulating and supportive environment.

I would like to express my gratitude to my thesis committee: Dr. Michael Cima, Dr.

Zhigang He and Dr. Srinivasan Mukundan. Their time, invaluable feedback and genuine

interest in my work is truly appreciated. Their feedback has led to form more

collaborations and pursue interdisciplinary projects that included MRI. Discussions with

all three members have given me broader view of my project and its impact on the patients

suffering from spinal cord injury. It was my honor to have them on my committee.

I would like to thank my collaborators for their assistance in various aspects of the

thesis project. Dr. Motoichi Kurisawa at A*STAR in Singapore provided the Gtn-HPA

conjugate to fabricate various formulations of the Gtn-HPA. I am grateful to Dr. Yaotang

Wu and Michael Marcotrigiano for their technical assistance with MRI at SAIL facility in

Children’s Hospital. I would like to acknowledge Dr. Srinivasan Mukundan, Dr. Samuel

Patz, Dr. Sharon Peled and Tehya Johnson for their MRI expertise at SAIL facility at

Brigham and Women’s hospital. I am thankful to the staff, Diane Ghera and Sylvia Smith

at animal research facility (ARF) at Veterans Affairs Medical center, Jamaica Plains for

their assistance in the animal surgeries and animal care. I am thankful to Dr. Bertrand

Huber for his help with automated slide scanner at VA Jamaica Plains campus. This work

Page 5: An Injectable Gelatin-Based Conjugate Incorporating EGF

5

was possible due to the support of these collaborators. I would also like to thank anybody

who has directly and indirectly assisted me professionally with my research work.

The funding for this work was provided by the following sources: Veterans Affairs

administration, Harvard-MIT Division of Health, Science and Technology (HST) fellowship

and Neuroimaging Training Program (NTP) fellowship from NIBIB at NIH.

Personal acknowledgements:

I would like to dedicate this thesis work to my grandmother, Madhurika Shah. She

would have been the happiest person to see me finish my doctoral degree. As a teacher

herself, the seeds of curiosity and education were sown in me by her very early in my

childhood. Her smile and accommodating nature to all situations is very inspiring. Words

are not enough to express my gratitude to her – I miss you Ba!

I would like to thank my parents – Mitesh Shah and Dipti Shah – from the bottom

of my heart. They have supported all my endeavors from my childhood to today. My father

has worked very hard to ensure that we have all the comforts of the world. There have

been times when he had to stay away from the family as well. Moreover, his

inquisitiveness and dedication are an inspiration for me. He has left no stone unturned to

ensure we have the best of education. Thank you pappa! My mother is a very strong

woman – she has endured all the hardships with a smile! She has instilled discipline in

me which has been very helpful in my life as a doctoral student. Thank you mammy! Both

of you have always encouraged me to pursue my dreams and supported me in difficult

times. Your upbringing has made me who I am today! Thank you both for being my

parents.

A special thank you to my wife, Rinkal. Rinkal, my life has found a new meaning

due to your cheerfulness, support and loving attitude. You are the most selfless person

that I have known. We have been with each other and supported each other during the

various highs and lows of our life. Your unwavering support and words ‘everything will be

good’ have given me hope during the tough times during my last two years of PhD. I

cherish our discussions and all the fun times that we have spent together. I look forward

to our journey together. I am extremely proud of your achievements in such a short time

and look forward to celebrating our graduations together. I love you!

I would like to thank my brother, Namank. Namank, you truly are an inspiration to

many. You have set an example of how an ideal life should be. You have excelled in all

aspects of your life. You have been extremely supportive during my years in the PhD.

You have taken on more responsibility so that I can focus on my studies. You have

provided continued technical guidance as necessary, especially when my laptop crashed.

You have saved me from panic attacks. Moreover, I truly cherish our car rides together. I

will always remember your support and feedback during my qualifying exam preparation.

Now that I will be more available, I look forward to hanging out with you and Priya.

Page 6: An Injectable Gelatin-Based Conjugate Incorporating EGF

6

I would like to also acknowledge continued support of my extended family

members here in the USA and in India. Your company and conversations have made my

time as graduate student more enjoyable. I would like to thank all my friends at MIT,

Cityview and in India. I sincerely appreciate your love and support. I will leave MIT with

fond memories, which shall be cherished throughout my life.

In the end, I would like to express my immense gratitude to revered Pandurang

Shastri Athavale, inspirer of the Swadhyay movement all across the globe. Pujniya

Dadaji, your thoughts have guided me and molded my life for the better. You have given

me the tools to develop myself in all the aspects of my life. Your path has taken me closer

to God. You are truly responsible for giving me a comprehensive view of the world –

external and internal. I can’t imagine the state of my life had you not come into my life. I

will give my best to live according to your teachings with the guidance of Aadarniya Didiji.

Jay Yogeshwar!

Page 7: An Injectable Gelatin-Based Conjugate Incorporating EGF

7

Table of Contents

Acknowledgments ________________________________________________________ 4

Table of Contents _________________________________________________________ 7

Chapter 1

1. Introduction ___________________________________________________________ 11

1.1. Spinal cord injury – clinical motivation

1.2. Hypothesis and specific aims

1.3. Summary of the chapters of the thesis

1.4. References

Chapter 2

2. Clinical motivation and background

2.1. Introduction to spinal cord injury _________________________________________ 17

2.1.1. Prevalence of spinal cord injury

2.1.2. Clinical presentation of SCI and functional impairments

2.1.3. Clinical management and current treatment

2.1.4. Anatomy and pathophysiology of SCI

2.1.5. Cavitary lesions – potential for regeneration and recovery

2.1.6. Pre-clinical SCI models – an overview

2.2. Tissue engineering approaches in SCI ____________________________________ 34

2.2.1. Three pillars of tissue regeneration – an overview

2.2.2. Scaffolding materials for SCI

2.2.2.1. Natural vs. synthetic materials

2.2.2.2. Pre-formed vs. injectable biomaterials

2.2.2.3. Injectable biomaterials for SCI

2.2.2.4. Criteria for biomaterials for cavitary lesions

2.2.3. Biomolecules for SCI – an overview

2.2.3.1. Neurotrophins and growth factors

2.2.3.2. ECM and scar modifying molecules

2.2.4. Cell-based treatment for SCI – an overview

2.3. References _________________________________________________________ 46

Chapter 3

3. Rationale and development of the formulation with Gtn-HPA matrix with EGF and

SDF-1α for spinal cord injury rat model evaluation

3.1. Biomaterial: Gelatin-hydroxyphenyl propionic acid (Gtn-HPA) __________________ 55

3.1.1. Introduction

3.1.2. Synthesis of Gtn-HPA

3.1.3. Mechanical properties of Gtn-HPA

3.1.4. Prior in vitro studies with Gtn-HPA

3.1.5. Gtn-HPA: promising biomaterial for spinal cord injury repair

3.2. Growth factor: Epidermal growth factor (EGF) ______________________________ 63

3.2.1. Introduction

3.2.2. Prior relevant in vitro studies with EGF

Page 8: An Injectable Gelatin-Based Conjugate Incorporating EGF

8

3.2.3. Prior in vivo studies incorporating EGF in SCI models

3.3. Chemokine: Stromal-cell derived factor-1α (SDF-1α) _________________________ 67

3.3.1. Introduction

3.3.2. Prior relevant in vitro studies with SDF-1α

3.3.3. Prior in vivo studies incorporating SDF-1α in SCI models

3.4. Two formulations of Gtn-HPA for injection in the SCI models ___________________ 71

3.5. Release profile of EGF from Gtn-HPA formulations for four weeks ______________ 72

3.5.1. Methods

3.5.2. Results and discussion

3.6. References _________________________________________________________ 74

Chapter 4

4. Investigating the effects of Gtn-HPA matrix with EGF and SDF-1α in a T8 2 mm complete resection rat model of spinal cord injury after four weeks 4.1. Introduction and clinical motivation _______________________________________ 79 4.2. Overall goal and hypotheses of the chapter ________________________________ 80

4.3. Methods ____________________________________________________________ 81

4.3.1. Gtn-HPA gel fabrication

4.3.2. Animal surgical procedure and Gtn-HPA injection

4.3.3. Experimental design

4.3.4. Animal care post-surgery

4.3.5. Behavioral assessment – BBB scale

4.3.6. Animal sacrifice and harvesting of the spinal cord

4.3.7. Tissue embedding procedure and cryosection

4.3.8. Histology and histomorphometric analysis

4.3.9. Immunofluorescent staining and analysis

4.3.10. Statistical Analysis

4.4. Results _____________________________________________________________ 90

4.4.1. Qualitative observations of the animals and Gtn-HPA injection

4.4.2. Functional evaluation using BBB scale

4.4.3. Gross histological outcomes

4.4.4. Histomorphometric evaluation of the injury site

4.4.5. Immunohistochemistry evaluation of the injury site

4.4.5.1. GFAP – astroytes

4.4.5.2. Iba1 – microglia/macrophages

4.4.5.3. NeuN – mature neurons

4.4.5.4. MMP2 – Gtn-HPA degradation

4.4.5.5. EGF – Gtn-HPA incorporated EGF

4.4.5.6. Nestin – neural stem cells

4.5. Discussion _________________________________________________________ 107

4.5.1. Functional recovery assessment

4.5.2. Histological evaluation of the injury site

4.5.3. Immunohistochemical evaluation of the injury site

4.6. References _________________________________________________________117

Page 9: An Injectable Gelatin-Based Conjugate Incorporating EGF

9

Chapter 5

5. Developing an MRI protocol to visualize the Gtn-HPA matrix in the spinal cord

5.1. Introduction and motivation ___________________________________________ 121

5.2. Overall goal and hypotheses of the chapter ______________________________ 123

5.3. Approach 1: T2-weighted MRI _________________________________________ 124

5.3.1. Introduction

5.3.2. Methods

5.3.2.1. Surgery procedure and experimental design

5.3.2.2. Sample preparation and acquisition of MRI

5.3.2.3. Image analysis

5.3.3. Results

5.4. Approach 2: Magnetic resonance elastography (MRE) ______________________ 128

5.4.1. Introduction

5.4.2. Methods

5.4.3. Results

5.5. Approach 3: T1-weighted Inversion Recovery (T1IR) ________________________ 130

5.5.1. Introduction

5.5.2. Preliminary phantom work - Relaxometry

5.5.2.1. Methods

5.5.2.2. Results

5.5.3. Preliminary phantom work – T1IR

5.5.3.1. Methods

5.5.3.2. Results

5.5.4. Ex vivo spinal cord T1IR – Protocol 1.0

5.5.4.1. Methods – surgical procedure and MRI acquisition

5.5.4.2. Results

5.5.5. Ex vivo spinal cord T1IR – Protocol 2.0

5.5.5.1. Methods - surgical procedure and MRI acquisition

5.5.5.2. Results

5.6. Summary and discussion _____________________________________________ 140

5.7. References ________________________________________________________ 144

Chapter 6

6. Investigating the effects of Gtn-HPA matrix with EGF and SDF-1α in a T8 2-mm hemi-

resection rat model of spinal cord injury after four weeks

6.1. Introduction and clinical motivation ______________________________________ 147

6.2. Overall goal and hypotheses of the chapter _______________________________ 149

6.3. Methods ___________________________________________________________ 150

6.3.1. Gtn-HPA gel fabrication

6.3.2. Animal surgical procedure and Gtn-HPA injection

6.3.2.1. Improvement in the dural replacement technique

6.3.3. Experimental design

6.3.4. Animal care post-surgery and bladder function evaluation

6.3.5. Behavioral assessment – BBB scale

6.3.6. Animal sacrifice and harvesting of the spinal column

6.3.7. MRI ex vivo scan of the spinal column

Page 10: An Injectable Gelatin-Based Conjugate Incorporating EGF

10

6.3.8. Gtn-HPA volume determination using MRI images

6.3.9. Cryoprotection, tissue embedding and cryosection

6.3.10. Histology and histomorphometric analysis

6.3.11. Spared tissue area and injury area quantification

6.3.12. Immunofluorescent staining and analysis

6.3.13. Statistical analysis

6.4. Results ____________________________________________________________ 163

6.4.1. Surgical observations and Gtn-HPA injection

6.4.2. Functional evaluation – bladder function

6.4.3. Functional evaluation – motor recovery (BBB scoring)

6.4.4. Spared tissue and injury area quantification

6.4.5. Spared tissue correlation with functional evaluation

6.4.6. Gtn-HPA gel volume measurement using MRI images

6.4.7. Gross histological outcomes

6.4.7.1. Control group – no injection

6.4.7.2. 8 wt% Gtn-HPA gel groups

6.4.7.3. EGF in suspension group

6.4.7.4. 12 wt% Gtn-HPA groups

6.4.8. Histomorphometric evaluation of the injury site

6.4.9. Immunohistochemistry evaluation of the injury site

6.4.9.1. GFAP – astrocytes

6.4.9.2. Iba1 - resident and/or activated microglia

6.4.9.3. CD68 – macrophages/monocytes

6.4.9.4. NeuN – mature neurons

6.4.9.5. MMP2 – Gtn-HPA gel degradation

6.4.9.6. EGF – Gtn-HPA incorporated EGF

6.4.9.7. Nestin – neural stem cells

6.4.9.8. vWF – endothelial cells

6.5. Discussion ________________________________________________________ 195

6.5.1. Functional evaluation

6.5.1.1. Bladder function – urine retention volume

6.5.1.2. Motor function – BBB scale

6.5.2. Visualizing Gtn-HPA using MRI

6.5.3. Histological evaluation of the injury site

6.5.4. Immunohistochemical evaluation of the injury site

6.6. References ________________________________________________________ 210

Chapter 7

7. Conclusions __________________________________________________________ 215

Chapter 8

8. Limitations and Future directions

8.1. Limitations of the thesis ______________________________________________ 218

8.2. Future directions ___________________________________________________ 221

Appendix ______________________________________________________________ 223

Page 11: An Injectable Gelatin-Based Conjugate Incorporating EGF

11

Chapter 1:

Introduction

Page 12: An Injectable Gelatin-Based Conjugate Incorporating EGF

12

1.1. Spinal cord injury – clinical motivation

Spinal cord injury (SCI) is a devastating disease with significant loss of neuronal function.

Any damage – mostly mechanical - to the spinal cord tissue resulting in temporary or permanent

changes in sensory, motor or autonomic function is defined to be SCI [1]. According to WHO,

roughly 250,000 new SCI cases are registered every year throughout the globe [2]. In the USA,

SCI currently affects 288,000 patients with around 17,700 new cases every year. 78% patients

are male with average age for the onset of injury 43 years. The leading causes of traumatic SCI

are automobile acccidents and falls that constitute roughly 70% of SCI patients [3]. In terms of

anatomy, injury in upper thoracic region could results in complete loss of function in all four limbs

(tetraplegia) while SCI at lower thoracic and lumbar level could lead to paraplegia depending upon

the severity of the injury. It is shocking that 67% of patients suffer from tetaplegia leading to

complete bed-rest [4]. As a result, they suffer secondary complications such as pressure ulcers

(52%) and depression (22.2%) [5, 6]. The lifetime costs could be as high as $5 million for

tetraplegic patients. It is reported that less than 1% of the patients experience complete gain in

the function after hospital discharge [3]. Clinically, three major approaches that mitigate the

secondary damage caused by the primary injury are taken – the cord is decompressed as a result

of pinching of bone fragments, acute anti-inflammatory agents are administered and rehabilitation

efforts are made for functional recovery. Although clinical ‘treatment’ of the SCI include

management of the symptoms, the root cause – degeneration of the neural tissue – is not

addressed.

SCI initiates a complex cascade of pathophysiological responses that hinder possibility of

any intrinsic regeneration of the severed axons. Various molecules that are released as a result

of myelin degradation. These molecules that are recruited to the injury site are inhibitors of

regeneration [7]. Moreover, the parenchyma and stroma of the tissue are lost post injury, often

resulting in a fluid-filled cavitary defect. Our thesis is that, an intrinsic regenerative process will be

facilitated by filling the fluid-filled cavity with a biomaterial matrix (also referred to as a scaffold)

capable of serving as a stroma to enable population of the defect with endogenous neural

reparative cells. Since the surgical intervention necessary of placing pre-formed scaffolds in the

injury site would cause further damage, we propose to develop a minimally invasive injectable

matrix that is permissive of endogenous cell migration, proliferation, and differentiation. Moreover,

injectable materials can also conform to the heterogeneous cavities post SCI, which are of

variable size and shape. In addition, these injectable biomaterials can incorporate signaling

molecules and factors that could promote a reparative/regenerative tissue response, eventually

functional recovery.

Page 13: An Injectable Gelatin-Based Conjugate Incorporating EGF

13

1.2. Hypothesis and specific aims

The overall goal of this thesis was to improve functional outcome after SCI through a new

treatment employing the tools of tissue engineering and regenerative medicine: a biomaterial

scaffold and recombinant regulatory molecules (viz., a growth factor and chemokine). While it

may not be possible to fully regenerate a spinal cord after its injury, the goal was to improve short-

and long-term outcomes, respectively, by: interrupting degenerative processes in the acute phase

post-SCI; and populating the defect with cells capable of re-establishing neural circuitry. To this

end, this thesis investigated a novel gelatin-based conjugate (gelatin-hydroxyphenyl propionic

acid; Gtn-HPA) capable of undergoing covalent cross-linking in vivo after being injected as a

liquid. Gtn-HPA incorporating epidermal growth factor (EGF) and/or stromal cell-derived factor -

1α (SDF-1α) was evaluated for promoting tissue healing and functional recovery using a

standardized 2-mm surgical excision SCI rat model, four weeks after injection. Specifically, the

supposition was that EGF and SDF-1α would act to reduce degenerative processes acutely post-

SCI and recruit endogenous neural stem cells to repopulate the defect in order to promote a

wound healing response for potential regeneration of certain neural processes. Moreover, the

goal was also to evaluate the neuroprotective effect and short-term functional recovery outcomes

four weeks post injection of Gtn-HPA, alone and incorporating EGF and SDF-1α.

Our principal hypothesis is that, when injected immediately after injury, the gelatin-based

gel (Gtn-HPA) incorporating EGF and SDF-1α results in: short term functional improvement (at

four weeks) post injury; and recruitment of endogenous neural stem cells (NSCs) into the gel-

filled lesion.

The specific aims of the thesis were as follows:

1) To develop and study the release of EGF from 8% and 12% Gtn-HPA for injection into in vivo

SCI model evaluated in specific aims 2 and 3, and to develop a magnetic resonance imaging

(MRI) protocol to visualize the Gtn-HPA gel, and to distinguish it from the surrounding tissue.

2) To investigate the potential of EGF and SDF-1α delivered by 8% Gtn-HPA matrix in two

different rodent SCI models (2 mm complete resection and 2 mm hemi-resection) four weeks post

injection.

3) To investigate the effects of 12% Gtn-HPA/EGF with an improved dural replacement technique

in a 2 mm hemi-resection model four weeks post injection.

Page 14: An Injectable Gelatin-Based Conjugate Incorporating EGF

14

1.3. Summary of the chapters of the thesis

This thesis is divided into eight chapters – the contents of which is summarized below:

Chapter 2 provides clinical motivation and a detailed background related to SCI. It discusses the

prevalence of the SCI and the functional deficits that the patients suffer post SCI followed by the

current practices of the clinical management after SCI. The chapter describes the temporal and

anatomical changes at the molecular level after the injury leading the formation of spinal cord

cavities usually seen in the patients. After describing relevant pre-clinical models to study SCI, a

detailed background into the prior regenerative research work both in vitro and in vivo to promote

regeneration of the defect lesion after SCI is discussed.

Chapter 3 gives background and rationale for the specific tissue engineering tools that were

employed in this thesis work – Gtn-HPA biomaterial and the factors EGF and SDF-1α. By way of

background, several prior in vitro studies characterizing the Gtn-HPA are presented, and the

results from the current ELISA assessment for EGF release from Gtn-HPA are presented over

four weeks are described as evaluated in the first specific aim. The chapter also presents the

history of the prior biomaterials studied in our laboratory for SCI regeneration. Importantly, the

rationale and the formulations of Gtn-HPA that were evaluated in Chapter 4 and Chapter 6 are

discussed.

Chapter 4 describes the evaluation of Gtn-HPA incorporating EGF and SDF-1α in a standardized

2-mm complete transection model of SCI for evaluation of second specific aim. These

experiments were guided by the optimizing the formulation of the Gtn-HPA that was described in

Chapter 3. The animals were assessed each week for their motor function followed by their

histological and immunohistochemical assessment of the injury site. The learnings from this

Chapter were applied in Chapter 6. In addition, a challenge regarding visualizing the Gtn-HPA in

a non-destructive manner was overcome in the work described in Chapter 5.

Chapter 5 describes various approaches that were taken to develop a noninvasive and non-

destructive imaging protocol to visualize the Gtn-HPA gel as per the first specific aim. The results

from three MRI-based approaches – T2-weighted imaging, Magnetic resonance elastography and

T1-weighed inversion recovery – are presented. The final protocol that was able to unambiguously

distinguish the Gtn-HPA based on the T1-weighted inversion recovery method is presented. All

the animals evaluated in Chapter 6 in hemi-resection model were imaged using this protocol.

Chapter 6 describes a comprehensive in vivo evaluation of two different – 8% and 12% - Gtn-

HPA formulations incorporating EGF and SDF-1α in a standardized 2-mm hemi-resection SCI

Page 15: An Injectable Gelatin-Based Conjugate Incorporating EGF

15

model as evaluated for specific aim 2 and 3. The chapter also described various surgical

improvements that were made based on the outcomes of Chapter 4. The formulation of the Gtn-

HPA was chosen based on the findings from Chapter 4. Histologic, immunohistochemical and

functional evaluation was performed on the rats post SCI. Specifically, motor function and bladder

function was evaluated to gain greater insights into the effects of Gtn-HPA/EGF matrix.

Furthermore, all animals were scanned using MRI-based T1-weighted inversion recovery protocol

developed in Chapter 5.

Chapter 7 details the conclusions that are supported based on the findings of the thesis work.

Chapter 8 discusses the limitations of this work and provides thoughts on the future investigation

in regard to the work presented in this thesis.

At the end, Appendix includes protocols and relevant information for meticulous planning of the

future experiments.

1.4. References

1. Staff, M.C. Spinal cord injury. 2014 10/8/2014 [cited 2014 12/20/2014]; Available from: http://www.mayoclinic.org/diseases-conditions/spinal-cord-injury/basics/definition/con-20023837.

2. Spinal cord injury. 2014 11/2013 [cited 2014 12/19]; Available from: http://www.who.int/mediacentre/factsheets/fs384/en/.

3. Facts and Figures at a Glance, in SCI Data Sheet. 2018, University of Alabama at Birmingham: National SCI Statistical Center. p. 2.

4. Ramer, L.M., M.S. Ramer, and E.J. Bradbury, Restoring function after SCI: towards clinical translation of experimental strategies. Lancet Neurol, 2014. 13(12): p. 1241-1256.

5. McKinley, W.O., et al., Long-term medical complications after traumatic SCI: A regional model systems analysis. Archives of Physical Medicine and Rehabilitation, 1999. 80(11): p. 1402-1410.

6. Williams, R. and A. Murray, Prevalence of Depression After SCI: A Meta-Analysis. Arch Phys Med Rehabil, 2014.

7. Schwab, M.E. and S.M. Strittmatter, Nogo limits neural plasticity and recovery from injury. Curr Opin Neurobiol, 2014. 27: p. 53-60.

Page 16: An Injectable Gelatin-Based Conjugate Incorporating EGF

16

Chapter 2:

Clinical motivation and background

Page 17: An Injectable Gelatin-Based Conjugate Incorporating EGF

17

2.1. Introduction to spinal cord injury

2.1.1. Prevalence of spinal cord injury: SCI is a debilitating condition that greatly affects

the quality of life for the patients. SCI is a clinically complex and life-changing condition. The

incidence of SCI is also high with roughly 17,700 new cases every year in the US - equivalent to

54 new cases per million in the USA [1]. As per the World Health Organization, SCI affects nearly

500,000 people worldwide and higher incidence of SCI of around 80 new cases per million

compared to the USA, which is a conservative estimate given the poor reporting of the cases in

various countries [2]. Leading causes of SCI are motor vehicle accidents (38%), falls (31%),

violence (13%), and sports (8%) [1]. These causes lead to differing level and extent of the SCI

depending upon the location and injury to the cord. Most severe form of the injury is injury in the

cervical region with severe mechanical insult, for example, resulting from a diving accident. Such

injury to the spinal cord would lead to complete paralysis of four limbs – complete tetraplegia.

Spinal cord injuries are typically categorized into four types – complete tetraplegia, incomplete

tetraplegia, complete paraplegia and incomplete paraplegia. Most SCI patients suffer from

incomplete tetraplegia (47%) followed by incomplete paraplegia (20%). Considering these two

most prevalent types of injury, lifetime costs for the patient could be as high as 4 million dollars if

they are injured at a young age [2]. It is reported that less than 1% patients experienced complete

recovery after the SCI. Furthermore, it is interesting to note that the average age of SCI has

increased from 29 in 1970s to 43 in 2018. There are two age groups where peaks are observed

in the incidence rates of the SCI – younger age group (20-29 years) due to motor accidents and

sports while the older age group (65-75 years) due to falls in the home or hospitals [3]. Therefore,

sincere efforts to prevent falls have led to reduced incidence of falls in the hospitals. As these

numbers suggest, there could be direct and indirect costs to an individual due to SCI. Direct costs

include health care costs, rehabilitation costs, more expensive transportation modes and special

needs. Indirect costs include reduced life expectancy, reduced productivity, stress, societal

challenges and disability. In addition to these costs, there are notable clinical functional

impairments that drastically reduce the quality of life for the patient.

2.1.2. Clinical presentation of the SCI and functional impairments: SCI is a debilitating

condition that leads to significant loss in the neuronal function resulting in autonomic, motor and

sensory impairments. Traumatic SCI usually starts with an abrupt mechanical insult resulting in

fracture or dislocation to the vertebrae intruding into the spinal canal. This process initiates a

cascade of pathophysiological changes at the site of the injury depending upon the severity of the

SCI. This leads to irreversible damage of the axons and break of the neuronal circuitry prompting

Page 18: An Injectable Gelatin-Based Conjugate Incorporating EGF

18

loss of function for the patient. As a result, SCI patients suffer from multiple adverse secondary

consequences that drastically reduce their quality of life [4].

In the acute phase of the SCI, the level and the severity of the injury determines the clinical

presentations of the SCI. Level of the SCI could be classified into two – tetraplegia and paraplegia.

Typically, cervical injuries would lead to tetraplegia affecting the functions of the neck, trunk, upper

and lower limbs. On the other hand, paraplegic injury refers to loss of function in the lower limbs

and the trunk without affecting the upper limbs. Severity of the injury is categorized in two types

– complete and incomplete. A complete SCI would lead to a complete loss of motor and sensory

below the level of the injury while an incomplete SCI retains some residual function below the

level of the injury [2]. Therefore, depending upon the level and the severity of the SCI, relevant

clinical management is necessary in the acute phase of the SCI, specifically cardiovascular and

pulmonary symptoms, which could be life-threatening [5]. The direct mechanical insult results in

secondary damages to the cord ensuing in functional loss. Most common secondary

complications include pressure ulcers, incontinence, pneumonia, autonomic dysreflexia, urinary

tract infections, loss of sexual function, fractures and depression-related disorders [6].

Acute clinical manifestations post SCI include pathologies in the cardiovascular, nervous

and respiratory systems. Cervical injuries cause loss of the sympathetic tone leading to imbalance

between the sympathetic and parasympathetic nervous system known as neurogenic shock.

Neurogenic shock results from the interruption of sympathetic tone due to disruption in

Figure 2.1 Prevalence of SCI with respect to age and gender. Reference - WHO Report on SCI, 2014

Page 19: An Injectable Gelatin-Based Conjugate Incorporating EGF

19

supraspinal control, and an intact parasympathetic influence via the vagus nerve, leading to an

imbalance in the autonomic control. Immediately, patient would suffer from bradycardia and

hypotension. Bradycardia is reported in between 64 to 77 percent of patients with cervical SCI.

Hypotension leads to further secondary damage in the cord due to reduced blood flow to the cord.

In addition, cervical injuries cause pathologies in the cardiovascular system – hypotension,

instability in the blood pressure and abnormalities in the heart rate [7, 8]. Autonomic dysreflexia

is another commonly reported clinical outcome post high cervical SCI. Autonomic dysreflexia is

characterized by extreme hypertension, in which the systolic blood pressure could reach up to

300 mm Hg accompanied by headache and slow heart rate. Untreated autonomic dysreflexia

could lead to intercranial hemorrhage, seizures, cardiac arrythmia and even death [9]. In addition

to the autonomic and cardiovascular system, respiratory system is also clinically affected due to

cervical injuries. In fact, respiratory complications are leading cause of death after SCI. Studies

have found that 67% of acute SCI patients experience severe respiratory complications within the

first days after the injury – namely atelectasis (36.4%), pneumonia (31.4%), and respiratory failure

(22.6%) [10]. People with an SCI above C3 may require constant mechanical ventilation or

implantation of a phrenic or diaphragm pacemaker to maintain adequate breathing [3, 5].

Chronic clinical presentations post SCI include pressure ulcers, loss of sexual function,

renal abnormalities, bladder and bowel incontinence and depression-related disorders (Table

2.1). Pressure ulcers are most common long term complication post SCI, especially in tetraplegic

patients followed by autonomic dysreflexia and renal abnormalities [11]. Moreover, loss of bladder

and bowel control affect the patients on a day-to-day basis. Due to SCI, patient loses voluntary

control on the sphincter resulting in recurrent and spontaneous voiding of the bladder. The onset

of symptoms is also dependent on the type of neuron that is affected to cause incontinence of the

bladder – upper or lower motor neuron because the management would depend upon the type of

the injury causing loss of bladder control. Similarly, bowel movements are also recurrent with loss

of control. It is shown that these symptoms could lead to social anxiety and isolation due to the

nature of the symptoms [12]. Another chronic clinical outcome of the SCI is spasticity. Around

70% of the patients develop varying degrees of spasticity that could cause significant disabilities.

Spasticity results because of the inability to control the muscles. Usually, muscles become very

stiff as a result of continuous contraction. This is due to imbalance in signals from the brain and

Page 20: An Injectable Gelatin-Based Conjugate Incorporating EGF

20

the cord. For example, if an upper motor neuron is severed, electrical signal does not reach the

synapse between the upper and lower motor neuron near the spinal cord roots. However, lower

motor neuron continuously is supplied with signal without any inhibition resulting in ‘permanent’

contraction of the muscles. Spasticity restrict the range of motion and thereby hinder daily

functions of the patient. Therefore, spasticity is primarily focused on by the rehabilitation care

team to making the patients more independent with their daily works, e.g. picking up a spoon,

opening a lid of the bottle and similar actions [13, 14].

Figure 2.2 Common acute and chronic clinical presentation of SCI patients. Reference – Expert Chikitsa, http://wiki.expertchikitsa.com

Table 2.1 Prevalence of various chronic secondary complications after SCI. Reference – McKinley, et al. Arch of Phys Med and Rehab, 1999. 80(11): p. 1402-1410.

Page 21: An Injectable Gelatin-Based Conjugate Incorporating EGF

21

All the above clinical outcomes were directly related to the primary and secondary

damages to the tissue (Figure 2.2). However, chronic clinical manifestations of the SCI also

include anxiety, societal isolation, post-traumatic stress disorder (PTSD) and depression. Given

how the quality of life is reduced and prognosis of their injury, it is not surprising that many patients

undergo psychological stress. Patients need to be given proper care and attention to deal with

this stress and anxiety that could come with the injury. It is reported that 20-30% of patients show

symptoms of clinically significant symptoms of depression [15]. Taking into consideration all the

clinical symptoms that patients experience, they have very little life satisfaction after the injury.

Although many patients do cope up well with their injury and new accommodations, better clinical

management and treatments are needed for the patients. Currently, SCI clinical management

involves management of the symptoms but not the root cause – degeneration of the spinal cord.

2.1.3. Clinical management and current treatment: SCI is managed by an integrated

team consisting for surgeons, physicians, rehabilitation therapist, occupational therapist,

recreational therapist, speech and hearing therapist, and many more depending upon the

symptoms of the patients. Therefore, professional societies have laid down management

guidelines. Initial care highlights the importance of addressing the airways, breathing and

circulation (ABC) first to ensure no life-threatening episode occurs [16]. In order to prevent further

spinal damage, patient’s spine is immobilized by the means of rigid cervical collar and backboard

for transfers. Once the patient is conscious and responding, a motor and sensory examination is

performed as per the American Spinal Injury Association (ASIA) impairment scale to assess the

level and extent of the injury (Table 2.2). The findings would be used for further care and

management [17]. Hypotension is corrected to systolic pressure of more than 90 mm Hg and

hypovolemia is treated on a case-by-case basis. Patients are kept in ICU setting for continuous

monitoring of cardiac, hemodynamic and respiratory indicators [16]. Once the patient is stable

and conscious, either CT and MRI imaging is performed to confirm the location and extent of the

SCI. X-rays are not recommended because they have been shown to miss nearly 6% of the

injuries [18].

After the appropriate imaging modality, efforts are focused on neuroprotection of the

healthy tissue to minimize further secondary damage. Neuroprotection includes efforts to

preserve the spared tissue after the injury from further secondary damage. Studies have shown

correlation between the amount of spared tissue and functional outcome. Given that

neuroprotective efforts are employed in the acute phase, we aim to evaluate the neuroprotective

effects of acute Gtn-HPA injection immediately after the creation of SCI cavity in our surgical

Page 22: An Injectable Gelatin-Based Conjugate Incorporating EGF

22

excision models discussed later. The first clinical treatment considered is early surgical

decompression. After SCI, continuous mechanical compression of the cord due to pinching of a

broken bone or an intact bone could significantly impair blood flow to the healthy tissue, thereby

resulting in ischemia and in increased zone of neuronal injury. Early surgical decompression

relives the compression improving the blood supply to the healthy neural tissue and thereby

preventing further expansion of the neural injury. Early surgical decompression has shown to be

very effective for improving pathological and functional outcomes by 31% in the pre-clinical

studies [19]. Patients who underwent early surgical decompression within 24 hours of injury

showed improvement of at least 2 ASIA grades within 6 months of injury compared to patients

undergoing the decompression later in the acute care [20]. However, depending on the level of

trauma, the type of SCI and risk-benefits analysis, not all patients are considered for the early

surgical decompression. For example, patients with central cord injury are typically not considered

for the surgical decompression due to poor surgical outcomes. Moreover, surgical decompression

could involve breaching the dura that surrounds the spinal cord thereby making patients

vulnerable for more infection-related clinical problems [21].

ASIA grade Clinical state

ASIA Grade A

(complete)

Complete lack of motor and sensory function below the level of

injury and no sacral functions

ASIA Grade B

(incomplete)

Loss of motor function but preserved sensory function below the

level of the injury and sacral functions

ASIA Grade C

(incomplete)

Motor function is preserved below the level of injury with more

than half of key muscles below the injury level has muscle

strength of less than 3

ASIA Grade D

(incomplete)

Motor function is preserved below the level of injury with at least

half of key muscles below the injury level having muscle strength

of more than 3

ASIA Grade E Normal motor and sensory functions

Table 2.2 ASIA grade categories for classification of the SCI. Clinical management is dependent on the

ASIA scale assessment. Reference – American Spinal Injury Association

The second clinical treatment with the goal of neuroprotection involves the use of steroids

to minimize the neuroinflammation. Methylprednisolone is a synthetic glucocorticoid that

upregulates the release of anti-inflammatory cytokines and reduces the oxidative stress to the

injured tissue. Methylprednisolone have shown to neutralize acute inflammation and reduce

Page 23: An Injectable Gelatin-Based Conjugate Incorporating EGF

23

swelling to reduce injury to remaining tissue in addition to decreasing lipid peroxidation. Therefore,

methylprednisolone was recommended for the acute administration within 48 hours of the injury

in the 1990s and 2000s [22, 23]. However, case reports and studies showed that

methylprednisolone increased the risk of infections thereby outweighing the neurological benefits

it provided if it was administered with high dose for 48 hours. Current guidelines do not

recommend the use of methylprednisolone after 8 hours of the injury either in 24-hour dose

protocol or 48-hour dose protocol. A 2012 review meta-analysis from six important randomized

clinical trials found that patients receiving methylprednisolone within 8 h of injury had a 4-point

greater ASIA motor score improvement [24]. Therefore, current guidelines recommend IV

administration of methylprednisolone only if it can be received within 8 hours of the traumatic

injury for the duration of 24 hours [25].

The third clinical treatment aimed at protecting the neural tissue from further damage is

blood pressure augmentation. Reduced blood flow and local edema due to neuroinflammation

contributes to ischemia in the area immediately surrounding the injury. This could further worsen

the functional sparing after the SCI. Therefore, blood pressure augmentation is a viable strategy

to protect the perilesional neural tissue from further degeneration by enhancing blood perfusion

in the area. The technique aims to maintaining the mean arterial pressure (MAP) to be greater

than 80-90 mm Hg for 7 days post injury. This method has shown to be beneficial for long-term

ASIA grade improvement in the patients [26]. However, some patients are not considered for this

treatment if they cannot be immobilized significantly. Additionally, systemic hypothermia was

considered as a therapy to reduce the rate of spread of the neuronal damage as a neuroprotective

therapy. However, it is not shown to be highly effective at improving functional outcomes and

therefore, it is not recommended by the current guidelines [27].

Another long-term clinical treatment that could assist patient improve their quality of life by

augmenting their functional capabilities is rehabilitation. Various forms of rehabilitative care has

shown to be beneficial in improving functional outcomes by improving on the ASIA grade scale of

up to 2 grades [28]. Rehabilitation therapy range from bodyweight supported treadmill training,

functional electrical stimulation and bicycle training. Functional electrical stimulation (FES) is an

interesting method to send signals to the preserved neural circuit or to the muscles directly.

Transplantation of the electrodes near the nerves of interest have shown electrical activity near

the nerves and functional stimulation of the corresponding muscles. Overall, the type of

rehabilitative measures are dependent on the type of the injury, level of the injury and the current

function that patients have preserved [29].

Page 24: An Injectable Gelatin-Based Conjugate Incorporating EGF

24

Current clinical treatments are aimed at the clinical management of multiple symptoms in

the acute setting while rehabilitation being the clinical intervention in the chronic phase. Taking

all clinical treatments into consideration, these do not address the primary reason for the

functional impairment – degeneration of the neuronal tissue. Therefore, restorative and

regeneration approaches are the future in the context of developing effective clinical treatments

to improve the functional recovery in the patients following a SCI. Given the complexity of the SCI

affecting wide range of tissues and functions, better understanding of the pathophysiological

processes and the chronology of the injury progression is essential in designing restorative and

regenerative therapies for the SCI.

2.1.4. Anatomy and pathophysiology of the SCI: Spinal cord is an integral link

between the brain and the peripheral nervous system. It extends from the foramen magnum to

lumbar level vertebrae. Various ascending and descending tracts run along the length of spinal

cord. Ascending tracts carry sensory information such as pain, temperature, light touch,

proprioception of arms and legs to brainstem and cerebrum. Descending tracts carry motor and

autonomic information controlling the voluntary and involuntary movements of our body (Figure

2.3). Transverse section of the spinal cord shows that it is divided into two regions – gray matter

and white matter. H-shaped horns are gray matter while the surrounding area is the white matter.

Gray matter consists of the neuronal nuclei while the white matter consists of the axonal

ascending and descending tracts [10, 30]. At cellular level, spinal cord tissue consists of neuronal

parenchymal cells. Supportive cells, specifically known as glial cells, include astrocytes,

oligodendrocyte, and microglia. There is also a stem cell niche near the subependymal layer

surrounding the central canal and the meninges.

Typically, traumatic SCI is defined as neuronal damage to the cord that results in change

or loss of function. Traumatic SCI involves mechanical insult to the cord as a result of vehicular

accidents, falls, gun wounds or sports activities. The mechanical force initiates a cascade of

pathophysiological responses that causes critical neural degeneration – the process that leads to

break in the neuronal circuitry and thereby contributes to functional impairment. These functional

impairments manifest as clinical symptoms of the SCI. It is critical to review these

pathophysiological processes to develop various targets for the development of novel treatments

of SCI. Pathophysiological response of the SCI is considered to be biphasic – primary damage

and secondary damage [31].

Primary damage refers to the immediate mechanical damage to the cord. Most spinal cord

injuries are contusion injuries due to blunt force directly applied on the cord as a result of

Page 25: An Injectable Gelatin-Based Conjugate Incorporating EGF

25

dislocation of the spinal column and vertebrae. Laceration, shearing, sudden jerk due to varying

acceleration and deceleration and compression are other types of mechanical injuries to the cord.

Cause of the SCI determines the type of primary damage caused to the cord [32]. For example,

a knife or gun wound would cause shearing or laceration type of primary injury while an accident

would cause contusion injury. Sheer force from the primary injury severs the axons, breaches the

blood-brain barrier (BBB), causes spinal and neurogenic shock and disrupts the vasculature.

Primary injury triggers the wave of the secondary damage [33].

The secondary damage deals with the cascade of cellular and molecular mechanisms in

response to the primary injury after the traumatic event at the injury site. The secondary damage

could be categorized into four phases based on the chronology – immediate (0-2 hours), acute

Figure 2.3 Anatomy of the spinal cord (a) Spinal cord has multiple ascending and descending axonal tracts either originating or reaching the brain responsible for transmitting functional information (b) Spinal cord is well protected by the spinal column and perfused by the spinal arteries (c) Spinal cord classified into cervical, thoracic, lumbar and sacral segments.

Image modified from Ahuja, CS, et al, Nat Rev Dis Primers, 3: 1017. Apr 2017.

Page 26: An Injectable Gelatin-Based Conjugate Incorporating EGF

26

(2-48 hours), subacute (48 hours – 2 weeks), intermediate (2 weeks – 6 months), and chronic

(beyond 6 months) (Figure 2.4) [34]. Immediate phase (0-2 hours) starts at the time of the injury

and runs parallel with the primary injury phase after the traumatic event. This includes immediate

death of the neurons and glial cells due to the mechanical disturbance. Moreover, disruption of

the blood vessel membrane leads to the hemorrhage of the parenchyma of the cord, especially

of the gray matter of the spinal cord. As a result, cells immediately start undergoing necrotic cell

death due to local ischemia. Microglial activation due to the hemorrhage upregulates the

proinflammatory cytokines such as IL-1β and TNF-α. These cytokines further increased vascular

permeability through the acute and subacute phase up to about 2 weeks [35]. Extracellular

glutamate levels also increase – in certain cases to levels which are determinantal to the

functioning of neurons [36, 37]. Acute phase (2-48 hours) pathophysiology is very critical to the

clinical management and acute therapies because patients would be presented in the ER room

in this phase [34, 38]. Acute phase is characterized by sustained hemorrhage, progressive

edema, neuroinflammation and resulting ischemia. Increased cell permeabilization permits the

movement of vasoactive peptides, cytokines and inflammatory cells such as neutrophils and

microglia. Endogenous microglia are continually activated while other inflammatory cells such as

astrocytes, neutrophils, T cells and monocytes invade the injury site [39]. These mechanisms

initiate production of free radicals, ionic imbalance and immune-associated neurotoxicity. Ionic

imbalance, especially that of calcium, is the most common reason for the necrotic and apoptotic

cell death. Free radicals released in this phase initiate the process of lipid peroxidation – oxidative

degradation of the lipids - thereby marking the beginning of the demyelination and

oligodendrocyte cell death [23]. Furthermore, glutamate levels keep rising in this phase and are

not reabsorbed by the astrocytes increasing glutamate levels result in excitotoxicity for

surrounding neurons [40, 41]. Continuous hypotension and reduced blood flow cause severe

prolonged ischemia that contributes to cell swelling and overall, swelling of the cord. Swelling

applies more pressure on the peri-lesional healthy tissue and therefore spread the injury through

multiple spinal cord segment [42]. Given that most pathological processes initiate in the acute

phase, most restorative and pharmacological approaches target these pathological processes to

improve functional outcomes and thereby, protecting the further spread of the injury. Several

potential molecules are currently in clinical trials and have shown promise in the initial Phase I

and II clinical trials.

Subacute phase (48 hours – 2 weeks) is characterized by continuation of all the

pathological processes from the acute phase and their consequences. The blood brain barrier is

reestablished while the edema is being resolved at the injury site. The invaded inflammatory and

Page 27: An Injectable Gelatin-Based Conjugate Incorporating EGF

27

immune cells aid the ongoing the inflammatory response and contribute to the apoptosis of

oligodendrocytes and neurons. Phagocytic cells such as neutrophils and macrophages start to

Figure 2.4 Key pathological events after a contusive spinal cord injury. The pathophysiological process is divided into five phases - (a) - Immediate (0-2 hours) and acute phase (2-48 hours) (b) subacute phase (48 hours - 2 weeks) and (c) Intermediate (2 weeks – 6 months) and chronic phase (beyond 6 months). Reference - Ahuja, CS, et al, Nat Rev Dis Primers, 3: 1017. Apr 2017.

Page 28: An Injectable Gelatin-Based Conjugate Incorporating EGF

28

clear the debris from the injury site – phagocytic activity is highest in this phase. However, they

also release free radicals, ATP and neurotransmitters that propagate the injury to nearby tissues

[43]. Clearing the myelin debris is believed to be an important process in promoting some axonal

growth because the molecules from the myelin degradation such as myelin-associated proteins

(MAP) and Nogo are inhibitory to the axonal cone regeneration [44]. The clearing of the debris

leaves behind a cystic cavity that would be surrounded by the reactive astrocytes to form a glial

scar. Towards the end of the subacute phase, astrocytes around the cavity become proliferative

and reactive. Most cell-based therapies target their intervention in the subacute phase because

a very dense glial scar is formed in the intermediate phase making it difficult for the exogenous

and endogenous cells to migrate to the injury site [45].

Intermediate phase (2 weeks – 6 months) involves the maturation of the astrocyte-

mediated glial scar and cavity formation. This phase also sees the attempts made by the tissue

to regenerate the severed axons, remyelination and remodeling of the neuronal circuitry.

However, it is insufficient to cause any functional improvement in the patient. Astrocytic glial scar

surrounds the cystic cavity that houses anti-regenerative molecules barrier to the axonal growth.

In the subacute phase and intermediate phase, activated microglia, macrophages and astrocytes

release various ECM molecules such as chondroitin sulfate proteoglycan (CSPG) [46, 47]. These

CSPG molecules along with reactive astrocytes processes and macrophages create a dense

interwoven barrier around the cavity. This barrier is chemically and physically impenetrable by the

surrounding cells or molecules and therefore impedes any regenerative processes such as axonal

cone sprouting and neurite outgrowth [48]. Moreover, Wallerian degeneration continues through

the intermediate phase as well degenerating distal axons. Numerous ECM proteins are laid down

by the fibroblasts and astrocytes, specifically fibrous connective tissue and thick collagen deposits

[49]. Chronic phase (beyond 6 months) witnesses limited Schwann cell remyelination and minor

axonal outgrowth. Neural plasticity plays an important role in regaining of function through

rehabilitation and function-specific training. The perilesional astrogliosis continues to form a

mature glial scar that is inhibitory of any potential regeneration [50]. Wallerian degeneration and

demyelination of the axons continues even in this phase and could last up to one-two years post

injury. Phagocytic processes are still ongoing and it could take similar time for severed axons with

their debris to be cleared from the injury site [51, 52]. In most patients, the lesion has stabilized

but the duration could vary from patient to patient. However, unfortunately, in nearly 30% of the

SCI patients, these lesions do not remain static and further grow into syrinxes, which can cause

delayed neuronal dysfunction and functional impairment. These syrinxes, also known as

Page 29: An Injectable Gelatin-Based Conjugate Incorporating EGF

29

Syringomyelia, are described by extension of fluid-filled cysts through multiple spinal cord

segments in the chronic phase of the injury [53, 54].

2.1.5. Cavitary lesions – potential for regeneration and recovery: Interestingly,

despite the similar pathophysiological processes, spontaneous regeneration is observed in the

peripheral nervous system to a certain degree. Pre-formed guidance channels and suturing of the

nerves have shown the elongation of axons to appreciate distances to promote recovery [55].

After PNS injury, Wallerian degeneration follows resulting in the disintegration of the distal axons

while the proximal axons sprout, elongate and re-establish synaptic connection to the nerve or

the muscles [56]. Schwann cells remyelinate the axons, guide the axons, secrete growth factors

and synthesize the necessary ECM molecules to rebuild the tissue [57]. However, in CNS, the

equivalent oligodendrocytes are few in number compared to the axons and therefore are not able

to contribute significantly in the myelination of spared axons [58]. In addition, many of the

oligodendrocytes die after the injury reducing the support for axonal regeneration and

spontaneous recovery. In fact, the neural progenitor stem cells are reported to commit to

oligodendrocyte precursor cells after injury [59]. Moreover, the degraded myelin from the severed

axons and demyelinated axons as a result of spinal cord injury release potent growth inhibitors

such as Nogo, myelin-associated glycoproteins, and oligodendrocyte myelin glycoprotein [60].

SCI injury phases Key pathological processes

Immediate phase (0-2 hours)

• Mechanical insult to the cord

• Severing of the axons

• Microglial activation and release of proinflammatory cytokines

• Hemorrhagic necrosis

Acute phase (2-48 hours)

• Edema & ionic imbalance

• Glutamate-mediated excitotoxicity

• Hemorrhage, ischemia and free radical production

• Lipid peroxidation & early demyelination (oligodendrocyte death)

• Neutrophil invasion and activated microglia

• Increased permeability through the blood-brain barrier (BBB)

Subacute phase (48 hours – 2 wks)

• Macrophage infiltration and reactive astrocytes

• Edema resolution but continued cell death

• Start of glial scar formation and repair of the blood-brain barrier

Intermediate phase (2 wks - 6 months)

• Cyst and cavity formation

• Maturation of glial scar

Chronic phase (beyond 6 months)

• Continued demyelination due to Wallerian degeneration

• Functional gain due to spared axons and plasticity

• Spinal cord tissue repair

• Cavity extension into syrinxes and syringomyelia

Table 2.3 Key pathophysiological processes summarized with their chronological phases after a SCI

Page 30: An Injectable Gelatin-Based Conjugate Incorporating EGF

30

These molecules break down the cytoskeleton of the regenerating axonal growth cones inhibiting

their regeneration [61]. Typically in PNS, these molecules are rapidly cleared by macrophages

but in CNS, it takes a long time for microglia to clear the debris [50]. In addition to the inhibitory

molecules, CSPGs and reactive astrocytes create a physical barrier for the which are barriers to

spontaneous regeneration. The pathophysiological response after CNS injury typically leads to

the formation of cyst-filled cavity surrounded by a very dense glial scar by these reactive

astrocytes [62]. There are beneficial and harmful effects of the cavity and the glial scar that is

formed [63]. Glial scar limits the spread of the secondary damage to the nearby healthy tissue

because the cystic fluid contains molecules that are anti-regeneration and would induce injury

processes to the nearby tissue. Therefore, the glial scar protects the neural tissue from ongoing

damage [64]. However, the contents of the cystic fluid and dense nature of the glial scar

constituting chondroitin sulphate proteoglycan (CSPGs) are inhibitory to the migration of any

reparative cells inside the cavity [65]. These are several intrinsic and extrinsic factors that lead to

the failure of regeneration in the CNS.

While there are significant barriers to spontaneous regeneration after SCI, research during

the last three decades have shown promise for repair. In fact, regeneration of numerous tracts

might not be necessary to promote functional recovery after SCI. There are various approaches

that are used to promote regeneration after SCI (Figure 2.5). These include inhibition of pathways

to lead to disintegration of the cytoskeleton of the axonal cones, remyelination of the spared

axons, recruitment of endogenous neural progenitor stem cells to the injury site using chemotactic

factors, implanting exogeneous cells that secrete growth-stimulating factors and re-establishing

neural connections lost as a result of the injury [17]. A number of pre-clinical studies have shown

functional recovery and injury site reorganization after an early intervention before the glial scar

is formed [66-69]. These studies indicate that redundancy in the nervous system and neural

plasticity can organize the re-wiring of the axons to promote recovery. Moreover, some of the

studies employing rat and monkey SCI models have shown axonal sprouting and axonal

regeneration after SCI [70-73]. Various potential therapeutics such as anti-NogoA and Cethrin

target the pathway of growth-inhibitory molecules to promote axonal growth [74, 75] . Therefore,

there is potential for regeneration in the injury site after SCI. If the local environment in the injury

site can be modulated using signaling molecules, if the tissue can be filled with regeneration-

Page 31: An Injectable Gelatin-Based Conjugate Incorporating EGF

31

promoting stroma and if the injury site is repopulated with the lost cells, there is sufficient evidence

that there is potential in regeneration of the spinal cord tissue following SCI. Since there are no

well-established and validated in vitro models that could be used to study the potential repair after

SCI, most investigations focus on the in vivo investigation of proposed treatment and therapy in

relevant animals models [76]. The choice of animal models is critical to evaluating the type of

therapy and the anticipated outcomes of the treatment.

2.1.6. Pre-clinical SCI models – an overview: SCI models have been pivotal in

investigating the efficacy of various proposed therapeutic interventions and in obtaining useful

information regarding the pathophysiological processes after SCI. The first animal SCI model was

weight-drop contusion model developed in 1911 [77]. Rats are the most commonly used species

for studying SCI because of several factors. Although there are some differences between the

pathophysiological progression of the injury post SCI, they very closely resemble the human SCI

[45]. Rats have demonstrated similar histological, behavioral and electrophysiological outcome

measures compared to humans following SCI [78]. Moreover, rats are easier to handle, relatively

inexpensive and readily available making them a preferred choice over other SCI models.

Although nonhuman primate SCI models better approximate human SCI, rats are preferred in

Figure 2.5 Various neuroregenerative strategies after spinal cord injury. Figure Reference - Ahuja, CS, et al, Nat Rev Dis Primers, 3: 1017. Apr 2017.

Page 32: An Injectable Gelatin-Based Conjugate Incorporating EGF

32

preliminary proof of principle investigations [79]. Positive findings then drive the investigation into

higher animal models such as primates, pigs and dogs.

Considering the incredibly multisystem complex nature of SCI, there are various SCI

models available for pre-clinical evaluation based on the mechanism of the injury. They are

classified as contusion, compression, dislocation, chemical and transection models. Contusion

models are the most clinically relevant SCI models as most injuries inflict a transient mechanical

force on the cord without breaching the dura [80]. In a typical contusion model, a transient force

is applied on the cord exposed by laminectomy. This mechanical force displaces and damages

the cord. The transient force is applied using several mechanisms – weight-drop method, air

pressure and electromagnetic impactor. Various devices such as NYU-MASCIS, IH impactor,

OSU impactor and air gun impactor have been developed to perpetrate a controlled and

consistent force on the spinal cord [81]. However, the lesion created is not reproducible and lacks

the volume to inject a biomaterial. Moreover, there is technical problems with the clamping

technique especially in the rodent model [82]. Compression models involve compressing the

spinal cord for a prolonged period. Fracture dislocations and burst fractures primarily result in the

compression type injuries. Clip compression model is the most widely used SCI compression

model. One of the foremost advantages of the clip compression model is that it can be adapted

to any region in the cord. Along with clips, balloon and forceps compression models have also

been used. Although compression models are less expensive to implement, the compression

parameters are typically not recorded making it difficult to compare across several studies using

the compression model. Moreover, it does not create a cavity into which a biomaterial can be

injected [83]. Chemically-induced SCI models are different because they were developed to study

the specific facet of the secondary injury mechanism post SCI. Chemically-induced SCI models

are primarily used to evaluate the molecular mechanisms involved post SCI and to evaluate the

effect of proposed therapeutic agents on these molecular mechanisms. Although these are very

specific to the nature of the mechanism, they do not approximate the overall effect of SCI.

Moreover, the location and the method of delivery of these chemicals to induce SCI varies in the

experimental procedures throughout the field. They do not create a reproducible injury for valid

comparison between the groups and with other models [84].

In the context of tissue engineering, transection models have been widely utilized to

investigate the effects of various biomaterials, neurotrophic factors and implanted cells to study

neuronal regeneration in the injury site [85]. Transection models ensures total loss of the neuronal

circuitry at the site of injury making them ideal setups to investigate axonal regeneration and

Page 33: An Injectable Gelatin-Based Conjugate Incorporating EGF

33

functional recovery. It is important to note that these transection models are not ideal to study the

pathophysiological processes because they are less common in clinical settings [81]. There are

two types of transection models – complete transection and partial transection. Complete

transaction involves complete loss of neuronal circuitry at the site of the injury with disassociation

between the rostral and caudal segments of the injury. Moreover, complete section models better

mimic the most severe clinical ‘complete SCI’ inducing complete break of the neuronal circuitry

as observed in most violence-related SCIs such as gun and knife wounds [86]. The most important

advantage of the complete resection model is the creation of a reproducible standardized defect

at a pre-defined location on the cord. In addition, complete resection is a model that has been

used to study the ‘bridging gap’ therapies such as scaffolds and similar devices [87]. Partial

resection is yet another type of transection model that has been applied in various ways. One

method involves severing of a particular spinal tract of interest. Specifically, transection of

corticospinal tract has been widely studied using the partial transection SCI model in mice, rats

and monkeys [88-90]. The second method involves severing of all the axons in one half of the

cord known as hemi-resection SCI model. This type of model allow for comparison with the

contralateral side for the effects of the proposed treatment [91]. Overall, these surgical excision

models are ideal for studying the outcome of injectable biomaterial therapies because it creates

a standardized unambiguous defect at the pre-defined location in the cord. Additionally, the cavity

provides a voluminous injection of injectable gels proposed to replace the stroma of the injured

site. In addition, these models offer advantage to visually confirm the injection of the injectable

therapeutics into the cavity, which is not achievable with any other pre-clinical SCI models. As

discussed, these models are widely used in all animal models from mice and rats to pigs and

monkeys. The typical outcomes measures for transection models include behavioral assessment,

imaging and histological evaluation of the injury site along with the immunohistochemical

investigation. For the advantages that transaction models provide to study axonal regeneration

and functional recovery as a result of biomaterial-based intervention, we will employ both 2-mm

complete and hemi-resection rat SCI model to investigate the effect of our proposed treatment –

injectable gelatin-based matrix incorporating epidermal growth factor (EGF) and stromal cell-

derived factor-1α (SDF-1α). However, prior promising work in the context of tissue engineering

approaches to treat SCI commends a review.

Page 34: An Injectable Gelatin-Based Conjugate Incorporating EGF

34

2.2. Tissue engineering approaches in SCI

2.2.1. Three pillars of tissue regeneration – an overview: Tissue engineering has shown

great promise in providing a systematic approach to promote regeneration in the injured tissue. It

is an extremely interdisciplinary field that has applications cutting across various industries

including food industry. Tissue engineering utilizes biomaterials, biochemical signals, physical

signals and cells, alone or in combination, to generate tissues that mimic the native tissue [92].

Clinically, organ transplantation is the gold standard to replace the ailing organ. However, limited

availability and rejection has led to finding alternatives. Early applications of tissue engineering

aimed at maintaining, restoring and improving the function of damaged tissue has been

encouraging in various organ systems [93]. Considering these advances, tissue engineering

approaches have become popular clinically as well. Traditional medical tools are not capable of

providing a functional restoration after injuries, especially in case of CNS injuries. The first tissue

engineered skin products were introduced in 1970s and 1980s [94]. Tissue engineering typically

involves a triad – three pillars for regeneration of the tissue. Scaffolds, signals and cells form the

‘central dogma’ of the field of tissue engineering. Depending upon the goal of the specific

application and the complexity of the tissue in consideration, one or all three approaches in

combination are essential for success in tissue engineering [95].

In the context of SCI, all three pillars are essential because of the complex organization

of the spinal cord tissue. Post SCI, there is a significant cell death of different types of cells. As a

result, signaling molecules that promote regeneration and repair the tissue are ineffective.

Moreover, the stroma is completely lost due to pathological process post injury. Therefore, all

three aspects are critical to promote regeneration in the spinal cord after injury. In fact, all three,

either alone or in combination, have proved useful in promoting regeneration and functional

recovery after SCI [96]. Historically, however, regeneration was thought to be impossible in CNS.

In ‘Edwin Smith Papyrus’ book, an Egyptian medical textbook dating 1700 BC, SCI was thought

to be ‘ailment not to be treated’ [97]. As a result of Wallerian degeneration, distal nerves

degenerate and lose their connection to the target. Therefore, true regeneration involves three

major steps – 1) axons proximal to the injury has to grow and infiltrate through the cyst-like

environment at the injury site, 2) axons has to grow all the way to the correct target, and 3) axon

has to synapse correctly at the target. In addition, new glial cells have to establish contact with

nerve axons for true regeneration. Currently, research in SCI regeneration is at the initial stages.

In 1981, David, et al, showed that it was indeed possible to achieve significant axon elongation

using a peripheral nerve graft in SCI if a permissive environment is provided [98]. Researchers

Page 35: An Injectable Gelatin-Based Conjugate Incorporating EGF

35

have come a long way in SCI regeneration by means of three pillars of Tissue engineering –

scaffold, soluble factors and cells.

2.2.2. Scaffolding materials for SCI: Various biomaterials, both individually or in

combination with factors and cells, have been tested for promoting regeneration and repair post

SCI. Biomaterials provide passive structural support, bioactive modulation of matrix, delivery

vehicle for factors and attachment surface for cells. In addition, cystic cavitation created as a

result of the pathophysiologic progression completely destroys the stroma in the injured tissue,

thereby inhibiting any support to potentially growing axons post SCI. Therefore, biomaterials have

emerged as a novel strategy to fill the cavitary lesions to provide a multifunctional substitute of

extracellular matrix molecules [99]. Generally, natural polymers such as collagen, laminin, PLGA,

and chitosan have been widely researched with cells and factors, which have shown decrease in

the cyst volume, increase in angiogenesis and axonal regrowth [99, 100]. For example, when

equine collagen alone was implanted in hemi-resection SCI model, it showed increase in the

number of axons [101]. Usually, pre-formed scaffolds are not suitable for SCI due to large variation

in injury size and volume. So, injectable gels offer a minimally invasive novel way to deliver

biomaterial matrix accounting for injury specifics. A detailed introduction to the various

biomaterials evaluated to promote SCI follows.

2.2.2.1. Natural vs. synthetic materials: Prior to evaluating a material for scaffold

fabrication for SCI, consideration need to be made whether to employ a natural or a synthetic

material. Natural materials offer a range of advantages over the use of synthetic polymers. First,

natural polymers consist of moieties such as RGD sequences that can aid in cell adhesion. In

addition, viability of cells grown on natural biomaterials is generally higher than that of synthetic

polymers [102]. Moreover, they can be designed to match the stiffness of the host tissue and are

degradable by the body [103]. Natural materials typically have high water content aiding in cellular

migration, diffusion of particles and molding into various forms [99]. In addition, they are readily

available from various sources such as plants and animals. However, one of the limitations of

using natural biomaterials is a potential transmission of diseases from the host source. Usually,

natural materials provide minimal control on the important properties such as degradation rate

and mechanical properties. Although synthetic biomaterials provide various advantages such as

easily controlled customized mechanical properties and easier sterilization techniques, they are

usually non-biocompatible eliciting a severe immune reaction. Moreover, they would need to be

functionalized extensively to gain ligands for cell adhesion. Considering the complex nature of the

injury site in SCI, synthetic polymers would induce further secondary damage would not be

Page 36: An Injectable Gelatin-Based Conjugate Incorporating EGF

36

beneficial. Therefore, most spinal cord applications utilize natural biomaterials to provide a

supportive and bioactive scaffold for cellular migration and proliferation post SCI [104-106].

2.2.2.2. Pre-formed vs. injectable biomaterials: Another consideration to be made is

whether a pre-formed scaffold or an injectable biomaterial would be beneficial and feasible for

SCI. Typically, contusion type injuries are most common in clinical setting. These contusions

create an irregularly shaped cavity heterogeneous in both shape and the volume. Pre-formed

scaffolds of specific size and dimensions would not conform to the type of cavitary lesions

observed in the spinal cord. Although pre-formed scaffold such as guidance channels have shown

great promise in pre-clinical models in several SCI studies using both natural and synthetic

polymers, it is becoming increasing evident that injectable biomaterials are more feasible to

implant the biomaterial in a minimally invasive manner to minimize additional damage to the

unaffected spinal cord tissue [107]. Moreover, injectable gels can conform to the irregularly

shaped defect and can completely fill the defect to build a better interface with the surrounding

host tissue – extremely useful in the context of CNS injuries. In addition, injectable gels can also

be functionalized to deliver encapsulated factors and transplant cells in the injury site. Although

there is reduced control on the porosity of the injectable gels, the materials can be chosen such

that it can be degraded by the migratory cells by releasing ECM enzymes such as matrix

metalloproteinases (MMPs) [108, 109].

2.2.2.3. Injectable biomaterials for SCI: Various injectable biomaterials have been

developed and designed to be evaluated in pre-clinical SCI models. Transaction and contusion

models are typically utilized to be able to inject into a cavity created by these models [110].

Collagen-based injectable biomaterials are probably the most widely studied in the context

of SCI. Collagen is a naturally occurring most abundant ECM molecule [111]. It has highly tunable

mechanical properties to be fabricated into an injectable gel. The gelation of collagen-based

scaffolds could depend upon the pH, temperature and the chemical cross-linker, if any [112]. In

the context of SCI repair, it has been utilized as hydrogel, aligned fibers, drug-delivery vehicle,

and cell-transplantation substrate [113]. Considering that collagen degrades relatively quickly

compared to the timeline for axonal regeneration in SCI, it is usually formulated with other

biomaterials or linkers to reduce the degradation rate. In a study by Marchand et al, self-

assembling glyoxal-linked collagen gel into a thoracic hemi-resection model. Collagen-glyoxal gel

persisted for 8 weeks in the injury site. However, glyoxal induced an inflammatory response.

Nevertheless, axonal regeneration was observed in the treatment groups three months are

injection into the hemi-resection cavity [114]. In a recent study, Li et al implanted a collagen

Page 37: An Injectable Gelatin-Based Conjugate Incorporating EGF

37

scaffold with cetuximab in the lower thoracic 4-mm hemi-resection model. They reported that

collagen-cetuximab scaffolds encouraged neuronal differentiation while decreasing the

differentiation of the NPCs into astrocytes. Overall, the scaffolds promoted axonal regeneration

and functional recovery 12 weeks after surgery [115].

Laminin is a glycoprotein that has capability to self-assemble by controlling the pH,

temperature and concentration [116]. Laminin signals through a wide variety of integrins and

therefore has been shown to induce neurite outgrowth in several studies [117, 118]. A study by

Menezes et al acutely injected polylaminin in the injury site. The results showed improved motor

function in three different SCI models - partial transaction, complete transection and compression.

Moreover, neurons retrogradely labeled were identified in the spina cord tissue near the injury

indicates axonal growth through the injury site. In addition, laminin also contributed to mitigate the

inflammatory response after the injury.

Fibronectin is another glycoprotein evaluated for SCI repair. Since fibronectin is not

capable of undergoing gelation on its own despite environmental modifications, it is usually

combined with other ECM molecules such as fibrin to fabricate a long-lasting scaffold. In a study

by King et al, combination of fibrin and fibronectin was used to fabricate an injectable hydrogel.

Fibrin is shown to degrade very quickly in vivo as early as one week [119]. However, the

combination gel persisted for four weeks after injection into a dorsal hemi-resection cavity. By one

week, the gel showed good integration with the surrounding tissue and promoted deposition of

laminin. After four weeks, the combination gel supported growth of axons and invasion of

Schwann cells and blood vessels in the injury site. Moreover, the combination gel showed greater

viability of neurons compared to the controls and gels alone [120].

Hyaluronan/hyaluronic acid is a naturally occurring non-sulfated glucosamine in the CNS

[121]. Given its greater degradation rate, it is typically combined with other factors and materials.

For example, hyaluronic acid was cross-linked with different concentration of thiols in a in vitro

and in vivo evaluation by Horn et al. Although there were promising findings in vitro of increased

neurite outgrowth in these gels compared to the fibrin, there were no differences among any

assessment parameters in vivo [122]. However, another study that cross-linked hyaluronic acid

with IKVAV peptide delivering BDNF showed neuronal regeneration after six weeks in a clip

compression model of SCI. These hydrogels were injected into the intrathecal space near the

injury. Compared to all other groups, animals injected with BDNF-containing hydrogels showed

the greatest improvement in the functional recovery assessed using the well-established BBB

scale [123].

Page 38: An Injectable Gelatin-Based Conjugate Incorporating EGF

38

Few non-mammalian biomaterials such as agarose, chitosan, and alginate have been

studied in few studies for treatment of SCI. Few studies have shown neuronal survival in vitro and

promotion of potential axonal regeneration in vivo, however to a minor degree [124, 125]. For

example, BDNF-loaded microtubules were prepared to be delivered using an agarose gel in a

hemi-resection models of SCI in a study by Jain et al. The agarose gel was persistent even after

6 weeks of injection in the injury site. Although there were some encouraging findings of reduced

inflammatory response and presence of neurons at the border outlining the defect, it did not show

axonal regeneration even after eight weeks. The authors hypothesized that the persistence of

agarose after eight weeks was impinging on the new tissue formation [124]. There are two major

limitations of these polymers – they typically elicit a inflammatory response due to non-

biocompatibility and do not have the ligand-associated interactions with the host tissue [99]. Given

a wide range of injectable biomaterials available and evaluated for spinal cord repair, certain

criteria must be established for ideal injectable biomaterial for SCI.

2.2.2.4. Criteria for biomaterials for cavitary lesions: Several criteria in designing an

ideal biomaterial for cavitary lesions of the spinal cord are listed based on the desired

characteristics for SCI repair [103, 109]. This comprehensive list was based on review papers

and discussions from our laboratory. Figure 2.6 gives a good schematic view of the considerations

for biomaterial apt for SCI repair.

1) It should be biocompatible without inducing more than typical inflammatory response

2) It should be injectable so that it can be delivered to the injured cavity in minimally invasive

manner to abate surgical damage during implantation

3) It should be able to gelate in situ to conform to the heterogeneous spinal cord cavities

4) It should be non-swelling not to apply pressure to the surrounding tissue, which is known to

initiate further secondary damage after the injury

5) It should have similar mechanical properties as that of the spinal cord to minimize secondary

damage and non-shrinking so that it can integrate well with the surrounding tissue

6) The constituents of the biomaterial should not induce cytotoxicity in the tissue

7) It should serve as a viable replacement of the disappearance of stroma lost due to SCI and be

permissive of cellular migration, proliferation and differentiation into the defect site to promote

neural repair

8) It should have tunable physical properties to control gelation time, stiffness and degradation

characteristics. Specifically, for experimental models, it should be able to complete gelation in

two-four minutes to visually confirm the gelation in the surgical hemi-resection model

Page 39: An Injectable Gelatin-Based Conjugate Incorporating EGF

39

9) It should have chemical moieties such as RGD sequences as ligand for cell integrins

10) It should be able to persist in vivo for timeline of our experiment (i.e. four weeks) and longer

11) It should be able to incorporate and locally deliver signaling molecules, growth factors, and

cells in the injury site without the particular use of vehicles

In the context of this work, we believe a gelatin-based matrix – gelatin-hydroxyphenyl

propionic acid (Gtn-HPA) meets all the desired criteria. A detailed discussion of the properties

and benefits of Gtn-HPA is presented in Chapter 3.

Figure 2.6 Schematic of various properties of a desired scaffold for SCI repair

Reference – Straley et al J Neurotrauma. 2010 Jan; 27(1): 1–19.

Page 40: An Injectable Gelatin-Based Conjugate Incorporating EGF

40

2.2.3. Biomolecules for SCI – an overview: In addition to replacing the cystic cavity and

degraded stroma in the injured spinal cord tissue with injectable biomaterials, biomolecules are

locally delivered in the injury site. Biomolecules play a variety of role after SCI. These factors can

enhance survival of the neurons, promote myelination of spared axons, alter the phenotype of

stem cells to differentiate into neuronal and glial fate, promote axonal regrowth and promote

plasticity [126]. Moreover, these factors can induce the migration, proliferation and differentiation

of endogenous and exogenously transplanted cells in the injury site. In additions, few

biomolecules aid functional recovery by means of modifying the ECM and the glial scar. These

factors, alone or in combination with scaffolds and cells, have been shown to improve

morphological and functional outcomes in pre-clinical models of SCI [127].

2.2.3.1. Neurotrophins and growth factors: A wide range of growth factors have been

evaluated in the spinal cord models to assess the therapeutic potential in promoting functional

recovery and regeneration in the SCI. Some of these have been or are currently being evaluated

in clinical trials [128].

Brain-derived neurotrophic factor (BDNF) is one of the most extensively studied factors in

pre-clinical models of SCI. BDNF has been reported to play a role in neuroprotection, axonal

sprouting, neurogenesis, remyelination and adaptive plasticity after SCI [129]. In a study by

Blesch et al, administering BDNF in a transient manner was shown to increase the number of

spared axon and induce regrowth of the axons in the injury site indicating that sustained delivery

of these factors is essential [130]. Moreover, when BDNF was delivered using collagen matrix,

improved functional recovery along with axonal regeneration of CST axons was observed by Han

et al in a complete transection model of SCI [131]. Several studies have shown that BDNF

increases the myelination of axons and confers neuroprotection to the healthy spinal cord tissue

in rat or cat models [132, 133].

Another neurotrophin to be widely studied in experimental SCI models is NT-3. Typically,

NT-3 is delivered near the injury site acutely. It has been delivered with a wide range of scaffolds

ranging from collagen, fibrin and chitosan to hydrogels [127]. Most studies show axonal sprouting

in the injured axons after injection of NT-3 at the injury site mostly in the thoracic models of SCI

[134]. In a study by Paiantino et al, an injectable hydrogel incorporating NT-3 was implanted in

the 1.5-mm hemisection model of SCI. The results showed increased axonal growth of CST axons

and better functional recovery in the treatment groups compared to the controls [135]. In addition,

a number of studies showed myelinating and neuroprotective effects of NT-3 suggesting NT-3 to

be a promising biomolecule to promote repair in the injured spinal cord tissue after SCI [136, 137].

Page 41: An Injectable Gelatin-Based Conjugate Incorporating EGF

41

Various other notable neurotrophins and growth factors include nerve growth factor (NGF),

NT-4/5, fibroblast growth factor (FGF), Epidermal growth factor (EGF), Glial cell-derived

neurotrophic factor, IGF-1 and platelet-derived growth factor (PDGF) [127]. Martens et al pursued

an in vivo investigation in a mouse model where they injected a combination of EGF and FGF-2

in the fourth ventricle in the brain and central canal of adult male CD1 mice. Just six days post

injection of EGF and FGF, greater proliferating cells were observed in both the cervical spinal

cord and near fourth ventricle in the brain. Specifically, in the spinal cord, these cells were in the

ependymal and subependymal tissue closer to the central canal. Indeed, these cells were nestin

positive indicating the presence of neural progenitor stem cells. The results from this study

indicated that endogenous neural progenitor stem cells near the central canal in the spinal cord

proliferate in response to EGF and FGF-2 [138]. In a more comprehensive in vivo study by

Hamann et al investigating the same EGF and FGF-2 combination in a clip compression rat

model, EGF and FGF-2 were delivered in a highly concentrated collagen solution to localize the

growth factors in the injury site. Interestingly, they labeled the injected EGF and FGF-2 with

markers. The results indicated that EGF readily penetrated the injury site in the spinal cord while

FGF-2 was predominantly observed in the dura. Furthermore, significantly less cavitations were

seen in the growth factors group indicating that EGF promoted tissue sparing including greater

white matter sparing in the injury site. Moreover, as reported by Kojima et al, they also observed

greater BrdU+ cells in the EGF and FGF-2 group than the controls in the ependymal region [69].

A study by Johnson et al investigated the combined effect of PDGF and NT-3 delivered using a

fibrin scaffold in subacute model of SCI. These increased the migration of SCIs to the injury site

and promoted differentiation of SCIs into neurons [119]. Despite abundant studies with many

factors, ideal combination of the factors needs to be tailored to the anticipated outcome after SCI.

2.2.3.2. ECM and scar modifying molecules: Chondroitinase ABC (chABC) is an

enzyme that breaks down the chondroitin sulphate proteoglycans (CSPGs) that form the physical

barrier around fluid-filled cystic cavity. Therefore, to promote migration of factors and endogenous

cells into the cavity site, chABC, alone or in combination, has been evaluated in various spinal

cord models [139]. In a recent canine SCI model study by Hu et al, administration of chABC

significantly improved the forelimb and hindlimb coordination after 6 months compared to the

control group [140]. Moreover, Alluin et al investigated a combinatorial approach of delivering

chABC with cocktail of factors containing FGF-2, EGF and PDGF-AA in a clip spinal compression

model. This combination was injected into the injury site via subarachnoid catheter connected to

an osmotic pump. The combination showed reduced GFAP presence in the perilesional area

indicating reduction of astrogliosis near the injury site. Interestingly, even the macrophages

Page 42: An Injectable Gelatin-Based Conjugate Incorporating EGF

42

response was attenuated in the defect area due to the administration of the combination of chABC

and growth factors. One of the key findings was collateral sprouting of the corticospinal tract using

BDA tracing. There was greater spared tissue observed in the cord due to implantation of the

chABC with growth factors compared to the controls. Lastly, this intervention also shown improved

functional recovery suggesting augmentation of neuronal plasticity [141]. Overall, biomolecules

with different effects modulate the injury site environment and are critical to regeneration in SCI.

In addition to biomolecules, repopulating the injury site with exogeneous cells has also proven to

be beneficial as described next.

2.2.4. Cell-based treatment for SCI – an overview: Cell implantation could confer notable

activities in the injury site that could promote neural regeneration and functional recovery. First,

cells can secrete neurotrophic factors at the injury site to initiate signaling pathways that can lead

to neural recovery. Second, cells can synthesize extracellular matrix to replace the lost stroma

due to the pathophysiological cascade post injury and to provide a scaffold to regenerating axons.

Third, cells can repopulate the lost cell types such as neurons, oligodendrocytes and endothelial

cells. The cells to be implanted should meet general criteria as follows – cells should show

proliferative ability in vitro, should exhibit good viability and function retention, should be safe for

injection meaning they should not elicit any responses that can lead to cancer-like tissue formation

and should be readily available or isolated using established procedure [142]. Different types of

cells meet the general criteria and have been studied extensively for treatment of SCI.

Embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) are an attractive

choice for cell implantation because they have the ability to differentiate into various neural cell

types that include oligodendrocytes, astrocytes and functional neurons [143]. Both cell types have

shown tremendous therapeutic benefit in vivo after SCI. In an interesting experiment, Bottai et al

injected non-differentiated ESCs through the tail vein acutely in mouse SCI model. They observed

a significant improvement of functional recovery BBB scores in the treatment groups compared

to the controls. Moreover, they reported reduction in the number of invading macrophages and

neutrophils suggesting that the cells were able to modulate inflammatory response post SCI. The

authors also speculated that the injection of ESCs could potentially preserve greater spared tissue

through reducing the inflammatory response. No teratoma formation was observed in the injury

site in the treatment groups [144]. Besides injecting ESCs or iPSCs directly, another strategy is

to transplant cells of specific fate derived from them. For example, Niapour et al implanted ESC

derived neural stem cells (NSCs) along with Schwann cells (SCs) in a contusion model of SCI.

They hypothesized that ESC-derived NSCs would lead to greater neuronal differentiation. Their

Page 43: An Injectable Gelatin-Based Conjugate Incorporating EGF

43

results showed significant motor function recovery due in the experimental groups when

compared with the controls. Moreover, experimental groups showed significantly increased

expression of TUJ1 and MAP2 biomarkers, and reduced GFAP signal five weeks post

implantation. Another strategy is to transplant the cells in a multifunctional scaffold. Hatami et al

employed a rat hemi-resection model to implant ESC-derived NPCs encapsulated in collagen

scaffolds into the cavity created by hemi-resection. They showed that transplanted ESC-derived

NPCs differentiated into neurons and glial cells in vivo. Moreover, the intervention promoted

significant hindlimb functional recovery and sensory responses in the groups treated with cells

compared to the controls [145].

Mesenchymal stem cells (MSCs) are yet another preferred cell types for SCI studies

considering that they could be readily isolated from the patients and preserved with minimal loss

in the potency [146]. MSCs have been shown to differentiate into all the cell types of neural tissue

under preferred differentiating medium [147]. Moreover, transplanted MSCs are capable of

guiding axonal growth, promote axonal cone growth, mitigate the extent of demyelination and

reduce anti-inflammatory molecules in the injury site in SCI [148]. One goal for using MSCs for

SCI is to mitigate the inflammatory response thereby conferring greater neuroprotection to the

unaffected tissue. For example, Nakajima et al transplanted bone-marrow derived MSCs (bMSCs)

in a rat contusion model and evaluated their effect up to five weeks post transplantation. After

transplantation of bMSCs into the contusion epicenter, the bMSCs significantly upregulated the

levels of IL-4 and IL-13 while downregulated the levels of TNF-alpha and IL-6. These processes

led to phenotype change in macrophages from M1 to M2. The authors suggest that this

mechanism led to more preserved axons, greater myelin sparing and reduced scar formation.

Moreover, the treatment groups also exhibited better locomotion recovery compared to the

controls [149]. However, there was no indication of differentiation of these MSCs into neurons –

a limitation observed in several studies [150, 151]. MSCs have been shown to secrete various

growth factors such as GDNF and NGF, guide axonal elongation, decrease demyelination and

promote axonal regeneration [152]. Moreover, MSCs have been shown to test positive for

neuronal markers NeuN confirming neuronal differentiation [153]. Also, Cho et al showed that

MSCs were able to improve functional recovery and neural regeneration along with neural

differentiation [154]. No adverse effects were reported when bone-marrow derived MSCs

(bmMSCs) were injected into rats in various SCI models [152, 155].

Neural stem cells (NSCs) have been found to exist in the subependymal layer around the

central canal and meninges in spinal cord. Moreover, NSCs are capable of self-renewal and

Page 44: An Injectable Gelatin-Based Conjugate Incorporating EGF

44

generating various neural fate cells such as neurons, oligodendrocytes and astrocytes both in

vitro and in vivo. Therefore, NSCs are thought to be important for regeneration after SCI [156,

157]. NSCs from human fetal spinal cord were grafted into the lumbar cord of adult nude rats by

Yan et al. They reported differentiation of NSCs into neurons, axon regeneration, and formation

of synapses with host motor neurons. Interestingly, the differentiated neurons from NSCs were

able to integrate with host neuronal circuitry pointing to the potential axonal regeneration and

recovery of function at anatomical level [158]. In yet another interesting approach using NSCs by

Yasuda et al, NSCs were preconditioned with the ability for remyelination. The results were

significant and observed that at the end of experiment, the remyelination potential of NSCs

increased and was probably linked with the observed improved motor and electrophysiological

functional recovery in treatment groups [159]. In a very recent study, neural stem cells (NSCs)

enriched with interneurons were transplanted at T3-T4 level of spinal cord. They showed that the

transplanted cells survived, differentiated and integrated with the host spinal cord tissue at the

injury site. Electromyography results suggested improved recovery one month post the

implantation. Moreover, animals that received donor cells showed significantly greater motor

functional improvement than the controls and point to the important role of NSCs in re-establishing

neuronal circuits in the injured adult spinal cord [160]. In the context of our work, we propose to

deliver EGF and SDF-1α to recruit the endogenous NSCs at the injury site to promote functional

recovery and regeneration after SCI. It has been shown that EGF induces migration and

proliferation of NSCs both in vitro and in vivo [161, 162].

Olfactory ensheathing cells (OECs) and Schwann cells (SCs) have also been transplanted

in SCI models and have shown to promote neuroprotection, expression of the neurotrophic

factors, remyelination and axonal regeneration [152]. Richter et al demonstrated remyelination,

reduced scar formation, and better axonal regeneration after transplantation of lamina propria-

derived OECs (LP-OECs) in a T3 compression SCI model [163]. Moreover, Zhang et al. also

transplanted LP-OECs in a chronically contused SCI model. They observed greater axonal

regeneration, remyelination and reduced lesion volume after scar ablation. However, they did not

observe any improvement in the motor functional recovery [164]. Since Schwan cells play a critical

role in promoting regeneration in the peripheral nervous system, SCs have been transplanted in

the SCI models to promote remyelination of the spared axons and improvement in the functional

recovery [57]. An interesting approach was taken by Agudo et al by transplanting the Schwann

cell precursors (SCPs) immediately after inducing a crush SCI. The authors presented that they

observed proliferation of SCPs and SCPs were instrumental in promoting angiogenesis.

Remarkably, the transplanted groups with SCPs had reduced glial scar formation. Moreover,

Page 45: An Injectable Gelatin-Based Conjugate Incorporating EGF

45

anterograde tracing revealed regeneration of BDA-labeled CST axons beyond the injury site

[165]. Overall, cells transplanted in the lesion cavity especially in scaffolds offer an innovative

approach to repair and regeneration after SCI. Given the complexity of the spinal cord tissue after

an injury, a combinatorial approach with all three pillars of tissue engineering should be evaluated

for synergistic effects of the treatment.

Page 46: An Injectable Gelatin-Based Conjugate Incorporating EGF

46

2.3. References

1. Facts and Figures at a Glance, in SCI Data Sheet. 2018, University of Alabama at Birmingham: National Spinal Cord Injury Statistical Center. p. 2.

2. Chan, M., International Perspectives on Spinal Cord Injury. 2013, World Health Organization. 3. Spinal Cord Injury. 2014 11/2013 [cited 2014 12/19]; Available from:

http://www.who.int/mediacentre/factsheets/fs384/en/. 4. McKinley, W.O., et al., Long-term medical complications after traumatic spinal cord injury: A

regional model systems analysis. Archives of Physical Medicine and Rehabilitation, 1999. 80(11): p. 1402-1410.

5. Hagen, E.M., Acute complications of spinal cord injury. World Journal of Orthopedics, 2015. 6(1): p. 17-23.

6. Munce, S.E., et al., Impact of quality improvement strategies on the quality of life and well-being of individuals with spinal cord injury: a systematic review protocol. Syst Rev, 2013. 2: p. 14.

7. Rouanet C, R.D., Rocha E, Gagliardi V, Silva GS, Traumatic spinal cord injury: current concepts and treatment update. Arquivos de Neuro-Psiquiatria, 2017. 75(6).

8. Krassioukov, A., Autonomic function following cervical spinal cord injury. Respiratory Physiology & Neurobiology, 2009. 169(2): p. 157-164.

9. Claydon, V.E. and A.V. Krassioukov, Orthostatic Hypotension and Autonomic Pathways after Spinal Cord Injury. Journal of Neurotrauma, 2006. 23(12): p. 1713-1725.

10. Kirshblum, S.C., et al., 1. Etiology, classification, and acute medical management. Archives of Physical Medicine and Rehabilitation, 2002. 83: p. S50-S57.

11. Gélis, A., et al., Pressure ulcer risk factors in persons with SCI: part I: acute and rehabilitation stages. Spinal Cord, 2008. 47: p. 99.

12. Burns, A.S., D.A. Rivas, and J.F. Ditunno, The Management of Neurogenic Bladder and Sexual Dysfunction After Spinal Cord Injury. Spine, 2001. 26(24S): p. S129-S136.

13. Hiersemenzel, L.-P., A. Curt, and V. Dietz, From spinal shock to spasticity. Neuronal adaptations to a spinal cord injury, 2000. 54(8): p. 1574-1582.

14. Rekand, T., E.M. Hagen, and M. Gronning, Spasticity following spinal cord injury. Tidsskr Nor Laegeforen, 2012. 132(8): p. 970-3.

15. Post, M.W.M. and C.M.C. van Leeuwen, Psychosocial issues in spinal cord injury: a review. Spinal Cord, 2012. 50: p. 382.

16. Resnick, D.K., Updated Guidelines for the Management of Acute Cervical Spine and Spinal Cord Injury. Neurosurgery, 2013. 72 Suppl 2: p. 1.

17. Ahuja, C.S., et al., Traumatic Spinal Cord Injury-Repair and Regeneration. Neurosurgery, 2017. 80(3s): p. S9-s22.

18. Ryken, T.C., et al., Radiographic Assessment. Neurosurgery, 2013. 72(suppl_3): p. 54-72. 19. Batchelor, P.E., et al., Meta-analysis of pre-clinical studies of early decompression in acute spinal

cord injury: a battle of time and pressure. PloS one, 2013. 8(8): p. e72659-e72659. 20. Wilson, J.R., et al., Early versus late surgery for traumatic spinal cord injury: the results of a

prospective Canadian cohort study. Spinal Cord, 2012. 50(11): p. 840-3. 21. Schneider, R.C., G. Cherry, and H. Pantek, The syndrome of acute central cervical spinal cord

injury; with special reference to the mechanisms involved in hyperextension injuries of cervical spine. J Neurosurg, 1954. 11(6): p. 546-77.

22. Scholtes, F., G. Brook, and D. Martin, Spinal cord injury and its treatment: current management and experimental perspectives. Adv Tech Stand Neurosurg, 2012. 38: p. 29-56.

23. Christie, S.D., et al., Duration of lipid peroxidation after acute spinal cord injury in rats and the effect of methylprednisolone. Neurosurg Focus, 2008. 25(5): p. E5.

Page 47: An Injectable Gelatin-Based Conjugate Incorporating EGF

47

24. Bracken, M.B., Steroids for acute spinal cord injury. Cochrane Database Syst Rev, 2012. 1: p. Cd001046.

25. Fehlings, M.G., et al., A Clinical Practice Guideline for the Management of Acute Spinal Cord Injury: Introduction, Rationale, and Scope. Global spine journal, 2017. 7(3 Suppl): p. 84S-94S.

26. Wilson, J.R., N. Forgione, and M.G. Fehlings, Emerging therapies for acute traumatic spinal cord injury. CMAJ : Canadian Medical Association journal = journal de l'Association medicale canadienne, 2013. 185(6): p. 485-492.

27. Levi, A.D., et al., Clinical outcomes using modest intravascular hypothermia after acute cervical spinal cord injury. Neurosurgery, 2010. 66(4): p. 670-7.

28. McDonald, J.W., et al., Late recovery following spinal cord injury. Case report and review of the literature. J Neurosurg, 2002. 97(2 Suppl): p. 252-65.

29. Ho, C.H., et al., Functional electrical stimulation and spinal cord injury. Phys Med Rehabil Clin N Am, 2014. 25(3): p. 631-54, ix.

30. Nolte, J., The Human Brain: An Introduction to Its Functional Anatomy. 5th ed. 2001: Elsevier Health Sciences. 650.

31. Norenberg, M.D., J. Smith, and A. Marcillo, The pathology of human spinal cord injury: defining the problems. J Neurotrauma, 2004. 21(4): p. 429-40.

32. Baptiste, D.C. and M.G. Fehlings, Pharmacological approaches to repair the injured spinal cord. J Neurotrauma, 2006. 23(3-4): p. 318-34.

33. Ditunno, J.F., et al., Spinal shock revisited: a four-phase model. Spinal Cord, 2004. 42(7): p. 383-95.

34. Rowland, J.W., et al., Current status of acute spinal cord injury pathophysiology and emerging therapies: promise on the horizon. Neurosurg Focus, 2008. 25(5): p. E2.

35. Noble, L.J. and J.R. Wrathall, Distribution and time course of protein extravasation in the rat spinal cord after contusive injury. Brain Res, 1989. 482(1): p. 57-66.

36. Pineau, I. and S. Lacroix, Proinflammatory cytokine synthesis in the injured mouse spinal cord: multiphasic expression pattern and identification of the cell types involved. J Comp Neurol, 2007. 500(2): p. 267-85.

37. Wrathall, J.R., Y.D. Teng, and D. Choiniere, Amelioration of functional deficits from spinal cord trauma with systemically administered NBQX, an antagonist of non-N-methyl-D-aspartate receptors. Exp Neurol, 1996. 137(1): p. 119-26.

38. Bramlett, H.M. and W.D. Dietrich, Progressive damage after brain and spinal cord injury: pathomechanisms and treatment strategies, in Progress in Brain Research, J.T. Weber and A.I.R. Maas, Editors. 2007, Elsevier. p. 125-141.

39. Fleming, J.C., et al., The cellular inflammatory response in human spinal cords after injury. Brain, 2006. 129(Pt 12): p. 3249-69.

40. Ahuja, C.S., A.R. Martin, and M. Fehlings, Recent advances in managing a spinal cord injury secondary to trauma. F1000Res, 2016. 5.

41. Mautes, A.E., et al., Vascular events after spinal cord injury: contribution to secondary pathogenesis. Phys Ther, 2000. 80(7): p. 673-87.

42. Tator, C.H. and M.G. Fehlings, Review of the secondary injury theory of acute spinal cord trauma with emphasis on vascular mechanisms. J Neurosurg, 1991. 75(1): p. 15-26.

43. Kwon, B.K., et al., Pathophysiology and pharmacologic treatment of acute spinal cord injury. Spine J, 2004. 4(4): p. 451-64.

44. Donnelly, D.J. and P.G. Popovich, Inflammation and its role in neuroprotection, axonal regeneration and functional recovery after spinal cord injury. Exp Neurol, 2008. 209(2): p. 378-88.

Page 48: An Injectable Gelatin-Based Conjugate Incorporating EGF

48

45. Hagg, T. and M. Oudega, Degenerative and spontaneous regenerative processes after spinal cord injury. J Neurotrauma, 2006. 23(3-4): p. 264-80.

46. Yiu, G. and Z. He, Glial inhibition of CNS axon regeneration. Nature Reviews Neuroscience, 2006. 7: p. 617.

47. Puckett, W.R., et al., The astroglial response to Wallerian degeneration after spinal cord injury in humans. Exp Neurol, 1997. 148(2): p. 424-32.

48. Kakulas, B.A., A review of the neuropathology of human spinal cord injury with emphasis on special features. J Spinal Cord Med, 1999. 22(2): p. 119-24.

49. Cregg, J.M., et al., Functional regeneration beyond the glial scar. Experimental Neurology, 2014. 253: p. 197-207.

50. Buss, A., et al., Sequential loss of myelin proteins during Wallerian degeneration in the human spinal cord. Brain, 2005. 128(Pt 2): p. 356-64.

51. Coleman, M.P. and V.H. Perry, Axon pathology in neurological disease: a neglected therapeutic target. Trends Neurosci, 2002. 25(10): p. 532-7.

52. Ehlers, M.D., Deconstructing the axon: Wallerian degeneration and the ubiquitin-proteasome system. Trends Neurosci, 2004. 27(1): p. 3-6.

53. Heiss, J.D., et al., Pathophysiology of primary spinal syringomyelia. J Neurosurg Spine, 2012. 17(5): p. 367-80.

54. Stoodley, M.A., Pathophysiology of syringomyelia. J Neurosurg, 2000. 92(6): p. 1069-70; author reply 1071-3.

55. Fu, S.Y. and T. Gordon, The cellular and molecular basis of peripheral nerve regeneration. Mol Neurobiol, 1997. 14(1-2): p. 67-116.

56. Tucker, B.A. and K.M. Mearow, Peripheral sensory axon growth: from receptor binding to cellular signaling. Can J Neurol Sci, 2008. 35(5): p. 551-66.

57. Jessen, K.R. and R. Mirsky, The repair Schwann cell and its function in regenerating nerves. The Journal of physiology, 2016. 594(13): p. 3521-3531.

58. Michalski, J.-P. and R. Kothary, Oligodendrocytes in a Nutshell. Frontiers in cellular neuroscience, 2015. 9: p. 340-340.

59. Soundarapandian, M.M., et al., Zfp488 promotes oligodendrocyte differentiation of neural progenitor cells in adult mice after demyelination. Scientific reports, 2011. 1: p. 2-2.

60. Lee, J.K. and B. Zheng, Role of myelin-associated inhibitors in axonal repair after spinal cord injury. Experimental neurology, 2012. 235(1): p. 33-42.

61. Hu, F. and S.M. Strittmatter, The N-terminal domain of Nogo-A inhibits cell adhesion and axonal outgrowth by an integrin-specific mechanism. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2008. 28(5): p. 1262-1269.

62. Leal-Filho, M.B., Spinal cord injury: From inflammation to glial scar. Surgical neurology international, 2011. 2: p. 112-112.

63. Fehlings, M.G. and G.W. Hawryluk, Scarring after spinal cord injury. J Neurosurg Spine, 2010. 13(2): p. 165-7; discussion 167-8.

64. Hu, R., et al., Glial scar and neuroregeneration: histological, functional, and magnetic resonance imaging analysis in chronic spinal cord injury. J Neurosurg Spine, 2010. 13(2): p. 169-80.

65. Fitch, M.T. and J. Silver, CNS injury, glial scars, and inflammation: Inhibitory extracellular matrices and regeneration failure. Exp Neurol, 2008. 209(2): p. 294-301.

66. Zendedel, A., et al., Stromal cell-derived factor-1 alpha (SDF-1alpha) improves neural recovery after spinal cord contusion in rats. Brain Res, 2012. 1473: p. 214-26.

67. Bethea, J.R., et al., Systemically administered interleukin-10 reduces tumor necrosis factor-alpha production and significantly improves functional recovery following traumatic spinal cord injury in rats. J Neurotrauma, 1999. 16(10): p. 851-63.

Page 49: An Injectable Gelatin-Based Conjugate Incorporating EGF

49

68. Erceg, S., et al., Transplanted oligodendrocytes and motoneuron progenitors generated from human embryonic stem cells promote locomotor recovery after spinal cord transection. Stem Cells, 2010. 28(9): p. 1541-9.

69. Jimenez Hamann, M.C., C.H. Tator, and M.S. Shoichet, Injectable intrathecal delivery system for localized administration of EGF and FGF-2 to the injured rat spinal cord. Exp Neurol, 2005. 194(1): p. 106-19.

70. Ajay, B., et al., Mechanically engineered hydrogel scaffolds for axonal growth and angiogenesis after transplantation in spinal cord injury. Journal of Neurosurgery: Spine, 2004. 1(3): p. 322-329.

71. Darian-Smith, C., Primary afferent terminal sprouting after a cervical dorsal rootlet section in the macaque monkey. J Comp Neurol, 2004. 470(2): p. 134-50.

72. Jaerve, A., et al., Differential effect of aging on axon sprouting and regenerative growth in spinal cord injury. Experimental Neurology, 2011. 231(2): p. 284-294.

73. Lee, J.K. and B. Zheng, Axon regeneration after spinal cord injury: insight from genetically modified mouse models. Restor Neurol Neurosci, 2008. 26(2-3): p. 175-82.

74. Gonzenbach, R.R., et al., Delayed anti-nogo-a antibody application after spinal cord injury shows progressive loss of responsiveness. J Neurotrauma, 2012. 29(3): p. 567-78.

75. Fehlings, M.G., et al., Rho Inhibitor VX-210 in Acute Traumatic Subaxial Cervical Spinal Cord Injury: Design of the SPinal Cord Injury Rho INhibition InvestiGation (SPRING) Clinical Trial. J Neurotrauma, 2018. 35(9): p. 1049-1056.

76. Weightman, A.P., et al., An in vitro spinal cord injury model to screen neuroregenerative materials. Biomaterials, 2014. 35(12): p. 3756-65.

77. Allen, A., Surgery of experimental lesion of spinal cord equivalent to crush injury of fracture dislocation of spinal column: A preliminary report. Journal of the American Medical Association, 1911. LVII(11): p. 878-880.

78. Metz, G.A., et al., Validation of the weight-drop contusion model in rats: a comparative study of human spinal cord injury. J Neurotrauma, 2000. 17(1): p. 1-17.

79. Iwanami, A., et al., Establishment of graded spinal cord injury model in a nonhuman primate: the common marmoset. J Neurosci Res, 2005. 80(2): p. 172-81.

80. Young, W., Spinal cord contusion models. Prog Brain Res, 2002. 137: p. 231-55. 81. Kwon, B.K., T.R. Oxland, and W. Tetzlaff, Animal models used in spinal cord regeneration

research. Spine (Phila Pa 1976), 2002. 27(14): p. 1504-10. 82. Cheriyan, T., et al., Spinal cord injury models: a review. Spinal Cord, 2014. 52(8): p. 588-95. 83. Mahdi, S.-A., Animal Models in Traumatic Spinal Cord Injury in Topics in Paraplegia, R.-M. Vafa,

Editor. 2014. 84. Kjell, J. and L. Olson, Rat models of spinal cord injury: from pathology to potential therapies.

Disease Models & Mechanisms, 2016. 9(10): p. 1125-1137. 85. Xiaofei Wang, X.-M.X., Spinal Cord Lateral Hemisection and Implantation of Guidance Channels,

in Animal Models of Acute Neurological Injuries, Z.C.X. J. Chen, X.-M. Xu, J.H. Zhang, Editor. 2009.

86. Heimburger, R.F., Return of function after spinal cord transection. Spinal Cord, 2005. 43(7): p. 438-40.

87. Lai, B.Q., et al., Graft of a tissue-engineered neural scaffold serves as a promising strategy to restore myelination after rat spinal cord transection. Stem Cells Dev, 2014. 23(8): p. 910-21.

88. Nakagawa, H., et al., Reorganization of corticospinal tract fibers after spinal cord injury in adult macaques. Scientific reports, 2015. 5: p. 11986-11986.

89. Du, K., et al., Deletion Promotes Regrowth of Corticospinal Tract Axons 1 Year after Spinal Cord Injury. The Journal of Neuroscience, 2015. 35(26): p. 9754-9763.

Page 50: An Injectable Gelatin-Based Conjugate Incorporating EGF

50

90. Chen, Q. and H.D. Shine, Neuroimmune processes associated with Wallerian degeneration support neurotrophin-3-induced axonal sprouting in the injured spinal cord. J Neurosci Res, 2013. 91(10): p. 1280-91.

91. Alilain, W.J., et al., Functional regeneration of respiratory pathways after spinal cord injury. Nature, 2011. 475(7355): p. 196-200.

92. Berthiaume, F., T.J. Maguire, and M.L. Yarmush, Tissue engineering and regenerative medicine: history, progress, and challenges. Annu Rev Chem Biomol Eng, 2011. 2: p. 403-30.

93. Langer, R. and J.P. Vacanti, Tissue engineering. Science, 1993. 260(5110): p. 920-6. 94. Bell, E., et al., Living tissue formed in vitro and accepted as skin-equivalent tissue of full

thickness. Science, 1981. 211(4486): p. 1052-4. 95. Hasan, R.M.A., Introduction to Tissue Engineering, in Tissue Engineering for Artificial Organs, A.

Hasan, Editor. 2017. 96. Ji, W., et al., Tissue engineering is a promising method for the repair of spinal cord injuries

(Review). Experimental and therapeutic medicine, 2014. 7(3): p. 523-528. 97. van Middendorp, J.J., G.M. Sanchez, and A.L. Burridge, The Edwin Smith papyrus: a clinical

reappraisal of the oldest known document on spinal injuries. European Spine Journal, 2010. 19(11): p. 1815-1823.

98. David, S. and A.J. Aguayo, Axonal elongation into peripheral nervous system "bridges" after central nervous system injury in adult rats. Science, 1981. 214(4523): p. 931-3.

99. Haggerty, A.E. and M. Oudega, Biomaterials for spinal cord repair. Neurosci Bull, 2013. 29(4): p. 445-59.

100. Krishna, V., et al., Biomaterial-based interventions for neuronal regeneration and functional recovery in rodent model of spinal cord injury: a systematic review. J Spinal Cord Med, 2013. 36(3): p. 174-90.

101. Petter-Puchner, A.H., et al., The long-term neurocompatibility of human fibrin sealant and equine collagen as biomatrices in experimental spinal cord injury. Exp Toxicol Pathol, 2007. 58(4): p. 237-45.

102. Mano, J.F., et al., Natural origin biodegradable systems in tissue engineering and regenerative medicine: present status and some moving trends. J R Soc Interface, 2007. 4(17): p. 999-1030.

103. Macaya, D. and M. Spector, Injectable hydrogel materials for spinal cord regeneration: a review. Biomed Mater, 2012. 7(1): p. 012001.

104. Dmitry, T., et al., Injectable Extracellular Matrix Hydrogels as Scaffolds for Spinal Cord Injury Repair. Tissue Engineering Part A, 2016. 22(3-4): p. 306-317.

105. Macaya, D., Biomaterials-Tissue Interaction of an Injectable Collagen-Genipin Gel in a Rodent Hemi-Resection Model of Spinal Cord Injury in Health, Science and Technology. 2013, Massachusetts Institute of Technology: Massachusetts Institute of Technology. p. 210.

106. Trimaille, T., V. Pertici, and D. Gigmes, Recent advances in synthetic polymer based hydrogels for spinal cord repair. Comptes Rendus Chimie, 2016. 19(1): p. 157-166.

107. Straley, K.S., C.W. Foo, and S.C. Heilshorn, Biomaterial design strategies for the treatment of spinal cord injuries. J Neurotrauma, 2010. 27(1): p. 1-19.

108. Hong, L.T.A., et al., An injectable hydrogel enhances tissue repair after spinal cord injury by promoting extracellular matrix remodeling. Nature communications, 2017. 8(1): p. 533-533.

109. Spector, M. and T.C. Lim, Injectable biomaterials: a perspective on the next wave of injectable therapeutics. Biomed Mater, 2016. 11(1): p. 014110.

110. Tsintou, M., K. Dalamagkas, and A.M. Seifalian, Advances in regenerative therapies for spinal cord injury: a biomaterials approach. Neural regeneration research, 2015. 10(5): p. 726-742.

111. Di Lullo, G.A., et al., Mapping the ligand-binding sites and disease-associated mutations on the most abundant protein in the human, type I collagen. J Biol Chem, 2002. 277(6): p. 4223-31.

Page 51: An Injectable Gelatin-Based Conjugate Incorporating EGF

51

112. Wood, G.C. and M.K. Keech, The formation of fibrils from collagen solutions. 1. The effect of experimental conditions: kinetic and electron-microscope studies. Biochem J, 1960. 75: p. 588-98.

113. Klapka, N. and H.W. Muller, Collagen matrix in spinal cord injury. J Neurotrauma, 2006. 23(3-4): p. 422-35.

114. Marchand, R. and S. Woerly, Transected spinal cords grafted with in situ self-assembled collagen matrices. Neuroscience, 1990. 36(1): p. 45-60.

115. Li, X., et al., Promotion of neuronal differentiation of neural progenitor cells by using EGFR antibody functionalized collagen scaffolds for spinal cord injury repair. Biomaterials, 2013. 34(21): p. 5107-16.

116. Freire, E., et al., Structure of laminin substrate modulates cellular signaling for neuritogenesis. J Cell Sci, 2002. 115(Pt 24): p. 4867-76.

117. Menezes, K., et al., Polylaminin, a polymeric form of laminin, promotes regeneration after spinal cord injury. Faseb j, 2010. 24(11): p. 4513-22.

118. Plantman, S., et al., Integrin-laminin interactions controlling neurite outgrowth from adult DRG neurons in vitro. Mol Cell Neurosci, 2008. 39(1): p. 50-62.

119. Johnson, P.J., S.R. Parker, and S.E. Sakiyama-Elbert, Fibrin-based tissue engineering scaffolds enhance neural fiber sprouting and delay the accumulation of reactive astrocytes at the lesion in a subacute model of spinal cord injury. J Biomed Mater Res A, 2010. 92(1): p. 152-63.

120. King, V.R., et al., The use of injectable forms of fibrin and fibronectin to support axonal ingrowth after spinal cord injury. Biomaterials, 2010. 31(15): p. 4447-56.

121. Bignami, A., M. Hosley, and D. Dahl, Hyaluronic acid and hyaluronic acid-binding proteins in brain extracellular matrix. Anat Embryol (Berl), 1993. 188(5): p. 419-33.

122. Horn, E.M., et al., Influence of cross-linked hyaluronic acid hydrogels on neurite outgrowth and recovery from spinal cord injury. J Neurosurg Spine, 2007. 6(2): p. 133-40.

123. Park, J., et al., Nerve regeneration following spinal cord injury using matrix metalloproteinase-sensitive, hyaluronic acid-based biomimetic hydrogel scaffold containing brain-derived neurotrophic factor. J Biomed Mater Res A, 2010. 93(3): p. 1091-9.

124. Jain, A., et al., In situ gelling hydrogels for conformal repair of spinal cord defects, and local delivery of BDNF after spinal cord injury. Biomaterials, 2006. 27(3): p. 497-504.

125. Lee, H., R.J. McKeon, and R.V. Bellamkonda, Sustained delivery of thermostabilized chABC enhances axonal sprouting and functional recovery after spinal cord injury. Proceedings of the National Academy of Sciences of the United States of America, 2010. 107(8): p. 3340-3345.

126. Hodgetts, S.I. and A.R. Harvey, Neurotrophic Factors Used to Treat Spinal Cord Injury. Vitam Horm, 2017. 104: p. 405-457.

127. Harvey, A.R., et al., Neurotrophic factors for spinal cord repair: Which, where, how and when to apply, and for what period of time? Brain Research, 2015. 1619: p. 36-71.

128. Ahuja, C.S. and M. Fehlings, Concise Review: Bridging the Gap: Novel Neuroregenerative and Neuroprotective Strategies in Spinal Cord Injury. Stem Cells Transl Med, 2016. 5(7): p. 914-24.

129. Kovalchuk, Y., K. Holthoff, and A. Konnerth, Neurotrophin action on a rapid timescale. Current Opinion in Neurobiology, 2004. 14(5): p. 558-563.

130. Blesch, A. and M.H. Tuszynski, Transient growth factor delivery sustains regenerated axons after spinal cord injury. J Neurosci, 2007. 27(39): p. 10535-45.

131. Han, S., et al., The collagen scaffold with collagen binding BDNF enhances functional recovery by facilitating peripheral nerve infiltrating and ingrowth in canine complete spinal cord transection. Spinal Cord, 2014. 52(12): p. 867-73.

Page 52: An Injectable Gelatin-Based Conjugate Incorporating EGF

52

132. Ollivier-Lanvin, K., et al., Either brain-derived neurotrophic factor or neurotrophin-3 only neurotrophin-producing grafts promote locomotor recovery in untrained spinalized cats. Neurorehabil Neural Repair, 2015. 29(1): p. 90-100.

133. McTigue, D.M., et al., Neurotrophin-3 and brain-derived neurotrophic factor induce oligodendrocyte proliferation and myelination of regenerating axons in the contused adult rat spinal cord. J Neurosci, 1998. 18(14): p. 5354-65.

134. Bradbury, E.J., et al., NT-3 promotes growth of lesioned adult rat sensory axons ascending in the dorsal columns of the spinal cord. Eur J Neurosci, 1999. 11(11): p. 3873-83.

135. Piantino, J., et al., An injectable, biodegradable hydrogel for trophic factor delivery enhances axonal rewiring and improves performance after spinal cord injury. Exp Neurol, 2006. 201(2): p. 359-67.

136. Gu, Y.L., et al., Neurotrophin expression in neural stem cells grafted acutely to transected spinal cord of adult rats linked to functional improvement. Cell Mol Neurobiol, 2012. 32(7): p. 1089-97.

137. Ruitenberg, M.J., et al., NT-3 expression from engineered olfactory ensheathing glia promotes spinal sparing and regeneration. Brain, 2005. 128(Pt 4): p. 839-53.

138. Martens, D.J., R.M. Seaberg, and D. Van Der Kooy, In vivo infusions of exogenous growth factors into the fourth ventricle of the adult mouse brain increase the proliferation of neural progenitors around the fourth ventricle and the central canal of the spinal cord. European Journal of Neuroscience, 2002. 16(6): p. 1045-1057.

139. Zhao, R.R. and J.W. Fawcett, Combination treatment with chondroitinase ABC in spinal cord injury--breaking the barrier. Neurosci Bull, 2013. 29(4): p. 477-83.

140. Hu, H.Z., et al., Therapeutic efficacy of microtube-embedded chondroitinase ABC in a canine clinical model of spinal cord injury. Brain, 2018. 141(4): p. 1017-1027.

141. Alluin, O., et al., Examination of the combined effects of chondroitinase ABC, growth factors and locomotor training following compressive spinal cord injury on neuroanatomical plasticity and kinematics. PLoS One, 2014. 9(10): p. e111072.

142. Lu, P., Chapter 1 - Stem cell transplantation for spinal cord injury repair, in Progress in Brain Research, S.B. Dunnett and A. Björklund, Editors. 2017, Elsevier. p. 1-32.

143. Miura, K., et al., Variation in the safety of induced pluripotent stem cell lines. Nat Biotechnol, 2009. 27(8): p. 743-5.

144. Bottai, D., et al., Embryonic stem cells promote motor recovery and affect inflammatory cell infiltration in spinal cord injured mice. Exp Neurol, 2010. 223(2): p. 452-63.

145. Hatami, M., et al., Human embryonic stem cell-derived neural precursor transplants in collagen scaffolds promote recovery in injured rat spinal cord. Cytotherapy, 2009. 11(5): p. 618-30.

146. Kotobuki, N., et al., Cultured Autologous Human Cells for Hard Tissue Regeneration: Preparation and Characterization of Mesenchymal Stem Cells from Bone Marrow. Artificial Organs, 2004. 28(1): p. 33-39.

147. Pittenger, M.F., et al., Multilineage potential of adult human mesenchymal stem cells. Science, 1999. 284(5411): p. 143-7.

148. Malgieri, A., et al., Bone marrow and umbilical cord blood human mesenchymal stem cells: state of the art. Int J Clin Exp Med, 2010. 3(4): p. 248-69.

149. Nakajima, H., et al., Transplantation of mesenchymal stem cells promotes an alternative pathway of macrophage activation and functional recovery after spinal cord injury. J Neurotrauma, 2012. 29(8): p. 1614-25.

150. Mothe, A.J., et al., Intrathecal transplantation of stem cells by lumbar puncture for thoracic spinal cord injury in the rat. Spinal Cord, 2011. 49(9): p. 967-73.

151. Gu, W., et al., Transplantation of bone marrow mesenchymal stem cells reduces lesion volume and induces axonal regrowth of injured spinal cord. Neuropathology, 2010. 30(3): p. 205-17.

Page 53: An Injectable Gelatin-Based Conjugate Incorporating EGF

53

152. Li, J. and G. Lepski, Cell Transplantation for Spinal Cord Injury: A Systematic Review. BioMed Research International, 2013. 2013: p. 32.

153. Mezey, E., et al., Turning blood into brain: cells bearing neuronal antigens generated in vivo from bone marrow. Science, 2000. 290(5497): p. 1779-82.

154. Cho, S.R., et al., Functional recovery after the transplantation of neurally differentiated mesenchymal stem cells derived from bone barrow in a rat model of spinal cord injury. Cell Transplant, 2009. 18(12): p. 1359-68.

155. Kim, J.-W., et al., Bone Marrow–Derived Mesenchymal Stem Cell Transplantation for Chronic Spinal Cord Injury in Rats: Comparative Study Between Intralesional and Intravenous Transplantation. Spine, 2013. 38(17): p. E1065-E1074 10.1097/BRS.0b013e31829839fa.

156. Hamilton, L.K., et al., Cellular organization of the central canal ependymal zone, a niche of latent neural stem cells in the adult mammalian spinal cord. Neuroscience, 2009. 164(3): p. 1044-56.

157. Reubinoff, B.E., et al., Neural progenitors from human embryonic stem cells. Nat Biotechnol, 2001. 19(12): p. 1134-40.

158. Yan, J., et al., Extensive neuronal differentiation of human neural stem cell grafts in adult rat spinal cord. PLoS Med, 2007. 4(2): p. e39.

159. Yasuda, A., et al., Significance of remyelination by neural stem/progenitor cells transplanted into the injured spinal cord. Stem Cells, 2011. 29(12): p. 1983-94.

160. Zholudeva, L.V., et al., Transplantation of Neural Progenitors and V2a Interneurons after Spinal Cord Injury. J Neurotrauma, 2018.

161. Tureyen, K., et al., EGF and FGF-2 infusion increases post-ischemic neural progenitor cell proliferation in the adult rat brain. Neurosurgery, 2005. 57(6): p. 1254-63; discussion 1254-63.

162. Lee, D.C., et al., Isolation of neural stem/progenitor cells by using EGF/FGF1 and FGF1B promoter-driven green fluorescence from embryonic and adult mouse brains. Mol Cell Neurosci, 2009. 41(3): p. 348-63.

163. Richter, M.W., et al., Lamina propria and olfactory bulb ensheathing cells exhibit differential integration and migration and promote differential axon sprouting in the lesioned spinal cord. J Neurosci, 2005. 25(46): p. 10700-11.

164. Zhang, S.X., et al., Scar ablation combined with LP/OEC transplantation promotes anatomical recovery and P0-positive myelination in chronically contused spinal cord of rats. Brain Res, 2011. 1399: p. 1-14.

165. Agudo, M., et al., Schwann cell precursors transplanted into the injured spinal cord multiply, integrate and are permissive for axon growth. Glia, 2008. 56(12): p. 1263-70.

Page 54: An Injectable Gelatin-Based Conjugate Incorporating EGF

54

Chapter 3:

Rationale and development of the

formulation of Gtn-HPA matrix

formulation with EGF and SDF-1α for

spinal cord injury rat model evaluation

Page 55: An Injectable Gelatin-Based Conjugate Incorporating EGF

55

3.1. Biomaterial: Gelatin-hydroxyphenyl propionic acid (Gtn-HPA)

3.1.1. Introduction: Various biomaterials, both individually or in combination with factors

and cells, have been tested for promoting regeneration and repair post SCI. Biomaterials provide

passive structural support, bioactive modulation of matrix, delivery vehicle for factors and

attachment surface for cells. Endogenous cells such as neural stem cells (NSCs) and glial cells

need a supportive matrix to migrate into the defect. Therefore, natural polymers such as collagen,

laminin, PLGA, and chitosan have been widely researched with cells and factors, which have

shown decrease in the cyst volume, increase in angiogenesis and axonal regrowth [1, 2]. For

example, when equine collagen alone was implanted in hemi-resection SCI model, it showed

increase in the number of axons [3]. Generally, pre-formed scaffolds are not suitable for SCI due

to large variation in injury size and volume. So, injectable gels offer a minimally invasive novel

way to deliver biomaterial matrix accounting for injury specifics conferring greater preservation of

the spared tissue.

The general design criteria of the injectable matrix for SCI applications could be as follows:

1) can undergo gelation in vivo within 1-2 minutes; 2) modulus matches that of spinal cord; 3)

have tunable physical properties to control degradation rate; and 4) should be able to persist for

timeline of our experiment and longer [4]. There various hydrogels that are based on the HRP-

based crosslinking mechanism by decomposing hydrogen peroxide (H2O2) that offer independent

tuning of properties of the gel [5, 6]. Considering these criteria, we propose to use a conjugated

gelatin (Gtn)-hydroxyphenyl propionic acid (HPA) gel that is capable of being cross-linked by

peroxidase and hydrogen peroxide, which allow for independent tuning of rate and degree of

cross-linking, respectively. This independent tunability is essential in obtaining an optimal

degradation rate, and in matching the mechanical properties (shear modulus) of the gel and of

the spinal cord tissue [7]. Moreover, arginine-glycine-aspartic acid (RGD peptide sequences) are

important moiety for cell attachment, cell anchorage, cell spreading and cell signaling [8]. Since

gelatin-based hydrogels such as Gtn-HPA comprises of abundance of RGD sequences that aid

in establishing ECM-like material for the migrating cells, Gtn-HPA is a promising biomaterial for

tissue repair applications [9].

3.1.2. Synthesis of Gtn-HPA: Gtn-HPA hydrogels were developed by Wang et al with the

intent to be utilized as scaffolds for tissue engineering applications [10]. Fabricating Gtn-HPA gel

is two step procedure: First, a lyophilized conjugate of gelatin (Gtn) and 3-(4-hydroxyphenyl)

propionic acid (HPA) is prepared from raw materials (Figure 3.1A). Second, Gtn-HPA matrix is

prepared by covalently cross-linking the Gtn-HPA conjugate with horseradish peroxidase (HRP)

Page 56: An Injectable Gelatin-Based Conjugate Incorporating EGF

56

and hydrogen peroxide by an HRP-catalyzed crosslinking mechanism (Figure 3.1B). The gelatin

content, HRP & H2O2 concentration is chosen based on the application and desired properties of

the Gtn-HPA hydrogel in the second step.

Synthesis of Gtn-HPA conjugate: Gtn–HPA conjugates were prepared by a general

carbodiimide/active ester-mediated coupling reaction in distilled water. 3.32 g of HPA was

dissolved in 250 ml of mixture prepared with 3:2 ratio of distilled water and N,N-

dimethylformamide (DMF) respectively. To this, 3.20 g of N-hydroxysuccinimide and 3.82 g of 1-

ethyl-3-(3-dimethylaminopropyl)-carbo-diimide hydrochloride were added. The reaction mixture

was stirred at room temperature for 5 hours, and the pH of the mixture was kept at 4.7. Then, 150

ml of 6.25 wt% gelatin aqueous solution was added to the reaction mixture and continuously

stirred overnight at room temperature at pH 4.7. Subsequently, the solution was dialyzed against

Figure 3.1 Schematic of the two-step fabrication of Gtn-HPA gel (A) A non-crosslinked Gtn-HPA conjugate is prepared using gelatin (Gtn) and 3-(4-hydroxyphenyl) propionic acid (HPA). Gtn-HPA conjugate can be lyophilized to be used for later experiments (B) Gtn-HPA gel is prepared by enzymatic HRP-catalyzed crosslinking mechanism using horseradish peroxidase (HRP) and hydrogen peroxide (H2O2). Figure courtesy – Hu et al Biomaterials, 2009. 30(21): 3523-3531

Page 57: An Injectable Gelatin-Based Conjugate Incorporating EGF

57

100 mM sodium chloride solution for 2 days, 25% ethanol in distilled water for 1 day and distilled

water for 1 day, successively. The purified solution was lyophilized to obtain the Gtn–HPA

conjugate [10]. Lyophilized Gtn-HPA conjugate was obtained from Dr. Motoichi Kurisawa, Agency

of Science, Technology and Research (A*STAR), Singapore.

Synthesis of Gtn-HPA hydrogels: Gtn-HPA gels were prepared by first preparing a Gtn-HPA

conjugate solution. In this example, procedure to prepare 8% Gtn-HPA is discussed (Table 3.1).

12% Gtn-HPA solution was prepared by adding PBS to the lyophilized Gtn-HPA conjugate to a

final concentration of 120 mg/ml in PBS. The resulting solution was kept at 37 °C for an hour for

complete dissolution of the Gtn-HPA conjugate to prepare a homogeneous 12% Gtn-HPA

conjugate solution. 12% Gtn-HPA, 1x PBS, HRP and H2O2 was added in volumetric ratio of

20:8:1:1 to a final concentration of 8% Gtn-HPA, 0.1 U/ml HRP and 6.8 mM H2O2 in the specified

order. HRP and H2O2 solutions were kept at 4°C and away from the direct light.

3.1.3. Mechanical properties of Gtn-HPA: Mechanical properties should be considered

while developing an injectable hydrogel for in vivo applications because modulus mismatch has

been shown to induce strains at the interface impeding integration of the gel with the native tissue

[11]. Therefore, one of the noteworthy advantages of the Gtn-HPA gels is that the mechanical

properties are tunable based on the concentration of H2O2 and HRP used for synthesis of the

Gtn-HPA gel. Prior work in our laboratory showed that the gelation rate and shear modulus of the

Gtn-HPA matrix is independently tunable by controlling the concentration of HRP and H2O2

respectively (Figure 3.2AB) [12]. Moreover, concentration of H2O2 and the gelatin content of the

Gtn-HPA affects the degradation rate of the Gtn-HPA gels – an important property for the tissue

engineering applications (Figure 3.2C). Greater the H2O2 concentration, higher is the stiffness and

therefore slower is the rate of degradation of the Gtn-HPA gel. Increase in the concentration of

HRP reduces the gelation time. These studies and parameters were important in determining the

final Gtn-HPA formulation that was evaluated in vivo in the SCI models for this thesis work.

Formulation Gtn-HPA HRP H2O2 Gelation time and estimated modulus

8% Gtn-HPA 80 mg/ml 0.1 U/ml 6.8 mM ~ 3 min & ~ 2000 Pa

12% Gtn-HPA 120 mg/ml 0.1 U/ml 10.2 mM ~ 4 min & ~ 3000 Pa

Table 3.1 Gtn-HPA formulation with their constituent concentrations

Page 58: An Injectable Gelatin-Based Conjugate Incorporating EGF

58

3.1.4. Prior in vitro studies with Gtn-HPA: Given the promise of Gtn-HPA hydrogels with

valuable tunable properties, various in vitro evaluation of the Gtn-HPA with different cell types

have assessed the permissiveness of the Gtn-HPA for cellular migration and effect on the fates

of the cells cultured onto the 2D or 3D Gtn-HPA hydrogels. Wang et al reported that the stiffness

of the Gtn-HPA gel strongly directed cell adhesion, proliferation and differentiation of the human

mesenchymal stem cells (hMSCs) for both 2D and 3D cell culture [10, 13]. The stiffer Gtn-HPA

showed greater spreading area, more stable cell attachment, faster migration and better

organized cytoskeleton than the softer Gtn-HPA gel. hMSCs cultured on the softer Gtn-HPA gel

were reported to express neurogenic markers while the stiffer Gtn-HPA gel upregulated the

myogenic markers in hMSCs without addition of any chemical or differentiation cues. Similar

evaluation was performed by Wang et al with human fibroblasts (HFF-1) to study the effect of the

stiffness of the Gtn-HPA gel. In 2D environment, the proliferation of HFF-1 increased when

cultured on stiffer Gtn-HPA gel. However, the proliferation decreased with stiffer 3D environment

[14]. The phenomenon of the stiffness-dependent differentiation of stem cells and proliferation of

cells have been widely accepted in the tissue engineering research as exhibited by Gtn-HPA [15].

Wang et al presented encouraging findings of preliminary in vivo evaluation of the Gtn-HPA

Figure 3.2 Dependence of HRP and H2O2 concentration on the gelation time, stiffness and degradation of the Gtn-HPA gels. Insert in C plots the rate of degradation (%/hr).

Figure modified from Lim, TC et al Biomaterials, 2012. 33(12): 3446-3455

Page 59: An Injectable Gelatin-Based Conjugate Incorporating EGF

59

hydrogels with varying stiffness for applications in cartilage repair in a rabbit model. The results

showed stiffness-dependent cartilage repair. Gtn-HPA hydrogels with medium stiffness were

shown to have greater hyaline cartilage formation compared to low or high stiffness Gtn-HPA

matrices. Moreover, medium stiffness Gtn-HPA hydrogels also promoted better integration with

the surrounding unaffected tissue promoting osteochondral repair [16]. Various studies have since

employed Gtn-HPA in several applications including microfluidic cell culture, bone regeneration

and personalized tissue engineering [17-19].

Prior work in our laboratory by Lim et al commenced in vitro evaluation to study the effects

of Gtn-HPA hydrogels on the migration and differentiation of adult neural stem cells (aNSCs).

Results showed high viability of the aNSCs on the Gtn-HPA matrix with good cell adhesion. Gtn-

HPA promoted proliferation and migration of aNSCs. Interestingly, viability of the aNSCs was

higher than collagen scaffolds even with as high concentration as 500 µM H2O2 indicating that

Gtn-HPA cross-linking mechanism aided in increasing the resistance to oxidative stress by

preconditioning aNSCs to sub-cytotoxic levels of oxidative stress. Remarkably, Gtn-HPA

modulated the differentiation of aNSCs towards astrocytic and neuronal lineage – however, with

mixed conditions for astrocytic and neuronal fate, Gtn-HPA showed augmented levels of βIII

tubulin compared to GFAP indicating that Gtn-HPA promoted differentiation of aNSCs towards

neuronal fate [7, 20]. Moreover, SDF-1α encapsulated in Gtn-HPA promoted chemotactic

recruitment of adult rat hippocampal neural progenitor cells into the Gtn-HPA gel confirming that

Gtn-HPA is permissive of cellular migration in addition to providing a cytocompatible matrix [12].

3.1.5. Gtn-HPA - promising biomaterial for spinal cord injury repair: Several promising

biomaterials were evaluated in our laboratory for tissue repair and functional recovery after spinal

cord injury in rodent models. Spilker et al used pre-formed collagen and collagen-GAG tubes in a

T7-T10 5-mm complete transection rodent model to evaluate tissue regeneration after spinal cord

injury. Although the collagen-based tubes resulted in oriented axonal and connective tissue

constituents with reduced scar formation, much of the collagen-based tube was absorbed by 30

days. Additionally, the collagen-based matrices and tubes did not induce migration of endogenous

reparative cells. Moreover, pre-fabricated tubes would not be clinically feasible given that the

spinal cord cavities are irregular in shape and heterogeneous in volume [21, 22]. Subsequently,

Cholas et al evaluated pre-formed collagen scaffolds delivering chondroitinase ABC (chABC) and

mesenchymal stem cells (MSCs) in a 3-mm and 5-mm hemi-section model of spinal cord injury.

The collagen-based scaffold with the therapeutic agents induced angiogenesis, cellular infiltration

and improved functional recovery in collagen-treated animals. However, there were two major

Page 60: An Injectable Gelatin-Based Conjugate Incorporating EGF

60

limitations of the study – the collagen scaffolds resorbed almost completely by four weeks and

pre-formed scaffolds would not be clinically viable to treat spinal cord injury [23]. However, it

became increasingly clear that pre-formed scaffolds are not suitable for future evaluation.

Therefore, Macaya et al evaluated a promising injectable collagen-genipin gel incorporating

fibroblast growth factor (FGF-2) in a 3-mm hemi-resection model of spinal cord injury in Lewis

rats based on the exciting in vitro results [24]. The in vivo evaluation after four weeks post

implantation of the injectable collagen-genipin gel displayed greater astrocyte migration into the

defect site, some GAP-43 positive axons, and correlation between the functional score and

spared tissue. However, collagen-genipin gel degraded as early as two weeks after implantation

limiting the long-term delivery of FGF-2 for any potential therapeutic effect. Increasing the genipin

concentration, to obtain greater stiffness of the collagen-genipin gel with greater cross-linking,

was not feasible because of toxicity [25]. Due to these limitations and our prior work with

biomaterials for spinal cord injury repair, we devised specific comprehensive criteria for the

biomaterial to be evaluated in vivo for this thesis work as follows [26] and Gtn-HPA met these

criteria making it a promising biomaterial for spinal cord injury applications:

1) It should be biocompatible without inducing more than typical inflammatory response

2) It should be injectable so that it can be delivered to the injured cavity in minimally invasive

manner to abate surgical damage during implantation

3) It should be able to gelate in situ to conform to the heterogeneous spinal cord cavities

4) It should be non-swelling not to apply pressure to the surrounding tissue, which is known to

initiate further secondary damage after the injury

5) It should have similar mechanical properties as that of the spinal cord to minimize secondary

damage and non-shrinking so that it can integrate well with the surrounding tissue

6) The constituents of the biomaterial should not induce cytotoxicity in the tissue

7) It should serve as a viable replacement of the disappearance of stroma lost due to spinal cord

injury and be permissive of cellular migration, proliferation and differentiation into the defect site

to promote neural repair

8) It should have tunable physical properties to control gelation time, stiffness and degradation

characteristics. Specifically, for experimental models, it should be able to complete gelation in

two-four minutes to visually confirm the gelation in the surgical hemi-resection model

9) It should have chemical moieties such as RGD sequences as ligand for cell integrins

10) It should be able to persist in vivo for timeline of our experiment (i.e. four weeks) and longer

11) It should be able to incorporate and locally deliver signaling molecules, growth factors, and

cells in the injury site without the particular use of vehicles

Page 61: An Injectable Gelatin-Based Conjugate Incorporating EGF

61

Based on the promising in vitro and preliminary in vivo evaluation of Gtn-HPA matrices

in the intracerebral hemorrhagic (ICH) stroke rodent model, we examined the Gtn-HPA in greater

detail to assess if it would be a potential candidate biomaterial that can overcome the limitations

of previous work and meet the devised criteria above. We found Gtn-HPA to be very promising

given its properties. First, both in vitro and in vivo evaluation of the Gtn-HPA have found it be a

biocompatible material that does not elicit a severe inflammatory reaction. Moreover, it is an

injectable biomaterial capable of undergoing covalent cross-linking after being injected as a liquid

[10, 20]. Therefore, it can provide conformational filling of the heterogenous spinal cord injury

cavitary lesions. In addition, HRP and H2O2 allow for independent tuning of rate and degree of

cross-linking, respectively by choosing a desired concentration. This independent tunability is

essential in obtaining an optimal degradation rate, and in matching the physical properties (elastic

modulus) of the gel and of the spinal cord tissue [7]. The shear modulus of spinal cord is reported

to be an average of 1500-2500 Pa [25]. The concentration of 1.7 mM H2O2 per 2wt% Gtn-HPA

gel would fabricate Gtn-HPA in the range of reported values of the spinal cord (Figure 3.2B).

Moreover, desired gelation time of two-to-four minutes was achieved using HRP concentration

between 0.1 and 0.15 U/ml using quick bench-top Gtn-HPA fabrication. Furthermore, Gtn-HPA

has been shown to be permissive of cellular migration as reported in several studies [7, 12, 20].

Gtn-HPA was found to be non-shrinking and non-swelling based on the prior in vitro evaluation

by Wang et al [10]. In vivo evaluation of Gtn-HPA for intracerebral hemorrhagic stroke model

showed persistence of the Gtn-HPA gel two weeks after implantation in our laboratory work.

Therefore, these collective prior studies show that the Gtn-HPA meets all the criteria of an ideal

biomaterial for implantation in the cavity created in the spinal cord injury to assess the tissue

repair, regeneration and functional recovery in SCI model (Table 3.2). Our work will evaluate the

persistence of the Gtn-HPA for four weeks in 2-mm complete and hemi-resection spinal cord

injury rat models. Considering the results from prior in vitro and in vivo evaluation of Gtn-HPA, we

propose to evaluate the effect of injectable Gtn-HPA gel incorporating epidermal growth factor

(EGF) alone or in combination with stromal cell-derived factor (SDF-1α) because it is a very

promising biomaterial for applications in spinal cord injury.

Page 62: An Injectable Gelatin-Based Conjugate Incorporating EGF

62

Property Function Gtn-HPA characteristics

Reference

Porosity Tunable porosity, cell encapsulation and migration

Tailorable porosity, hMSCs in 3D Gtn-HPA culture, monocytes

Al-Abboodi et al, 2014 [19]; Wang LS et al, 2010 [10]; Bystronova et al, 2018 [27]

Tunable mechanical properties

Control of stiffness to match the host tissue modulus and gelation rate

Concentration of H2O2 (stiffness) and HRP (gelation time)

Lim TC et al, 2012 [7]

Degradation rate Persistence for long duration in vivo

Concentration of H2O2 and % gelatin content

Wang LS et al, 2010 [13]

Maintain cell viability

Provide non-cytotoxic environment

High viability for hMSCs, HFF-1, aNSCs

Wang LS et al, 2012 [14]; Lim TC et al, 2013 [7, 12]

Permit cell adhesion

Promote cell proliferation and migration

RGD sequences Hu et al, 2009 [9]

Permissive of cell migration, proliferation and differentiation

Repopulate the cells and promote tissue repair and regeneration of the tissue

Stiffness-dependent differentiation of hMSCs, migration of aNSCs

Wang et al, 2012 [28], Lim et al, 2013 [12]

Shrinking Integrate well with surrounding host tissue

Non-shrinking Lim TC, 2014 [20]

Swelling Minimize additional pressure on the host tissue

Non-swelling Wang LS, 2010 [10]

Oxidative stress Reduced oxidative stress to minimize cytotoxicity

Confers resistance to oxidative stress

Lim TC et al, 2012 [7]

Formulations Fabrication in different forms

Electrospun fibers, hydrogels

Hu et al 2009 [9]; Wang LS et al

In vivo injectable implantation

Pre-clinical and clinical applications

Bone regeneration, cartilage repair, spinal cord injury, stroke injury, retinal pathologies

Chun et al, 2016 [18]; Wang et al, 2014 [16]; Lim TC, 2014 [20]; Shah, A 2018 (unpublished), Love, C (ongoing), Colombe, P (ongoing)

Table 3.2 Gtn-HPA exhibits various important properties making it an ideal biomaterial for spinal cord application

Page 63: An Injectable Gelatin-Based Conjugate Incorporating EGF

63

3.2. Growth factor: Epidermal growth factor (EGF)

3.2.1. Introduction: Epidermal growth factor (EGF) is shown to induce cell growth,

proliferation, differentiation and signaling modulation in mesodermal and ectodermal cells.

Specifically, in addition to other organs such as kidney, thyroid gland and pancreas, it is also

produced in the nervous system. EGF has been isolated from central nervous system (CNS),

peripheral nervous system (PNS) and cerebrospinal fluid (CSF) although the concentration of

EGF varies depending on the location in the nervous system [29]. Immunohistochemistry

demonstration of EGF and EGFR showed the presence of both EGF and EGFR in spinal cord

tissues specifically in the anterior horn cells [30]. CNS has abundance of epidermal growth factor

receptors (EGFR) and has been shown to play critical role in the development of the brain. EGFR

is activated predominantly by EGF but also by other factors such as heparin-binding epidermal

growth factor (hb-EGF) and transforming growth factor-α (TGF- α) [31]. EGF and its family of

proteins have been shown to play a critical role during embryonic development of the neural

tissue. Neuregulin-1 (NRG1) contains EGF-like domain and is widely studied and characterized

to play a vital role during neural development. Interestingly, NRG1-mutated mice show profound

changes in the neural development at various temporal stages [32]. Moreover, NRG1 is released

from the mature neurons to promote the formation and maintenance of glial cells, which are

essential in directing the migration of neurons from the ventricular regions to the developing

tissue. Furthermore, study by Nelson et al reported that the high concentrations of EGF leads to

greater growth and survival of neurogenic radial glial cells in human neurosphere culture including

higher expression of glial markers [33].

EGF and NRG1 plays a major role in axon guidance, axon sprouting, axonal myelination

and synapse formation [32, 34, 35]. EGF aids enhanced neural cell proliferation in the sub-

ventricular zone in brain. Moreover, EGF has shown great promise by promoting neurogenesis in

brain interneurons as well [36]. In an in vitro study, cells isolated from embryonic brain and adult

sub-ventricular zone differentiated into radial glial cells and supported neuronal cell migration [37].

One of the most profusely reported roles of EGF in the nervous system regulating the cell

recruitment, proliferation and eventual differentiation of neural stem cells (NSCs). NSCs are highly

responsive to EGF receptors stimulating their proliferation [38]. Due to its pivotal role in neural

tissue development and growth, EGF has been studied in SCI experimental models to recruit

endogenous NSCs and promote tissue repair post SCI. For example, intrathecal administration

of EGF resulted in significantly greater ependymal cell proliferation in the central

canal immediately rostral and caudal to the lesion and greater white matter sparing compared to

controls [39]. Several other studies have also shown axonal regeneration and functional recovery

Page 64: An Injectable Gelatin-Based Conjugate Incorporating EGF

64

in SCI using EGF [40, 41]. Therefore, EGF was considered as a potential factor to be incorporated

in Gtn-HPA.

3.2.2. Prior relevant in vitro studies with EGF: There are several relevant and promising

in vitro investigations that prompted the delivery of EGF in the injury site created by the spinal

cord pre-clinical rat model in our work. There is evidence that multipotent neural stem cells are

present in the ependymal layer near the central canal of the adult mammalian spinal cord. These

cells were confirmed to be NSCs by Nestin, Sox2, and CD-133 [42]. In addition, these cells are

highly responsive in proliferation and migration to EGF treatment [43]. Several studies have

supported the findings of these studies. Based on the prior work demonstrating higher levels of

EGF in the region of injury after spinal cord injury [44], Kang et al studied the effect of EGF on the

proliferation of neural progenitor cells isolated from adult mice spinal cord. As early as 3-6 days

of culture in EGF-containing medium, the authors reported significantly greater proliferation of

neural progenitor cells compared to the controls indicating that EGF truly promotes the cellular

growth of neural progenitor cells present in the spinal cord via JAK2/STAT3 and MAPK signaling

pathway. These pathways are reported to be of critical importance in the regeneration after spinal

cord injury in various experimental models [45].

Lam et al studied the effect of EGF on the neural differentiation of embryonic stem cells

and axon growth in ESCs cultured on PLLA nanofibrous scaffolds [46]. They also assessed the

effect of adsorbed and soluble EGF. The results showed that EGF treatment significantly

increased the expression of neuronal markers NES and NEF3 in ESC-derived neural cells

compared to the controls suggesting that EGF promotes differentiation into the neuronal lineage.

In addition, EGF promoted greater axonal outgrowth from the surface of the nanofibers when they

were functionalized with heparin compared to the controls. Moreover, directly adsorbed EGF on

the nanofibers were not effective in promoting axonal growth on the nanofibers. These results

suggest that soluble EGF delivery would be effective in Gtn-HPA compared to its delivery in a

vehicle to which EGF would be adsorbed. Yet another interesting in vitro assessment of the role

of EGF was done by Garcez et al. Neural crest stem cells were isolated and were exposed to

EGF for six days. After 6 days of culture, EGF seemed to increase the number of melanocytes

and neurons in the neural crest stem cell culture [47]. These results further support that the theory

that EGF promotes cell growth, proliferation, migration of the neural stem cells and differentiation

of these neural stem cells into neuronal lineage. Therefore, we propose to deliver soluble EGF

via Gtn-HPA in the injury site in our 2-mm complete and hemi-resection preclinical spinal cord

injury rat model to recruit endogenous neural stem cells into the injury site and promote their

Page 65: An Injectable Gelatin-Based Conjugate Incorporating EGF

65

differentiation into neurons and glial cells. Our proposition also found rationale from the results of

various in vivo studies with EGF.

3.2.3. Prior in vivo studies incorporating EGF in SCI models: Encouraging

preliminary in vitro studies investigating the role of EGF have led to pre-clinical examination of

the role of EGF in promoting tissue repair and recovery post spinal cord injury. EGF has been

shown to have neuroprotective effect in spinal cord after SCI in a rat model [48]. Moreover, it has

shown promise as a therapeutic agent for preventing blood-spinal cord barrier interruption after

spinal cord injury [49]. Considering these beneficial effects of EGF, a recent study by Ozturk et al

investigated the effect of EGF in a rat model of spinal cord injury in the acute and subacute phase

by delivering EGF in the injury site with a sterilized pipette with a Hamilton syringe. Injection of

EGF was shown to reduce the apoptosis-related marked in the injury site as early as 8 days post

injury. Moreover, EGF also increased the levels of antioxidants suggesting that EGF plays a role

in regulating apoptosis and promoting reduction in the oxidative stress in the injury site, thereby

promoting reparative and regenerative mechanism. Lastly, EGF also showed better functional

recovery than the controls. However, 8-day timepoint is too early to evaluate functional recovery

[50]. Recently, Sun et al explored the benefits that tacrolimus (FK506) could provide in tissue

repair after spinal cord injury. Tacrolimus is an FDA-approved immunosuppressant and typically

used to mitigate the acute rejection after transplant surgeries. Remarkably, tacrolimus increased

the levels of EGF expression both in vitro and in vivo, which led to longer neurite length suggesting

that EGF plays a vital role in promoting neurite outgrowth after injury. Moreover, better functional

recovery was observed suggesting that EGF plays a key role in tissue repair post SCI [51]. Alluin

et al investigated a combinatorial approach with EGF as one of the key growth factors in the

cocktail of factors containing FGF-2, PDGF-AA, and chondroitinase ABC (chABC) in a clip spinal

compression model. This combination was injected into the injury site via subarachnoid catheter

connected to an osmotic pump. The combination showed reduced GFAP presence in the

perilesional area indicating reduction of astrogliosis near the injury site. Interestingly, even the

macrophages response was attenuated in the defect area due to the administration of the

combination of chABC and growth factors. One of the key findings was collateral sprouting of the

corticospinal tract using BDA tracing. There was greater spared tissue observed in the cord due

to implantation of the chABC with growth factors compared to the controls. Lastly, this intervention

also shown improved functional recovery suggesting augmentation of neuronal plasticity [40].

Given the complexity of the damage after spinal cord, more combinatorial studies such as by

Alluin et al should be investigated to promote repair after spinal cord injury.

Page 66: An Injectable Gelatin-Based Conjugate Incorporating EGF

66

Based on the finding that EGF and FGF-2 induces proliferation of the stem cells residing

near fourth ventricle in the brain and spinal cord, Martens et al pursued an in vivo investigation in

a mouse model. They injected a combination of EGF and FGF-2 in the fourth ventricle in the brain

and central canal of adult male CD1 mice. Just six days post injection of EGF and FGF, greater

proliferating cells were observed in both the cervical spinal cord and near fourth ventricle in the

brain. Specifically, in the spinal cord, these cells were in the ependymal and subependymal tissue

closer to the central canal. Indeed, these cells were nestin positive indicating the presence of

neural stem cells. The results from this study indicated that endogenous neural stem cells near

the central canal in the spinal cord proliferate in response to EGF and FGF-2 [52]. Niche of these

neural progenitor cells has previously been shown as well [42]. Combination of EGF and FGF-2

was investigated by Kojima et al in an adult rat spinal cord. EGF and FGF-2 was intrathecally

delivered over 14 days using an osmotic pump. EGF groups showed the two highest BrdU+ cells

in the ependymal tissue were found in EGF alone and EGF+FGF-2 group. These results suggest

that EGF has a mitogenic effect on the migration and proliferation of the ependymal precursor

cells, which are shown to be closely related to neural stem cells. As anticipated, these cells also

stained positive for nestin confirming that these ependymal precursor cells are neural stem cells

(NSCs) [41]. In a more comprehensive in vivo study by Hamann et al investigating the same EGF

and FGF-2 combination in a clip compression rat model. EGF and FGF-2 were delivered in a

highly concentrated collagen solution to localize the growth factors in the injury site. Interestingly,

they labeled the injected EGF and FGF-2 with markers. The results indicated that EGF readily

penetrated the injury site in the spinal cord while FGF-2 was predominantly observed in the dura.

Furthermore, significantly less cavitations were seen in the growth factors group indicating that

EGF promoted tissue sparing including greater white matter sparing in the injury site. Moreover,

as reported by Kojima et al, they also observed greater BrdU+ cells in the EGF and FGF-2 group

than the controls in the ependymal region [39].

All these studies validated a critical mitogenic role of EGF in proliferation of the neural

stem progenitor cells found near the central canal. However, ability of EGF to chemotactically

recruit endogenous NSCs has not been studied in vivo. Prior work in our laboratory showed

chemoattractive effect of EGF towards NSCs in vitro. Injection of Gtn-HPA/EGF in the

intracerebral hemorrhage model of the stroke showed the ability of EGF to repopulate the injury

area with NSCs (manuscript in publication). Considering encouraging results of the in vivo

evaluation of EGF as discussed above, we propose to provide evidence of EGF-induced migration

of the NSCs to the injury site in a rodent SCI model.

Page 67: An Injectable Gelatin-Based Conjugate Incorporating EGF

67

3.3. Chemokine: Stromal cell-derived factor-1α (SDF-1α)

3.3.1. Introduction: SDF-1α, also known as CXCL12, was one of the first α-chemokines to

be discovered and has the closest homology to β-chemokines. SDF-1α is regarded as a

chemoattractant that is synthesized by bone marrow stromal cell lines. CXCR4, one of the

receptors for SDF-1α, along with SDF-1α have been highly expressed in a developing embryo

indicating the crucial role SDF-1α plays in the development process [53]. SDF-1α has been widely

studied because of its ample secretion and rich expression in nearly all the organs. SDF-1α has

been shown to be a potent chemoattractant for monocytes, T cells, pre-B cells, hematopoietic

progenitor cells, and dendritic cells. SDF-1α has been shown to play a prominent role in various

processes such as implantation of the embryo, germ cell migration, and cerebellar development

[54]. Specifically, exploratory research over the last two decades have shown critical role of SDF-

1α/CXCR4 signaling pathway in the development of cerebellum, dorsal root ganglion and cerebral

cortex [55, 56]. The common aspect of SDF-1α in all these processes defined the role of SDF-1α

in the migration of cells such as interneurons and neuronal precursors. Strangely, Tiveron et al

demonstrated that the motor neuron progenitor cells of the cortex express SDF-1α to regulate the

migration of the interneurons near the subventricular zone [57].

In addition to the migration of precursors and interneurons, SDF-1α has also been shown

to act as a cue for axonal guidance during neural development. Lieberam et al showed that the

gradient of SDF-1α directs the axonal guidance in embryonic development of the spinal cord. The

growth cones are guided towards their final peripheral targets based on the varying expression of

SDF-1 along the axon path. Specifically, initial axonal trajectory of motor neurons towards their

peripheral target in the spinal cord is regulated by SDF-1α. The absence of SDF-1α signal was

reported to drive the axon dorsally as opposed to ventrally [58]. In addition to guiding the axons,

SDF-1α modulates the concentrations of known growth cone repellants such as semaphorins in

the growth cones of target axons of neurons [59]. These studies suggest that SDF-1α controls

numerous facets of CNS development. Furthermore, there is a growing evidence that SDF-1α is

found to play important role in axonal pathfinding, outgrowth and branching. In addition, SDF-1α

seem to be involved in promoting nerve regeneration by regulating synaptic function and

influencing the release of cytokines for tissue repair after CNS injury [60, 61]. In the context of our

work, the primary role of SDF-1α in influencing the migration, proliferation and differentiation of

neural stem progenitor cells (NSCs) found in the ependymal region around the central canal of

the spinal cord was of utmost promise. Validating its role, CXCR4 is strongly expressed in

ependymal layer of the central canal in spinal cord post SCI [62].

Page 68: An Injectable Gelatin-Based Conjugate Incorporating EGF

68

3.3.2. Prior relevant in vitro studies with SDF-1α: Various in vitro investigations have

shown the promise of SDF-1α to recruit interneurons, NSCs and other glial cells. Several studies

have shown the chemotactic response of interneurons towards SDF-1α in a chain migration model

[63, 64]. Specifically, Lysko et al performed an insightful in vitro mouse explant system experiment

studying the migration, speed of migration and branching of the interneurons when they were

exposed to SDF-1α. They were able to show that decreasing the SDF-1α levels led to slower

migration rate and decrease in the number of migratory interneurons. Moreover, decreasing SD-

1α also led to profuse branching of interneurons suggesting that the directional guidance was

affected as well. Moreover, the results also showed that SDF-1α is necessary for faster and proper

migration of interneurons [65]. In addition to interneurons, glial cells such as oligodendrocyte

precursor cells (OPCs) and astroglia are also responsive to the SDF-1α. Kadi et al investigated

the effects of various growth factors in the migration regulation of OPCs 48 hours after being

exposed to the mitogens. They used OPC-like cell line known as Oli-neu. A proliferation

developed by the authors showed that SDF-1α induced proliferation and migration of Oli-neu cells

compared to the untreated controls. Interestingly, the concentration of SDF-1α had significant

effect on the proliferation of these cells. As a validating method, they also blocked the CXCR4

receptors in these cells and exposed these cells to SDF-1α. This modification reduced the

proliferation rate of the cells indicating the role of SDF-1α/CXCR4 mediated chemoattraction of

OPCs. In addition to proliferation, SDF-1α also increased the levels of myelin basic protein

expression suggesting that it promotes myelin sheath formation in vitro [66]. Odemis et al

investigated the role of SDF-1α in the chemotaxis of cortical astroglia. The authors reported

significantly higher migration index using SDF-1α mediated by IL-6 and CAMP. The authors

demonstrated that IL-6 and cAMP increased the CXCR4 expression on the cell surfaces of cortical

astroglia making them more responsive to SDF-1α. This could be a possible mechanism by which

IL-6 and cAMP induces SDF-1α-dependent chemotaxis of the astrocytes after injury in the

nervous tissue [67]. In the context of our work, we aim was to recruit endogenous neural stem

cells to the spinal cord injury site to promote repair and regeneration post SCI. Several in vitro

studies have shown strong mitogenic and chemoattractive role of SDF-1α towards NSCs [60, 68,

69].

A recent study by Stewart et al engineered the mesenchymal stem cells (MSCs) to

overexpress SDF-1α. In a transwell migration assay, these SDF-1α-expressing MSCs were

seeded in the migration chamber. The results showed significantly higher migration of the neural

stem cells into the well seeded with SDF-1α-expressing MSCs compared to the normal MSCs-

seeded chamber. This study indicates the chemotactic response of NSCs towards SDF-1α [70].

Page 69: An Injectable Gelatin-Based Conjugate Incorporating EGF

69

These results were supported by the findings in our laboratory by Lim, TC. In a 3D migration

assay, cell-free SDF-1α incorporated Gtn-HPA gel was placed in the core surrounded by aNSCs

seeded in the type I collagen gel annulus. After 7 days of culture and migration, Gtn-HPA/SDF-

1α cores showed increased accumulation of the NSCs at the interface between the Gtn-HPA and

type I collagen construct. Moreover, administration of AMD3100, a well-known antagonist of the

CXCR4, in the medium showed reduction in the number of cells accumulated at the interface.

Both individual cell migration and chain migration was observed in the assay. In addition, SDF-

1α also induced increased migration distance of the neural stem cells into the Gtn-HPA core. The

mitogenic role of SDF-1α was also confirmed with greater proliferation of aNSCs observed in the

Gtn-HPA/SDF-1α core [20, 71]. Considering the multifunctional role of SDF-1α in attraction,

proliferation and differentiation of various cells such as OPCs, NSCs and other glial cells that are

important for neural tissue repair, we were encouraged to evaluate these characteristics of SDF-

1α in vivo in our spinal cord injury model. Our proposition is that sustained localized delivery of

SDF-1α will enhance the recruitment of the precursor cells by EGF to the injury site repopulating

the injury site with reparative cells that can differentiate into neuronal and glial fate and can

stimulate the release of pro-regenerative molecules.

3.3.3. Prior in vivo studies incorporating SDF-1 α in SCI models: SDF-1α is expressed

in the adult spinal cord predominantly in dorsal corticospinal tract and the meninges. Moreover,

its receptor, CXCR4, has high levels of expression in the ependymal layers of the central canal

probably in the neural stem progenitor cells niche [62]. There is strong evidence that the levels of

SDF-1α and its receptor, CXCR4, drastically increase as early as 2 days after inducing a traumatic

spinal cord injury in rats. These upregulated levels can also be detected 42 days post SCI [72].

Similar upregulation of SDF-1α is also observed after stroke injury and ischemic injuries of the

central nervous system pointing to the ubiquitous role of SDF-1α in the repair and potential

regenerative response in the CNS [73]. After spinal cord injury, several cells such as reactive

astrocytes, macrophages, endothelial progenitors, oligodendrocyte precursor cells and neural

progenitor cells exhibit increased expression of SDF-1α and its receptor CXCR4 [68]. Given the

pervasive role of SDF-1α, a wide range of pre-clinical studies have investigated the role of SDF-

1α in promoting repair and functional recovery after spinal cord injury. Jaerve et al infused SDF-

α in the dorsal hemi-resection injury site via a intrathecal catheter placed in the subarachnoid

space. The catheter was connected to a osmotic minipump that was loaded with 10µM SDF-1α

ensuring sustained local release. Infusion took place over 7 days and then the catheter was

removed. Five weeks after the injury, they showed rostral axonal regrowth and reduced scarring

due to the infusion of SDF-1α. They employed both young and geriatric rats to show that the

Page 70: An Injectable Gelatin-Based Conjugate Incorporating EGF

70

axonal regeneration capacity is not hampered in geriatric rats suggesting that although the axonal

sprouting might be diminished in geriatric rats, the potential of regeneration in CST fibers was

comparable between the young and the geriatric rats [74]. Stewart et al utilized another approach

to evaluate the effect of SDF-1α in a spinal cord injury model. They transplanted engineered SDF-

1α-overexpressing MSCs into the spinal cord injury created in a contusion model. The result

showed significantly greater functional recovery assessed using the BBB scale in SDF-1α/MSCs

group compared to the untreated controls. Moreover, intervention of SDF-1α/MSCs reduced the

astrogliosis at the border outline the defect suggesting a scar-inhibitory role of SDF-1α. Moreover,

the injured area was also reduced in SDF-1α/MSCs groups suggesting a neuroprotective effect

of SDF-1α after spinal cord injury. Probably as a result of neuroprotection of unaffected tissue,

greater white matter sparing was observed in the SDF-1α/MSCs group a mm away from the injury

site. Interestingly, SDF-1α also promoted greater presence of GAP43-positive axons in the injury

site – marker for axonal cone growth. These collective results direct potential role of SDF-1α in

neuroprotection, functional recovery and axonal regeneration after spinal cord injury [70].

The role of SDF-1α was also reported in another great examination by Zendedel et al. In

the study, SDF-1α was intrathecally administered via an osmotic pump for 14 days. After 28 days,

rats with SDF-1α concentration of 500-1000 ng/ml showed significantly higher behavioral scores,

decreased numbers of apoptotic cells, increased astroglia and microglia responses, and induced

angiogenesis by intrathecal injection of SDF-1α in SCI contusion model suggesting the

recruitment of various vital cells of the spinal cord tissue repopulating the injury site [75].

Moreover, encouraging in vivo studies in our laboratory have shown increased presence of the

neural stem cells in the stroke lesion in the groups where SDF-1α was locally delivered using Gtn-

HPA matrix in an intracerebral hemorrhagic model of stroke. Significantly high DCX+ cells were

observed around the Gtn-HPA gel. In the same evaluation, Gtn-HPA/ SDF-1α promoted

neovascularization as well as neutrophilic response [20]. Given the prominent role of SDF-1α

reported in promoting repair in spinal cord injury and central nervous system, we propose to utilize

the synergistic effects of EGF and SDF-1α to attract endogenous stem cells and glial cells in

hemi-resection model of SCI.

Page 71: An Injectable Gelatin-Based Conjugate Incorporating EGF

71

3.4. Two formulations of Gtn-HPA for injection in the SCI models: Based on the rationale

provided for Gtn-HPA, EGF and SDF-1α in this chapter, our goal was to evaluate the synergistic

effects of EGF and SDF-1α locally delivered by an injectable Gtn-HPA that is capable of

undergoing covalent crosslinking in situ after being injected as a liquid in the cavity created by

surgical excision rat model of SCI after four weeks. Based on the positive results from the prior

work in the intracerebral hemorrhagic stroke model of stroke injecting Gtn-HPA with SDF-1α

encapsulated in polyelectrolyte complex nanospheres (PCN), we chose 8% Gtn-HPA gel to be

injected into the 2-mm complete and hemi-resection models of spinal cord injury. Since a

preliminary study showed significant inflammatory response around the Gtn-HPA probably due to

the PCNs, we wished to evaluate the effect of incorporating soluble growth factors EGF and SDF-

1α without any delivery vehicle.

The first formulation was as follows: 20 µl of 8% Gtn-HPA + 6 µg/rat EGF + 6 µg/rat SDF-

1α was injected in the 2-mm complete resection SCI model. The methods and the results are

described in Chapter 4 of this thesis. However, we did not see the persistence of Gtn-HPA in all

the animals after four weeks. Several reasons were considered for the disappearance of the Gtn-

HPA gel from the injury site: 1) the degradation rate of 8% Gtn-HPA is high leading to complete

disappearance by four weeks, 2) the Gtn-HPA gel was dislodged from the injury site due to

mechanical forces experienced by the Gtn-HPA gel during the movements post-operation, 3)

significant migration of the cells around the Gtn-HPA gel led to more secretion of degrading

enzymes such as MMP2, and 4) Injection of Gtn-HPA into a bleeding cavity would lead to

disruption of gelation procedure of the Gtn-HPA and therefore, the stiffness of resultant Gtn-HPA

is not as anticipated. Since we were able to see substantial Gtn-HPA in the groups with Gtn-HPA

incorporating both EGF and SDF-1α, the third reason was unlikely. Therefore, we designed a

second formulation of the Gtn-HPA gel to address these concerns. The gelatin content was

increased to 12% to reduce the degradation rate of the Gtn-HPA gel.

The second formulation was designed as follows: 20 µl of 12% Gtn-HPA + 6 µg/rat EGF

+ 6 µg/rat SDF-1α was injected in the 2-mm complete resection SCI model. The dosage of the

growth actors was kept the same based on the evaluation in both the models. The second

formulation along with an improved dural replacement technique (to better contain the Gtn-HPA

gel in the cavity) as described in Chapter 6 was evaluated in the 2-mm hemi-resection model. As

a result, all the animals in the groups showed presence of Gtn-HPA gel in the injury site. In an

attempt to characterize the release of growth factor from these Gtn-HPA formulations, release

profile of EGF from 8% and 12% Gtn-HPA was examined.

Page 72: An Injectable Gelatin-Based Conjugate Incorporating EGF

72

3.5. Release profile of EGF from Gtn-HPA formulations for four weeks

3.5.1. Methods: Given the encouraging results from prior work with EGF delivery in

intracerebral hemorrhage stroke model, studying the release of EGF from Gtn-HPA was

important. Four experimental groups were studied for the release of EGF with n = 4 for each

group: 1) 8% Gtn-HPA only, 2) 12% Gtn-HPA only, 3) 8% Gtn-HPA/EGF, and 4) 12% Gtn-

HPA/EGF. For the EGF groups, 6 µg of EGF was encapsulated in both 8% and 12% Gtn-HPA

formulations. The concentrations of each constituent are shown in Table 3.2.

Formulation Gtn-HPA HRP H2O2 EGF

8% Gtn-HPA 80 mg/ml 0.1 U/ml 6.8 mM 0

8% Gtn-HPA/EGF 80 mg/ml 0.1 U/ml 6.8 mM 6 µg

12% Gtn-HPA 120 mg/ml 0.1 U/ml 10.2 mM 0

12% Gtn-HPA/EGF 120 mg/ml 0.1 U/ml 10.2 mM 6 µg

Table 3.3 Experimental groups for EGF release study with their concentration constituents

After fabrication of the Gtn-HPA gel in 2 ml cryotubes, gels were submerged in 1.5 ml PBS release

buffer. 0.75 ml of supernatant was collected at the following time points to study the release of

EGF: 1 hr, 1d, 2d, 3d, 5d, 7d, 9d, 12d, 15d, 18d, 22d and 28d followed by addition of 0.75 ml PBS.

four-week endpoint was chosen because of the four-week evaluation timepoint in the in vivo

evaluation. After four weeks, Gtn-HPA gel was degraded using Collagenase IV solution for 3

hours at 37 °C. The concentration of EGF at these time points was assessed using EGF ELISA

kit (R&D Systems). For unknown samples, 1:600 dilution was used for all the time points.

3.5.2. Results and discussion: The standard curve fit was calculated using 4-parameter

logistic regression curve. The concentration and cumulative EGF release amount over four weeks

were calculated based on the best fit equation using the 4-parameter standard curve for all four

groups. We observed that no EGF was released from the 8% and 12% Gtn-HPA groups as

expected. However, the results show that there was no statistical difference in the amount of EGF

released up to 9 days possibly due to the initial burst effect typically observed in the drug-delivery

systems (Figure 3.3). However, greater EGF was released from 8% Gtn-HPA after 9 days

compared to the 12% Gtn-HPA gel (Two-way ANOVA, p<0.05). However, the differences in the

amount of EGF was small probably due to the ability of small molecules such as EGF to diffuse

from the dense Gtn-HPA network. The EGF release curve indicates that Gtn-HPA is able to locally

deliver EGF in a sustained manner. Moreover, after four weeks, all EGF is not released for both

8% and 12% confirming the presence of EGF in the degraded gel after four weeks. Further in vivo

Page 73: An Injectable Gelatin-Based Conjugate Incorporating EGF

73

evaluation employing immunohistochemistry was performed to validate the presence of EGF after

four weeks.

Figure 3.3 Release profile of EGF from 8% and 12% Gtn-HPA gel over four-week duration to mimic the in vivo evaluation. Greater EGF was released from 8% Gtn-HPA after 9 days. However, the difference of EGF released was small between 8% and 12% groups after four weeks

Page 74: An Injectable Gelatin-Based Conjugate Incorporating EGF

74

3.6. References

1. Haggerty, A.E. and M. Oudega, Biomaterials for spinal cord repair. Neurosci Bull, 2013. 29(4): p. 445-59.

2. Krishna, V., et al., Biomaterial-based interventions for neuronal regeneration and functional recovery in rodent model of spinal cord injury: a systematic review. J Spinal Cord Med, 2013. 36(3): p. 174-90.

3. Petter-Puchner, A.H., et al., The long-term neurocompatibility of human fibrin sealant and equine collagen as biomatrices in experimental spinal cord injury. Exp Toxicol Pathol, 2007. 58(4): p. 237-45.

4. Macaya, D. and M. Spector, Injectable hydrogel materials for spinal cord regeneration: a review. Biomed Mater, 2012. 7(1): p. 012001.

5. Bae, J.W., et al., Horseradish peroxidase-catalysed in situ-forming hydrogels for tissue-engineering applications. Journal of Tissue Engineering and Regenerative Medicine, 2015. 9(11): p. 1225-1232.

6. Li, Y., J. Rodrigues, and H. Tomas, Injectable and biodegradable hydrogels: gelation, biodegradation and biomedical applications. Chem Soc Rev, 2012. 41(6): p. 2193-221.

7. Lim, T.C., et al., The effect of injectable gelatin-hydroxyphenylpropionic acid hydrogel matrices on the proliferation, migration, differentiation and oxidative stress resistance of adult neural stem cells. Biomaterials, 2012. 33(12): p. 3446-55.

8. Hwang, D.S., S.B. Sim, and H.J. Cha, Cell adhesion biomaterial based on mussel adhesive protein fused with RGD peptide. Biomaterials, 2007. 28(28): p. 4039-4046.

9. Hu, M., et al., Cell immobilization in gelatin-hydroxyphenylpropionic acid hydrogel fibers. Biomaterials, 2009. 30(21): p. 3523-31.

10. Wang, L.S., et al., Injectable biodegradable hydrogels with tunable mechanical properties for the stimulation of neurogenesic differentiation of human mesenchymal stem cells in 3D culture. Biomaterials, 2010. 31(6): p. 1148-57.

11. Straley, K.S., C.W. Foo, and S.C. Heilshorn, Biomaterial design strategies for the treatment of spinal cord injuries. J Neurotrauma, 2010. 27(1): p. 1-19.

12. Lim, T.C., et al., Chemotactic recruitment of adult neural progenitor cells into multifunctional hydrogels providing sustained SDF-1alpha release and compatible structural support. Faseb j, 2013. 27(3): p. 1023-33.

13. Wang, L.S., et al., The role of stiffness of gelatin-hydroxyphenylpropionic acid hydrogels formed by enzyme-mediated crosslinking on the differentiation of human mesenchymal stem cell. Biomaterials, 2010. 31(33): p. 8608-16.

14. Wang, L.S., J.E. Chung, and M. Kurisawa, Controlling fibroblast proliferation with dimensionality-specific response by stiffness of injectable gelatin hydrogels. J Biomater Sci Polym Ed, 2012. 23(14): p. 1793-806.

15. Guilak, F., et al., Control of stem cell fate by physical interactions with the extracellular matrix. Cell stem cell, 2009. 5(1): p. 17-26.

16. Wang, L.S., et al., Modulation of chondrocyte functions and stiffness-dependent cartilage repair using an injectable enzymatically crosslinked hydrogel with tunable mechanical properties. Biomaterials, 2014. 35(7): p. 2207-17.

17. Al-Abboodi, A., et al., In situ generation of tunable porosity gradients in hydrogel-based scaffolds for microfluidic cell culture. Adv Healthc Mater, 2014. 3(10): p. 1655-70.

18. Chun, Y.Y., et al., A Periosteum-Inspired 3D Hydrogel-Bioceramic Composite for Enhanced Bone Regeneration. Macromol Biosci, 2016. 16(2): p. 276-87.

Page 75: An Injectable Gelatin-Based Conjugate Incorporating EGF

75

19. Al-Abboodi, A., et al., Injectable 3D hydrogel scaffold with tailorable porosity post-implantation. Adv Healthc Mater, 2014. 3(5): p. 725-36.

20. Lim, T.C., Injectable Chemokine-Releasing Gelatin Matrices for Enhancing Endogenous Regenerative Responses in the Injured Rat Brain, in Health, Science and Technology. 2014, Massachusetts Institute of Technology: Cambridge, MA.

21. Spilker, M.H., et al., The Effects of Collagen-Based Implants on Early Healing of the Adult Rat Spinal Cord. Tissue Engineering, 1997. 3(3): p. 309-317.

22. Spilker, M.H., et al., The effects of tubulation on healing and scar formation after transection of the adult rat spinal cord. Restor Neurol Neurosci, 2001. 18(1): p. 23-38.

23. Cholas, R., H.P. Hsu, and M. Spector, Collagen scaffolds incorporating select therapeutic agents to facilitate a reparative response in a standardized hemiresection defect in the rat spinal cord. Tissue Eng Part A, 2012. 18(19-20): p. 2158-72.

24. Macaya, D.J., et al., Astrocyte infiltration into injectable collagen-based hydrogels containing FGF-2 to treat spinal cord injury. Biomaterials, 2013. 34(14): p. 3591-602.

25. Macaya, D., Biomaterials-Tissue Interaction of an Injectable Collagen-Genipin Gel in a Rodent Hemi-Resection Model of Spinal Cord Injury in Health, Science and Technology. 2013, Massachusetts Institute of Technology: Massachusetts Institute of Technology. p. 210.

26. Spector, M. and T.C. Lim, Injectable biomaterials: a perspective on the next wave of injectable therapeutics. Biomed Mater, 2016. 11(1): p. 014110.

27. Bystroňová, J., et al., Creating a 3D microenvironment for monocyte cultivation: ECM-mimicking hydrogels based on gelatine and hyaluronic acid derivatives. RSC Advances, 2018. 8(14): p. 7606-7614.

28. Wang, L.S., et al., Enzymatically cross-linked gelatin-phenol hydrogels with a broader stiffness range for osteogenic differentiation of human mesenchymal stem cells. Acta Biomater, 2012. 8(5): p. 1826-37.

29. Plata-Salaman, C.R., Epidermal growth factor and the nervous system. Peptides, 1991. 12(3): p. 653-63.

30. Birecree, E., L.E. King, and L.B. Nanney, Epidermal growth factor and its receptor in the developing human nervous system. Developmental Brain Research, 1991. 60(2): p. 145-154.

31. Carmen Estrada, A.V., Epidermal Growth Factor Receptor in the Adult Brain, in The Cell Cycle in the Central Nervous System, D. Janigro, Editor. 2006, Humana Press: Totowa, New Jersey. p. 265-280.

32. Mei, L. and W.-C. Xiong, Neuregulin 1 in neural development, synaptic plasticity and schizophrenia. Nature reviews. Neuroscience, 2008. 9(6): p. 437-452.

33. Nelson, A.D., M. Suzuki, and C.N. Svendsen, A high concentration of epidermal growth factor increases the growth and survival of neurogenic radial glial cells within human neurosphere cultures. Stem Cells, 2008. 26(2): p. 348-55.

34. Flames, N., et al., Short- and long-range attraction of cortical GABAergic interneurons by neuregulin-1. Neuron, 2004. 44(2): p. 251-61.

35. Garcia-Alonso, L., S. Romani, and F. Jimenez, The EGF and FGF receptors mediate neuroglian function to control growth cone decisions during sensory axon guidance in Drosophila. Neuron, 2000. 28(3): p. 741-52.

36. Teramoto, T., et al., EGF amplifies the replacement of parvalbumin-expressing striatal interneurons after ischemia. J Clin Invest, 2003. 111(8): p. 1125-32.

37. Gregg, C. and S. Weiss, Generation of functional radial glial cells by embryonic and adult forebrain neural stem cells. J Neurosci, 2003. 23(37): p. 11587-601.

38. Lillien, L. and H. Raphael, BMP and FGF regulate the development of EGF-responsive neural progenitor cells. Development, 2000. 127(22): p. 4993-5005.

Page 76: An Injectable Gelatin-Based Conjugate Incorporating EGF

76

39. Jimenez Hamann, M.C., C.H. Tator, and M.S. Shoichet, Injectable intrathecal delivery system for localized administration of EGF and FGF-2 to the injured rat spinal cord. Exp Neurol, 2005. 194(1): p. 106-19.

40. Alluin, O., et al., Examination of the combined effects of chondroitinase ABC, growth factors and locomotor training following compressive spinal cord injury on neuroanatomical plasticity and kinematics. PLoS One, 2014. 9(10): p. e111072.

41. Kojima, A. and C.H. Tator, Intrathecal administration of epidermal growth factor and fibroblast growth factor 2 promotes ependymal proliferation and functional recovery after spinal cord injury in adult rats. J Neurotrauma, 2002. 19(2): p. 223-38.

42. Hamilton, L.K., et al., Cellular organization of the central canal ependymal zone, a niche of latent neural stem cells in the adult mammalian spinal cord. Neuroscience, 2009. 164(3): p. 1044-56.

43. Weiss, S., et al., Multipotent CNS Stem Cells Are Present in the Adult Mammalian Spinal Cord and Ventricular Neuroaxis. The Journal of Neuroscience, 1996. 16(23): p. 7599-7609.

44. Ahn, Y.H., G. Lee, and S.K. Kang, Molecular insights of the injured lesions of rat spinal cords: Inflammation, apoptosis, and cell survival. Biochem Biophys Res Commun, 2006. 348(2): p. 560-70.

45. Kang, M.K. and S.K. Kang, Interleukin-6 induces proliferation in adult spinal cord-derived neural progenitors via the JAK2/STAT3 pathway with EGF-induced MAPK phosphorylation. Cell Prolif, 2008. 41(3): p. 377-92.

46. Lam, H.J., et al., In vitro regulation of neural differentiation and axon growth by growth factors and bioactive nanofibers. Tissue engineering. Part A, 2010. 16(8): p. 2641-2648.

47. Garcez, R.C., et al., Epidermal growth factor (EGF) promotes the in vitro differentiation of neural crest cells to neurons and melanocytes. Cell Mol Neurobiol, 2009. 29(8): p. 1087-91.

48. Sun, D., et al., The Effect of Epidermal Growth Factor in the Injured Brain after Trauma in Rats. Journal of Neurotrauma, 2010. 27(5): p. 923-938.

49. Zheng, B., et al., Epidermal growth factor attenuates blood-spinal cord barrier disruption via PI3K/Akt/Rac1 pathway after acute spinal cord injury. Journal of Cellular and Molecular Medicine, 2016. 20(6): p. 1062-1075.

50. Ozturk, A.M., et al., Epidermal growth factor regulates apoptosis and oxidative stress in a rat model of spinal cord injury. Injury, 2018. 49(6): p. 1038-1045.

51. Cai, J., et al., Analysis of FK506-mediated functional recovery and neuroprotection in a rat model of spinal cord injury indicates that EGF is modulated in astrocytes. Experimental and therapeutic medicine, 2018. 16(2): p. 501-510.

52. Martens, D.J., R.M. Seaberg, and D. Van Der Kooy, In vivo infusions of exogenous growth factors into the fourth ventricle of the adult mouse brain increase the proliferation of neural progenitors around the fourth ventricle and the central canal of the spinal cord. European Journal of Neuroscience, 2002. 16(6): p. 1045-1057.

53. Mithal, D.S., G. Banisadr, and R.J. Miller, CXCL12 signaling in the development of the nervous system. Journal of neuroimmune pharmacology : the official journal of the Society on NeuroImmune Pharmacology, 2012. 7(4): p. 820-834.

54. Yu, L., et al., Identification and expression of novel isoforms of human stromal cell-derived factor 1. Gene, 2006. 374: p. 174-9.

55. Stumm, R.K., et al., CXCR4 regulates interneuron migration in the developing neocortex. J Neurosci, 2003. 23(12): p. 5123-30.

56. Belmadani, A., et al., The chemokine stromal cell-derived factor-1 regulates the migration of sensory neuron progenitors. J Neurosci, 2005. 25(16): p. 3995-4003.

Page 77: An Injectable Gelatin-Based Conjugate Incorporating EGF

77

57. Tiveron, M.C., et al., Molecular interaction between projection neuron precursors and invading interneurons via stromal-derived factor 1 (CXCL12)/CXCR4 signaling in the cortical subventricular zone/intermediate zone. J Neurosci, 2006. 26(51): p. 13273-8.

58. Lieberam, I., et al., A Cxcl12-CXCR4 chemokine signaling pathway defines the initial trajectory of mammalian motor axons. Neuron, 2005. 47(5): p. 667-79.

59. Chalasani, S.H., et al., A chemokine, SDF-1, reduces the effectiveness of multiple axonal repellents and is required for normal axon pathfinding. J Neurosci, 2003. 23(4): p. 1360-71.

60. Cheng, X., et al., The Role of SDF-1/CXCR4/CXCR7 in Neuronal Regeneration after Cerebral Ischemia. Frontiers in neuroscience, 2017. 11: p. 590-590.

61. Lazarini, F., et al., Role of the alpha-chemokine stromal cell-derived factor (SDF-1) in the developing and mature central nervous system. Glia, 2003. 42(2): p. 139-48.

62. Tysseling, V.M., et al., SDF1 in the dorsal corticospinal tract promotes CXCR4+ cell migration after spinal cord injury. J Neuroinflammation, 2011. 8: p. 16.

63. Sanchez-Alcaniz, J.A., et al., Cxcr7 controls neuronal migration by regulating chemokine responsiveness. Neuron, 2011. 69(1): p. 77-90.

64. Lewellis, S.W. and H. Knaut, Attractive guidance: how the chemokine SDF1/CXCL12 guides different cells to different locations. Seminars in cell & developmental biology, 2012. 23(3): p. 333-340.

65. Lysko, D.E., M. Putt, and J.A. Golden, SDF1 regulates leading process branching and speed of migrating interneurons. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2011. 31(5): p. 1739-1745.

66. Kadi, L., et al., Differential effects of chemokines on oligodendrocyte precursor proliferation and myelin formation in vitro. J Neuroimmunol, 2006. 174(1-2): p. 133-46.

67. Odemis, V., et al., Interleukin-6 and cAMP induce stromal cell-derived factor-1 chemotaxis in astroglia by up-regulating CXCR4 cell surface expression. Implications for brain inflammation. J Biol Chem, 2002. 277(42): p. 39801-8.

68. Jaerve, A., J. Schira, and H.W. Müller, Concise Review: The Potential of Stromal Cell-Derived Factor 1 and Its Receptors to Promote Stem Cell Functions in Spinal Cord Repair. Stem Cells Translational Medicine, 2012. 1(10): p. 732-739.

69. Filippo, T.R.M., et al., CXCL12 N-terminal end is sufficient to induce chemotaxis and proliferation of neural stem/progenitor cells. Stem Cell Research, 2013. 11(2): p. 913-925.

70. Stewart, A.N., et al., SDF-1 overexpression by mesenchymal stem cells enhances GAP-43-positive axonal growth following spinal cord injury. Restor Neurol Neurosci, 2017. 35(4): p. 395-411.

71. Lim, T.C., et al., The effect of injectable gelatin-hydroxyphenylpropionic acid hydrogel matrices on the proliferation, migration, differentiation and oxidative stress resistance of adult neural stem cells. Biomaterials, 2012. 33(12): p. 3446-3455.

72. Knerlich-Lukoschus, F., et al., Chemokine expression in the white matter spinal cord precursor niche after force-defined spinal cord contusion injuries in adult rats. Glia, 2010. 58(8): p. 916-31.

73. Hill, W.D., et al., SDF-1 (CXCL12) is upregulated in the ischemic penumbra following stroke: association with bone marrow cell homing to injury. J Neuropathol Exp Neurol, 2004. 63(1): p. 84-96.

74. Jaerve, A., et al., Differential effect of aging on axon sprouting and regenerative growth in spinal cord injury. Experimental Neurology, 2011. 231(2): p. 284-294.

75. Zendedel, A., et al., Stromal cell-derived factor-1 alpha (SDF-1alpha) improves neural recovery after spinal cord contusion in rats. Brain Res, 2012. 1473: p. 214-26.

Page 78: An Injectable Gelatin-Based Conjugate Incorporating EGF

78

Chapter 4:

Investigating the effects of Gtn-HPA

matrix with EGF and SDF-1α in a T8 2

mm complete resection rat model of

spinal cord injury after four weeks

Page 79: An Injectable Gelatin-Based Conjugate Incorporating EGF

79

4.1. Introduction and clinical motivation

There are various animal species and models that are used to study functional recovery

and repair after spinal cord injury (SCI). Rat models are typically used to assess potential

treatments for spinal cord injury due to their closely described pathophysiology post SCI, easiness

for post-operative care, well established functional recovery analysis procedures, low cost and

low incidence of infections [1]. Well-established and validated animals models include contusion-

compression models, ischemic models, full or partial transection models and chemical models [2].

The specific SCI models are chosen based on the therapy being evaluated and the clinical

motivation for the evaluation. As reported in the introduction, there are various causes for the

spinal cord injury – vehicular accidents, falls, violence and sports activities. Violence-related

injuries account for as high as 14% of total spinal cord injuries in the USA [3]. The causes range

from gun wounds, knives, and Improvised Explosive devices (IEDs). Typically, violence-related

injuries result in complete or partial transection of the cord. These incidences are higher in the

veterans’ population. Out of 250,000 people living with spinal cord injury, 42,000 were veterans

as per the 2009 report of the Veterans Affairs (VA) Department – a staggering 17%. This makes

VA the single-largest network to provide quality care to the veterans suffering from spinal cord

injury across the United States. Therefore, it was critical for us to evaluate our proposed treatment

in the context of military applications – motivation for our work in this chapter [4, 5].

Given the incidence of spinal cord injury in veterans population and major cause of these

injuries being the gun wounds, knife laceration, and IEDs, we were motivated to evaluate the

effect of EGF and SDF-1α locally delivered by the biomaterial Gtn-HPA in promoting spinal cord

injury repair and functional recovery in a 2 mm T8 complete resection rat model. The formulation

of the Gtn-HPA gel and the dosage of the EGF and SDF-1α was chosen based on the in vitro

studies presented in Chapter 3. The evaluation of the performed after four weeks of inducing the

complete resection spinal cord injury and injecting the Gtn-HPA with appropriate growth factors.

The results presented in this chapter provided proof of principle for further testing and were

promising to deduce significant conclusions about the proposed therapy. The work gave a

valuable insight into the interaction of the proposed biomaterial and its interactions with the

surrounding tissue at the injury site. Learnings from this study were incorporated in designing a

second generation Gtn-HPA gel formulation and incorporating more assessment methods to

reliably evaluate the injury site with its population, which are discussed in Chapter 5 and Chapter

6.

Page 80: An Injectable Gelatin-Based Conjugate Incorporating EGF

80

4.2. Overall goal and hypotheses of this chapter

The overall goal of this chapter was to investigate the effect of epidermal growth factor

(EGF) and stromal cell-derived factor-1α (SDF-1α) locally delivered by injectable gelatin-

hydroxyphenyl propionic acid (Gtn-HPA) in promoting healing and functional recovery in a 2 mm

standardized T8 full-resection rat model of SCI.

Chapter-specific hypotheses:

1) Gtn-HPA gel is able to fill in the cavity created by the 2 mm resection during injection.

2) Gtn-HPA is biocompatible and does not elicit major inflammatory response.

3) 8% Gtn-HPA is present in the injury site after four weeks and is permissive of cellular migration

(e. g. stiffness would not be very high to impede cellular migration)

4) EGF can recruit endogenous neural stem cells (NSCs) to the injury site while SDF-1α can

recruit endogenous glial cells to the injury site.

5) Better functional recovery is observed because of the intervention of Gtn-HPA injection with or

without EGF and SDF-1α.

6) SDF-1α enhances the response in all assessment metric compared to EGF delivery.

Page 81: An Injectable Gelatin-Based Conjugate Incorporating EGF

81

4.3. Methods

4.3.1. Gtn-HPA gel fabrication: Raw material of gelatin-hydroxyphenylpropionic acid

conjugate was obtained from Dr. Motoichi Kurisawa at Agency of Science, Technology and

Research (A*STAR), which was prepared as per the protocol described previously [6]. Briefly,

Gtn–HPA conjugates were prepared by a general carbodiimide/active ester-mediated coupling

reaction in distilled water. 3.32 g of HPA was dissolved in 250 ml of mixture prepared with 3:2

ration of distilled water and N,N-dimethylformamide (DMF) respectively. To this, 3.20 g of N-

hydroxysuccinimide and 3.82 g of 1-ethyl-3-(3-dimethylaminopropyl)-carbo-diimide hydrochloride

were added. The reaction mixture was stirred at room temperature for 5 hours, and the pH of the

mixture was kept at 4.7. Then, 150 ml of 6.25 wt% Gtn aqueous solution was added to the reaction

mixture and continuously stirred overnight at room temperature at pH 4.7. Subsequently, the

solution was dialyzed against 100 mM sodium chloride solution for 2 days, 25% ethanol in distilled

water for 1 day and distilled water for 1 day, successively. The purified solution was lyophilized to

obtain the Gtn–HPA conjugate.

8 wt% Gtn-HPA gel was fabricated as per the procedure described in Chapter 3.

Specifically, 12% (w/v) Gtn-HPA conjugate solution was prepared fresh under sterile conditions

and kept on ice. Recombinant rat EGF (Peprotech, Cat# 400-25) and recombinant rat SDF-1α

solutions (Peprotech, Cat# 400-32A) were aliquoted as per the recommended concentration of

1.0 mg/ml and stored at -20°C. Depending upon the experimental group, EGF and SDF-1α were

added to the freshly prepared 12% Gtn-HPA conjugated solution at room temperature. The final

dosage of the EGF and SDF-1α was 6 µg/rat. To this mixture, 2 µl of Horseradish peroxidase

(HRP) (Wako Pure Chemical Industries, Japan) was added with final HRP concentration of 0.1

U/ml. Then, 2 µl of H2O2(Sigma Aldrich) was added to give final H2O2 concentration of 6.8 mM.

Addition of H2O2 initiated the cross-linking and gelation process. The solution was mixed

thoroughly for homogenous gelation and immediately injected into the cavity before the solution

turned into a hydrogel with an appropriate gelation time.

4.3.2. Animal surgical procedure and Gtn-HPA gel injection: Adult female Lewis rats

(Charles River Laboratories, Willington, MA) weighing 201-225 grams were chosen as the rat

species in this study based on the previous work in our laboratory. Animals were given at least

48 hours to acclimatize before the major survival surgery of inducing a spinal cord injury. Before

the surgery, animals were singly housed in a cage as per the housing requirements after the

spinal cord injury to minimize the pain. Survival surgery was performed as per the approval from

Veterans Affairs Boston Healthcare system Institutional Animal Care and Use Committee

Page 82: An Injectable Gelatin-Based Conjugate Incorporating EGF

82

(IACUC) on the protocol #378-J. Animal surgeries were performed at the Animal Research

Facility located in the research building of the Veterans Affairs Jamaica Plains, MA campus.

Animal surgery was performed by the skilled and experienced Dr. Hu-ping Hsu,

Department of Orthopedic surgery, Brigham and Women’s Hospital. The complete resection

model is based on the work done by Cholas et al [7]. After recording their weight immediately

before the surgery, Lewis rats were anesthetized using 2-4 % isoflurane in oxygen through a nose

mask for induction followed by 1-3 % isoflurane for maintenance of anesthesia, during the surgical

procedure. On average, the procedure lasted for 30-45 mins and isoflurane was administered for

the entire duration. After confirming that the rat is unconscious, rats were placed in the prone

position on a flat operating board and all the limbs were gently fixed using rubber bands to stabilize

the rat. The hair on the back of the rat were shaved and the skin was thoroughly cleaned with

betadine. A 4-cm skin incision was made on the thoracic spine portion. The skin and musculature

overlying the thoracic spine was incised along the midline and retracted laterally from the vertebral

column. A laminectomy was performed removing the dorsal aspect of T7-T10 using small bone

rongeurs and microscissors, exposing approximately 10 mm of spinal cord. Bleeding from the

muscle was controlled by Gelfoam® (Pfizer, Andover, MA). The dura will be incised in the midline

and the dorsal spinal artery will be coagulated with a bipolar cautery. A sterile 2 mm plastic

template was placed in the center of the exposed spinal cord and a lateral complete resection of

the spinal cord was created by making two lateral cuts (2 mm apart) using ultra fine surgical

scissors using the 2 mm template as guide (Figure 4.1). A 2 mm gap was created by removing

the tissue between the two lateral cuts. Bleeding was controlled using Gelfoam® placed into the

defect site. Once hemostasis was achieved, the Gelfoam® was removed from the defect and Gtn-

HPA gel was immediately injected that was freshly fabricated immediately before the injection. In

control animals [N], no treatment was injected into the defect. However, 8 wt% Gtn-HPA gel was

injected in the experimental groups as per the dosage mentioned above. In 8wt% Gtn-HPA gel

only [G8] group, Gtn-HPA gel was fabricated and injected without addition of any factors.

However, appropriate growth factors were added to the Gtn-HPA conjugate solution prior to the

gelation for local delivery of EGF and SDF-1α [GE8 and GES8]. In all the groups that involved

injecting the Gtn-HPA gel into the cavity, 20 µl of the 8 wt% Gtn-HPA gel, with or without the

factors were injected into the cavity using a pipette such that gelation will occur rapidly within in a

few minutes to confine the material to the lesion site. The injury was closed after visual

confirmation of complete gelation and filling of the complete 2 mm defect with the Gtn-HPA gel.

Following gel injection, a thin collagen membrane was placed over the top of the spinal cord

wound as dural replacement, to maintain positioning of the scaffold and separate the wound site

Page 83: An Injectable Gelatin-Based Conjugate Incorporating EGF

83

from the overlying tissue. Following treatment and placement of the collagen membrane, the

overlying musculature was closed using 4-0 vicryl sutures (Johnson and Johnson, Sommerville,

NJ) and the skin was closed with wound clips.

4.3.3. Experimental design: There were three experimental groups evaluated in this study

with the fourth group being the control group that received no injection treatment. The animals

were randomly assigned to the groups and were blinded to the surgeon during injection. I was

blinded to the treatment group and animals were identified using their serial number only. The

following table shows the groups and the number of animals in each group.

Experimental treatment Group Duration of

evaluation

Study size

(n)

Control – no injection N 4 weeks 5

8 wt% Gtn-HPA gel only G8 4 weeks 6

8wt% Gtn-HPA gel with EGF GE8 4 weeks 3

8wt% Gtn-HPA gel with EGF and SDF-1α GES8 4 weeks 5

Table 4.1. Experimental Design for the study to evaluate the Gtn-HPA along with EGF and SDF-1a in 2

mm complete section rat model

Figure 4.1. (left) 2 mm complete resection cavity created by laminectomy at T8 level shown with the 2 mm template after achieving hemostasis (right) Cavity injected with the Gtn-HPA gel and covered with collagen membrane as dural replacement

Page 84: An Injectable Gelatin-Based Conjugate Incorporating EGF

84

4.3.4. Animal care post-surgery: Post-surgery, animals were carefully brought to the pre-

OR room and placed in a heated cage with constant supply of oxygen at 1 L/min to maintain body

temperature and breathing. Rats were subcutaneously injected with warm 3 ml of lactated

Ringer’s solution to compensate for the blood loss during the surgery. They were subcutaneously

administered Meloxicam (1 mg/kg) as an analgesic and Cephazolin (35 mg/kg) as an antibiotic to

manage pain and infections, respectively. Rats were monitored for breathing every ten minutes

for an hour after surgery, every half an hour subsequent two hours and every hour for the last

three hours. Continuous oxygen supply was maintained until they were transported to their

housing room in the animal research facility. 6 hours post-surgery, their bladders were expressed

manually and placed in the cage with wood-chip bedding with free access to food and water. Dry

food was kept on the flood of the cage in the initial days for easier access. Rats were kept in a

light/dark automated cycle in the housing room as per the research facility guidelines. Due to the

loss of bladder control, their bladders were manually emptied using Crede’s technique twice a

day with utmost care to minimize the pain. Bladder volume was recorded in three categories –

small, medium and large during manual expression. Rats were subcutaneously administered

analgesic Meloxicam (1 mg/kg) once every 22-24 hours for four days post-surgery while antibiotic

Cephazolin (35 mg/kg) was administered twice in a day for a week post-surgery. They were also

administered lactated Ringer’s solution as necessary based on their hydration status. One week

after the surgery, suture clips were removed after confirming suturing of the skin. Animals were

anesthetized using 2-3% isoflurane for painless removal of the wound clips. They were weighed

once a week and their record was meticulously kept in the housing room as per the VA animal

research facility requirements.

4.3.5. Behavioral assessment – BBB scale: In order to assess the functional recovery of

the animals after the surgery, open-field locomotor test was conducted once every week until

sacrifice. Animals walked freely on the table with absorbent chute. Their movement, especially of

the hindlimbs, was digitally video recorded for three minutes. The animals were scored using the

well-established 21-point Basso-Beattie-Bresnahan (BBB) locomotor rating scale [8]. The scale

ranged from 0 (complete paralysis) to 21 (normal gait) as listed in Table 4.2. The evaluator was

blinded to the treatment group during scoring of the both hindlimbs. Separate BBB score was

assigned to the left and the right hindlimb each week.

Page 85: An Injectable Gelatin-Based Conjugate Incorporating EGF

85

BBB score Description

0 No observable hindlimb movement

1 Slight movement (<50% of the range of motion) of one or two of the three joints – hip, knee and ankle

2 Extensive movement (>50% of range of motion) of one joint and slight movement of another joint

3 Extensive movement of the two joints

4 Slight movement of all three hindlimb joints

5 Slight movement of the two joints and extensive movement of the third joint

6 Extensive movement of two joints and slight movement of the third joint

7 Extensive movement of all three hindlimb joints

8 Sweeping movement (rhythmic movement of all three hindlimb joints) without any weight bearing (elevation of the hindlimb during movement) by the hindlimbs

9 Plantar paw placement with some weight bearing in stationary phase or minor weight bearing with dorsal stepping of the paw

10 Occasional (<50% of the times) weight bearing with plantar placement of the paw without any hindlimb-forelimb (HL-FL) coordination

11 Occasional weight bearing with plantar placement of the paw and occasional FL-HL coordination

12 Frequent (>50% of the times) weight wearing with plantar placement of the paw with occasional FL-HL coordination

13 Frequent weight bearing with the plantar placement of the paw and frequent Hl-FL coordination

14 Consistent (~100%) weight-bearing with plantar steps, consistent FL-HL coordination with rotational movement of the paw during stepping

15 Consistent FL-HL coordination with consistent dragging of the toes and rotational movement of the paw during stepping

16 Consistent FL-HL coordination with frequent dragging of the toes and rotational movement of the paw during stepping

17 Consistent FL-HL coordination with occasional dragging of the toes and rotational movement of the paw during stepping

18 Consistent FL-HL coordination with no dragging of the toes with parallel paw position during liftoff and initial contact

19 Consistent FL-HL coordination, no dragging of the toes, parallel paw position during liftoff and initial contact and tail is down part or all the time

20 Consistent FL-HL coordination, no dragging of the toes, parallel paw position during liftoff and initial contact and tail is consistently up

21 Coordinated gait, consistent toe clearance, predominant paw position is parallel during movement and tail consistent up with full trunk stability

Table 4.2 BBB scale implemented to assess the functional recovery for the rats after the spinal cord injury with the BBB score and corresponding functional description of the hindlimb

Adapted from the Basso et al, J Neurotrauma 1995. 12(1): p. 1-21.

4.3.6. Animal sacrifice and harvesting of the spinal cord: After four weeks, animals were

sacrificed by transcardial perfusion. Rats were administered a dose of 150 mg/kg of sodium

pentobarbital and anesthetized. The level of anesthesia was checked by pinching their tail.

Quickly, a thoracotomy was performed to expose the heart and the needle attached to the

peristaltic pump was inserted into the left ventricle and into the ascending aorta. The pump was

turned on at the speed of four initially and then the right atrium was quickly cut to allow open

Page 86: An Injectable Gelatin-Based Conjugate Incorporating EGF

86

system for the flow of perfusion solution. The speed of the pump was increased to 7. A total of

120 ml of heparinized saline (20 U/ml) was circulated through the animal, followed by 120 ml of

4% paraformaldehyde (PFA) without introducing any bubbles into the line. Both the solutions were

kept on ice during the perfusion. After completing the sacrifice, the spinal column along with the

musculature was cut from the neck region to pelvic region. The entire spine was placed in 50 ml

4% PFA at 4°C overnight. The following day, the spine was trimmed of the overlying musculature

while leaving the vertebral bone and fascia over the defect intact. The trimmed spine was kept in

4% PFA for additional 2 days at 4°C. After three days, they were kept in 60 ml of PBS for three

days at 4°C. Then, the spinal cord was carefully isolated and harvested from the spinal column

using bone rongeurs, surgical scissors, and a scalpel.

4.3.7. Tissue embedding procedure and cryosection: Immediately after harvesting the

cord from the spine column, the spinal cord with the injury site was embedded in Tissue Tek OCT

compound solution kept in a rectangular tinfoil mold (Figure 4.2). The spinal cord was then flash-

frozen in the OCT compound solution using isopentane cooled in liquid nitrogen. The frozen block

with proper labeling with animal information was kept on dry ice until they were stored at -80°C.

The blocks were serially cryosectioned using Leica Cryostat CM 3050 (Leica Biosystems,

Wetzelar, Germany). 30 µm thick coronal sections of the spinal cord were cut and were mounted

on a pre-treated Superfrost Gold glass microscope slides (Fisher Scientific, Hampton, NH).

4.3.8. Histology and histomorphometric analysis: Tissue sections were warmed to room

temperature and then baked at 60 °C for 3 hours prior to histological staining to improve the

adhesion to the glass slides. The tissue sections roughly in the middle of the coronal plane were

chosen to best represent the tissue morphology for each rat. Then, after they reached room

temperature, tissue sections were rehydrated in PBS for 45 mins to dissolve the OCT on the

slides. Tissue sections were stained with Hematoxylin and Eosin (H&E) method to visualize the

Figure 4.2 Embedding the spinal cord (left) gross image of the Spinal cord with the injury site carefully harvested from the spinal column (middle) spinal cord kept in Tissue Tek OCT compund prior to flash-freezing in isopentane cooled in liquid nitrogen (right) Frozen block with spinal cord mounted on the object holder in the cryostat for coronal cryosectioning

Page 87: An Injectable Gelatin-Based Conjugate Incorporating EGF

87

tissue morphology and the implanted gel at the injury site. After being rehydrated in PBS, the

tissue sections were permeabilized in cold methanol for 10 min followed by immersion in water

for 5 min, rinsed in PBS for 10 min and again immersed in water for 5 min. Then, they were stained

with hematoxylin Gill #2 solution for 4 min followed by thorough wash in running tap water for 5

min. Then, the tissue sections were decolorized with 3 quick drops in acid alcohol and then were

thoroughly washed in running tap water for 5 min. Furthermore, they were stained with eosin for

40 seconds. Tissue sections were dehydrated with 100% alcohol (2 x 3 min) and processed in

xylene (2 x 3 min). Coverslip was mounted on using cytoseal and then the tissue sections were

air dried in a chemical hood overnight.

The brightfield microscopy images were acquired using Olympus microscope. In addition,

Zeiss automated slide scanner was used to scan the entire tissue sections (Zeiss, Germany). For

each animal, gel area and cellular migration area immediately next to the gel islands were

calculated using freehand option in ImageJ (NIH, Bethesda, MD). Two analysis methods were

used to validate the findings. The first method (Method 1 – total area method) measured the total

Gtn-HPA gel area in all the gel islands and total cellular migration immediately around all the gel

islands in the injury cavity were measured in ImageJ using ROI manager tool. The second method

(Method 2 – standardized ROI method) measured the Gtn-HPA gel area and cellular migration

immediately surrounding the gel in three standardized 0.3 mm2 ROI at the interface using ImageJ

in three randomly chosen images. Average gel area and migration area from each animal was

calculated using both the methods and compared among the groups. All histological analyses

were conducted blindly with respect to the treatment group.

4.3.9. Immunofluorescent staining and analysis: Tissue sections were stained with a

biochemical marker to obtain more specific details regarding the cellular population in the injury

site and immediately surrounding the gel islands. Tissue sections were warmed to room

temperature and then baked at 60 °C for 3 hours prior to immunofluorescent staining to improve

the adhesion to the glass slides. Primary antibodies, dilutions and their antigen retrieval were

performed as shown in Table 4.3. Then, after they reached room temperature, tissue sections

were rehydrated in PBS for 45 mins to dissolve the OCT on the slides. Subsequently, antigen-

retrieval was performed for the tissue sections in low pH citrate buffer (Vector laboratories, CA,

USA) at 97 °C for 20 min. Then, tissue sections were cooled to room temperature in the antigen

retrieval solution and then immersed in 0.3% Triton-X in PBS for 30 min. Next, they were washed

with quick dips in PBS and then permeabilized in cold methanol for 10 min followed by PBS wash

for 5 min. Then, the slides were cleaned and marked with an IHC marker around the section to

contain the staining solutions to the sections. The tissue sections were blocked with serum-free

Page 88: An Injectable Gelatin-Based Conjugate Incorporating EGF

88

protein block (Dako, Agilent, USA) for 1 hr. Then, they were washed with 0.05% Tween20 in PBS

solution (5 min) and then in PBS (2 x 5 min). Then, they primary antibodies diluted in Dako

antibody diluent solution were added to the sections in a black staining box with ample moisture.

The slides were then kept overnight at 4 °C. Next day, slides were washed with 0.05% Tween20

in PBS solution (5 min) and then in PBS (2 x 10 min) followed by addition of secondary antibodies

diluted in Dako antibody diluent. The slides were incubated with secondary antibodies for 2 hours

at room temperature in the black staining box. Subsequently, they were washed with 0.05%

Tween20 in PBS solution (5 min) and then in PBS (2 x 10 min). After 3 quick dips in deionized

water, coverslips were mounted with DAPI-containing Fluoro Gel-II mounting medium (Electron

Microscopy Sciences, Hatfield, PA) and stored at 4 °C.

Fluorescent images were acquired using epifluorescence microscope (Olympus BX60)

and confocal laser microscope (Nikon). The images were analyzed using ImageJ to quantify the

signal from the biomarker of interest. Three random images were selected and analyzed with a

standardized 0.3 mm2 ROI. Images were converted to 8-bit images followed by a standard

threshold optimized for each antibody. Then, the percent area covered by the fluorescent signal

in the standardized 0.3 mm2 ROI was measured using ImageJ. The average percent area was

calculated and used for analysis for comparison among the treatment groups.

Primary antibody Dilution Antigen retrieval

Secondary antibody Cells/molecules identified

Rabbit poly clonal anti-GFAP (abcam #ab7260)

1:400 Citrate buffer pH 6 (97 °C at 20 min)

Dylight 488 Donkey anti-rabbit IgG (Jackson Immunoresearch)

Astrocytes

Rabbit monoclonal anti-NeuN (abcam #ab177487)

1:400 Citrate buffer pH 6 (97 °C at 20 min)

Cy3 Red Donkey anti-rabbit IgG (Jackson Immunoresearch)

Mature neurons

Rabbit polyclonal anti-Iba1 (Wako #019-19741)

1:400 Citrate buffer pH 6 (97 °C at 20 min)

Dylight 488 Donkey anti-rabbit IgG (Jackson Immunoresearch)

Microglia/ macrophages

Mouse monoclonal anti-Nestin (Millipore #MAB353)

1:200 Citrate buffer pH 6 (97 °C at 20 min)

Cy3 Red Donkey anti-mouse IgG (Jackson Immunoresearch)

Neural stem cells

Rabbit polyclonal anti-EGF (Peprotech #500-P277)

1:500 Citrate buffer pH 6 (97 °C at 20 min)

Dylight 488 Donkey anti-rabbit IgG (Jackson Immunoresearch)

Rat EGF

Mouse monoclonal anti-MMP2 (Millipore #MAB3308)

1:300 Citrate buffer pH 6 (97 °C at 20 min)

Cy3 Red Donkey anti-mouse IgG (Jackson Immunoresearch)

Gelatinase MMP2

Table 4.3 Immunofluorescent staining of the cells and molecules with their staining protocol

Page 89: An Injectable Gelatin-Based Conjugate Incorporating EGF

89

4.3.10. Statistical analysis: Before the experiment, our power calculation for sample size

determination is based on the desire to determine significant a 30% difference in a selected

outcome variable with a 15% standard deviation, and with α=0.05 and β=0.05. The statistical

significance among the experimental groups for histomorphometric and IHC staining results were

determined by one-way ANOVA analysis using Tukey Post-hoc tests. Non-parametric data such

as BBB scores were analyzed for statistical significance using the Krusal-Wallis Test, Fisher’s

Exact Test or Chi squared test. SPSS 25 (IBM) and Graphpad Prism 8.0 (Graphpad) were used

for the statistical analysis.

Page 90: An Injectable Gelatin-Based Conjugate Incorporating EGF

90

4.4. Results

4.4.1. Qualitative observations of the animals and Gtn-HPA injection: The animals

tolerated the surgery well and had no major complications immediately after the surgery. One

animal from the Gtn-HPA with EGF (GE8) group could not survive due to surgical complications.

Gtn-HPA was injected into the cavity after achieving hemostasis. However, achieving complete

hemostasis was difficult despite the use of Gelfoam® in 6 out of 19 animals. We observed some

bleeding from the cavity during the Gtn-HPA gel injection. Visually, the Gtn-HPA gel achieved

complete gelation in 2-3 minutes. Post-operation, animals gained consciousness in roughly half

an hour after the surgery and were completely paralyzed in their hindlimbs. They had lost the

ability to empty their bladder due to loss of micturition reflex indicated by their full bladders, which

were manually expressed twice a day. All animals showed good markers of recovery in terms of

weight gain and rehydration. No animals were suspected of any infections after surgery. By four

weeks, they had partial return of the function in their hindlimb movement and bladder function.

They were well acclimatized with the open field locomotor test to assess functional recovery.

4.4.2. Functional evaluation using BBB scale: In order to assess the functional recovery

after the spinal cord injury, animals underwent open field locomotor test in which they freely

walked on the table. Prior to surgery, all animals displayed normal gait. However, immediately

after the injury, they suffered from complete paralysis for both the right and left hindlimb. Even

after one week, most animals were scored to be zero (complete paralysis) on the BBB scale in

both hindlimbs. During the course of four weeks, animals showed modest improvement in both

left and right hindlimb (Figure 4.3). As can be seen in the Figure 4-3, the average BBB scores

were less than 7 for all the four groups. After four weeks, all four groups showed functional

improvement compared to the first week. However, the treatment groups containing the Gtn-HPA

gel with the growth factors EGF and SDF-1α did not show better functional recovery compared to

the control (Krusal-Wallis test). To account for the improvement after four weeks from the first

week, each rat was treated as its own control and the improvement in the BBB score was

calculated to assess whether the treatment groups showed better improvement in the functional

score. Although the trend suggests that the treatment groups show more improvement in the BBB

score compared to the control especially in the left hindlimb, there was no statistical significance

observed (Figure 4.4). In addition, the improvement in the BBB scores were categorized in two

types – poor improvement (∆BBB = 0-2) and good improvement (∆BBB = 2+). In order to evaluate

the effect of our intervention of injecting Gtn-HPA gel, the number of animals showing poor and

good improvement were tabulated for both right and left hindlimb after four weeks in form of a

contingency table (Table 4.4). Interestingly, all the rats showed poor improvement in the control

Page 91: An Injectable Gelatin-Based Conjugate Incorporating EGF

91

group while 10/14 rats showed good improvement after the injection of the gel in the cavity for the

left hindlimb. Fisher’s exact test analysis showed significant association between the degree of

improvement and the type of intervention in the left hindlimb indicating that injection of gel shows

better improvement in the BBB score compared to the control (** p<0.01).

Figure 4.3 Animals showed functional recovery after the spinal cord injury in both the right and left hindlimbs (Increase in the BBB score indicates functional recovery). However, treatment groups did not have better recovery compared to the control group after four weeks

Figure 4.4 Improvement in the BBB score was plotted in both the right and left hindlimbs after four weeks using each rat as its own control. Trend seem to indicate more improvement in the treatment groups with Gtn-HPA injection, but no statistical significance was observed. Mean ± SEM

Page 92: An Injectable Gelatin-Based Conjugate Incorporating EGF

92

Table 4.4 Contingency table showing the number of animals showing poor and good improvement in the BBB score after four weeks in the controls versus the treatment group. Fisher’s exact test showed that there is better improvement in function when gel is injected in the cavity in the left hindlimb recovery

4.4.3. Gross histological outcomes: After four weeks, histological images showed

important characteristics of the injury site and the response to the Gtn-HPA gel. Across all the

groups, complete resection of the unaffected tissue was observed in histology and the injury site

was either filled with the fibrous tissue or Gtn-HPA gel. In the control [N] group where no material

was injected in the cavity site, the cavity was covered with dense fibrous tissue with empty areas,

which were presumably filled with cystic fluid prior to the tissue processing (Figure 4.5A).

Moreover, ferritin deposits were seen in the injury site indicative of blood clots. The injury site and

the unaffected spinal cord tissue was visibly different in terms of the morphology (Figure 4.6A).

On both proximal and dorsal ends of the injury, semi-circular stumps were observed, which are

typically observed in response to the severed axons [9]. Normal histological features of white and

gray matter were observed in the unaffected tissue proximal and distal to the injury site. In the

Gtn-HPA only [G8] group, injury site was substantially filled with the Gtn-HPA gel as opposed to

the dense fibrous tissue observed in the control group images after four weeks (Figure 4.5B).

However, dense fibrous tissue was also observed in the areas where the Gtn-HPA gel was not

present. The typical porous fibrillar structure of the Gtn-HPA gel was observed with weakly eosin

stained Gtn-HPA matrix as reported previously [10]. Moreover, usually Gtn-HPA gel was observed

in ‘gel islands’ throughout the injury site and not as a one contiguous structure. Gtn-HPA gel was

not physically continuous with the surrounding tissue. Implanted Gtn-HPA matrix was mostly void

of any cellular infiltration. However, some cells were seen at the boundary of the gel island (Figure

Intervention in the cavity

Poor improvement (∆BBB = 0-2)

Good improvement (∆BBB = 2+)

LEFT

Control – no injection 5 0

Treatment – gel injection 4 10

RIGHT

Control – no injection 2 3

Treatment – gel injection 4 10

Page 93: An Injectable Gelatin-Based Conjugate Incorporating EGF

93

4.6B). The morphology of the Gtn-HPA with EGF [GE8] and Gtn-HPA with EGF + SDF-1α [GES8]

Figure 4.5 Representative H&E stained images of the four groups – (A) Control [N] (B) Gtn-HPA only [G8], (C) Gtn-HPA with EGF [GE8], and (D) Gtn-HPA with EGF and SDF-1α [GES8] showing Gtn-HPA islands and cellular migration around the Gtn-HPA islands in the EGF-delivered groups (i.e. GE and GES8). Scale bar = 1000 µm

gel

gel

gel

A

B

C

D

Gtn

-HP

A o

nly

[G8]

Co

ntr

ol

[N]

Gtn

-HP

A +

EG

F

[GE8

] G

tn-H

PA

+ E

GF

+ SD

F-1 α

[GES

8]

Page 94: An Injectable Gelatin-Based Conjugate Incorporating EGF

94

sections was very similar (Figure 4.5C-D). Like the Gtn-HPA gel only group, histological sections

of these group showed the substantial presence of the Gtn-HPA gel in the injury site in form of

gel islands. However, contrary to the Gtn-HPA gel only group, there was a well-defined physically

continuous interface that was established with the surrounding tissue in the injury site as shown

by high magnification micrographs. In addition, qualitatively, many endogenous cells infiltrated

the injury site and migrated immediately surrounding the gel islands (Figure 4.6C-D). This was

evident by a blue band as a result of hematoxylin stained nuclei of the cells immediately around

the gel islands. The migratory pattern of the cells did not seem to be unidirectional – cells migrated

all around the gel islands. Remarkably, there were few sections in which cells were captured

A: Control [N] B: Gtn-HPA only [G8]

C: Gtn-HPA + EGF [GE8] D: Gtn-HPA+EGF+SDF-1α [GES8]

Figure 4.6 Micrographs showing important observations of the four groups (A) Control - dense fibrous tissue (B) Gtn-HPA only - morphology of Gtn-HPA, 'gel island', minimal cellular migration around the island, poor interface (C) & (D) presence of the Gtn-HPA gel, greater cellular migration around the gel, well-defined interface with the surrounding tissue, and migration of cells through the gel. Scale bar = 100 µm

Page 95: An Injectable Gelatin-Based Conjugate Incorporating EGF

95

through the gel islands in a somewhat linear arrangement, which could give insights into the

formation of the gel islands and degradation of the gel (Figure 4.6C).

It is important to note that the Gtn-HPA gel was not observed in in the injury site in all the

gel-injected groups. Table 4.5 shows the study size and number of animals that showed the

presence of the gel in the injury site. There are several reasons for the absence of the gel in

certain animals including incomplete gelation due to bleeding during injection, displacement of

the Gtn-HPA gel from the cavity due to animal movement or due to poor containment of the gel

by the collagen fascia membrane and degradation of the gel by four weeks.

Experimental treatment Group Duration of

evaluation

Study

size (n)

# of animals showing

presence of the gel

Control – no injection N 4 weeks 5 N/A

8 wt% Gtn-HPA gel only G8 4 weeks 6 5

8wt% Gtn-HPA gel with EGF GE8 4 weeks 3 3

8wt% Gtn-HPA gel with EGF

and SDF-1α GES8 4 weeks 5 4

Table 4.5. Table comparing the number of animals exhibiting the presence of the Gtn-HPA gel in the

injury site after four weeks in H&E staining

4.4.4. Histomorphometric evaluation of the injury site: Based on the observation that

there was substantial presence of the gel and cellular migration immediately around the gel

islands, there were two methods employed to measure the Gtn-HPA gel area and cellular

migration area immediately surrounding the gel islands in the injury site. The histological sections

evaluated to measure the areas represented the midplane coronal section of the cord.

The first method (Method 1 – total area method) measured the total Gtn-HPA gel area in

all the gel islands and total cellular migration immediately around all the gel islands in the injury

site in ImageJ using ROI manager tool. The average Gtn-HPA gel area was measured to be 0.25

± 0.10 mm2 for the Gtn-HPA only group [G8], 0.32 ± 0.18 mm2 for the Gtn-HPA with EGF group

[GE8] and 0.65 ± 0.36 mm2 for the Gtn-HPA with EGF and SDF-1α group [GES8]. Presence of

the growth factors did not affect the area of the gel present in the injury site (One-Way ANOVA,

Tukey post-hoc test, p > 0.05) (Figure 4.7A). The average cellular migration area immediately

surrounding the gel islands was measured to be 0.07 ± 0.01 mm2 for the Gtn-HPA only group

[G8], 0.49 ± 0.33 mm2 for the Gtn-HPA with EGF group [GE8] and 0.69 ± 0.10 mm2 for the Gtn-

HPA with EGF and SDF-1α group [GES8]. Local delivery of the EGF and SDF-1α with the Gtn-

HPA matrix significantly increased the number of the migrated cells immediately surrounding the

Page 96: An Injectable Gelatin-Based Conjugate Incorporating EGF

96

gel islands compared to Gtn-HPA gel only group (One-Way ANOVA, Tukey post-hoc test, p <

0.05) (Figure 4.7B).

The second method (Method 2 – standardized ROI method) measured the Gtn-HPA gel

area and cellular migration immediately surrounding the gel in the standardized 0.3 mm2 ROI at

the interface using ImageJ in three randomly chosen images. The average Gtn-HPA gel area was

measured to be 0.062 ± 0.009 mm2 for the Gtn-HPA only group [G8], 0.062 ± 0.013 mm2 for the

Gtn-HPA with EGF group [GE8] and 0.053 ± 0.016 mm2 for the Gtn-HPA with EGF and SDF-1α

group [GES8]. Presence of the growth factors did not affect the area of the gel present in the

injury site (One-Way ANOVA, Tukey post-hoc test, p > 0.05) (Figure 4.8A). The average cellular

migration area immediately surrounding the gel islands was measured to be 0.039 ± 0.002 mm2

for the Gtn-HPA only group [G8], 0.112 ± 0.009 mm2 for the Gtn-HPA with EGF group [GE8] and

0.115 ± 0.013 mm2 for the Gtn-HPA with EGF and SDF-1α group [GES8]. Similar to the results

observed in Method 1 (total area measurement), local delivery of the EGF alone or EGF and SDF-

1α in combination with the Gtn-HPA matrix significantly increased the presence of the migrated

cells immediately surrounding the gel islands by 175% compared to Gtn-HPA gel only group

(One-Way ANOVA, Tukey post-hoc test, p < 0.05). However, the addition of SDF-1α did not

further increase the number of migratory cells compared to EGF alone (Figure 4.8B).

A B

Figure 4.7 Histomorphometric analysis performed on the entire injury site to calculate the total gel area and cellular migration immediately surrounding the gel based on Method 1 (total area method). (A) Delivering growth factors with the Gtn-HPA gel did not affect the extent of the gel present while (B) EGF and SDF-1α presence in the gel recruited significantly greater number of cells immediately surrounding the gel in the injury cavity four weeks post injection. * p < 0.05, Tukey post-hoc test and the bars showing Mean ± SEM

Page 97: An Injectable Gelatin-Based Conjugate Incorporating EGF

97

A B

Figure 4.8 Histomorphometric analysis performed in three random standardized ROI (0.3 mm2) to calculate the total gel area and cellular migration immediately surrounding the gel based on Method 2 (standardized area method). (A) Delivering growth factors with the Gtn-HPA gel did not affect the extent of the gel present while (B) EGF alone and EGF and SDF-1α in combination delivered with the recruited significantly greater number of cells immediately surrounding the gel in the injury cavity four weeks post injection. *** p < 0.001, Tukey post-hoc test and the bars showing Mean ± SEM

Page 98: An Injectable Gelatin-Based Conjugate Incorporating EGF

98

4.4.5. Immunohistochemistry evaluation of the injury site: H&E staining provided general

morphology of the injury site. Next step was to identify the cell population that migrated into the

injury site and surrounding the gel islands. Sections were stained with various primary antibody

markers, results from which are reported below.

4.4.5.1. GFAP – Astrocytes: GFAP was used as a marker to study the response of the

astrocytes in and around the injury site. Numerous GFAP+ cells were observed around the injury

site and at the interface between the unaffected tissue and the injury site upon the qualitative

examination of the images (Figure 4.9). There was high density of the GFAP+ cells bordering the

injury site. The density of astrocytes was very high near the interface between the unaffected

B

C

D

Figure 4.9 Representative GFAP stained sections of the four groups (A) Control [N] (B) Gtn-HPA gel only [G8] (C) Gtn-HPA with EGF [GE8], and (D) Gtn-HPA with EGF and SDF-1α [GES8]. Scale bar = 1500 µm

A

Interface

Page 99: An Injectable Gelatin-Based Conjugate Incorporating EGF

99

tissue and the injury site, while it decreased towards the proximal and distal unaffected tissue.

Besides at the interface, the astrocytes were sparsely present in the injury site. Some GFAP+

cells had migrated into the injury site and had long thin cells with their astrocytic processes. In

order to assess the presence of astrocytes, GFAP+ cells were quantified at two different locations

– in the injury site and at the interface between the unaffected tissue and the injury site (Figure

4.9BC). Quantitative analysis of the area covered by GFAP+ cells in the injury site indicated no

significant difference among the treatment groups. However, the quantitative analysis of the area

covered by GFAP+ cells at the border of the unaffected tissue and the injury site showed

significant decrease in the presence of astrocytes in the three gel-injected groups (i.e. G8, GE8

and GES8) compared to the control [N] as shown in Figure 4.10 (One-Way ANOVA, Tukey post-

hoc test, p < 0.05). Furthermore, cells migrated immediately surrounding the gel islands did not

stain positive for the GFAP indicating that the cells recruited by EGF and SDF-1α possibly were

not astrocytes (Figure 4.11).

Figure 4.10. Quantitative analysis was performed to measure the %area covered by GFAP+ cells (astrocytes) at two locations - at the interface between the injury site and the unaffected tissue & in the injury site. Injection of Gtn-HPA gel with or without EGF and SDF-1α showed significantly less presence of astrocytes at the interface between the injury site and unaffected tissue while there were no differences among the groups regarding astrocytes present in the injury site. **** p <0.0001 and bars represent Mean ± SEM

Page 100: An Injectable Gelatin-Based Conjugate Incorporating EGF

100

4.4.5.2. Iba-1 – microglia/macrophages: Sections were stained with Iba-1 biomarkers to

study the presence of the microglia and macrophages population. Iba-1+ cells were present

throughout the injury site across all the groups. They were sparsely present in the unaffected

tissue of the spinal cord in its typical morphology – small ramified cells with processes (Figure

4.12A). However, Iba1+ cells in the injury site had different distinct morphology – these cells were

enlarged spherical cells without any processes (Figure 4.12C). This enlarged morphology

suggests that these cells are activated microglia [11]. Interestingly, Iba1+ cells at the border of

the injury site showed both the types of morphology of Iba1+ cells (Figure 4.12B). The difference

in the morphology of the Iba1+ cells in the unaffected tissue and the injury site was consistent

across all the treatment groups. Quantitative analysis of the Iba1+ cells in the injury site showed

that the percent stained area occupied by Iba1+ activated microglia/macrophages was similar

across all the treatment groups (Figure 4.13).

Gtn-HPA+EGF+SDF-1α

*

* *

* *

*

Gtn-HPA+EGF+SDF-1α

Figure 4.11 The cells that have migrated immediately surrounding the gel block stained negative for GFAP indicating that these cells are not astrocytes (Top) GFAP stained section showing the presence of the gel marked by white arrow and the cellular migration around the gel block marked by red asterisks. (Bottom) The same animal H&E showing the presence of the gel and cellular migration for comparison with the GFAP stained section. Scale bar = 1000 µm

Page 101: An Injectable Gelatin-Based Conjugate Incorporating EGF

101

4.4.5.3. NeuN – mature neurons: NeuN is a neuron-specific nuclear protein and is widely

used as a marker for mature neurons. NeuN staining was performed to assess if any newly

generated mature neurons are present either in the injury site or at the border of the injury site.

The staining protocol worked very well, and it showed cell bodies of the neurons stained with

NeuN. NeuN+ cells were present in the gray matter of the unaffected tissue on the proximal and

distal side of the injury site across all the treatment groups. However, very few NeuN+ cells were

present at the interface and arguably no NeuN+ cells were observed in the injury site. This result

was consistent across all the treatment groups (Figure 4.14).

Figure 4.12. Representative micrographs showing the Iba1+ cells present (A) in unaffected tissue with small ramified cells with processes (B) at the interface showing two types of morphologies - small ramified cells and enlarged spherical cells and (C) in injury site exhibiting enlarged spherical cells morphology. Scale bars = 200 µm

unaffected interface Injury site

A B C

Figure 4.13 Quantitative analysis of Iba1+ microglia/macrophages: percent area covered by the microglia/macrophages in the injury site is similar across all four groups. Bars represent Mean ± SEM

Page 102: An Injectable Gelatin-Based Conjugate Incorporating EGF

102

4.4.5.4. MMP2 – Gtn-HPA gel degradation: It was of interest to study the mode of

degradation of the gel and Matrix metalloproteinases (MMPs) are speculated to be the enzymes

responsible for the degradation of ECM proteins. Representative sections were stained with

different anti-MMP antibodies. It was observed that there was strong signal of MMP2 presence at

the border of the Gtn-HPA gel for both Gtn-HPA gel only group [G8] and Gtn-HPA with EGF [GE8]

NeuN/DAPI NeuN/DAPI NeuN unaffected tissue near the interface injury site

G

ES

8

G

E8

G8

N

Figure 4.14. NeuN staining to stain for mature neurons for all four treatment groups - Control (N), Gtn-HPA gel only (G8), Gtn-HPA+EGF (GE8), and Gtn-HPA+EGF+SDF-1α (GES8). The image grid shows the representative images at three different locations – unaffected tissue, near the interface and injury site. There were no mature neurons present in the injury site while neuronal cell bodies were stained well in the unaffected gray matter. Scale bars = 200 µm

Page 103: An Injectable Gelatin-Based Conjugate Incorporating EGF

103

group (Figure 4.15). This result indicates that MMP2 is one of the molecules responsible for the

degradation of the Gtn-HPA gel in vivo. It is interesting to note the higher cellular density around

the Gtn-HPA gel for Gtn-HPA+EGF [GE8] group compared to Gtn-HPA only [G8] group, as noted

by the DAPI staining. This result is consistent with the observations noted in the H&E staining.

4.4.5.5. EGF – Gtn-HPA incorporated EGF: EGF was locally delivered to the injury site

using Gtn-HPA as a delivery biomaterial. The proposition is that as the gel degrades, EGF will be

released creating a concentration gradient to recruit endogenous neural progenitors and glial cells

in the injury site. In order to assess the presence of EGF in the Gtn-HPA gel, few representative

sections were stained with anti-EGF antibody, we found that wherever we see the presence of

the gel, EGF gave strong signal in the Gtn-HPA gel for the Gtn-HPA+EGF [GE8] group. However,

Gtn-HPA gel in the Gtn-HPA gel only group [G8] stained negative for the EGF, confirming the

presence of the EGF in the gel delivered with EGF four weeks post-injection (Figure 4.16). EGF

staining could be used as a method to reliably stain the Gtn-HPA gel in the EGF delivered groups.

MMP2 DAPI MMP2/DAPI

Gtn-HPA + EGF

gel

Gtn-HPA

gel

Gtn

-HP

A g

el

on

ly

Gtn

-HP

A g

el

+ E

GF

Figure 4.15 Representative MMP2 stained micrographs showing the strong signal of MMP2 at the border (marked by white dotted line) of the Gtn-HPA gel in the injury site as discovered in both the Gtn-HPA gel only [G8] and Gtn-HPA gel+EGF [GE8] group. Scale bars = 200 µm

Gtn-HPA + EGF

gel

Gtn-HPA

gel

Page 104: An Injectable Gelatin-Based Conjugate Incorporating EGF

104

4.4.5.6. Nestin – Neural stem cells: Nestin is a type VI intermediate filament protein found

in neural stem cells. Nestin staining on the sections showed greater Nestin staining in the cells

present around the Gtn-HPA gel in the both the groups where chemoattractants, i.e. EGF and

SDF-1α were delivered with the Gtn-HPA gel [GE8 & GES8] compared to the Gtn-HPA gel only

group [G8] (Figure 4.17B-D). Control group [N] did not show the presence of Nestin positive cells

in the injury site (Figure 4.17A). Nestin staining showed elongated cells oriented perpendicular to

the border of the Gtn-HPA gel indicating the radial migratory pattern of the cells (Figure 4.17CD).

These Nestin positive cells were present predominantly in the cells migrated immediately

surrounding the Gtn-HPA gel in the injury site. It is important to note that these cells around the

gel were GFAP negative (Figure 4.11) and Nestin positive indicating that these are NSCs and not

astrocytes. Quantitative analysis of the Nestin positive cells showed that Gtn-HPA with EGF [GE8]

and Gtn-HPA with EGF+SDF-1α [GES8] groups showed significantly greater presence of NSCs

immediately surrounding the Gtn-HPA gel (One-Way ANOVA, Tukey post-hoc test, p < 0.05).

However, there was no statistical difference between GE8 and GES8 groups indicating that SDF-

1α did not enhance the recruitment of NSCs to the injury site. There was no statistical significance

between the Nestin positive cells between the control [N] and Gtn-HPA gel only [G] group

indicating that the gel alone is not able to recruit these cells to the injury site (Figure 4.18).

EGF DAPI EGF/DAPI

Gtn

-HP

A g

el

+ E

GF

G

tn-H

PA

ge

l o

nly

Gtn-HPA

Gtn-HPA + EGF

Figure 4.16 Representative EGF stained micrographs showing the strong signal of EGF in the Gtn-HPA gel in the Gtn-HPA gel+EGF [GE8] group while Gtn-HPA gel stained negative in the Gtn-HPA gel only group [G8]. Scale bars = 200 µm

Page 105: An Injectable Gelatin-Based Conjugate Incorporating EGF

105

A B

C D

Gtn-HPA [G8]

Gtn-HPA + EGF

GES8

Figure 4.17 Representative Nestin stained sections show that the (A) control and (B) Gtn-HPA gel only [G8] did not show Nestin+ cells in the injury site while (C) Gtn-HPA+EGF [GE8] and (D) Gtn-HPA+EGF+SDF-1α [GES8] showed significantly higher presence of Nestin+ cells around the Gtn-HPA gel islands (marked by white dotted line). Scale bars = 200 µm

GE8

Page 106: An Injectable Gelatin-Based Conjugate Incorporating EGF

106

Figure 4.18 Quantitative analysis of the Nestin+ cells showed that EGF and SDF-1 delivered by the Gtn-HPA gel were able to recruit significantly greater number of neural stem cells to the Gtn-HPA gel islands compared to the control and Gtn-HPA gel only group. Moreover, there was no significant difference between the GE8 and GES8 groups (***p<0.001, ****p<0.0001, Tukey post-hoc test; bars represent Mean ± SEM)

Page 107: An Injectable Gelatin-Based Conjugate Incorporating EGF

107

4.5. Discussion

In this chapter, we investigated the effects of injecting Gtn-HPA gel with or without EGF

and SDF-1α in a cavity created by complete resection of 2 mm of the spinal cord tissue at T8 level

in Lewis rat model. Since traumatic or contusion type of experimental model of spinal cord injury

do not completely sever the cord, surgical models such as complete or partial resection are

utilized to study the healing and neuronal regeneration in the field of tissue engineering [12].

Moreover, complete section models better mimic the most severe clinical ‘complete SCI’ inducing

complete break of the neuronal circuitry as observed in most violence-related SCIs such as gun

and knife wounds [13]. The most important advantage of the complete resection model is the

creation of a reproducible standardized defect at a pre-defined location on the cord. In addition,

complete resection is a model that has been used to study the ‘bridging gap’ therapies such as

scaffolds and similar devices [14]. Therefore, complete resection created a well-defined cavity in

which our proposed biomaterial Gtn-HPA can be injected. We studied four groups namely –

control [N], 8wt% Gtn-HPA gel only [G8], 8wt% Gtn-HPA gel with EGF [GE8], and 8wt% Gtn-HPA

gel with EGF and SDF-1α [GES8]. There were several assessment parameters we employed to

test our hypotheses mentioned in the ‘Overall goal and hypotheses’ section of this chapter.

4.5.1. Functional recovery assessment: Clinically, loss of function drastically affects the

quality of life of the patients. Therefore, improvement in the function is the most critical aspect of

pre-clinical studies that aim to treat SCI. Beattie, Basso, and Breshnahan (BBB) scale is a

validated, standardized and well-accepted behavioral scale that is used to assess the functional

integrity of rats’ hindlimbs post SCIs, usually with the underlying assumption that the cardiac and

respiratory functions are largely unaffected. BBB involves animals walking in an open field with

their hindlimb movement carefully observed. Various neurological functions such as joint

movements, weight bearing, fore/hind limb coordination, paw placement, and the directions of the

tail are judged based on the 21-point BBB scale [8, 15]. In our study, immediately after inducing

a 2 mm complete resection SCI at T8 level, both the right and left hindlimbs were completely

paralyzed. Even after a week, most animals were scored 0 (complete paralysis of hindlimbs) on

the BBB scale for both the hindlimbs. Rats showed very modest spontaneous recovery in the

function in the initial weeks, which is attributed to recovery from the spinal shock and ‘re-wiring’

to facilitate neuronal signal transmission [16]. After four weeks, animals in all four groups saw

modest recovery improving the BBB score by 2 (control) and 3-5 (for the treatment groups). The

extent of the recovery in BBB scores observed after four weeks post complete resection SCI in

our study is consistent with similar studies [1, 17-19]. For example, human dental pulp-derived

Page 108: An Injectable Gelatin-Based Conjugate Incorporating EGF

108

stem cells were implanted after inducing a complete transection injury in rats. After four weeks,

controls showed improvement in BBB score of 3 points while the treatment groups showed

improvement in the BBB score of 4-6 [18]. Although the trend suggested better improvement in

the treatment groups, our study did not indicate any statistical differences among the groups in

the functional recovery using the BBB scale after four weeks (Figure 4.3). Future study should

focus on repetition of the study with the same number of animals in each group. In order to

account for the differences in the baseline scores of the animals, each rat was treated as its own

control to compare the differences in the scores at 1 (acute) and 4 (chronic) weeks post injury. It

is important to consider that since the BBB scale is not linear, improvement from a score of 2 to

4 is different than the improvement in the score from 4 to 6 even though both would show net

improvement by two points. Interestingly, the number of animals that improved by more than 2

points were greater in the treatment groups compared to the control and were statistically

significant for the left hindlimb but not for the right hindlimb (Table 4.4). However, it is difficult to

pinpoint the exact underlying mechanism due to which more animals showed better improvement

in the Gtn-HPA gel injected groups compared to the controls, given highly sophisticated nature of

the neural tissue after an injury. Various explanations of recovery include the greater sparing of

the axons, intrinsic changes in the circuitry termed as neuroplasticity, and stimulation of the lower

motor neurons not directly affected by the injury to the upper motor neuron [20, 21]. Most probably

explanation is the neuroprotection of further secondary damage caused by the injury due to

implanted Gtn-HPA matrices. It could explain the better net improvement in BBB scores in the

Gtn-HPA injected groups compared to the controls observed in this study. This phenomenon has

been observed in various studies where the implantation of a biomaterial has shown

neuroprotective effects in preventing further degradation of the unaffected spinal cord tissue and

thereby showing better functional recovery in biomaterial treated groups [22-24]. However, longer

time period assessments with more animals would be required to assess potentially significant

benefit of the Gtn-HPA injection to improve functional recovery after complete resection SCI. In

addition, other behavioral tests and electrophysiological measurements should be considered for

functional assessment for future study.

4.5.2. Histological evaluation of the injury site: Histological evaluation was performed four

weeks after inducing SCI in coronal sections. In all the groups, it was evident that the surgery had

indeed created a complete break in the neuronal circuitry. The histological sections displayed

typical characteristics of central gray matter in both hemispheres in both the caudal and rostral

sides of the injury site. The length of the injury site was more than 2 mm supporting the well-

Page 109: An Injectable Gelatin-Based Conjugate Incorporating EGF

109

known phenomenon that the axons retract in both directions after complete transection and there

is secondary damage to the unaffected tissue progressing the injury [25, 26].

Although no treatment was applied to the control group [N], the injury site was filled two

kinds of morphology – ‘empty spaces’ and dense collagen fibrous tissue (Figure 4.5A). The empty

spaces probably represent the cystic cavitations, which are formation of fluid-filled cysts in

response to the injury. These cavities are usually filled with macrophages and the growth-

inhibiting molecules deposited in response to degradation of myelin [27]. These cavitations are

consistent in the chronology of the pathophysiological response because these cysts are

developed in the subacute and intermediate phase of the response – which coincides with the

four week evaluation of our study [28, 29]. In addition, ferritin deposits were observed in the control

group indicating the remains of the blood clot that was formed due to the bleeding either from

cord parenchyma or from surrounding muscles after creating the surgical injury. It is possible that

the blood clot provided a temporary matrix to the infiltrating inflammatory cells especially because

the dura membrane was compromised due to the surgical nature of the model. Both the

morphologies – cysts and dense fibrous collagen tissue is inhibitory to the cellular migration into

the cavity site for any potential reparative response. Therefore, goal of this study to replace the

growth-inhibitory environment with that of a biomaterial that would be permissive of the cellular

infiltration without eliciting major inflammatory response. Therefore, our proposed therapy

involves injecting a Gtn-HPA matrix in the defect site with chemoattractants, EGF and SDF-1α to

recruit endogenous neural stem cells and glial cells to the injury site.

After four weeks, Gtn-HPA gel covered the predominant area in the injury site in all the

treatment groups. It is important to note that this is the first time it has been shown that 8wt% Gtn-

HPA gel is present in the injury site after four weeks. This is a significant finding in the context of

previous work done in our laboratory with injectable materials for spinal cord repair. Previous work

aimed to develop injectable matrices to promote regenerative response in our laboratory included

the evaluation of collagen-genipin gels [30]. The results indicated that most of the injected

collagen-genipin gel was degraded as early as two weeks after injection. It is essential for the

biomaterial to persist for longer duration to facilitate the infiltration of cells into the defect, better

cell-substrate signaling, remodel the underlying ECM and deliver signaling molecules over long

period of time [31]. The presence of the Gtn-HPA matrix in most animals in all the treatment

groups was assuring that the Gtn-HPA could serve as a reliable scaffold to promote regenerative

response after spinal cord injury. However, we did not see the presence of the Gtn-HPA gel in all

the animals from a treatment group. We believe there are few potential explanations that account

Page 110: An Injectable Gelatin-Based Conjugate Incorporating EGF

110

for the absence of the Gtn-HPA gel in two out of 13 animals that were injected with Gtn-HPA

matrix in this study (Table 4.5). First, bleeding in the cavity site during the injection could cause

incomplete gelation or more porous Gtn-HPA gel leading to a less stiff Gtn-HPA gel than

anticipated. This would result in faster degradation of the gel in vivo. Second, the animals were

quick to gain consciousness after the surgery and did movements that could lead to dislodging of

the Gtn-HPA gel from the cavity because the dural replacement collagen membrane was placed

to cover the injury site and was not sutured to the cord. Third, some animals could show more

severe inflammatory response due to which the Gtn-HPA gel degraded quickly. These potential

explanations were considered in designing a second generation Gtn-HPA gel and in better

containing the gel in the cavity site.

Another interesting feature of the morphology of the Gtn-HPA gel was the shape of the

Gtn-HPA gel after four weeks. It was predominantly present in the injury site in form of multiple

‘islands’ of varying sizes (Figure 4.5B-C) as opposed to a continuous block (Figure 4.5D). These

islands could mimic the small cavitary lesions, which are common clinical manifestation of the

spinal cord injury due to various disease progressions [32]. We believe there could be several

mechanisms that could leads to the formation of ‘gel islands. First, heterogeneity in the modulus

of the Gtn-HPA gel during gelation could contribute to mechanical differences in the gel. These

differences could result in varying rates of the gel degradation leading to the break in the

continuous matrix resulting in formation of gel islands over time. Second, the number and the

types of cells present around the Gtn-HPA during acute phase of the pathophysiological response

could affect the extent of cellular migration into the Gtn-HPA matrix. For example, macrophages

are shown to occupy the injury site as early as a week after the spinal cord injury and release

various inflammatory cytokines and ECM degrading molecules, especially Matrix

metalloproteinases (MMPs) [33, 34]. MMP2 is a zinc-based gelatinase and we observed strong

MMP2 signal at the border of the Gtn-HPA gel islands in this study (Figure 4.15). Therefore, it is

possible that activated macrophages released significant amount of MMP2 at the border of the

continuous Gtn-HPA block resulting in formation of gel islands during the four weeks. Third,

animal movement immediately after the surgery and within initial weeks could mechanically break

the Gtn-HPA into physically separate entities leading to the formation of gel islands. More detailed

investigation for shorter evaluation duration should be done for future studies. For a proof of

principle, histological evaluation will be performed at earlier time points in few animals to

investigate the phenomenon of formation of gel islands in the study described in Chapter 6.

Page 111: An Injectable Gelatin-Based Conjugate Incorporating EGF

111

Another feature important to discuss is the ability of the Gtn-HPA to establish a continuous

physical interface with the surrounding tissue to re-establish the lost framework due to the injury.

We observed that although Gtn-HPA alone was largely not able to establish an interwoven

physical structure with the surrounding tissue, Gtn-HPA with EGF [GE8] and Gtn-HPA with EGF

and SDF-1α [GES8] was able to form a well-defined continuous interface with the surrounding

tissue re-establishing continuous tissue throughout the injury site (Figure 4.6) This is in

accordance with the findings in various studies that matrices without any form of the molecular

cues failed at establishing a well-defined continuous framework with the surrounding tissue in

tissue repair approach [35, 36]. Therefore, signaling molecules such as EGF and SDF-1α are

critical to include in the injection of Gtn-HPA matrices.

One of the hypotheses of the study was that EGF and SDF-1α would promote the

recruitment of different cells types such as the endogenous neural stem cells and other glial cells

to the injury site. Qualitatively, greater number of cells had migrated to surround the Gtn-HPA gel

islands in both Gtn-HPA with EGF [GE8] and Gtn-HPA with EGF and SDF-1α [GES8] group

compared to the Gtn-HPA gel only [G8] group (Figure 4.5 & 4.6). Moreover, this finding in our

study is in accordance with the conclusion that Gtn-HPA gel is permissive of the cellular migration

both in vitro and in vivo brain ICH model as reported based on previous work in our laboratory

[10] . Quantitative analysis of the Gtn-HPA gel area and cellular migration immediately

surrounding the Gtn-HPA gel islands showed that EGF and SDF-1α were able to recruit

significantly greater number of cells to the Gtn-HPA gel islands compared to the Gtn-HPA gel only

group (Figure 4.7 & 4.8). We validated this result with two different quantification methods

measuring two parameters – gel area and cellular migration area immediately surrounding the

gel. The first method accounted for the all the gel islands and cellular migration immediately

around all of them present throughout the injury site. The second method measured both the

parameters in randomly chosen images in the injury site. Statistical analyses of these parameters

in both the methods showed significantly greater recruitment of cells by EGF and SDF-1α

compared to the control and Gtn-HPA only group (Figure 4.7 & 4.8). Therefore, we deduce that

Gtn-HPA gel alone would not be able to recruit cells to the injury site and therefore, signaling

molecules, i.e. EGF and SDF-1α, are essential. This is a critical result supporting our hypothesis

for this study. However, GES8 group did not show significantly greater cellular migration

compared to the GE8 groups indicating that presence of SDF-1α did not enhance the recruitment

of cells to the injury site. Subsequent to the promising histological observations and conclusions,

our goal was to identify the cells that migrated around the gel islands and in the injury site to study

the cell population of the injury site.

Page 112: An Injectable Gelatin-Based Conjugate Incorporating EGF

112

4.5.3. Immunohistochemical evaluation of the injury site: Several biomarkers were

stained to identify the presence and the extent of various cell types and important molecules in

the context of this study. Specifically, GFAP (astrocytes), Iba1 (microglia/macrophages), NeuN

(mature neurons), Nestin (neural stem cells), EGF and MMP2 (Gtn-HPA gel degradation) were

stained to obtain a comprehensive composition of the injury site post Gtn-HPA implantation with

or without EGF and/or SDF-1α.

Astrocytes have been shown to play a multi-functional role in the context of repair and

regeneration after spinal cord injury. The multifaceted astrocytic response contributes to

neuroprotection of the spinal cord while inhibiting axonal regeneration [37]. Astrocytes are

typically categorized into three types in the CNS – naïve astrocytes, reactive astrocytes and scar-

forming astrocytes [38]. Naïve astrocytes help in maintaining the neuronal parenchyma and the

blood-brain barrier while the reactive astrocytes assist in limiting the spread of inflammatory cells

in response to injury promoting remodeling of the ECM and repair of the tissue [39]. These

reactive astrocytes convert to scar-forming astrocytes based on the environmental cues. They

form the scar by adhering to each other and releasing molecules like CSPG to produce a

penetrating glial scar that impedes axonal regeneration after spinal cord injury [40, 41]. We

observed the presence of GFAP+ cells in the unaffected tissue and in the injury site but

predominantly at the interface between the unaffected tissue and the injury site. Quantification of

GFAP+ cells showed no statistical differences between the groups in the injury site (Figure 4.10).

Moreover, the cellular migration immediately around the Gtn-HPA gel islands stained negative for

the GFAP indicating that those cells are probably not reactive astrocytes because GFAP is

considered to be the hallmark biomarker for reactive astrocytes (Figure 4.11) [42]. However,

quantification of the GFAP+ cells at the interface between the unaffected tissue and the injury site

showed significant reduction in the GFAP+ cells or astrocytes at the interface compared to the

control group indicating that implantation of Gtn-HPA matrix, with or without the growth factors,

into the cavity downregulates the astrocytic response to the injury thereby either delaying the

process of glial scar formation or reducing the extent of glial scar formation in the treatment groups

(figure 4.10). Interestingly, one hypothesis proposed by Macaya in his doctoral work stated that

the elongated astrocytic processes towards the center of the defect are growth-promoting while

the astrocytes bordering the injury site with dense processes are inhibitory [30]. Results of this

study show that although the treatment groups do not increase the presence of the growth

promoting astrocytes in the injury site, the treatment reduces the number of the inhibitory

astrocytes at the border. This is an important evidence to suggest that the treatment reduces the

glial astrogliosis, which could promote beneficial healing of the injury site.

Page 113: An Injectable Gelatin-Based Conjugate Incorporating EGF

113

Post-injury, the resident microglia migrate to the injury site while the monocytes

differentiate into macrophages in the acute phase of the pathophysiological process after SCI.

Since they are phagocytic in nature, they contribute to the debris clearing in the injury site to

mitigate the secondary damage [11, 43]. Macrophages have been classified into two phenotypes

– M1 and M2. M1 is shown to be promoting the release of pro-inflammatory factors while M2

contributes in wound healing and repair. Interestingly, M1 have been shown to have detrimental

effect in the spinal cord injury but M2 macrophages have shown to promote regenerative

response after spinal cord injury [44]. Iba1 is a microglia/macrophage specific marker that was

used to assess the inflammatory response to the Gtn-HPA gel. The morphology of the Iba1+ cells

varied depending upon the location on the section (Figure 4.12). Unaffected tissue had typical

ramified phenotype cells with its processes indicating that these are resident microglia. Injury site

predominantly had enlarged macrophages indicating that these were the activated microglia or

macrophages responding to the injury and contributing the clearing the debris from the myelin

degradation and other necrotic tissue. Interface between the unaffected tissue and the injury site

had combination of both phenotypes possibly indicating that the resident microglia are being

recruited to the injury site in response to the cytokines released in the injury site. This observation

is consistent with the mechanism of the activation of microglia after spinal cord injury [45].

Quantitative analysis did not show statistical differences among the number of

microglia/macrophages present in the injury site among all four treatment groups (Figure 4.13).

This indicates that injection of Gtn-HPA, with or without the factors, did not elicit a severe

inflammatory response providing an in vivo evidence that Gtn-HPA is a biocompatible biomaterial.

Gtn-HPA and its constituents were well tolerated by the rats in the study. This is an important

finding in vivo for the Gtn-HPA material injected for the first time in the spinal cord injury models.

NeuN is a very specific biomarker for nuclear protein expressed in mature neurons. It also

has been used to study the differentiation of stem cells into neurons [46, 47]. Sections from all

treatment were stained with NeuN to assess whether any neuronal lineage was promoted by the

potential regenerative effects of the migrated cells in the injury site. Across all the treatment

groups, we found that there were no NeuN+ cells in the injury site while we were able to visualize

characteristic neuronal cell body stained with NeuN in the unaffected tissue and at the interface

of the unaffected tissue and the injury site. Qualitatively, the number of NeuN+ cells present in

the unaffected tissue were higher than at the interface (Figure 4.14). This finding is consistent

with our expected outcome of NeuN staining, especially four weeks after Gtn-HPA matrix

injection. However, future studies should quantitatively investigate any differences among the

Page 114: An Injectable Gelatin-Based Conjugate Incorporating EGF

114

groups for the presence of neuronal differentiation in the injury site and near the interface for

longer duration post Gtn-HPA matrix injection.

Gtn-HPA is a gelatin-based biomaterial and has excellent tunable properties to control the

gelation time and stiffness by varying the concentration of its constituents [6]. In our study, Gtn-

HPA was used in an in vivo application for the spinal cord injury for the first time. Therefore, it was

important to comment on the potential mode of degradation of the Gtn-HPA in vivo. Although Gtn-

HPA was injected in the brain ICH model, it’s in vivo degradation behavior was not studied

previously [10]. Matrix metalloproteinases (MMPs) were reported to be involved into the cellular

migration of neural progenitor cells [48]. Out of various MMPs that were tested, MMP2 provided

strong signal specifically at the border of the Gtn-HPA gel island in both treatment groups G8 and

GE8 (Figure 4.15). Since the MMP2 signal was strong only at the border of the gel island, we

hypothesize that MMP2 is responsible for the degradation of Gtn-HPA gel to promote cellular

infiltration into the Gtn-HPA matrix. Remarkably, MMP2 is a gelatinase A enzyme and therefore

could be the agent to degrade gelatin-based Gtn-HPA biomaterial in vivo. More detailed

investigation should be performed in the future studies to assess the role of MMP2 with Gtn-HPA.

In addition to MMP2, EGF was the other protein stained with few preliminary sections to confirm

the presence of EGF in the Gtn-HPA gel with EGF [GE8] group. The stained micrographs

confirmed the presence of EGF exclusively in the Gtn-HPA gel island. However, the Gtn-HPA gel

island in the Gtn-HPA gel only group [G8] stained negative for the EGF serving as a negative

control (Figure 4.16). Although no statement can be made about the bioactivity of the EGF after

four weeks because the sections were processed with fixation, it is encouraging that Gtn-HPA

matrix is able to deliver EGF locally to the injury site at least over four weeks. Furthermore, it

could also provide a reliable method to locate the presence of the Gtn-HPA gel in groups where

EGF was delivered with the Gtn-HPA matrix. Specificity of the staining is high such that EGF

stained positive on a very small gel island as well (Figure 4.16, bottom). Future studies should be

designed to assess the bioactivity of the EGF in vivo. For example, EGF delivered by the Gtn-

HPA gel could be pre-labeled with a fluorophore prior to injection and could be longitudinally

followed over four or more weeks and investigate if the signal is limited to the Gtn-HPA gel

location. This experiment could offer interesting insight into the degradation mechanism and EGF

release mechanism in vivo. These preliminary findings will be thoroughly validated in more

animals in Chapter 6 by comparing EGF expression in different groups.

The identity of the cells migrated immediately surrounding the gel islands and into the Gtn-

HPA gel was determined after sections were stained with Nestin to evaluate the presence of

Page 115: An Injectable Gelatin-Based Conjugate Incorporating EGF

115

neural stem cells in the injury site. Nestin is an intermediate type VI protein and is considered as

a biomarker for undifferentiated CNS cells prior to committing to a specific cell lineage, considered

to be the neural stem cells (NSCs) [49]. It is reported that NSCs are found in a niche at the poles

of the spinal cord ependymal layer near the central canal. Similar niche is observed in the SVZ

region in the brain [50]. These niches stain positive for Nestin and negative for GFAP confirming

the presence of NSCs [51]. There is a panel of biomarkers that can be used to further confirm the

cells to be NSCs and include Vimentin, Mushashi-1, and Sox 2 (R&D systems, SC205). In our

study, cells migrated around gel islands stained positive for Nestin for both Gtn-HPA with EGF

[GE8] and Gtn-HPA with EGF and SDF-1α [GES8] (Figure 4.17). We did not encounter any

Nestin+ cells in the injury site for the control group. There were some cells that stained Nestin

positive surrounding the gel islands in the Gtn-HPA gel only [G8]. Quantitative analysis of the

Nestin+ cells showed significantly greater number of NSCs that have migrated to the gel islands

in GE8 and GES8 groups compared to the control and G8 group (Figure 4.18). These results

show that Gtn-HPA gel alone is not capable of recruiting important cells such NSCs while EGF

and SDF-1α were able to recruit the endogenous NSCs from the spinal cord ependymal layer

niche near the central canal. Furthermore, there was no difference between the GE8 and GES8

group indicating that SDF-1α either does not play a role in NSCs recruitment or does not enhance

the recruitment of NSCs to the injury site. This is consistent with the proposed hypothesis that

NSCs are highly responsive to EGF [52]. Hemann et al reported that intrathecal administration

of EGF resulted in significantly greater ependymal cell proliferation in the central

canal immediately rostral and caudal to the lesion compared to controls and in greater white

matter sparing [53]. Similar behavior of the NSCs was observed in our study in response to EGF

four weeks post Gtn-HPA with EGF injection. Future studies should include longer duration of

evaluation and it would be interesting to find out the identity of the migrated cells and assess if

they differentiate into neural or glial fate to promote regenerative response and improve functional

recovery after the spinal cord injury.

In summation, a 2 mm complete resection of the spinal cord tissue at T8 level created a

severe spinal cord injury completely disrupting the neuronal circuity. 8wt% Gtn-HPA persisted in

the injury site four weeks post injection and was permissive of cellular migration, making Gtn-HPA

a promising biomaterial for future studies. Injection of the Gtn-HPA matrix decreased the presence

of growth-inhibitory astrocytes at the border of the injury site thereby reducing the extent of glial

scar formation. Gtn-HPA is well tolerated and did not elicit a severe inflammatory response.

Injecting the 8wt% Gtn-HPA matrix with EGF alone or in combination with SDF-1α promoted

recruitment of endogenous neural progenitor cells to the gel islands in the injury site. However,

Page 116: An Injectable Gelatin-Based Conjugate Incorporating EGF

116

SDF-1α did not enhance the recruitment response. In a severe injury such as 2 mm complete

resection, injecting the Gtn-HPA matrix did not show better functional recovery compared to the

control four weeks post injection. However, left hindlimb shown better improvement in the

treatment groups between first and fourth week in the treatment groups compared to the controls.

The encouraging results in this study laid the foundation for designing a second generation Gtn-

HPA matrix formulation and for rigorous evaluation in a 2 mm hemi-resection animal model of SCI

in Chapter 6.

Page 117: An Injectable Gelatin-Based Conjugate Incorporating EGF

117

4.6. References

1. Lukovic, D., et al., Complete rat spinal cord transection as a faithful model of spinal cord injury for translational cell transplantation. Sci Rep, 2015. 5: p. 9640.

2. Cheriyan, T., et al., Spinal cord injury models: a review. Spinal Cord, 2014. 52(8): p. 588-95. 3. Facts and Figures at a Glance, in SCI Data Sheet. 2018, University of Alabama at Birmingham:

National Spinal Cord Injury Statistical Center. p. 2. 4. VA and spinal cord injury. 2008, Department of Veterans Affairs: Washington, DC. p. 6. 5. Ullrich, P.M., et al., Pain among veterans with spinal cord injury. Journal of rehabilitation research

and development, 2008. 45(6): p. 793-800. 6. Wang, L.S., et al., Injectable biodegradable hydrogels with tunable mechanical properties for the

stimulation of neurogenesic differentiation of human mesenchymal stem cells in 3D culture. Biomaterials, 2010. 31(6): p. 1148-57.

7. Cholas, R., H.P. Hsu, and M. Spector, Collagen scaffolds incorporating select therapeutic agents to facilitate a reparative response in a standardized hemiresection defect in the rat spinal cord. Tissue Eng Part A, 2012. 18(19-20): p. 2158-72.

8. Basso, D.M., M.S. Beattie, and J.C. Bresnahan, A sensitive and reliable locomotor rating scale for open field testing in rats. J Neurotrauma, 1995. 12(1): p. 1-21.

9. Menorca, R.M.G., T.S. Fussell, and J.C. Elfar, Nerve physiology: mechanisms of injury and recovery. Hand clinics, 2013. 29(3): p. 317-330.

10. Lim, T.C., Injectable Chemokine-Releasing Gelatin Matrices for Enhancing Endogenous Regenerative Responses in the Injured Rat Brain, in Health, Science and Technology. 2014, Massachusetts Institute of Technology: Cambridge, MA.

11. Zhou, X., X. He, and Y. Ren, Function of microglia and macrophages in secondary damage after spinal cord injury. Neural regeneration research, 2014. 9(20): p. 1787-1795.

12. Kwon, B.K., T.R. Oxland, and W. Tetzlaff, Animal models used in spinal cord regeneration research. Spine (Phila Pa 1976), 2002. 27(14): p. 1504-10.

13. Heimburger, R.F., Return of function after spinal cord transection. Spinal Cord, 2005. 43(7): p. 438-40.

14. Lai, B.Q., et al., Graft of a tissue-engineered neural scaffold serves as a promising strategy to restore myelination after rat spinal cord transection. Stem Cells Dev, 2014. 23(8): p. 910-21.

15. Scheff, S.W., D.A. Saucier, and M.E. Cain, A statistical method for analyzing rating scale data: the BBB locomotor score. J Neurotrauma, 2002. 19(10): p. 1251-60.

16. Batista, C.M., et al., Behavioral Improvement and Regulation of Molecules Related to Neuroplasticity in Ischemic Rat Spinal Cord Treated with PEDF. Neural Plasticity, 2014. 2014: p. 16.

17. Hwang, D., et al., Transplantation of human neural stem cells transduced with Olig2 transcription factor improves locomotor recovery and enhances myelination in the white matter of rat spinal cord following contusive injury. BMC Neuroscience, 2009. 10(1): p. 117.

18. Sakai, K., et al., Human dental pulp-derived stem cells promote locomotor recovery after complete transection of the rat spinal cord by multiple neuro-regenerative mechanisms. The Journal of clinical investigation, 2012. 122(1): p. 80-90.

19. Fouad, K., et al., Combining Schwann cell bridges and olfactory-ensheathing glia grafts with chondroitinase promotes locomotor recovery after complete transection of the spinal cord. J Neurosci, 2005. 25(5): p. 1169-78.

20. Muir, G.D. and A.A. Webb, Mini-review: assessment of behavioural recovery following spinal cord injury in rats. Eur J Neurosci, 2000. 12(9): p. 3079-86.

21. Maier, I.C. and M.E. Schwab, Sprouting, regeneration and circuit formation in the injured spinal cord: factors and activity. Philos Trans R Soc Lond B Biol Sci, 2006. 361(1473): p. 1611-34.

Page 118: An Injectable Gelatin-Based Conjugate Incorporating EGF

118

22. Kushchayev, S.V., et al., Hyaluronic acid scaffold has a neuroprotective effect in hemisection spinal cord injury. J Neurosurg Spine, 2016. 25(1): p. 114-24.

23. Faccendini, A., et al., Nanofiber Scaffolds as Drug Delivery Systems to Bridge Spinal Cord Injury. Pharmaceuticals (Basel), 2017. 10(3).

24. Kim, Y.C., et al., Transplantation of Mesenchymal Stem Cells for Acute Spinal Cord Injury in Rats: Comparative Study between Intralesional Injection and Scaffold Based Transplantation. J Korean Med Sci, 2016. 31(9): p. 1373-82.

25. Cheng, X.-L., et al., The longitudinal epineural incision and complete nerve transection method for modeling sciatic nerve injury. Neural regeneration research, 2015. 10(10): p. 1663-1668.

26. Kim, C.Y., PEG-assisted reconstruction of the cervical spinal cord in rats: effects on motor conduction at 1 h. Spinal cord, 2016. 54(10): p. 910-912.

27. van Niekerk, E.A., et al., Molecular and Cellular Mechanisms of Axonal Regeneration After Spinal Cord Injury. Molecular & cellular proteomics : MCP, 2016. 15(2): p. 394-408.

28. Fleming, J.C., et al., The cellular inflammatory response in human spinal cords after injury. Brain, 2006. 129(Pt 12): p. 3249-69.

29. Ahuja, C.S., et al., Traumatic spinal cord injury. Nat Rev Dis Primers, 2017. 3: p. 17018. 30. Macaya, D., Biomaterials-Tissue Interaction of an Injectable Collagen-Genipin Gel in a Rat Hemi-

Resection Model of Spinal Cord Injury in Health, Science and Technology. 2013, Massachusetts Institute of Technology: Massachusetts Institute of Technology. p. 210.

31. Kumar, P., Future biomaterials for enhanced cell-substrate communication in spinal cord injury intervention. Future science OA, 2017. 4(2): p. FSO268-FSO268.

32. Sundarakumar, D.K., et al., Evaluation of Focal Cervical Spinal Cord Lesions in Multiple Sclerosis: Comparison of White Matter–Suppressed T1 Inversion Recovery Sequence versus Conventional STIR and Proton Density–Weighted Turbo Spin-Echo Sequences. American Journal of Neuroradiology, 2016. 37(8): p. 1561-1566.

33. Campbell, E.J., et al., Neutral proteinases of human mononuclear phagocytes. Cellular differentiation markedly alters cell phenotype for serine proteinases, metalloproteinases, and tissue inhibitor of metalloproteinases. J Immunol, 1991. 146(4): p. 1286-93.

34. Elkington, P.T., J.A. Green, and J.S. Friedland, Analysis of matrix metalloproteinase secretion by macrophages. Methods Mol Biol, 2009. 531: p. 253-65.

35. Deguchi, K., et al., Implantation of a New Porous Gelatin–Siloxane Hybrid into a Brain Lesion as a Potential Scaffold for Tissue Regeneration. Journal of Cerebral Blood Flow & Metabolism, 2006. 26(10): p. 1263-1273.

36. Mooney, D.J. and H. Vandenburgh, Cell Delivery Mechanisms for Tissue Repair. Cell Stem Cell, 2008. 2(3): p. 205-213.

37. White, R.E., D.M. McTigue, and L.B. Jakeman, Regional heterogeneity in astrocyte responses following contusive spinal cord injury in mice. J Comp Neurol, 2010. 518(8): p. 1370-90.

38. Okada, S., et al., Astrocyte reactivity and astrogliosis after spinal cord injury. Neuroscience Research, 2018. 126: p. 39-43.

39. Herrmann, J.E., et al., STAT3 is a critical regulator of astrogliosis and scar formation after spinal cord injury. J Neurosci, 2008. 28(28): p. 7231-43.

40. Okada, S., et al., Conditional ablation of Stat3 or Socs3 discloses a dual role for reactive astrocytes after spinal cord injury. Nature Medicine, 2006. 12: p. 829.

41. Nathan, F.M. and S. Li, Environmental cues determine the fate of astrocytes after spinal cord injury. Neural regeneration research, 2017. 12(12): p. 1964-1970.

42. Sofroniew, M.V. and H.V. Vinters, Astrocytes: biology and pathology. Acta neuropathologica, 2010. 119(1): p. 7-35.

Page 119: An Injectable Gelatin-Based Conjugate Incorporating EGF

119

43. Wang, X., et al., Macrophages in spinal cord injury: phenotypic and functional change from exposure to myelin debris. Glia, 2015. 63(4): p. 635-51.

44. Kigerl, K.A., et al., Identification of two distinct macrophage subsets with divergent effects causing either neurotoxicity or regeneration in the injured mouse spinal cord. J Neurosci, 2009. 29(43): p. 13435-44.

45. Jin, X. and T. Yamashita, Microglia in central nervous system repair after injury. The Journal of Biochemistry, 2016. 159(5): p. 491-496.

46. Verdiev, B.I., et al., Molecular genetic and immunophenotypical analysis of Pax6 transcription factor and neural differentiation markers in human fetal neocortex and retina in vivo and in vitro. Bull Exp Biol Med, 2009. 148(4): p. 697-704.

47. Gusel'nikova, V.V. and D.E. Korzhevskiy, NeuN As a Neuronal Nuclear Antigen and Neuron Differentiation Marker. Acta naturae, 2015. 7(2): p. 42-47.

48. Lim, T.C., et al., Chemotactic recruitment of adult neural progenitor cells into multifunctional hydrogels providing sustained SDF-1alpha release and compatible structural support. Faseb j, 2013. 27(3): p. 1023-33.

49. Suzuki, S., et al., The neural stem/progenitor cell marker nestin is expressed in proliferative endothelial cells, but not in mature vasculature. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society, 2010. 58(8): p. 721-730.

50. Hamilton, L.K., et al., Cellular organization of the central canal ependymal zone, a niche of latent neural stem cells in the adult mammalian spinal cord. Neuroscience, 2009. 164(3): p. 1044-56.

51. Namiki, J., et al., Nestin protein is phosphorylated in adult neural stem/progenitor cells and not endothelial progenitor cells. Stem Cells Int, 2012. 2012: p. 430138.

52. Lillien, L. and H. Raphael, BMP and FGF regulate the development of EGF-responsive neural progenitor cells. Development, 2000. 127(22): p. 4993-5005.

53. Jimenez Hamann, M.C., C.H. Tator, and M.S. Shoichet, Injectable intrathecal delivery system for localized administration of EGF and FGF-2 to the injured rat spinal cord. Exp Neurol, 2005. 194(1): p. 106-19.

Page 120: An Injectable Gelatin-Based Conjugate Incorporating EGF

120

Chapter 5:

Developing an ex vivo MRI protocol to

visualize the Gtn-HPA matrix in the

spinal cord

Page 121: An Injectable Gelatin-Based Conjugate Incorporating EGF

121

5.1. Introduction and motivation

MRI plays a critical role in the clinical assessment and management of the spinal cord

injury clinically. Computed tomography (CT) is a widely used imaging modality to evaluate the

bone dislocations or fractures after a traumatic spinal cord injury. Although CT is a more

convenient and cost-effective method than MRI to assess the cause and the location of the injury,

it does not give information about the damage to soft tissues, such as spinal cord [1]. The advent

of MRI in 1980s led the neuroimaging community to realize the powerful capability of the MRI to

study the pathomorphic changes in the cord after injury [2]. In 1988, MRI was used to study

specific relevant pathophysiology for the diagnosis and prognosis after SCI. Specifically, for the

first time, MRI was utilized to identify and characterize the MRI signal patterns for hemorrhage in

the cord, edema in the cord and combination of both hemorrhage and edema [3]. Since then, MRI

has become the standard clinical tool used by the clinicians for neuroimaging. In current times,

MRI is widely used by the clinicians to evaluate the location of the injury, the severity of the injury,

extent of neurological deficit and presence of potential instability of the spine. MRI can serve as

a validation tool for the neurological deficit determined by the ASIA examination in the acute

setting [4]. However, there are certain trauma cases in which MRI is not advised due to logistical

issues such as patient stabilization and to prevent further injury [5]. With advances in the MRI

technology, many MRI sequences are available to the clinicians. Most widely used techniques

include sagittal T1 and T2 weighted imaging. However, short tau inversion recovery (STIR) and

gradient recall acquisition steady state (GRASS) area also implemented in applicable cases [4].

Although MRI is extensively used in the clinical setting, very few experimental SCI model

investigations include studying the injury using MRI. Histopathology is the most common method

used to assess the pre-clinical SCI models. However, histopathology techniques are labor-

intensive, invasive, and destructive for the cord. Moreover, it provides information in 2D plane

only after the duration of the experiment [6]. Longitudinal assessment in the same animal is

practically impossible with the conventional techniques. MRI has many advantages over the

conventional techniques that make it effective to study the experimental models of SCI. In addition

to being non-invasive and non-destructive, it also provides 3D volumetric information about the

morphology of the injured cord in a longitudinal manner such that each animal serves as its own

control [7]. Although in vivo MRI evaluation is difficult in rodent models, various studies have

shown the benefit of using MRI to study various histopathological findings such as edema,

hemorrhage, cyst localization and presence of biomaterials [7-10]. Some studies have utilized

Diffusion-weighted tensor imaging (DWI & DTI) to study the axonal structure, white matter sparing

Page 122: An Injectable Gelatin-Based Conjugate Incorporating EGF

122

after injury and demyelination process [11-13]. It is evident that MRI can provide critical

information after spinal cord injury in a reliable, noninvasive, non-destructive and longitudinal

manner. However, there are only few studies that utilize the MRI to learn the environments around

the implanted biomaterial post injection [14, 15]. One of the potential limitations is that water

accounts for as high as 60-70% of most biomaterials used in the SCI applications. Therefore, MRI

contrast agents such as Gadolinium-based contrast agents are suggested to be incorporated into

the biomaterial for better visualization. However, gadolinium-based contrast agents are relatively

large molecules that can easily affect the gelation procedure, porous structure, stiffness modulus

and other chemically important sites of the biomaterial [16]. Considering the disadvantages of the

contrast agents’ incorporation in Gtn-HPA matrix, I began investigation into developing an MRI-

based protocol to visualize the implanted Gtn-HPA matrix, to be able to differentiate the Gtn-HPA

matrix from the surrounding healthy tissue and to gain insights into the secondary damage due

to the induced SCI in our rodent model without the use of contrast agents. Specifically, in the

context of evaluating a biomaterial matrix such as Gtn-HPA, MRI can provide a powerful tool to

measure Gtn-HPA gel volume, lesion volume, extent of the injury, and study the changes in the

gel volume over the duration of the experiment.

We worked on three different MRI approaches that were promising to exclusively show

the implanted Gtn-HPA matrix in an MRI image. The first approach (Approach 1) employed T2

weighted imaging sequence to generate coronal images of the cord with defect area in the center

of the field of view. The second approach (Approach 2) aims to implement a novel Magnetic

Resonance Elastography (MRE) sequence to identify the Gtn-HPA gel based on the differences

in the mechanical properties of the injury area. The third approach (Approach 3) utilized a

systematic approach using T1-weighted Inversion recovery (T1IR) using various inversion times

(ITs) to generate separate images for the Gtn-HPA matrix and surrounding tissue at the same

location. All images captured were coronal slices with the injured site in the center of field of view.

In addition, evaluator was blinded to the treatment groups in the analysis for method 1 and 3.

Page 123: An Injectable Gelatin-Based Conjugate Incorporating EGF

123

5.2. Overall goal and hypotheses of this chapter

The overall goal of this chapter was to develop a reliable MRI protocol to visualize the

implanted Gtn-HPA matrix in the spinal cord cavity and distinguish the Gtn-HPA matrix from other

surrounding constituents without the use of contrast agents. In this chapter, the focus was to

achieve the goal in an ex vivo setting, which would lay the foundation for developing and

optimizing an in vivo protocol to visualize the Gtn-HPA matrix in longitudinal studies.

The chapter-specific hypotheses are:

1) MRI signatures (T1 or T2) of the Gtn-HPA matrix are different from the surrounding healthy

tissue, edema and hemorrhage to generate contrast using MRI

2) MRI generated images would be of high resolution and will exhibit enough contrast between

the Gtn-HPA matrix and other constituents to localize the defect and extent of the lesion.

3) Parameters such as Gtn-HPA matrix volume and lesion volume would be calculated using

directed analysis of the images generated from the protocol

4) The protocol would be feasible to be implemented in in vivo assessment for future studies

Page 124: An Injectable Gelatin-Based Conjugate Incorporating EGF

124

5.3. Approach 1: T2-weighted MRI

5.3.1. Introduction: Sagittal T2-weighted MRI is the most common MRI sequence ordered

by the physicians in the hospitals for spinal cord injury patients either in the acute or chronic

phase. T2 is one the key MRI signature that is utilized to generate contrast between different

constituents. Transverse relaxation time (T2) is the time constant that measures the time taken

for spinning protons to be completely out of phase perpendicular to the main magnetic field [17].

Essentially, T2 is a transverse magnetization decay time constant (Figure 5.1). In addition to the

key anatomical abnormalities, T2-weighted MRI images show bright signal for CSF, edema and

hemorrhage in the cord after the injury [18]. Given that Gtn-HPA matrix would have high water

content, we hypothesize that the matrix would give strong signal at the injury site in T2-weighted

MRI but would have different signal intensity for edema so that Gtn-HPA matrix can be

distinguished four weeks post injection.

5.3.2. Methods:

5.3.2.1. Surgery procedure and experimental Design: The surgical procedure was

analogous to the one described in detail in Chapter 6 for inducing a 2 mm left hemi-resection SCI,

except that in this procedure, a 1 mm left hemi-resection defect was created. Briefly, 12 Lewis

female rats were induced a left 1 mm hemi-resection SCI at T8 level following a laminectomy. 10

µl of 8 wt% Gtn-HPA matrix with EGF was injected into the cavity after achieving homeostasis

(Table 5.1). All the animal care guidelines post-operation were followed as per the

recommendations of Veterans Affairs Institute of Animal Care and Use Committee (VA IACUC)

under the protocol 335-J. They were administered antibiotics and analgesics as per the

requirements. Animals were sacrificed four weeks after implantation of the Gtn-HPA matrix by

transcardial perfusion of heparinized saline and 4% PFA both kept at 4°C.

Figure 5.1. (Left) T2 relaxation decay of the MRI signal in materials with a short and long T2 (Right) Signal intensities of various materials as observed in a T2 weighted MRI. Reference – mriquestions.com

Page 125: An Injectable Gelatin-Based Conjugate Incorporating EGF

125

Experimental treatment Group Study size (n) Duration of experiment

Control – no injection N 6 4 weeks

8wt% Gtn-HPA with EGF GE8 6 4 weeks

Table 5.1 Experimental groups investigated in the animal study with female Lewis rats

5.3.2.2. Sample preparation and acquisition of MRI: The spinal column was harvested

after the sacrifice and fixed in 4% PFA for 3 days at 4°C. Then, excess musculature around the

spinal column was removed. In order to prepare for ex vivo scan, the samples were immersed in

Fomblin oil to limit tissue dehydration and to generate better contrast. T2-weighted imaging was

performed at the Small Animal Imaging Laboratory (SAIL) at Brigham and Women’s Hospital

using Bruker Biospec 70/30 7.0T magnet with 30 cm bore horizontal magnet. 15 coronal MRI

slices were acquired with long TR (3500 ms) and long TE (80 ms) to produce a T2-weighted

image. Slice thickness was 250 µm and the scan time was 29 min per sample. The matrix size

was selected to be 192 x 192 with FOV being 25.6 x 19.2 mm. The imaging parameters were set

to optimize the visualization of Gtn-HPA gel.

5.3.2.3. Image analysis: RAW images were reconstructed using the inbuilt Bruker

software and were exported as DICOM files, which were later analyzed using ImageJ (NIH,

Bethesda, MD) to calculate the lesion volume using ROI manager. Study size was determined

based on the power calculation to determine significant a 30% difference in a selected outcome

variable with a 15% standard deviation, and with α=0.05 and β=0.05. The statistical significance

between the experimental groups was determined by one-tailed unpaired t-test because the

implantation of biomaterial has shown to reduce the lesion volume. Statistical analysis was

performed with GraphPad Prism (8.0).

5.3.3. Results: The protocol using approach 1 was able to generate high resolution T2

weighted images of the fixed spinal cord four weeks post injection. The resolution of the images

was measured to be 75 µm, considered to be high for similar applications. The images were able

to discern anatomical structures such as bones of the spinal column, vertebral discs, surrounding

musculature, and gray and white matter in the healthy cord in both the experimental groups –

control [N] and Gtn-HPA with EGF [GE8] (Figure 5.2 & 5.3). Importantly, we were able to identify

the injury location based on the expected intense signal due to edema and possibly the Gtn-HPA

gel. However, the images did not distinguish the implanted Gtn-HPA gel in the defect area from

the surrounding edema or fibrous connective tissue. Moreover, the morphology of the injury site

looked the same in terms of signal intensities looked the same in both groups. The images were

not able to distinguish between the fluid-filled injury site in controls [N] and Gtn-HPA matrix filled

Page 126: An Injectable Gelatin-Based Conjugate Incorporating EGF

126

Figure 5.3 Representative 8 MRI images for a rat from Gtn-HPA with EGF [GE8] group clearly showing the location of the injury and the extent of the secondary damage. However, this method was not able to exclusively visualize the Gtn-HPA gel in the injury site.

Figure 5.2 Representative 8 MRI images for a rat from Control [N] group clearly showing the location of the injury and the extent of the secondary damage.

Figure 5.4 Representative H&E images of the (left) control group and (right) Gtn-HPA with EGF group. Gtn-HPA matrix was present four weeks after implantation in the injury site Despite the presence of Gtn-HPA gel, the T2-weighted MRI images failed to show any differences between the injury sites of both groups. Scale bars = 300 µm, Images acquired using 10x objective.

Gtn-HPA

Page 127: An Injectable Gelatin-Based Conjugate Incorporating EGF

127

injury sites in the GE8 group. Histology of the Gtn-HPA with EGF group [GE8] did show the

presence of the Gtn-HPA confirming the presence of the gel in the injury site four weeks post

implantation (Figure 5.4, right). Nevertheless, these T2-weighted images were able to show the

extent of the injury and secondary damage. Therefore, we measured the lesion volume of each

sample using 9 best slices using ROI manager in ImageJ. The lesion area was manually marked

in each of the 9 slices. Total lesion area was calculated by adding the lesion areas of all 9 slices.

The lesion volume was calculated by multiplying the total determined lesion area with the slice

thickness of 250 µm. This was done for all the animals in both the treatment groups. An example

of the marking of the lesion area is shown in the figure 5.5, left. Quantitative analysis showed that

implantation of the Gtn-HPA matrix reduced the lesion volume and extent of secondary damage

compared to the controls (one-tailed t-test, p = 0.057) as shown in figure 5.5, right. Although the

p-value is not entirely less than 0.05, it is very close suggesting that the difference could be of

significance with more animals and with better method to measure the lesion area.

Lesion Volume = (Total lesion area) x 250µm (slice thickness)

Figure 5.5 (left) Lesion area from each of 9 slices were measured using ImageJ by manually drawing the area. The lesion volume was calculated by multiplying the total lesion area with the slice thickness (right) Implantation of the Gtn-HPA matrix reduced the lesion volume and extent of secondary damage compared to the controls, one-tailed t-test, p = 0.057. Although p<0.05 is not true, it is suggestive of real difference in the lesion volumes between the two groups. The result should be validated with other methods for future studies.

Page 128: An Injectable Gelatin-Based Conjugate Incorporating EGF

128

5.4. Approach 2: Magnetic resonance elastography (MRE)

5.4.1. Introduction: Magnetic resonance elastrography (MRE) is novel MRI technique to

compare the biomechanical properties of the soft tissues in a non-invasive and non-destructive

manner. Fundamentally, the working mechanism of MRE is a three-step process. First, shear

waves of a certain frequency are introduced in the sample and image acquisition is performed

simultaneously. Second, the MRI sequence is sensitized to detect the perturbations in the tissue

due to the shear waves. Third, the displacements measured by the MRI acquisition is then

converted into stiffness to generate elastogram of the region of interest in the tissue [19]. The

differentiating feature that generates these elastograms is the displacement affect that the shear

waves have on a soft tissue compared to on a hard tissue. Given that many pathologies involve

significant change in the stiffness of the tissues, MRE is emerging as a diagnostic method to

detect and characterize various disease states [20]. Physicians have used MRE to study chronic

liver diseases to detect liver fibrosis that significantly hardens the liver tissue. Singh et al reported

that MRE had high degree of accuracy in determining the different stages of liver fibrosis

compared to the gold standard of liver biopsy [21]. Considering the promise of MRE in staging

the liver disease for prognosis, MRE has been used in experimental and clinical setting to evaluate

its utility for assessing the degree of fibrosis in the spleen, kidney, breast and brain due to various

pathologic processes [19]. Interestingly, glioblastoma tumors have also been characterized by

MRE and were reported to have lower stiffness owing to softer necrotic tissues [22]. Therefore,

this approach looked very promising to visualize Gtn-HPA gel in the spinal cord tissue because

the stiffness of the Gtn-HPA would be different from the affected tissue and the healthy tissue.

The stiffness maps of the spinal cord could allow us to measure the gel volume and lesion volume.

5.4.2. Methods: Phantom work involving the 8wt% Gtn-HPA gel was performed at the Small

Animal Imaging Laboratory (SAIL) at Brigham and Women’s Hospital using Bruker Biospec 70/30

7.0T magnet with 30 cm bore horizontal magnet. 8 wt% Gtn-HPA gel block was fabricated as per

the dimensional requirement of MRE (15 mm x 15 mm x 5 mm). MRI magnet was customized

with special setup to introduce the shear waves into the Gtn-HPA sample during acquisition

(Figure 5.6). Images were reconstructed as per the method developed by Dr. Samuel Patz. Due

to technical issues such as breathing motion artifacts, CSF motion artifacts, and difficulty in

designing the apparatus for rats, no live animal MRE was performed.

Page 129: An Injectable Gelatin-Based Conjugate Incorporating EGF

129

5.4.3. Results: After overcoming several technical difficulties with the dimensions of gel block

because of its viscoelastic nature, we were able to scan 8wt% Gtn-HPA gel block on its own. MRE

elastogram was generated post reconstruction as shown in the figure 5.7. The elastogram shows

minor heterogeneity in the stiffness values in the block due to fabrication of the Gtn-HPA in such

large volume (~ 2000 µl).

Figure 5.6 Set-up for 8wt% Gtn-HPA gel block designed for MRE. (Left) The shear wave generator was connected to the coil apparatus by a yellow cylindrical tube to deliver shear waves to (right) the coil apparatus (close-up view) physically attached to the gel sample

Figure 5.7 Reconstructed images of the 8wt% Gtn-HPA gel block of size 15 x 15 x 5 mm. (Left) Normal proton density map of the block (right) MRE generated elastogram of the block indicating some minor heterogeneity in the stiffness of the Gtn-HPA gel block. Typically, Gtn-HPA would not be injected in such large volumes and it is homogeneous for the smaller volumes.

Page 130: An Injectable Gelatin-Based Conjugate Incorporating EGF

130

5.5. Approach 3: T1-weighted Inversion Recovery (T1IR)

5.5.1. Introduction: T1-weighted imaging is widely used in the clinical radiology practice to

detect soft tissue contrasts in the MRI to study the fine anatomical structures and pathologic

abnormalities of the brain. In addition, T1-weighted imaging is a preferred technique in cases

where gadolinium-based contrast agents are used. T1-weighted imaging is widely used in the

experimental research to measure the spinal cord volume affected due to an injury or pathological

process [23, 24]. T1-weighted MRI uses short TR and TE times to generate the image. However,

the image contrast of the T1-weighted images is poor and therefore, various modifications of T1-

weighted sequences are utilized by the clinicians and researchers. One of the most common

method is T1-weighted fluid-attenuated inversion recovery (FLAIR). FLAIR has very long TR and

TE to attenuate the signal from CSF. FLAIR provides a remarkable contrast between the edema,

normal parenchyma and normal tissues in the brain [24, 25]. Other T1-weighted methods for

various applications include T1 mapping for cardiovascular imaging, short tau inversion recovery

(STIR), T1-weighted inversion recovery (T1IR) and double inversion recovery (DIR). T1IR is

extremely useful for differentiating the elements that have different T1 times but not different

enough to generate enough contrast by T1-weighted imaging. Image contrast can be generated

by selecting an appropriate inversion time (IT) without compromising the signs of MR signal

intensities [26].

Inversion recovery sequences have shown to be better in determining spinal cord lesions

and discriminate between the edema and healthy tissue in 1980s [27, 28]. The fundamental

working mechanism of the inversion recovery involves selectively nulling the signal for specific

tissues such as fluid, CSF or fat. IR sequences introduce a 180° inversion RF pulse before any

MRI sequence after a pre-defined time known as inversion time (IT). This inversion RF pulse

inverts the direction of the longitudinal magnetization followed by continuation of the T1 recovery

(Figure 5.8). The optimization of the image involves carefully choosing an inversion time such that

the signal from the unwanted element such as CSF, fat or fluid can be nullified to generate an

image with good contrast for the tissues of interest [29]. Clinically, inversion recovery sequences

such as FLAIR are useful in differentiating between intracranial cystic lesions such as arachnoid

cyst and epidermoid cyst. Arachnoid cyst contains CSF meaning that it will be nullified by FLAIR

but epidermoid cyst containing cellular debris is not suppressed using the same sequences [30].

In the context of our work, T1 relaxation values for Gtn-HPA matrix and surrounding spinal

cord tissue was different prompting us to utilize T1-weighted imaging. However, conventional T1-

weighted imaging did not produce good contrast between the Gtn-HPA matrix and the

Page 131: An Injectable Gelatin-Based Conjugate Incorporating EGF

131

surrounding tissue. T1-weighted inversion recovery (T1IR) worked the best to achieve contrast to

scan rat spinal cords ex vivo four weeks after Gtn-HPA matrix injection.

5.5.2. Preliminary phantom work - Relaxometry

5.5.2.1. Methods: A systematic approach was taken to measure the T1 and T2

signatures of the normal spinal cord tissue and Gtn-HPA matrix. 600 µl of 8wt% Gtn-HPA gel was

fabricated. A healthy spinal cord tissue was isolated from the same anatomical location as that of

the injured site in our rodent model. Since our goal was to develop an ex vivo protocol to image

fixed spinal column in our animal studies, both the samples were fixed in 4% PFA overnight at

4°C. Bruker 7T Biospin 70/30 magnet with 30 cm bore was used to scan both the samples. The

MRI scanning was performed at Small Animal Imaging Laboratory (SAIL) at Boston Children’s

Hospital, Boston, MA. Bruker Relaxometry protocol was employed to measure the T1 and T2

properties of the Gtn-HPA gel and healthy spinal cord. Specific T1 and T2 were measured by

drawing an ROI in white matter of the cord, gray matter of the cord and the Gtn-HPA gel.

Figure 5.8. Comparison of the MRI sequence for conventional spin-echo and inversion recovery sequence highlighting the 180° pulse preceded by 90° pulse resulting in the inversion of the signal. Specific Inversion time can be selected to nullify signal from CSF, fat or tissue of interest. Reference – mriquestions.com

Page 132: An Injectable Gelatin-Based Conjugate Incorporating EGF

132

5.5.2.2. Results: The relaxometry protocol was able to measure the T1 and T2 values

for gray matter, white matter and Gtn-HPA gel using the MRI images generated from the scan

(Figure 5.9). T1 and T2 values for the white matter of the fixed cord were 963 and 32 ms

respectively. T1 and T2 values for the gray matter of the fixed cord were 920 and 39.8 ms

respectively. Lastly, T1 and T2 values for the fixed Gtn-HPA were 1840 and 36.5 ms respectively.

The phantom work gave us a satisfying explanation of the failure of T2-weighted MRE to

distinguish the Gtn-HPA from the spinal cord because the T2 values of the cord and Gtn-HPA

were found to be very similar. As far as T1 values are concerned, the values for gray and white

matter of the cord were very similar. However, the T1 value of the Gtn-HPA gel was double the

value the both elements of the spinal cord (Table 5.2). This result was very encouraging for us

because such difference in the T1 value could be exploited to distinguish between the Gtn-HPA

gel and the spinal cord tissue.

Table 5.3 T1 and T2 values of spinal cord tissue and Gtn-HPA gel

White matter Gray matter Gtn-HPA gel

T1 963 ms 920 ms 1840 ms

T2 32 ms 39.8 ms 36.5 ms

Table 5.2 T1 and T2 values of the white matter and gray matter of the cord and Gtn-HPA gel block. Although T2 values were similar, T1 value of the Gtn-HPA was twice that of the spinal cord.

Proton density image T1 Relaxometry T2 relaxometry

Figure 5.9 T1 and T2 relaxometry MRI images used to measure the T1 and T2 values of the 4% PFA fixed spinal cord (top) and Gtn-HPA gel (bottom).

Page 133: An Injectable Gelatin-Based Conjugate Incorporating EGF

133

5.5.3. Preliminary phantom work – T1IR

5.5.3.1. Methods: Based on the findings from the relaxometry phantom work, we

implemented T1-weighted imaging with various TR values but were unsuccessful in developing a

reliable protocol that achieved our objective. Next, we attempted T1-weighted Inversion recovery

with different inversion times (IT). Specimens used were the same as prepared in the ‘Methods’

section 5.5.2.1. Both the spinal cord and the Gtn-HPA gel block were scanned in the same plane.

Imaging parameters were set as follows: TR = 3500 ms, TE = 8.5 ms, and eight different IT = 300,

400, 450, 500, 600, 700, 800 and 900 ms. Multiple Inversion times were studied to find the

inversion time at which spinal cord signal could be completely nullified to exclusively scan only

the Gtn-HPA gel. Likewise, we were interested in finding an inversion time at which Gtn-HPA

signal could be completely nullified to exclusively scan only the spinal cord tissue.

5.5.3.2. Results: T1IR sequence generated individual images in the same plane

containing the spinal cord and the Gtn-HPA gel for eight different inversion times. The goal was

to find inversion time for which the signal from the spinal cord tissue could be nullified to

exclusively observe the Gtn-HPA gel signal. We found that with IT = 450 ms, the signal from the

spinal cord was completely nullified despite it being present in the same place. However, there

was strong signal from the Gtn-HPA block. In contrast, with IT = 800 ms, the signal from the Gtn-

300 ms 400 ms 450 ms 500 ms

600 ms 700 ms 800 ms 900 ms

Figure 5.10 T1-weighted inversion recovery (T1IR) images of the spinal cord and the Gtn-HPA gel in the same plane with different inversion times labeled above the respective image. It is evident that with IT = 450 ms, we only observe the signal from the Gtn-HPA gel nullifying the spinal cord tissue while with IT =800 ms, only the signal from the spinal cord tissue is observed nullifying the Gtn-HPA gel.

Page 134: An Injectable Gelatin-Based Conjugate Incorporating EGF

134

HPA gel was completely nullified while there was strong signal from the spinal cord tissue (Figure

5.10). Proton-density image was acquired on the same plane of scanning to state that both the

Gtn-HPA gel and the spinal cord was present in the plane during the scan with T1IR (Figure 5.11).

Figure 5.11. Proton-density image of the spinal cord and the Gtn-HPA gel during T1IR image acquisition confirming the presence of both samples in the same plane of acquisition

Page 135: An Injectable Gelatin-Based Conjugate Incorporating EGF

135

5.5.4. Ex vivo spinal cord T1IR - Protocol 1.0

5.5.4.1. Methods – surgical procedure and MRI acquisition: Based on the very

encouraging findings from the preliminary phantom work using T1-weighted inversion recovery

with 450 ms IT to visualize the Gtn-HPA gel and 800 ms IT to visualize the spinal cord tissue, we

optimized our protocol for ex vivo scanning of the harvested spinal columns with implanted Gtn-

HPA gels four weeks post injection.

Surgery procedure and sample preparation: The surgical procedure was identical to the

one described in detail in Chapter 6 for inducing a 2 mm left hemi-resection SCI. The same

animals were scanned using this protocol after sacrifice. Briefly, Lewis female rats were induced

a left 2 mm hemi-resection SCI at T8 level following a laminectomy. 20 µl of 8 wt% Gtn-HPA

matrix with EGF was injected into the cavity after achieving homeostasis. All the animal care

guidelines were followed as per the recommendations of Veterans Affairs Institute of Animal Care

and Use Committee (VA IACUC) under the protocol 378-J. They were administered antibiotics

and analgesics as per the requirements. Animals were sacrificed four weeks after implantation of

the Gtn-HPA matrix by transcardial perfusion with heparinized saline and 4% PFA both kept at

4°C. The spinal column was harvested after the sacrifice and fixed in 4% PFA for 3 days at 4°C

followed by 3 days in PBS at 4°C. Then, excess musculature around the spinal column was

removed. In order to prepare for ex vivo scan, the samples were immersed in Galden oil to limit

tissue dehydration and to generate better contrast.

Imaging parameters: T1IR imaging was performed at Small Animal Imaging Laboratory

(SAIL) at Boston Children’s Hospital in Boston using Bruker Biospec 70/30 7.0T magnet with 30

cm bore horizontal magnet. 16 coronal MRI slices each were acquired with long TR (3500 ms)

and short TE (8.5 ms) for both inversion times, 450 ms and 800 ms. Slice thickness was 220 µm

and the scan time was 34 mins per inversion time per sample totaling 68 min per sample. The

matrix size was selected to be 192 x 192 with FOV being 20 x 20 mm. All MRI safety precautions

and standards were followed as per the requirements set by the SAIL at Children’s Hospital.

5.5.4.2. Results and discussion: After localizing the injury area in the center of field of

view and setting proper geometry using quick localizer scans, the T1IR images at inversion time

of 450 ms showed strong signal indicating the presence of the Gtn-HPA gel in the animals injected

with the Gtn-HPA matrix. The final resolution of the images was 110 µm. The signal from the

spinal cord tissue was completely nullified. Simultaneously, T1IR images at inversion time of 800

ms showed signal from the fixed spinal cord tissue but the area of the Gtn-HPA gel was completely

dark. Images in the same geometrical location from both inversion times were complementary to

each other meaning the area with bright signal with 450 ms IT was dark in images with inversion

Page 136: An Injectable Gelatin-Based Conjugate Incorporating EGF

136

time of 800 ms (Figure 5.12). In the control groups where no Gtn-HPA gel was injected, the injury

site showed no signal at IT of 450 ms demonstrating that the protocol is sensitive only to the Gtn-

HPA matrix and not edema, hemorrhage or fibrous interstitial tissue. Gtn-HPA gel volume four

weeks post injury was calculated using these MRI images in ImageJ (results are reported in

Chapter 6). Moreover, Gtn-HPA signal was observed in one hemisphere of the spinal cord as

expected in this hemi-resection model. Future studies should employ the final protocol in

complete-resection models and assess whether the signal from the Gtn-HPA matrix is seen in

both hemispheres at the injury location.

Overall, Protocol 1.0 was able to exclusively show the presence of the Gtn-HPA matrix in

the injury site in one hemisphere of the cord using T1IR imaging sequence with IT of 450 ms.

Images with IT of 800 ms served as the evidence that indeed the signal from the Gtn-HPA was

nullified as predicted by the phantom work describing the signal intensities of the spinal cord and

Gtn-HPA matrix. However, one ambiguity arose from the CSF signal seen in the subarachnoid

spaces along the spinal cord. The signal intensity of the CSF was high in images with IT = 450

ms. Although the anatomical location of the CSF was different from the location of Gtn-HPA

matrix, our goal was to optimize the protocol to reduce ambiguity between the signal from the

Gtn-HPA matrix and CSF. Another improvement in the Protocol 1.0 that we aimed was to increase

the signal-to-noise ratio to produce better quality images.

Page 137: An Injectable Gelatin-Based Conjugate Incorporating EGF

137

IT = 800 ms IT = 450 ms

Gtn-HPA gel only

[G8]

Gtn-HPA with EGF

[GE8]

Control

[N]

Figure 5.22 Representative images from T1IR imaging of (top & middle) two rats showing the presence of the Gtn-HPA matrix in half of the injury site at IT = 450 ms while nullifying the Gtn-HPA signal with IT = 800 ms. However, (bottom) control group rat showing no signal at the injury site at both inversion times indicating that the protocol is sensitive to detecting the presence of Gtn-HPA matrix four weeks post injection

Page 138: An Injectable Gelatin-Based Conjugate Incorporating EGF

138

5.5.5. Ex vivo spinal cord T1IR - Protocol 2.0

5.5.5.1. Methods – surgical procedure and MRI acquisition: The goal was Protocol

2.0 was to improve upon the Protocol 1.0 by reducing ambiguity in the T1IR images between the

Gtn-HPA matrix and the CSF signal in the images with IT of 450 ms. The surgical procedure,

tissue preparation for MRI and the imaging parameters remain the same as described in the

‘Methods’ section of Protocol 1.0. Given that CSF has a very long T1, we theorized that longer

inversion times could nullify the CSF signal. Hence, we did T1IR imaging of the fixed spinal cords

at higher inversion times of 900, 950, 1000, 1100, 1200 and 1300 ms. In order to improve the

signal-to-noise ratio, number of averages to generate the image was doubled to theoretically

increase the signal-to-noise ratio by 40%. The scan time was increased significantly as a result

of doubling the number of averages. In the final Protocol 2.0, each sample was scanned with

three inversion times 450 ms, 800 ms and 1100 ms. The scan time of each inversion time was 68

mins as a result of doubling the averaging, increasing the total sample acquisition time to roughly

3 hours 30 mins.

5.5.5.2. Results and discussion: Out of all the inversion times investigated, 1100 ms

was chosen to scan all the future fixed spinal cords. Although T1IR images at 1100 ms did not

completely nullify the CSF signal, we were able to observe a distinctive feature between Gtn-HPA

and CSF. There was no signal from the CSF while the signal from Gtn-HPA was visible in all the

preliminary samples with Gtn-HPA implanted. Therefore, IT of 450 ms showed the presence of

Gtn-HPA and CSF, IT of 800 ms showed the presence of spinal cord tissue and IT of 1100 ms

showed the presence of spinal cord tissue and the gel but not CSF. Additional inversion time of

1100 ms cleared the ambiguity between the signal from the Gtn-HPA gel and CSF (Figure 5.13).

Moreover, doubling the number of averages during acquisition improved the quality of the images.

Protocol 2.0 provided reliable and definitive distinction of the Gtn-HPA gel from the surrounding

tissue and CSF. This protocol would serve as the foundation to optimize an in vivo protocol to

study the location of Gtn-HPA, volume of Gtn-HPA and the features of the surrounding tissue

near the injury site over the duration of the animal experiment. One of the most interesting

application of in vivo imaging would be to generate a degradation profile of Gtn-HPA matrix in

each rat by calculating the Gtn-HPA gel volume each week post injection. Future studies should

involve in vivo imaging of the rats from various SCI models to gain real-time insights into the

progression of the injury and the Gtn-HPA characteristics in a longitudinal manner.

Page 139: An Injectable Gelatin-Based Conjugate Incorporating EGF

139

IT = 800 ms IT = 450 ms IT = 1100 ms

G

tn-H

PA

+ E

GF

[G

E8

]

Gtn

-HP

A +

EG

F [G

E1

2]

Gtn

-HP

A g

el o

nly

[G

8]

Figure 5.33 Representative Ex vivo T1IR images from groups in which Gtn-HPA was injected in a left 2 mm hemi-resection cavity with three different inversion times (ITs) – 800 ms, 450 ms and 1100 ms. IT of 450 ms showed presence of Gtn-HPA gel and CSF while IT of 800 ms showed the presence of spinal cord tissue with dark regions where CSF and Gtn-HPA is present. IT of 1100 ms produced images that reliably differentiates between the Gtn-HPA gel and CSF because CSF appears dark while Gtn-HPA in the injury site shows greater signal.

Page 140: An Injectable Gelatin-Based Conjugate Incorporating EGF

140

5.6. Summary and discussion

The long-term goal of the work presented in this chapter was to develop tools and methods

to study the presence of implanted biomaterial (Gtn-HPA in our work) and the progression of the

injury in SCI animal models in a longitudinal manner. Conventionally, histological and

immunofluorescent techniques are used to study the general morphology and the specific cellular

composition with identifying the ECM. However, these conventional methods are invasive,

destructive for the sample and rely heavily on the expertise of the investigator. Technological

advances in the MRI techniques has made MRI an enticing technique to study various

pathological conditions. For example, the extent of remyelination after an intervention is studied

with the help of advanced MRI methods such as diffusion tensor imaging (DTI), Magnetization

transfer (MT) imaging and myelin water fraction imaging [31]. Few pre-clinical studies have also

developed parameters acquired in MRI to study and progression of the spinal cord injury in

patients with Multiple Sclerosis [32]. However, only few experimental SCI studies implanting

biomaterials at the location of the SCI employ MRI to study the progression of the disease, track

the presence of implanted biomaterial and correlate the histological findings with MRI [9, 10].

Zhang et al studied the effect of fibrin-based hydrogel in regenerating axons using tractography.

They reported that the hydrogel was able to promote the restoration of fibers post SCI using DTI

[33]. In the context of our work, our first goal was to develop an MRI-based protocol that reliably

discriminates between the Gtn-HPA matrix, surrounding spinal cord tissue and other pathologic

features. Three approaches were taken to achieve the goal – T2-weighted imaging, magnetic

resonance elastography and T1-weighted inversion recovery.

T2-weighted imaging was successful in discerning the healthy spinal cord tissue from that

of affected injury site (Figure 5.3). Intense bright signal marked the affected spinal cord tissue at

the injury site. It is known that high water content elements would show high signal intensity in a

T2-weighted image. Since the injury area is marked by the presence of fluid-filled cavities, necrotic

tissue and Gtn-HPA in the case of gel-injected groups, images indeed show high intensity signal

at the injury site. However, the radiologic signatures were the same for both control and gel-

injected groups suggesting that T2-weighted imaging was not capable of differentiating between

the Gtn-HPA matrix from the fluid-filled cavity and other pathologic elements at the injury site. The

presence of Gtn-HPA was confirmed with the histological processing of the same samples (Figure

5.4). Nonetheless, we were able to measure lesion volume from the T2-weighted MRI images.

This was a significant improvement over the conventional histopathology techniques, where we

would have to use all the sections of the sample to calculate the lesion volume. Using MRI, we

were able to obtain this critical information in a non-invasive and non-destructive method. The

Page 141: An Injectable Gelatin-Based Conjugate Incorporating EGF

141

quantitative analysis of the lesion volume showed that the tissue affected by the spinal cord injury

is reduced when Gtn-HPA treatment is injected after inducing a SCI (Figure 5.6). This is consistent

with the results from pre-clinical studies that claim the neuroprotective effects due to the

implantation of biomaterial into the cavity post SCI [34, 35]. Automated algorithms to identify the

injured area using MRI images with reduced slice thickness would be an interesting approach that

future studies can take, where image processing would be the central theme of the project. Given

that the slice thickness was comparatively high (250 µm), the measured lesion area is assumed

to be the same throughout the 250 µm, which might not be a good assumption. Therefore, smaller

slice thickness images can be taken to obtain a more accurate measurement of the lesion volume.

Overall, T2-weighted images gave critical insights into the injury area but was not able to

distinguish the Gtn-HPA matrix from the surrounding tissue.

Another promising approach that we attempted was the novel magnetic resonance

elastography (MRE). It has been clinically used to detect tissues with abnormal stiffness and have

applications in liver fibrosis, amyloid plaques, and studying tumors. We wanted to generate an in

vivo elastogram to differentiate Gtn-HPA matrix, interstitial fibrous tissue and surrounding healthy

tissue by exploiting the differences between their stiffness. MRE can be followed every week to

generate weekly elastogram to study the change in the modeling of the tissue properties at the

injury site and to study the changes in the volume of the Gtn-HPA matrix over the course of the

experiment. Our primary goal was to calibrate the MRE with our Gtn-HPA matrix such that the

values can be used in the reconstruction to accurately identify the presence of Gtn-HPA matrix.

Initial phantom work generated an elastogram for the Gtn-HPA matrix block, but the

reconstruction values were less reliable owing to heterogeneity in the sample (Figure 5.7). The

heterogeneity is attributed to the volume of the block fabricated to meet the dimensional

requirements of the MRE. There were other technical difficulties as well due to which we were not

able to work on the in vivo implementation of MRE to study the injury site after SCI. These include

anticipated motion artifacts from breathing and cardiac movement, motion artifacts from moving

CSF near the cord and the size of rat spinal cord. Moreover, the setup of the MRE was 3D printed

for mouse imaging in the collaboration facility. However, MRE is a promising technique that can

be applied well to our application. A project leader expert and interested primarily in the imaging

aspect of this work could gather the resources and expertise required to bring MRE to fruition.

The third approach with T1-weighted inversion recovery was the breakthrough protocol by

which we achieved our goal of developing a reliable method to distinguish Gtn-HPA gel from the

surrounding spinal cord tissue. The systematic approach was beneficial to developing T1IR based

protocol. First, T1 and T2 values of the spinal cord and the Gtn-HPA was measured. Our T2

Page 142: An Injectable Gelatin-Based Conjugate Incorporating EGF

142

values for the white and gray matter of the fixed spinal cords were measured to be 32 and 39.8

ms respectively (Table 5.2). The measured values agreed with previously T2 reported values for

fixed spinal cord tissues within assumed experimental error. Carvlin et al reported T2 values of

white and gray matter as 44 and 46 ms respectively for fixed rat spinal cords [36]. Our T1 values

for the white and gray matter of the fixed spinal cords were measured to be 963 and 920 ms

respectively. Carvlin et al reported T1 values of white and gray matter as 250 and 253 ms

respectively for fixed rat spinal cords using 1.5T magnet. According to the physics of MRI, T1

values increase as the field strength of the magnet increases. Doubling the magnet strength could

increase the T1 value by 25% [37]. Considering this phenomenon, the T1 values agree with the

modified reported values. Interestingly, the T2 values of the Gtn-HPA gel and the spinal cord were

very similar. It explained the reason behind not being able to discern the Gtn-HPA gel from the

surrounding tissue using T2-weighted imaging in approach 1. However, T1 value of the Gtn-HPA

gel was twice compared to the spinal cord tissues. This difference in T1 values proved to be the

turning point to achieve our objective. Phantom work using T1IR with inversion time (IT) of 450

ms completely nullified the spinal cord tissue signal while the Gtn-HPA signal intensity was high.

T1IR with IT of 800 ms did exactly the opposite. This systematic phantom work laid the foundation

for achieving our objective with ex vivo scanning of fixed spinal columns containing the defect

site. T1IR of fixed spinal columns showed Gtn-HPA exclusively with IT of 450 ms and spinal cord

tissue exclusively with IT of 800 ms (Figure 5.12). Moreover, the anatomical location of the Gtn-

HPA presence was also reassuring – the signal intensity was highest in the injury site only in one

half of the spinal cord as expected in a hemi-resection SCI rats in which Gtn-HPA was injected.

This protocol 1.0 was an important feat because it gave irrefutable evidence regarding the

presence of the Gtn-HPA gel. With these images, we were able to calculate the Gtn-HPA gel

volume present in the injury site four weeks post injection (Gel volume data are reported in

Chapter 6). However, T1IR images with IT of 450 ms also exhibited high signal intensity from the

CSF fluid in the subarachnoid space. Although the anatomical location of the Gtn-HPA gel and

CSF was different, our next objective was to reduce or remove ambiguity between the signal from

the Gtn-HPA gel and CSF. Protocol 2.0 included T1IR imaging with an additional IT of 1100 ms

that showed that the signal from CSF was still nullified while there was appreciable signal intensity

from the area depicting the presence of Gtn-HPA gel in the images with IT of 450 ms. This

additional set of images with IT of 1100 ms distinguished the Gtn-HPA from CSF both regarding

MRI signal intensity and anatomical location (Figure 5.13). Protocol 2.0 also produced clearer

images with double averaging increasing the totals can time to 3 hours 30 mins per sample.

Page 143: An Injectable Gelatin-Based Conjugate Incorporating EGF

143

We think ex vivo T1IR imaging with three ITs 450 ms, 800 ms and 100 ms coupled with

T2-weighted imaging of the spinal columns would give the most comprehensive view of the

characteristics and metrics of the Gtn-HPA gel and the injury site in our hemi-resection models.

Future studies could focus on validating the metric measured by MRI images with the histological

findings to assess if MRI metrics can be reliable and accurate predictors of the assessment

parameters conventionally obtained from the histopathological findings. For example, a validation

study comparing the Gtn-HPA gel volume measured with histology and MRI can be performed.

Moreover, future study can also investigate the percent volume acquired by the Gtn-HPA gel in

the lesion volume and can compare similar metrics measured by histological quantitative analysis.

If MRI seems to be a reliable and reproducible method to measure critical parameters such as

gel volume and lesion volume, in vivo implementation of the Protocol 2.0 should be implemented

to track the Gtn-HPA gel and its volume over the duration of the experiment in the longitudinal

manner. MRI could play a valuable role in giving important insights into the progression of the

injury, degradation behavior of Gtn-HPA gel and potential regenerative response in experimental

SCI models. Correlation between MRI metrics and functional recovery should be studied to

assess if MRI metrics could predict the functional recovery in the rats after SCI. It would drastically

improve the clinical role of MRI in predicting functional recovery in response to implantable

injectable biomaterials such as Gtn-HPA in promoting repair and functional recovery in patients.

Page 144: An Injectable Gelatin-Based Conjugate Incorporating EGF

144

5.7. References

1. Goldberg, A.L. and S.M. Kershah, Advances in imaging of vertebral and spinal cord injury. The journal of spinal cord medicine, 2010. 33(2): p. 105-116.

2. Perovitch, M., H. Wang, and S. Perl, The evolution of neuroimaging of spinal cord injury patients over the last decade. Paraplegia, 1992. 30(1): p. 39-42.

3. Kulkarni, M.V., et al., 1.5 tesla magnetic resonance imaging of acute spinal trauma. Radiographics, 1988. 8(6): p. 1059-82.

4. Bozzo, A., et al., The role of magnetic resonance imaging in the management of acute spinal cord injury. Journal of neurotrauma, 2011. 28(8): p. 1401-1411.

5. Sliker, C.W., S.E. Mirvis, and K. Shanmuganathan, Assessing cervical spine stability in obtunded blunt trauma patients: review of medical literature. Radiology, 2005. 234(3): p. 733-9.

6. Gonzalez-Lara, L.E., et al., In vivo magnetic resonance imaging of spinal cord injury in the mouse. J Neurotrauma, 2009. 26(5): p. 753-62.

7. Hsiao, J.K., et al., Magnetic nanoparticle labeling of mesenchymal stem cells without transfection agent: cellular behavior and capability of detection with clinical 1.5 T magnetic resonance at the single cell level. Magn Reson Med, 2007. 58(4): p. 717-24.

8. Fraidakis, M., et al., High-resolution MRI of intact and transected rat spinal cord. Exp Neurol, 1998. 153(2): p. 299-312.

9. Sundberg, L.M., J.J. Herrera, and P.A. Narayana, In vivo longitudinal MRI and behavioral studies in experimental spinal cord injury. Journal of neurotrauma, 2010. 27(10): p. 1753-1767.

10. Narayana, P.A., et al., Endogenous recovery of injured spinal cord: longitudinal in vivo magnetic resonance imaging. J Neurosci Res, 2004. 78(5): p. 749-59.

11. Zakszewski, E., et al., Diffusion imaging in the rat cervical spinal cord. Journal of visualized experiments : JoVE, 2015(98): p. 52390.

12. Zhang, J., M. Aggarwal, and S. Mori, Structural insights into the rodent CNS via diffusion tensor imaging. Trends in neurosciences, 2012. 35(7): p. 412-421.

13. Talbott, J.F., et al., Diffusion-Weighted Magnetic Resonance Imaging Characterization of White Matter Injury Produced by Axon-Sparing Demyelination and Severe Contusion Spinal Cord Injury in Rats. J Neurotrauma, 2016. 33(10): p. 929-42.

14. Karajanagi, S.S., et al., Assessment of canine vocal fold function after injection of a new biomaterial designed to treat phonatory mucosal scarring. Ann Otol Rhinol Laryngol, 2011. 120(3): p. 175-84.

15. Austin, J.W., et al., The effects of intrathecal injection of a hyaluronan-based hydrogel on inflammation, scarring and neurobehavioural outcomes in a rat model of severe spinal cord injury associated with arachnoiditis. Biomaterials, 2012. 33(18): p. 4555-64.

16. Rogosnitzky, M. and S. Branch, Gadolinium-based contrast agent toxicity: a review of known and proposed mechanisms. Biometals, 2016. 29(3): p. 365-76.

17. Radue, E.W., et al., Introduction to Magnetic Resonance Imaging for Neurologists. Continuum (Minneap Minn), 2016. 22(5, Neuroimaging): p. 1379-1398.

18. Kumar, Y. and D. Hayashi, Role of magnetic resonance imaging in acute spinal trauma: a pictorial review. BMC musculoskeletal disorders, 2016. 17: p. 310-310.

19. Mariappan, Y.K., K.J. Glaser, and R.L. Ehman, Magnetic resonance elastography: a review. Clinical anatomy (New York, N.Y.), 2010. 23(5): p. 497-511.

20. Low, G., S.A. Kruse, and D.J. Lomas, General review of magnetic resonance elastography. World J Radiol, 2016. 8(1): p. 59-72.

Page 145: An Injectable Gelatin-Based Conjugate Incorporating EGF

145

21. Singh, S., et al., Diagnostic performance of magnetic resonance elastography in staging liver fibrosis: a systematic review and meta-analysis of individual participant data. Clin Gastroenterol Hepatol, 2015. 13(3): p. 440-451.e6.

22. Schregel, K., et al., Characterization of glioblastoma in an orthotopic mouse model with magnetic resonance elastography. NMR Biomed, 2018. 31(10): p. e3840.

23. Kim, G., et al., T1- vs. T2-based MRI measures of spinal cord volume in healthy subjects and patients with multiple sclerosis. BMC neurology, 2015. 15: p. 124-124.

24. Hori, M., et al., T1-Weighted Fluid-Attenuated Inversion Recovery at Low Field Strength: A Viable Alternative for T1-Weighted Intracranial Imaging. American Journal of Neuroradiology, 2003. 24(4): p. 648-651.

25. Lee, J.K., et al., Usefulness of T1-weighted image with fast inversion recovery technique in intracranial lesions: comparison with T1-weighted spin echo image. Clin Imaging, 2000. 24(5): p. 263-9.

26. Hou, P., et al., Phase-sensitive T1 inversion recovery imaging: a time-efficient interleaved technique for improved tissue contrast in neuroimaging. AJNR Am J Neuroradiol, 2005. 26(6): p. 1432-8.

27. Bydder, G.M. and I.R. Young, MR imaging: clinical use of the inversion recovery sequence. J Comput Assist Tomogr, 1985. 9(4): p. 659-75.

28. White, S.J., et al., Use of fluid-attenuated inversion-recovery pulse sequences for imaging the spinal cord. Magn Reson Med, 1992. 28(1): p. 153-62.

29. Saranathan, M., et al., Physics for clinicians: Fluid-attenuated inversion recovery (FLAIR) and double inversion recovery (DIR) Imaging. J Magn Reson Imaging, 2017. 46(6): p. 1590-1600.

30. Lee, E.K., et al., Importance of Contrast-Enhanced Fluid-Attenuated Inversion Recovery Magnetic Resonance Imaging in Various Intracranial Pathologic Conditions. Korean J Radiol, 2016. 17(1): p. 127-41.

31. Harlow, D.E., J.M. Honce, and A.A. Miravalle, Remyelination Therapy in Multiple Sclerosis. Frontiers in neurology, 2015. 6: p. 257-257.

32. Gilli, F., et al., High-Resolution Diffusion Tensor Spinal Cord MRI Measures as Biomarkers of Disability Progression in a Rodent Model of Progressive Multiple Sclerosis. PLoS One, 2016. 11(7): p. e0160071.

33. Zhang, Z., et al., Effect of hierarchically aligned fibrin hydrogel in regeneration of spinal cord injury demonstrated by tractography: A pilot study. Sci Rep, 2017. 7: p. 40017.

34. Novikov, L.N., et al., A novel biodegradable implant for neuronal rescue and regeneration after spinal cord injury. Biomaterials, 2002. 23(16): p. 3369-76.

35. Wang, Y., et al., Effective improvement of the neuroprotective activity after spinal cord injury by synergistic effect of glucocorticoid with biodegradable amphipathic nanomicelles. Drug Deliv, 2017. 24(1): p. 391-401.

36. Carvlin, M.J., et al., High-resolution MR of the spinal cord in humans and rats. AJNR Am J Neuroradiol, 1989. 10(1): p. 13-7.

37. Maubon, A.J., et al., Effect of Field Strength on MR Images: Comparison of the Same Subject at 0.5, 1.0, and 1.5 T. RadioGraphics, 1999. 19(4): p. 1057-1067.

Page 146: An Injectable Gelatin-Based Conjugate Incorporating EGF

146

Chapter 6:

Investigating the effects of Gtn-HPA

matrix with EGF and SDF-1α in a T8 2

mm hemi-resection rat model of SCI

after four weeks

Page 147: An Injectable Gelatin-Based Conjugate Incorporating EGF

147

6.1. Introduction and clinical motivation

Traumatic spinal cord injury (SCI) is a life changing condition that causes severe disability,

dire reduction in the quality of life and high mortality [1]. Although the rehabilitation efforts assist

in improving the lives of the patients, the root cause of the problem – degeneration of the neuronal

tissue – is not addressed by the current clinical management. Therefore, various experimental

SCI models have been developed to understand the complex pathophysiological progression

after the injury and investigate the effects of potential therapies that could assist in unravelling a

promising treatment of SCI. Plethora of research approaches have shown promise in promoting

tissue repair and functional recovery after the SCI using the SCI models [2]. Various strategies

for SCI tissue repair such as pharmacological agents, exogeneous cells and growth factors have

been evaluated in pre-clinical studies and clinical studies for spinal cord repair and regeneration.

For the acute intervention, various pharmacological agents such as sodium channel blocker, anti-

NOGO antibodies, Rho-ROCK inhibitors, minocycline and myelin signal inhibitors to promote

neuroprotection and regenerative response after the injury [3]. Various cells such as induced

pluripotent stem cells (iPSCs), neural stem cells (NSCs), olfactory ensheathing cells (OECs),

schwann cells, and bone-marrow derived mesenchymal stem cells (bMSCs) have been implanted

after inducing SCI to investigate potential functional recovery in animals and regeneration of the

spinal cord tissue [4]. However, implanting cells in the spinal cord can be challenging and costly.

In addition, non-autologous cells also increases the risk of rejection worsening the injured tissue

after SCI [5]. Therefore, another approach is to deliver growth factors either in suspension,

intrathecally or in situ using a biocompatible biomaterial. The goal is to chemotactically recruit the

endogenous stem cells and glial cells to the injury site before the impenetrable glial scar is formed.

Therefore, these interventions are made either in the subacute or the intermediate phase of injury

progression after the SCI [6, 7]. Various neurotrophic factors such as nerve growth factor (NGF),

brain-derived neurotrophic factor (BDNF), fibroblast growth factor (FGF), neurotrophin-3 (NT-3)

and epidermal growth factor (EGF) have been tested in animals models and have shown

promising neuroprotection, tissue remodeling and functional recovery after SCI [8, 9].

There are several factors that directed the choice of the hemi-resection experimental

animal model – clinical motivation, the type of therapy being tested, and the anticipated effect of

the therapy. In our work, we were motivated to propose a therapy for small cavitary lesions in the

spinal cord. Incomplete SCI accounts for 68% spinal cord injuries [10]. These cavities do not

generally affect the entire spinal cord in the axial plane. Since our goal was not aimed at studying

a specific function or a spinal tract, we employed left hemi-resection rat model of SCI. Clinically,

this is comparable to Brown-Sequard syndrome in which the patients suffer from ipsilateral upper

Page 148: An Injectable Gelatin-Based Conjugate Incorporating EGF

148

motor neuron deficiencies and contralateral loss of temperature and pain [11]. Hemi-resection

models are widely used and considered to be ideal for studying axonal regeneration as a result

of growth-promoting trophic molecules, biomaterials implantation or cellular implantation [12].

Hemi-resection produces a standardized defect of preferred size and confirms the complete break

in the neuronal circuitry at the location of the cord. In addition, since the resection is made only

one side of the spinal cord, the contralateral side can serve as a control. The most important

advantage in the context of our work is that hemi-resection model creates a reproducible defect

in which biomaterial of a same volume can be injected along with growth-promoting factors such

as EGF and SDF-1α. Building on the results of the evaluation of the Gtn-HPA gel in a complete

resection model reported in Chapter 4 and a reliable MRI-based method T1-weighted inversion

recovery (T1IR) to visualize Gtn-HPA gel in the injury site as reported in Chapter 5, we were

interested in investigating the effects of delivering EGF and SDF-1α using injectable Gtn-HPA gel

in an 2 mm left lateral hemi-resection rat model. In this chapter, learnings from the Chapter 4 and

5 are considered to optimize the Gtn-HPA matrix to develop a second generation Gtn-HPA gel

for injecting into the injury site. In addition to testing the first generation 8wt% Gtn-HPA matrix,

we have evaluated a second generation 12wt% Gtn-HPA matrix to reduce the degradation rate

of the Gtn-HPA gel. Moreover, we have improved on the surgical parameters that could affect the

gelation and containment of the Gtn-HPA gel in the cavity created by a hemi-resection SCI.

Specifically, we have devised a better technique to achieve hemostasis after inducing SCI so that

Gtn-HPA gel can be injected into a ‘clean’ cavity devoid of blood. In addition, we modified the

dural replacement after injection of the Gtn-HPA matrix to better contain the biomaterial in the

cavity site. There were few procedural improvements done to improve the quality of histological

assessment of the spinal cord section. First, fixed cords were immersed in subsequent 15% and

30% sucrose solution to remove water molecules to prevent the formation of crystals during flash-

freezing, which can significantly alter the cellular architecture. Second, cryosection temperature

was optimized to reduce the horizontal and vertical tears in the section. Third,

immunohistochemistry protocol was optimized for several biomarkers, especially regarding the

antigen retrieval protocols. We believe that these methods and the second generation Gtn-HPA

gel is better equipped to improve the reparative and regenerative response in the injury site post

SCI. The results of the studies are promising and consistent throughout various assessment

parameters. Although the animal model and the proposed therapy has few limitations, findings

from this chapter provides a fundamental proof of principle demonstrating the benefit that local

delivery of EGF delivered using Gtn-HPA matrix provides in promoting functional recovery and

reparative response post SCI.

Page 149: An Injectable Gelatin-Based Conjugate Incorporating EGF

149

6.2. Overall goal and hypotheses of this chapter

The overall clinical goal of the project was to develop an injectable polymer incorporating

therapeutic molecules for the treatment of cavitary lesions that can result from SCI. The

supposition was that improved clinical outcome will result from enabling neural tissue to fill the

cavitary defect and to establish neural connections with the surrounding cord tissue. Moreover,

injection of the polymer into an acute lesion may mitigate the secondary damage.

The goal of this chapter was to evaluate a novel gelatin-based conjugate capable of

undergoing covalent cross-linking after being injected as a liquid. The gelatin hydrogel (Gtn-HPA)

will be tested alone and incorporating epidermal growth factor (EGF) and/or stromal cell-derived

factor - 1α (SDF-1α) in a T8 2 mm hemi-resection rat model to assess the tissue healing and

functional recovery four weeks post injection of the treatment.

Chapter-specific hypotheses:

We hypothesize that the gelatin-based gel (Gtn-HPA) incorporating chemoattracting

molecules, EGF and SDF-1α, leads to recruitment of endogenous neural stem cells into the gel-

filled lesion. Moreover, we hypothesize that the acute injection of the gel improves functional

recovery in the rats after four weeks. Specific hypotheses are:

1) Gtn-HPA is biocompatible and does not elicit major inflammatory response.

3) 8% and 12% Gtn-HPA gel are present in the injury site after four weeks and is permissive of

cellular migration (e. g. stiffness would not be very high to impede cellular migration)

4) EGF can recruit endogenous neural stem cells (NSCs) to the injury site while SDF-1α can

recruit endogenous glial cells to the gel-filled lesion.

5) Better functional recovery is observed as a result of Gtn-HPA injection with or without the EGF

and SDF-1α. The functional outcome is better compared to the complete resection model. There

is greater preservation of spared tissue in Gtn-HPA injected groups

6) SDF-1α enhances the response in all assessment metrics compared to EGF.

7) MRI aids visualization of Gtn-HPA matrix and quantification of the gel volume in the lesion

Page 150: An Injectable Gelatin-Based Conjugate Incorporating EGF

150

6.3. Methods

6.3.1. Gtn-HPA gel fabrication: Raw material of gelatin-hydroxyphenylpropionic acid

conjugate was obtained from Dr. Motoichi Kurisawa at Agency of Science, Technology and

Research (A*STAR), which was prepared as per the protocol described previously [13]. Briefly,

Gtn–HPA conjugates were prepared by a general carbodiimide/active ester-mediated coupling

reaction in distilled water. 3.32 g of HPA was dissolved in 250 ml of mixture prepared with 3:2

ration of distilled water and N,N-dimethylformamide (DMF) respectively. To this, 3.20 g of N-

hydroxysuccinimide and 3.82 g of 1-ethyl-3-(3-dimethylaminopropyl)-carbo-diimide hydrochloride

were added. The reaction mixture was stirred at room temperature for 5 hours, and the pH of the

mixture was kept at 4.7. Then, 150 ml of 6.25 wt% Gtn aqueous solution was added to the reaction

mixture and continuously stirred overnight at room temperature at pH 4.7. Subsequently, the

solution was dialyzed against 100 mM sodium chloride solution for 2 days, 25% ethanol in distilled

water for 1 day and distilled water for 1 day, successively. The purified solution was lyophilized to

obtain the Gtn–HPA conjugate.

8 and 12 wt% Gtn-HPA gel was fabricated as per the procedure described in Chapter 3

with few modifications. Specifically, 12% w/v (to prepare 8% Gtn-HPA gel) Gtn-HPA and 18% w/v

(to prepare 18% Gtn-HPA gel) conjugate solution was prepared fresh under sterile conditions and

kept on ice. Recombinant rat EGF (Peprotech, Cat# 400-25) and recombinant rat SDF-1α

solutions (Peprotech, Cat# 400-32A) were aliquoted as per the recommended concentration of

1.0 mg/ml and stored at -20°C. Depending upon the experimental group, EGF and SDF-1α were

added to the freshly prepared 12% or 18% Gtn-HPA conjugated solution at room temperature.

The final dosage of the EGF and SDF-1α was 6 µg/rat. To this mixture, 2 µl of Horseradish

peroxidase (HRP) (Wako Pure Chemical Industries, Japan) was added with final HRP

concentration of 0.1 U/ml. Then, 2 µl of H2O2 (Sigma Aldrich) was added to give final H2O2

concentration of 6.8 mM. Addition of H2O2 initiated the cross-linking and gelation process. The

solution was mixed thoroughly for homogenous gelation and immediately injected into the cavity

before the solution turned into a hydrogel with an appropriate gelation time.

6.3.2. Animal surgical procedure and Gtn-HPA injection: Adult female Lewis rats

(Charles River Laboratories, Willington, MA) weighing 201-225 grams were chosen as the rodent

species in this study based on the previous work in our laboratory. Animals were given at least

48 hours to acclimatize before the major survival surgery of inducing a SCI. Before the surgery,

animals were singly housed in a cage as per the housing requirements after the SCI to minimize

the pain. Survival surgery was performed as per the approval from Veterans Affairs Boston

Healthcare system Institutional Animal Care and Use Committee (IACUC) on the protocol #378-

Page 151: An Injectable Gelatin-Based Conjugate Incorporating EGF

151

J. Animal surgeries were performed at the Animal Research Facility (ARF) located in the research

building of the Veterans Affairs Jamaica Plains, MA campus. Animal surgery was performed by

the skilled and experienced Dr. Hu-ping Hsu, Department of Orthopedic surgery, Brigham and

Women’s Hospital. The hemi-resection model is based on the work done by Cholas et al [14].

After recording their weight immediately before the surgery, Lewis rats were anesthetized using

2-4 % isoflurane in oxygen through a nose mask for induction followed by 1-3 % isoflurane for

maintenance of anesthesia, during the surgical procedure. On average, the procedure lasted for

30-45 mins and isoflurane was administered for the entire duration. After confirming that the rat

is unconscious, rats were placed in the prone position on a flat operating board and all the limbs

were gently fixed using rubber bands to stabilize the rat. The hair on the back of the rat were

shaved and the skin was thoroughly cleaned with betadine. A 4-cm skin incision was made on

the thoracic spine portion. The skin and musculature overlying the thoracic spine was incised

along the midline and retracted laterally from the vertebral column. A laminectomy was performed

removing the dorsal aspect of T7-T10 using small bone rongeurs and microscissors, exposing

approximately 10 mm of spinal cord. Bleeding from the muscle was controlled by Gelfoam®®

(Pfizer, Andover, MA). The dura was incised in the midline and the dorsal spinal artery was

coagulated with a bipolar cautery. A sterile 2 mm plastic template was placed in the center of the

exposed spinal cord and a lateral hemi-resection of the spinal cord was created by making two

lateral cuts (2 mm apart) using scalpel #11 using the 2 mm template as guide (Figure 6.1A). After

the lateral cut, a midline cut was made by scalpel between the two lateral cut using dorsal spinal

vein as the anatomical marker for the midline. A 2 mm gap was created by removing the tissue

between the two lateral cuts. Bleeding was controlled using Gelfoam® placed into the defect site.

The Gelfoam® was kept for a total of 15 minutes changing the Gelfoam® every 5 minutes to

achieve better hemostasis (Figure 6.1B). Once hemostasis was achieved, the Gelfoam® was

removed from the defect and Gtn-HPA gel that was freshly fabricated immediately before the

injection was injected to fill the cavity (Figure 6.1C). In control animals [N], no treatment was

injected into the defect. However, 8 wt% Gtn-HPA gel was injected in the experimental groups as

per the dosage mentioned above. In 8wt% Gtn-HPA gel only [G8] group, Gtn-HPA gel was

fabricated and injected without addition of any factors. However, appropriate growth factors were

added to the Gtn-HPA conjugate solution prior to the gelation for local delivery of EGF and SDF-

1α [GE8 and GES8]. 12 wt% Gtn-HPA was injected into the Gtn-HPA with EGF group [GE12] and

Page 152: An Injectable Gelatin-Based Conjugate Incorporating EGF

152

Gtn-HPA with EGF with ‘tucked-in’ fascia group [GE12T]. In all the groups that involved injecting

the Gtn-HPA gel into the cavity, 20 µl of the 8 wt% or 12 wt% Gtn-HPA gel, with or without the

factors were injected into the cavity using a pipette such that gelation will occur rapidly within in a

Figure 6.1 Top – Schematic of the surgical procedure, Gtn-HPA injection and dural replacement placement (A) 2 mm left hemi-resection cavity created by laminectomy at T8 level shown with the 2 mm template. It is important to note that bleeding covers the entire cavity (B) Gelfoam is used as a hemostatic agent and is kept for a total of 15 minutes in the cavity to achieve hemostasis (C) Gtn-HPA is immediately injected into the cavity (D) Cavity is covered with collagen membrane as dural replacement followed by suturing the overlying musculature

A B

C D

A B

C

Gtn-HPA gel

Dural replacement (autologous fascia)

T8 Left lateral hemi-resection Acute Gtn-HPA injection

Inject 20 µl 2 mm

Histological plane of analysis

Page 153: An Injectable Gelatin-Based Conjugate Incorporating EGF

153

few minutes to confine the biomaterial to the lesion site. The injury was closed after visual

confirmation of complete gelation and filling of the 2 mm hemi-resection defect with the Gtn-HPA

gel. Following gel injection, a thin collagen membrane was placed over the top of the spinal cord

wound as dural replacement to maintain positioning of the scaffold and separate the wound site

from the overlying tissue (Figure 6.1D). Following treatment and placement of the collagen

membrane, the overlying musculature was closed using 4-0 vicryl sutures (Johnson and Johnson,

Sommerville, NJ) and the skin was closed with wound clips.

6.3.2.1. Improvement in the dural replacement technique: Based on our observations

regarding the presence of the Gtn-HPA gel in the injury cavity after four weeks in both complete

transection and hemi-resection model, we brainstormed various ways to better contain the Gtn-

HPA gel in the cavity. As a result, a modification was made in the placing of the dural replacement

membrane. Specifically, the collagen fascia was trimmed small enough only to cover the spinal

cord at the injury site (Figure 6.2A). In addition, the fascia membrane was ‘tucked-in’ around the

perimeter of the exposed cord (Figure 6.2B). The goal was to fully contain the gel in the cavity by

reducing the likelihood of the Gtn-HPA being dislodged from the cavity due to movement of the

rats after the surgery.

6.3.3. Experimental design: There were six experimental groups evaluated in this study

with the seventh group being the control group that received no injection treatment for the total

duration of four weeks (Table 6.1). The animals were randomly assigned to the groups and were

blinded to the surgeon during injection. I was blinded to the treatment group and animals were

identified using their serial number only. Various assessment parameters were employed to

evaluate the effect of the proposed therapy including the behavioral assessment,

histomorphometric analysis and immunohistochemical analysis. The time course of the

Figure 6.2 Modification in the dural replacement technique for group 12wt% Gtn-HPA with EGF (A) Collagen fascia dural replacement was trimmed small enough only to cover the spinal cord area (B) Schematic showing the ‘tucking-in’ of the fascia membrane around the surface of the spinal cord to better contain the Gtn-HPA in the cavity (C) Smaller fascia placed around the spinal cord cavity site

A B C

Page 154: An Injectable Gelatin-Based Conjugate Incorporating EGF

154

experiment describing the SCI injury, injection of the treatment, behavioral testing, manual

bladder expression and sacrifice are shown in figure 6.3. BBB scoring and the quantification

analyses were performed blinded to the treatment group.

Experimental Treatment Group Duration Study size (n)

Control – No injection N 4 weeks 6

EGF in suspension E 4 weeks 5

8 % Gtn-HPA gel only G8 4 weeks 6

8 % Gtn-HPA + EGF GE8 4 weeks 6

8 % Gtn-HPA + EGF + SDF-1α GES8 4 weeks 6

12 % Gtn-HPA + EGF GE12 4 weeks 6

12 % Gtn-HPA + EGF - ‘tucked-in’ dura GE12T Day 0 1

12 % Gtn-HPA + EGF - ‘tucked-in’ dura GE12T Day 1 3

12 % Gtn-HPA + EGF - ‘tucked-in’ dura GE12T Day 7 3

12 % Gtn-HPA + EGF - ‘tucked-in’ dura GE12T 4 weeks 6

Table 6.1 Experimental design to evaluate Gtn-HPA along with EGF and SDF-1α in a T8 2-mm left hemi-resection SCI rodent model

6.3.4. Animal care post-surgery and bladder function evaluation: Post-surgery, animals

were carefully brought to the pre-OR room and placed in a heated cage with constant supply of

oxygen at 1 L/min to maintain body temperature and breathing. Rats were subcutaneously

injected with warm 3 ml of lactated Ringer’s solution to compensate for the blood loss during the

Day 0 7 14 21 28

SCI & Gtn-HPA

injection Sacrifice

Bladder expression

Behavioral

assessment

Figure 6.3 Time course of the experiments detailing inducing SCI, Gtn-HPA injection, skin wound clip removal, manual bladder expression, behavioral testing by open field locomotor test and sacrifice after four weeks. Post-sacrifice harvested spinal columns were scanned with MRI. After MRI, spinal cord samples were prepared for histological and immunofluorescent analysis.

Skin Clip removal

Page 155: An Injectable Gelatin-Based Conjugate Incorporating EGF

155

surgery. They were subcutaneously administered Meloxicam (1 mg/kg) as an analgesic and

Cephazolin (35 mg/kg) as an antibiotic to manage pain and infections, respectively. Rats were

monitored for breathing every ten minutes for an hour after surgery, every half an hour subsequent

two hours and every hour for the last three hours. Continuous oxygen supply was maintained until

they were transported to their housing room in the animal research facility. 6 hours post-surgery,

their bladders were expressed manually and placed in the cage with wood-chip bedding with free

access to food and water. Dry food was kept on the flood of the cage in the initial days for easier

access. Rats were kept in a light/dark automated cycle in the housing room as per the research

facility guidelines. Due to the loss of bladder control, their bladders were manually emptied using

Crede’s technique twice a day throughout the duration of the experiment with utmost care to

minimize the pain (Figure 6.3). Bladder volume was recorded in three categories – small, medium

and large during manual expression. In addition, the first day any rat showed some sign of return

of bladder function was noted in the records. At the end of four weeks, frequency of animals in

each treatment group was categorized into three urine output volumes – small, medium and large.

Moreover, the number of animals in each group that showed at least partial return of the bladder

function was noted. The effect of treatment on the bladder function will be quantified using Chi-

square test.

Rats were subcutaneously administered analgesic Meloxicam (1 mg/kg) once every 22-

24 hours for four days post-surgery while antibiotic Cephazolin (35 mg/kg) was administered twice

in a day for a week post-surgery. They were also administered lactated Ringer’s solution as

necessary based on their hydration status. One week after the surgery, skin wound clips were

removed after confirming the suturing of the skin (Figure 6.3). Animals were anesthetized using

2-3% isoflurane for painless removal of the wound clips. They were weighed once a week and

were noted of any flinching behavior and other concerns. The record was meticulously kept in the

animal housing room as per the VA animal research facility requirements.

6.3.5. Behavioral assessment – BBB scale: In order to assess the functional recovery of

the animals after the surgery, open-field locomotor test was conducted once every week until

sacrifice (Figure 6.3). Animals walked freely on the table with absorbent chute. Their movement,

especially of the hindlimbs, was digitally video recorded for three minutes. The animals were

scored using the well-established 21-point Basso-Beattie-Bresnahan (BBB) locomotor rating

scale on live animals and recorded videos [15]. The scale ranged from 0 (complete paralysis) to

21 (normal gait) as listed in Table 6.2. The evaluator was blinded to the treatment group during

Page 156: An Injectable Gelatin-Based Conjugate Incorporating EGF

156

scoring of the both hindlimbs. Separate BBB score was assigned to the left and the right hindlimb

each week.

BBB score Description

0 No observable hindlimb movement

1 Slight movement (<50% of the range of motion) of one or two of the three joints – hip, knee and ankle

2 Extensive movement (>50% of range of motion) of one joint and slight movement of another joint

3 Extensive movement of the two joints

4 Slight movement of all three hindlimb joints

5 Slight movement of the two joints and extensive movement of the third joint

6 Extensive movement of two joints and slight movement of the third joint

7 Extensive movement of all three hindlimb joints

8 Sweeping movement (rhythmic movement of all three hindlimb joints) without any weight bearing (elevation of the hindlimb during movement) by the hindlimbs

9 Plantar paw placement with some weight bearing in stationary phase or minor weight bearing with dorsal stepping of the paw

10 Occasional (<50% of the times) weight bearing with plantar placement of the paw without any hindlimb-forelimb (HL-FL) coordination

11 Occasional weight bearing with plantar placement of the paw and occasional FL-HL coordination

12 Frequent (>50% of the times) weight wearing with plantar placement of the paw with occasional FL-HL coordination

13 Frequent weight bearing with the plantar placement of the paw and frequent Hl-FL coordination

14 Consistent (~100%) weight-bearing with plantar steps, consistent FL-HL coordination with rotational movement of the paw during stepping

15 Consistent FL-HL coordination with consistent dragging of the toes and rotational movement of the paw during stepping

16 Consistent FL-HL coordination with frequent dragging of the toes and rotational movement of the paw during stepping

17 Consistent FL-HL coordination with occasional dragging of the toes and rotational movement of the paw during stepping

18 Consistent FL-HL coordination with no dragging of the toes with parallel paw position during liftoff and initial contact

19 Consistent FL-HL coordination, no dragging of the toes, parallel paw position during liftoff and initial contact and tail is down part or all the time

20 Consistent FL-HL coordination, no dragging of the toes, parallel paw position during liftoff and initial contact and tail is consistently up

21 Coordinated gait, consistent toe clearance, predominant paw position is parallel during movement and tail consistent up with full trunk stability

Table 4.2 BBB scale implemented to assess the functional recovery for the rats after the SCI with the BBB score and corresponding functional description of the hindlimb

Adapted from the Basso et al, J Neurotrauma 1995. 12(1): p. 1-21.

6.3.6. Animal sacrifice and harvesting of the spinal column: After four weeks, animals

were sacrificed by transcardial perfusion. Rats were administered a dose of 150 mg/kg of sodium

pentobarbital and anesthetized. The level of anesthesia was checked by pinching their tail.

Quickly, a thoracotomy was performed to expose the heart and the needle attached to the

Page 157: An Injectable Gelatin-Based Conjugate Incorporating EGF

157

peristaltic pump was inserted into the left ventricle and into the ascending aorta. The pump was

turned on at the speed of 4 initially and then the right atrium was quickly cut to allow open system

for the flow of perfusion solution. The speed of the pump was increased to 7 after cutting the right

atrium. A total of 120 ml of heparinized saline (20 U/ml) was circulated through the animal,

followed by 120 ml of 4% paraformaldehyde (PFA) without introducing any bubbles into the line.

Both the solutions were kept on ice during the perfusion. After completing the sacrifice, the spinal

column along with the musculature was cut from the neck region to pelvic region. The entire spine

was placed in 50 ml 4% PFA at 4°C overnight. The following day, the spine was trimmed of the

overlying musculature while leaving the vertebral bone and fascia over the defect intact. The

trimmed spine was kept in 4% PFA for additional 2 days at 4°C. After three days, they were kept

in 60 ml of PBS for three days at 4°C.

6.3.7. MRI ex vivo scan of the spinal column: The excess musculature around the spinal

column was removed to prepare them for the MRI (Figure 6.4A). In order to prime the spinal

column for ex vivo scan, the samples were immersed in Galden oil to limit tissue dehydration and

to generate better contrast in a 15 ml Eppendorf tube. Galden oil is considered proton free and is

expected to generate no signal during MRI acquisition. Galden oil is not shown to affect the post-

processing methods including tissue embedding and histological processing [16]. T1-weighted

inversion recovery (T1IR) imaging was performed at Small Animal Imaging Laboratory (SAIL) at

Boston Children’s Hospital in Boston using Bruker Biospec 70/30 7.0T magnet with 30 cm bore

horizontal magnet. Once the sample was positioned in the coil, quick localization scans were

executed to assess the location of the injury. The images were used as a reference to define the

location and the geometry for T1IR scanning. Then, either protocol 1.0 or protocol 2.0 was

implemented depending upon the group as mentioned in the detailed imaging parameters below

for both protocols. All MRI safety precautions were followed as per the requirements of the SAIL

at Children’s Hospital, Boston, MA.

Imaging parameters Protocol 1.0: The MRI protocol 1.0 was implemented as described in Chapter

5 to scan N, G8, GE8 and GES8 groups as outlined in Table 6.2. Briefly, 16 coronal MRI slices

each were acquired with long TR (3500 ms) and short TE (8.5 ms) for both inversion times, 450

ms and 800 ms. Slice thickness was 220 µm and the scan time was 34 mins per inversion time

per sample totaling 68 min per sample. The matrix size was selected to be 256 x 256 with FOV

set to 20 x 20 mm.

Imaging parameters Protocol 2.0: The MRI protocol 2.0 was implemented as described in Chapter

5 to scan GE12 and GE12T groups as outlined in Table 6.2. Briefly, 16 coronal MRI slices each

Page 158: An Injectable Gelatin-Based Conjugate Incorporating EGF

158

were acquired with long TR (3500 ms) and short TE (8.5 ms) for three inversion times, 450 ms,

800 ms and 1100 ms. In addition, the averaging during acquisition was doubled than that of

protocol 1.0. Slice thickness was 220 µm and the scan time was 68 mins per inversion time per

sample totaling 3 hours 30 mins per sample. The matrix size was set to be 256 x 256 with FOV

set to 20 x 20 mm.

6.3.8. Gtn-HPA gel volume determination using MRI images: As described in Chapter

5, T1-weighted inversion recovery with inversion time (IT) of 450 ms was able to specifically

visualize the Gtn-HPA present in the injury site four weeks post injection. T1IR images with IT of

800 ms and 1100 ms were used to make accurate judgement about the presence of the Gtn-HPA

gel. Therefore, all the MRI slices obtained by T1IR imaging with IT of 450 ms were used to

calculate the area signaling the presence of Gtn-HPA in ImageJ using ROI manager feature for

each rat. Given that the slice thickness was known to be 220 µm, Gtn-HPA gel volume was

calculated by multiplying the total Gtn-HPA area in all the slices with the slice thickness (220 µm).

The quantification of the Gtn-HPA area was conducted blindly with respect to the treatment group.

6.3.9. Cryoprotection, tissue embedding and cryosection: After ex vivo scans, spinal

columns were wiped to remove as much Galden oil as possible. Then, spinal columns were

immersed in PBS overnight at 4°C. Next day, the spinal cord was carefully isolated and harvested

from the spinal column using bone rongeurs, surgical scissors, and a scalpel (Figure 6.4B). Then,

spinal cords were carefully placed in 15% sucrose solution in a 6 well plate until the cord sank.

Figure 6.4 (A) Gross image of the spinal column prepared for MRI ex vivo scan (B) Harvesting the spinal cord from the spinal column (C) gross image of the spinal cord showing the defect site filled with Gtn-HPA gel (D) spinal cord kept in Tissue Tek OCT compund prior to flash-freezing in isopentane cooled in liquid nitrogen (E) Frozen block with spinal cord mounted on the object holder in the cryostat for coronal cryosectioning

A B

C D E

Page 159: An Injectable Gelatin-Based Conjugate Incorporating EGF

159

Subsequently, the cords were kept in 30% sucrose solution in a 6 well plate until the cord sank.

Usually, it took a day for the cords to sink in 15% sucrose solution while it took roughly two days

for the cords to sink in 30% sucrose solution. After the cryoprotection method, the spinal cords

with the injury site were embedded in Tissue Tek OCT compound solution kept in a rectangular

tinfoil mold (Figure 6.4CD). The spinal cords were then flash-frozen in the OCT compound

solution using isopentane cooled in liquid nitrogen. The frozen block with proper labeling with the

animal information was kept on dry ice until they were stored at -80°C. The blocks were serially

cryosectioned using Leica Cryostat CM 3050 as shown in figure 6.4E (Leica Biosystems,

Wetzelar, Germany). 30 µm thick coronal sections of the spinal cord were cut and were mounted

on a pre-treated Superfrost Gold glass microscope slides (Fisher Scientific, Hampton, NH).

6.3.10. Histology and histomorphometric analysis: Tissue sections were warmed to room

temperature and then baked at 60 °C for 3 hours prior to histological staining to improve the

adhesion to the glass slides. The tissue sections roughly in the middle of the defect in coronal

plane were chosen to best represent the tissue morphology for each rat. Then, after they reached

room temperature, tissue sections were rehydrated in PBS for 45 mins to dissolve the OCT on

the slides. Tissue sections were stained with Hematoxylin and Eosin (H&E) method to visualize

the tissue morphology and the implanted gel at the injury site. After being rehydrated in PBS, the

tissue sections were permeabilized in cold methanol for 10 min followed by immersion in water

for 5 min, rinsed in PBS for 10 min and again immersed in water for 5 min. Then, they were stained

with hematoxylin Gill #2 solution for 4 min followed by thorough wash in running tap water for 5

min. Then, the tissue sections were decolorized with 3 quick drops in acid alcohol and then were

thoroughly washed in running tap water for 5 min. Furthermore, they were stained with eosin for

40 seconds. Tissue sections were dehydrated with 100% alcohol (2 x 3 min) and processed in

xylene (2 x 3 min). Coverslip was mounted on using cytoseal and then the tissue sections were

air dried in a chemical hood overnight.

The brightfield microscopy images were acquired using Olympus microscope. In addition,

Zeiss Axioscan.Z1 automated slide scanner was used to scan the entire tissue sections (Zeiss,

Germany). For each animal, gel area and cellular migration area immediately next to the gel

islands were calculated using ROI manager in ImageJ (NIH, Bethesda, MD). Two analysis

methods were used to validate the findings. The first method (Method 1 – total area method)

measured the total Gtn-HPA gel area in all the gel islands and total cellular migration immediately

around all the gel islands in the injury cavity were measured in ImageJ using ROI manager tool.

The second method (Method 2 – standardized ROI method) measured the Gtn-HPA gel area and

cellular migration immediately surrounding the gel in three random standardized 0.3 mm2 ROI at

Page 160: An Injectable Gelatin-Based Conjugate Incorporating EGF

160

the interface using ImageJ in three randomly chosen images. Average gel area and migration

area from each animal was calculated using both the methods and compared among the groups.

All histological analyses were conducted randomly and blindly with respect to the treatment group.

6.3.11. Spared tissue area and injury area quantification: Immediately after left hemi-

resection, the right hemisphere of the cord was kept intact. However, secondary damage after

the injury could lead to sparing of various degrees of the tissue in the right hemisphere of the

spinal cord. Using the H&E images, cross-sectional area of the spared tissue and cross-sectional

area of total lesion was calculated using ROI manager in ImageJ. Percent spared tissue and

percent injury area was calculated from these values for each animal. Next, percent spared tissue

and average BBB scores were correlated. The quantification of the spared tissue for each rat was

conducted randomly and blindly with respect to the treatment group.

6.3.12. Immunofluorescent staining and analysis: Tissue sections were stained with a

biochemical marker to gain more specific details regarding the cellular population in the injury site

and immediately surrounding the gel islands. Tissue sections were warmed to room temperature

and then baked at 60 °C for 3 hours prior to immunofluorescent staining to improve the adhesion

to the glass slides. Primary antibodies, dilutions and their antigen retrieval were performed as

shown in Table 6.3. Then, after they reached room temperature, tissue sections were rehydrated

in PBS for 45 mins to dissolve the OCT on the slides. Subsequently, antigen-retrieval was

performed for the tissue sections in low pH citrate buffer (Vector laboratories, CA, USA) at 97 °C

for 20 min. Then, tissue sections were cooled to room temperature in the antigen retrieval solution

and then immersed in 0.3% Triton-X in PBS for 30 min. Next, they were washed with quick dips

in PBS and then permeabilized in cold methanol for 10 min followed by PBS wash for 5 min. Then,

the slides were cleaned and marked with an IHC marker pen around the section to contain the

staining solutions to the sections. The tissue sections were blocked with serum-free protein block

(Dako, Agilent, USA) for 1 hr. Then, they were washed with 0.05% Tween20 in PBS solution (5

min) and then in PBS (2 x 5 min). Then, they primary antibodies diluted in Dako antibody diluent

solution were added to the sections in a black staining box with ample moisture. The slides were

then kept overnight at 4 °C.

The next day, slides were washed with 0.05% Tween20 in PBS solution (5 min) and then

in PBS (2 x 10 min) followed by addition of secondary antibodies diluted in Dako antibody diluent.

The slides were incubated with secondary antibodies for 2 hours at room temperature in the black

staining box. Subsequently, they were washed with 0.05% Tween20 in PBS solution (5 min) and

then in PBS (2 x 10 min). After 3 quick dips in deionized water, coverslips were mounted with

Page 161: An Injectable Gelatin-Based Conjugate Incorporating EGF

161

DAPI-containing Fluoro Gel-II mounting medium (Electron Microscopy Sciences, Hatfield, PA)

and stored at 4 °C.

Primary antibody Dilution Antigen retrieval

Secondary antibody

Cells/molecules identified

Rabbit poly clonal anti-GFAP (abcam #ab7260)

1:400 Citrate buffer pH 6 (97 °C at 20 min)

Dylight 488 Donkey anti-rabbit IgG

Astrocytes

Rabbit monoclonal anti-NeuN (abcam #ab177487)

1:400 Citrate buffer pH 6 (97 °C at 20 min)

Dylight 488 Donkey anti-rabbit IgG

Mature neurons

Rabbit polyclonal anti-Iba1 (Wako #019-19741)

1:400 Citrate buffer pH 6 (97 °C at 20 min)

Dylight 488 Donkey anti-rabbit IgG

Microglia/ macrophages

Mouse monoclonal anti-CD68 (BioRad # MCA341R)

1:200 Citrate buffer pH 6 (97 °C at 20 min)

Cy3 Red Donkey anti-mouse IgG

Macrophages

Mouse monoclonal anti-Nestin (Millipore #MAB353)

1:200 Citrate buffer pH 6 (97 °C at 20 min)

Cy3 Red Donkey anti-mouse IgG

Neural progenitor stem cells

Mouse monoclonal anti-Vimentin (Santa cruz #sc6260)

1:200 Citrate buffer pH 6 (97 °C at 20 min)

Cy3 Red Donkey anti-mouse IgG

Neural progenitor stem cells

Rabbit monoclonal Mushashi-1 (Abcam# ab52865)

1:100 Citrate buffer pH 6 (97 °C at 20 min)

Dylight 488 Donkey anti-rabbit IgG

Neural progenitor stem cells

Rabbit monoclonal anti-vWF (Abcam # ab6994)

1:500 High buffer pH 9 (97 °C at 20 min)

Dylight 488 Donkey anti-rabbit IgG

Endothelial cells

Rabbit polyclonal anti-EGF (Peprotech #500-P277)

1:500 Citrate buffer pH 6 (97 °C at 20 min)

Dylight 488 Donkey anti-rabbit IgG

Rat EGF

Mouse monoclonal anti-MMP2 (Millipore #MAB3308)

1:300 Citrate buffer pH 6 (97 °C at 20 min)

Cy3 Red Donkey anti-mouse IgG

Gelatinase MMP2

Table 6.3 Immunofluorescent staining of the cells and molecules with their staining protocol

Fluorescent images were acquired using epifluorescence microscope (Olympus BX60)

and confocal laser microscope (Nikon). The images were analyzed using ImageJ to quantify the

signal from the biomarker of interest. Three random images were selected and analyzed with a

standardized 0.3 mm2 ROI. Images were converted to 8-bit images followed by a standard

threshold optimized for each antibody. Then, the percent area covered by the fluorescent signal

in the standardized 0.3 mm2 ROI was measured using ImageJ. The average percent area was

calculated and used for analysis for comparison among the treatment groups.

Page 162: An Injectable Gelatin-Based Conjugate Incorporating EGF

162

6.3.13. Statistical analysis: Before the experiment, our power calculation for sample size

determination is based on the desire to determine significant a 30% difference in a selected

outcome variable with a 15% standard deviation, and with α=0.05 and β=0.05. The statistical

significance among the experimental groups for histomorphometric and IHC staining results were

determined by one-way ANOVA analysis using Tukey Post Hoc tests. Non-parametric data such

as BBB scores were analyzed for statistical significance using the Krusal-Wallis Test followed by

Dunn post-hoc test, Fisher’s Exact Test or Chi squared test. SPSS 25 (IBM) and Graphpad Prism

8.0 (Graphpad) were used for the statistical analysis.

Page 163: An Injectable Gelatin-Based Conjugate Incorporating EGF

163

6.4. Results

6.4.1. Surgical observations and Gtn-HPA injection: The animals tolerated the

surgery well and had no major complications immediately after the surgery. 2-3% isoflurane was

able to maintain the anesthesia level during the surgery. However, it was noted that compared to

using sodium pentobarbital as an anesthetic for surgical SCI, isoflurane showed more bleeding in

the rat and in the SCI cavity. The use of Gelfoam® for 15 minutes in three 5-min sequential

Gelfoam® insertions reduced the number of the animals that exhibited bleeding during the

injection of Gtn-HPA gel in the cavity. Achieving complete hemostasis was difficult in 10 out of 49

rats despite the use of Gelfoam® for 15 minutes. Once the hemostasis was established, 20 µl of

Gtn-HPA was injected into the cavity. Visually, both the 8% and 12% Gtn-HPA gel achieved

complete gelation in 2-3 minutes. 12% Gtn-HPA gel was more viscous due to the increased

gelation content and therefore care was taken to inject the liquid as early as possible into the

cavity. Complete gelation was confirmed in the remaining amount of the gel and was noted in the

surgical records.

Post-operation, animals gained consciousness in roughly half an hour after the surgery.

Immediately after gaining consciousness, rats seemed to be in shock and they frantically

observed the surroundings. These movements also included movement of the spine at the

surgical site. All the animals showed good markers of recovery in terms of weight gain and

rehydration within a week. No animals were suspected of any infections after surgery. In terms of

the function, they had lost the ability to empty their bladder due to loss of micturition reflex

indicated by their full bladders, which were manually expressed twice a day. Moreover, both their

hindlimbs were completely paralyzed after the induction of SCI. By four weeks, they had partial

return of the function in their hindlimb movement and bladder function. They were well

acclimatized with the open field locomotor test to assess functional recovery. In the next two

sections, bladder function and motor function were compared among the treatment groups to

investigate the effect of the intervention on improving the functional outcome after SCI.

6.4.2. Functional evaluation – bladder function: Neurogenic bladder dysfunction is a

significant outcome of the SCI that severely affects the standard of living of the SCI patients.

There are various treatment options available to the patient based on the comfort and

requirements based on the severity of the functional loss. It is one of the leading symptoms of

SCI that is managed clinically [17]. Therefore, it is of prime importance to study the effect of our

proposed therapy on the return of the urinary function post SCI. We recorded three different

parameters in the rats across all the treatment groups to evaluate the effect of the treatment on

Page 164: An Injectable Gelatin-Based Conjugate Incorporating EGF

164

urinary function – 1) number of animals that regained partial urinary function, 2) number of days

to regain partial urinary function and 3) urine retention volume during manual emptying for four

weeks.

Table 6.4 shows the results of the first two parameters – number of animals regaining

urinary function after four weeks and days required to regain partial urinary function. Qualitatively,

injection of Gtn-HPA with or without the factors increased the number of rats that showed partial

regain of urinary function (Table 6.4). However, the Chi-square analysis did not show a statistically

significant association between the treatment and total number of animals that showed partial

regain of urinary function after four weeks. Similarly, out of the animals that regained function, the

trend indicated that the injection of Gtn-HPA gel with or without factors promoted early regaining

of urinary function by 6 days compared to the controls (Table 6.4). However, one-way ANOVA

did not show statistically significant reduction in the days to regain partial urinary function.

Treatment # animals regaining

partial urinary function

# days to regain partial

urinary function

Control / EGF in suspension 5/11 21.25

8% Gtn-HPA gel only 4/6 21.5

8% Gtn-HPA + GFs 10/12 19.5

12% Gtn-HPA gel + GFs 11/12 18.2 Table 6.4 Trend shown in the table indicates that there is a statistically significant association between the treatment groups and number of animals that regain partial bladder function (Chi-square test). However, no significance was observed in the time to recovery among groups. However, the trend indicates an early recovery of bladder function in treatment groups.

The third parameter studied was the urine retentionvolume each time the bladder was

manually emptied using Crede’s maneuver. This parameter is clinically valuable tool and is

comparable to post-void residual volume (PVR). PVR is a noninvasively measured clinical metric

that is studied by the clinicians to evaluate voiding dysfunction. PVR is defined to be the residual

urine volume in the bladder after a voluntary void [18]. Similar measurements are made in the

urological animal models to study the effect of the treatment on the urinary function. Typically,

cystometric analysis involves use of catheterization to measure the urine retention volume [19].

However, catheterization was not performed to give free movement to the rats and to reduce the

chances of urinary infections. We applied a qualitative technique to measure the urine volume.

Urine retention volume was categorized as small, medium and large at each manual bladder

expression twice a day. In the beginning weeks, most all animals showed large residual urine

volumes while it decreased to medium and small in several animals. Table 6.5 shows the

frequency contingency table of the final average urinary output volume in the last four days with

Page 165: An Injectable Gelatin-Based Conjugate Incorporating EGF

165

the treatment groups. Qualitatively, controls had more animals with large volume bladders (4/6)

while the 12% Gtn-HPA with EGF groups had zero animals with large bladder volumes indicating

that there might be some dependence between the treatment groups and urine output volume. A

chi-square test of independence was performed to examine the relationship between the

treatment groups and urine volume output four weeks post injection. The relation between these

variables was significant (Χ2 = 13.068, p < 0.05) suggesting that injection of Gtn-HPA with EGF

promoted significantly better regaining of urinary function such that rats were voiding more volume

on their own four weeks after SCI.

Treatment / urine volume Small Medium Large

Control / EGF in suspension 2 4 5

8% Gtn-HPA gel only 1 2 3

8% Gtn-HPA gel + EGF 5 4 3

12% Gtn-HPA gel + EGF 7 5 0

Table 6.5 Contingency table showing the experimental groups and urine output volume after four weeks. Chi-square Analysis shows statistically significant association between the treatment groups and the urine output four weeks post SCI and injection (Chi-square analysis, Χ2 = 9.299, p < 0.05)

6.4.3. Functional evaluation – motor recovery (BBB scoring): In order to assess the

functional recovery after the SCI, animals underwent open field locomotor test in which they freely

walked on the long operation room table covered with absorbent chute. Prior to surgery, all

animals displayed normal gait. However, immediately after the hemi-resection injury, they

suffered complete paralysis for both the right and left hindlimb. BBB is a well-established 21-point

scale ranging from score of 0 (complete paralysis) to 21 (normal gait) of the hindlimbs. Each week

after the implantation, rats were scored using BBB scale in a randomized and blinded manner

with respect to the treatment group. There were three methods by which BBB scores were

analyzed – 1) BBB scores during four weeks of the experiment, 2) rate of improvement in the BBB

scores, and 3) number of animals showing either poor, good and better improvement.

During the course of the experiment, one week after inducing SCI and injecting the

treatment, all the groups showed some recovery in both hindlimbs. There was no difference

between the BBB scores among all the seven treatment groups. All the groups scored average

BBB score between 1 - 4 for both right and left hindlimb. Although no differences were noted

between the right and left hindlimb for the same group, left hindlimb scored less than right hindlimb

for the same group. By the end of four weeks, average BBB score for N and E groups was scored

between 4 – 6 while the average BBB score ranged from 9 to 11 for the Gtn-HPA treated groups

with factors (GE8, GES8, GE12 and GE12T) groups (Figure 6.5). Krusal-Wallis Test showed that

Page 166: An Injectable Gelatin-Based Conjugate Incorporating EGF

166

BBB scores were significantly dependent on the type of treatment four weeks post implantation

of the treatment (p < 0.01 for both hindlimbs). There were no differences among the treatment

groups at 1, 2 or 3-weeks post implantation. Subsequent Dunn post-hoc tests comparing four-

week scores showed that the groups delivering EGF and SDF-1α namely GE8, GES8, GE12 and

GE12T significantly recovered better than the control and EGF in suspension group for both left

Figure 6.5 Animals showed functional recovery after the SCI in both the right and left hindlimbs (Increase in the BBB score indicates functional recovery) during the four weeks of evaluation. Groups GE8, GES8, GE12 and GE12 showed better functional recovery compared to the controls and EGF in suspension group for both left and right hindlimbs four weeks after the implantation of the therapy. *p < 0.05, Krusal-Wallis Test followed by Dunn post-hoc test

Page 167: An Injectable Gelatin-Based Conjugate Incorporating EGF

167

and right hindlimbs. However, there were no differences between the G8, N and E group

suggesting that injection of Gtn-HPA alone does not assist in better functional recovery. In

addition, growth factor delivery in suspension is inefficient given that the growth factors would not

be present locally to confer any potential therapeutic effect and therefore requires delivery via a

biomaterial such as Gtn-HPA. Furthermore, we were interested in evaluating the effect of the left

hemi-section on the right and left hindlimb function. Since our model employed a left hemi-

resection, one hypothesis would be that left hindlimb would be severely affected compared to the

right hindlimb. However, 2-way ANOVA did not show the difference between the left and the right

hindlimb for each treatment group (p > 0.05) (Figure 6.6).

Next, in order to study the rate of improvement during the course of the experiment, each

rat was treated as its own control and rate of improvement was calculated for each rat by

subtracting the one-week score from four-week BBB score to examine the influence of the

treatment on the rate of improvement in the BBB score. After four weeks, quantitative analysis

showed that the rate of improvement was greater in the groups GE8, GES8, GE12 and GE12T

compared to N and E in the left hindlimb while rate of improvement was higher in the groups GE8,

GES8, GE12 and GE12T compared to E in the right hindlimb (Krusal Wallis Test, p < 0.05).

However, there were no differences in the rate of improvement between N, E and G8 groups

(Figure 6.7). These results indicate that administration of EGF and/or SDF-1α with Gtn-HPA gel

in the SCI cavity promotes faster improvement in the BBB scores compared to N and E.

Figure 6.6 There were no significant differences in the BBB scores for left and right hindlimb for the same treatment group four weeks after implantation. However, treatment groups with EGF and/or SDF-1α implanted with Gtn-HPA gel showed significantly better functional recovery compared to the control and EGF in suspension group. (Bars = Mean ± SEM, p < 0.05, Krusal-Wallis Test with Dunn post-hoc tests)

Page 168: An Injectable Gelatin-Based Conjugate Incorporating EGF

168

Third assessment parameter assessed was the number of animals that showed poor,

good and better improvement in the BBB scores by treating each animal as its own control. The

change in the BBB scores between first and fourth week was calculated. Three categories of

improvement were defined – poor (∆BBB = 0-4), good (∆BBB = 4-8), and better (∆BBB = 8+)

improvement. The frequency of rats in these categories are tabulated as shown in Table 6.6. No

rats from the control and Gtn-HPA gel only group [G8] showed more than 8 points improvement

in the score for both left and right hindlimb. However, Groups with EGF and/or SDF-1α delivered

by Gtn-HPA had 12/24 and 9/24 showed more than 8 points improvement in the BBB score in the

right and left hindlimb respectively. A chi-square test of independence was performed to examine

the relationship between the treatment groups and improvement in BBB scores four weeks post

injection. The relation between these variables was significant for the left hindlimb (Χ2 = 18.02, p

< 0.05) suggesting that injection of Gtn-HPA with EGF and/or SDF-1α promoted greater

improvement in the function compared to the controls. However, relation was not significant for

the improvement in the right hindlimb function.

Figure 6.7 Evaluation of the rate of improvement in the BBB score between week-1 and week-4 in both the hindlimbs. Quantitative analysis showed that the rate of improvement was higher in the groups GE8, GES8, GE12 and GE12T compared to N and E in the left hindlimb while rate of improvement was higher in the groups GE8, GES8, GE12 and GE12T compared to E in the right hindlimb. (*p<0.05, **p<0.01, Bars = Mean ± SEM

Page 169: An Injectable Gelatin-Based Conjugate Incorporating EGF

169

RIGHT (∆BBB) 0-4 4-8 8+

Control 1 5 0

8% Gtn-HPA gel only 1 5 0

8% Gtn-HPA + GFs 0 6 6

12% Gtn-HPA + GFs 0 6 6

LEFT (∆BBB) 0-4 4-8 8+

Control 4 2 0

8% Gtn-HPA gel only 3 3 0

8% Gtn-HPA + GFs 1 8 3

12% Gtn-HPA + GFs 0 6 6

Table 6.6 Tabulation of the improvement in the functional BBB scores from week 1 to week 4 was categorized in three levels – poor improvement (improvement of up to 4 points), moderate improvement (improvement by 4-8 points) and high improvement (improvement of more than 8 points). A chi-square test of independence showed that the relation between the treatment groups and extent of improvement was significant for the left hindlimb (Χ2 = 18.02, p < 0.05) suggesting that injection of Gtn-HPA with EGF and/or SDF-1α promoted greater improvement in the function compared to the controls. However, relation was not significant for the improvement in the right hindlimb function.

Spared tissue

Injury area

Figure 6.8 Gtn-HPA incorporating EGF and SDF-1α promoted greater preservation of the spared tissue and exhibited smaller injury area compared to control and EGF in suspension group. Scale bar = 2000 µm, bars = Mean ± SEM, One-way ANOVA, Tukey post-hoc test, * p<0.05, **p<0.01, ***p<0.001

Page 170: An Injectable Gelatin-Based Conjugate Incorporating EGF

170

6.4.4. Spared tissue and injury area quantification: Spared tissue was quantified based

on the histological images and was defined to be the unaffected tissue in the injury site continuous

with the rostral and caudal healthy spinal cord tissue. In addition, injury area was quantified to

assess if the injection of Gtn-HPA affects the extent of spread of the injury four weeks after SCI.

We found that Gtn-HPA incorporating EGF or SDF-1α significantly had lower injury area and

higher spared tissue compared to the control and EGF in suspension group (Figure 6.8). Gtn-

HPA gel alone was not sufficient promote greater sparing of the tissue near injury site.

6.4.5. Spared tissue correlation with functional evaluation: One potential explanation of

the better functional recovery in the treatment groups is attributed to the extent of spared spinal

cord tissue present in the injury site. A correlation analysis was undertaken between the percent

spared tissue and two functional evaluation – average BBB score of the left and right hindlimb at

weeks and days to regain urinary function. Spared tissue in the injury site was quantified using

H&E images of each animal. A reasonable positive correlation was observed between the percent

spared tissue in the injury site and the average BBB functional score (Figure 6.9). Correlation

analysis in Graphpad was found to be as r (41) = 0.7689, r2 = 0.5913, p < 0.001 suggesting that

there is a strong positive correlation between the spared tissue and functional recovery.

Figure 6.9 Correlation analysis between the percent spared tissue and average BBB functional score at four weeks showed variables to be positively correlated significantly r (41) = 0.7689, p < 0.001. The colorized data points show that treatment groups arguably promoted neuroprotection to exhibit greater spared tissue and therefore a higher functional score after four weeks

R² = 0.5913

0

2

4

6

8

10

12

14

16

0 10 20 30 40 50

BB

B s

co

re (

4 w

eek

s)

% spared tissue

Correlation between histology and function

N

E

G8

GE8

GES8

GE12

GE12T

Page 171: An Injectable Gelatin-Based Conjugate Incorporating EGF

171

Out of the animals that recovered partial return of bladder function, similar correlation

between the percent spared tissue and days to regain the partial urinary function was performed.

A correlation analysis in Graphpad was found to be r (30) = -0.1072, r2 = 0.01148, p > 0.05

suggesting that there was no correlation between the percent spared tissue and the time to regain

the bladder function (graph not shown).

6.4.6. Gtn-HPA gel volume measurement using MRI images: Ex vivo MRI images using

T1IR method were analyzed to measure the Gtn-HPA gel volume across all the treatment groups

in a randomized and blinded manner. All the measurements were normalized to the average

volume of the control group. The analysis showed that on average, the average Gtn-HPA volume

present in the injury site after four weeks was about 5.67 mm3 in all the treatment groups.

Quantification analysis showed that the Gtn-HPA gel volume was greater in the Gtn-HPA injected

groups compared to EGF in suspension (Figure 6.10). However, there were no differences

observed in the Gtn-HPA gel volume among the Gtn-HPA injected groups namely G8, GE8,

GES8, GE12 and GE12T suggesting that presence of EGF and SDF-1α had no effect on the Gtn-

HPA gel volume present in the injury site after four weeks (One-way ANOVA, p < 0.001).

Moreover, the results suggest that the T1IR technique is able to reliably distinguish the Gtn-HPA

gel in gel-containing groups from the non-gel containing groups namely E and N.

Figure 6.10 Gtn-HPA gel volume quantified using T1-weighted inversion recovery (T1IR) sequence normalized to the control volume. There were no differences among the Gtn-HPA gel injected groups indicating the presence of factors had no effect on the Gtn-HPA volume in the injury site after four weeks. However, the technique was sensitive in visualizing Gtn-HPA gel evident by negligible volume measured in the EGF in suspension and control groups. (Bars = Mean ± SEM, One-way ANOVA *p < 0.05, **p < 0.01, ***p<0.001)

Page 172: An Injectable Gelatin-Based Conjugate Incorporating EGF

172

6.4.7. Gross histological outcomes: General morphology and specific features regarding

the Gtn-HPA gel were studied using the H&E stained sections of all the rats across all the

treatment groups. As described in the experimental design section of the chapter, there were two

versions of the Gtn-HPA gel – 8 wt% Gtn-HPA and 12 wt% Gtn-HPA gel - based on the gelatin

content were evaluated in a 2 mm hemi-resection SCI model. After four weeks, histological

images showed important characteristics of the injury site and the response to the Gtn-HPA gel.

6.4.7.1. Control group – no injection: In the control [N] group, the injury site was

covered with dense fibrous tissue with empty regions, which were presumably filled with a cystic

fluid prior to the tissue processing (Figure 6.11A). Moreover, ferritin deposits were seen in the

injury site indicative of blood clots (Figure 6.12A). The injury site and the unaffected spinal cord

tissue was visibly different in terms of the morphology. Moreover, average percent spared tissue

in the injury site was 17.5% in the control animals. In the case where the injury has not caused

any secondary damage, we would expect to see 50% spared tissue given the right hemisphere

was the spinal cord in the injury site was not resected.

Page 173: An Injectable Gelatin-Based Conjugate Incorporating EGF

173

A

B

C

D

Gtn

-HP

A o

nly

[G8]

Co

ntr

ol

[N]

Gtn

-HP

A +

EG

F

[GE8

] G

tn-H

PA

+ E

GF

+ SD

F-1

α

[GES

8]

Figure 6.11. Representative H&E stained images of the four groups – (A) Control [N] (B) Gtn-HPA only [G8], (C) Gtn-HPA with EGF [GE8], and (D) Gtn-HPA with EGF and SDF-1α [GES8]. Scale bar = 2000 µm

Page 174: An Injectable Gelatin-Based Conjugate Incorporating EGF

174

6.4.7.2. 8 wt% Gtn-HPA gel groups: Normal histological features of white and gray

matter were observed in the unaffected tissue proximal and distal to the injury site. In the 8% Gtn-

HPA only [G8] group, injury site was substantially filled with the Gtn-HPA gel as opposed to the

dense fibrous tissue observed in the control group images after four weeks (Figure 6.12B).

However, dense fibrous tissue was also observed in the areas where the Gtn-HPA gel was not

present. The typical porous fibrillar structure of the Gtn-HPA gel was observed with weakly eosin

stained Gtn-HPA matrix as reported previously [20] (Figure 6.12B-D). Moreover, usually Gtn-HPA

gel was observed in ‘gel islands’ throughout the injury site and not as a one contiguous structure.

Gtn-HPA gel was not physically continuous with the surrounding tissue. Implanted Gtn-HPA

matrix was mostly void of any cellular infiltration around the gel islands. However, some cells were

A: Control [N] B: 8% Gtn-HPA only [G8]

C: 8% Gtn-HPA + EGF [GE8] D: 8%Gtn-HPA+EGF+SDF-1α[GES8]

Figure 6.12 Micrographs showing important observations of the four groups (A) Control - dense fibrous tissue (B) Gtn-HPA only - morphology of Gtn-HPA', minimal cellular migration around the island, poor interface (C) & (D) presence of the Gtn-HPA gel ‘islands’, greater cellular migration around the gel islands, well-defined interface with the surround tissue and migrated cells. Scale bar = 200 µm

Page 175: An Injectable Gelatin-Based Conjugate Incorporating EGF

175

seen at the boundary of the gel island (Figure 6.12B). Spared tissue was present in the G8

sections with average percent spared tissue being 30.0% of the injury site.

The morphology of the Gtn-HPA with EGF [GE8] and Gtn-HPA with EGF + SDF-1α [GES8]

sections was very similar (Figure 6.11C-D). Like the Gtn-HPA gel only group, histological sections

of these group showed the substantial presence of the Gtn-HPA gel in the injury site in form of

gel islands. However, contrary to the Gtn-HPA gel only group, there was a well-defined physically

continuous interface that was established with the surrounding tissue in the injury site as shown

by high magnification micrographs. In addition, qualitatively, many endogenous cells infiltrated

the injury site and migrated immediately surrounding the gel islands (Figure 6.12C-D). This was

evident by a blue band as a result of hematoxylin stained nuclei of the cells immediately around

the gel islands. The migratory pattern of the cells did not seem to be unidirectional – cells migrated

all around the gel islands. Remarkably, there were few sections in which cells were captured

through the gel islands in a somewhat linear arrangement, which could give insights into the

formation of the gel islands and degradation of the gel (Figure 6.12C). Spared tissue was present

in both the groups and average percent spared tissue was quantified to be 31.8% and 30.3% in

GE8 and GES8 respectively.

It is important to note that the Gtn-HPA gel was not observed in in the injury site in all the

gel-injected groups. Table 6.7 shows the study size and number of animals that showed the

presence of the gel in the injury site. Only 2/6 rats in the GES8 showed presence of the Gtn-HPA

gel and therefore was not accounted for histomorphometric analysis. There are several reasons

for the absence of the gel in certain animals including incomplete gelation due to bleeding during

injection, displacement of the Gtn-HPA gel from the cavity due to animal movement or due to poor

containment of the gel by the collagen fascia membrane and degradation of the gel by four weeks.

Experimental treatment Group Duration of

evaluation

Study

size (n)

# of animals showing

presence of the gel

Control – no injection N 4 weeks 6 N/A

8 wt% Gtn-HPA gel only G8 4 weeks 6 4

8 wt% Gtn-HPA gel with

EGF GE8 4 weeks 6 4

8 wt% Gtn-HPA gel with

EGF and SDF-1α GES8 4 weeks 6 2

Table 6.7. Table comparing the number of animals exhibiting the presence of the Gtn-HPA gel in the injury site after four weeks in H&E staining

Page 176: An Injectable Gelatin-Based Conjugate Incorporating EGF

176

6.4.7.3. EGF in suspension group [E]: The morphology of the EGF in suspension

group was similar to the control group. Dense fibrous tissue with empty areas covered the injury

site (Figure 6.12, left). The empty areas could be attributed to the cystic fluid present in the injury

site. Moreover, ferritin deposits were seen in the injury site indicative of blood clots (Figure 6.13,

right). The injury site and the unaffected spinal cord tissue was visibly different in terms of the

morphology with white and gray matter identifiable in the unaffected tissue. Moreover, average

percent spared tissue in the injury site was 15.5% in the animals with EGF delivered in PBS

suspension.

6.4.7.4. 12wt% Gtn-HPA groups: As shown in Table 6.7, all the rats in the 8wt% Gtn-

HPA injected groups did not show the presence of the gel in the injury site after four weeks. Two

strategies were evaluated – increasing the gelatin content (8 to 12%) and improving the

placement of the dural replacement to better contain the gel in the cavity. Furthermore, early

timepoints of Day 0, Day 1 and Day 7 were evaluated after implanting the 12% Gtn-HPA gel with

EGF [GE12T] to ensure that the Gtn-HPA gel was implanted properly and was not dislodged. The

H&E image of rat sacrificed on the same day (Day 0) showed the Gtn-HPA gel as a contiguous

block present in the left hemi-section defect with some degree of blood from hemorrhage (Figure

6.14A). It is important to note that there is no physical interface between the Gtn-HPA gel and

the surrounding tissue. Similar morphological features were observed in the H&E image of the rat

sacrificed the next day (Day 1). Gtn-HPA is present as a contiguous block in the left hemi-section

defect area with some degree of blood from hemorrhage after one day (Figure 6.14B). Moreover,

there is no physical interface between the Gtn-HPA gel and the nearby tissue. For both Day 0

Figure 6.13 (left) Representative general morphology of the EGF in suspension group – minimal spared tissue and empty areas are visible in the injury site, scale bar = 1000 µm (right) a micrograph showing dense fibrous tissue with ferritin deposits, scale bar = 200 µm

Page 177: An Injectable Gelatin-Based Conjugate Incorporating EGF

177

and Day 1 images, right hemisphere of the cord is intact. However, there are notable different

features in Day 7 histology compared to Day 0 and Day 1. There is a continuous physical interface

between the Gtn-HPA gel and the surrounding tissue in the injury site (Figure 6.14C). The Gtn-

HPA gel is present in form of gel ‘islands’ by one week. These observations suggest that the

C, Day 7

B, Day 1

A, Day 0

Figure 6.14 Representative images from the implantation of Gtn-HPA with EGF [GE12T] at Day 0, Day 1 and Day 7. (A) Day 0 section shows the presence of the Gtn-HPA gel as a contiguous block in the left hemisphere of the cord and intact right hemisphere tissue. Blood is also observed in the section (B) Day 1 section shows similar morphology as that of Day 0 with blood deposit. Gtn-HPA is seen as a contiguous block and no interface between the Gtn-HPA gel and the surrounding tissue is established (C) Day 7 morphology is strikingly different than Day 0 and Day 1. Gtn-HPA is present in form of gel islands and physically continuous interface is existent around gel islands. Scale bars = 1000 µm

*

*

* *

Page 178: An Injectable Gelatin-Based Conjugate Incorporating EGF

178

better placement of the fascia membrane by ‘tucking-in’ the dural replacement membrane helps

containing the implanted Gtn-HPA gel in the hemi-resection cavity after one week.

Two experimental groups were evaluated implanting 12wt% Gtn-HPA gel in the hemi-

resection defect four weeks after injection – 12% Gtn-HPA gel with EGF [GE12] and 12% Gtn-

HPA gel with EGF with ‘tucked-in’ dura [GE12T]. The morphology of the both the groups were

A: GE12

B: GE12T

C: 12% Gtn-HPA with EGF [GE12] D: 12% Gtn-HPA with EGF + ‘tucked-in’ dura [GE12T]

Figure 6.15 Representative injury site showing the presence of Gtn-HPA with cellular migration around the gel islands in (A) GE12, Scale bar = 1000 µm and (B) GE12T group, scale bar = 2000 µm. High-magnification micrographs of (C) GE12 and (D) GE12T group showing features of the Gtn-HPA gel with cellular migration around the gel. Scale bar for C-D = 200 µm

Page 179: An Injectable Gelatin-Based Conjugate Incorporating EGF

179

very similar given that the treatment composition was the same for both groups. However, the

procedure and size of the fascia membrane as dural replacement was different in GE12T group.

The Gtn-HPA gel was present in form of islands and there was immense cellular migration around

these gel islands in the injury site similar to the GE8 and GES8 groups (Figure 6.15A-D). This

suggests that Gtn-HPA with higher gelatin content was permissive of the cellular migration and

EGF is able to recruit endogenous cells to the gel islands. Immunohistochemical assessment

would aid in identifying these cells. Interestingly, all animals (6/6) in the GE12T group showed the

presence of the Gtn-HPA in the injury site, reaffirming that the ‘tucking-in’ procedure improved

the containment of the Gtn-HPA in the hemi-resection cavity. Moreover, there was a continuous

physical framework between the gel islands and the surrounding tissue in the injury site. Percent

spared tissue was quantified to be 37.2% and 34.4% in GE12 and GE12T group respectively.

6.4.8. Histomorphometric evaluation of the injury site: The Gtn-HPA gel area and

cellular migration area around the gel was measured by two method. The histological sections

evaluated to measure the areas represented the midplane coronal section of the cord. GES8

group was not included in the analysis because only 2 out of 6 animals showed presence of the

gel in the injury site four weeks post implantation of the Gtn-HPA gel.

The first method (Method 1 – total area method) measured the total Gtn-HPA gel area in

all the gel islands and total cellular migration immediately around all the gel islands in the entire

injury site in ImageJ using ROI manager tool. Presence of the growth factors did not affect the

area of the gel present in the injury site for both 8% and 12% Gtn-HPA gel formulations (One-

Way ANOVA, Tukey post-hoc test, p > 0.05) (Figure 6.16A). Local delivery of the EGF with the

Gtn-HPA matrix significantly increased the number of the migrated cells immediately surrounding

the gel islands compared to Gtn-HPA gel only group for both 8% and 12% Gtn-HPA gel

formulations except for GE12 group (One-Way ANOVA, Tukey post-hoc test, p < 0.05) (Figure

6.16B).

The second method (Method 2 – standardized ROI method) measured the Gtn-HPA gel

area and cellular migration immediately surrounding the gel in the standardized 0.3 mm2 ROI at

the interface using ImageJ in three randomly chosen images. Presence of the growth factors did

not affect the area of the gel present in the injury site for both 8% and 12% Gtn-HPA gel

formulations (One-Way ANOVA, Tukey post-hoc test, p > 0.05) (Figure 6.16C). Like the results

observed in Method 1 (total area measurement), local delivery of the EGF with the Gtn-HPA matrix

significantly increased the number of the migrated cells immediately surrounding the gel islands

compared to Gtn-HPA gel only group for both 8% and 12% Gtn-HPA gel formulations (One-Way

ANOVA, Tukey post-hoc test, p < 0.001) (Figure 6.16D). These results are consistent with the

Page 180: An Injectable Gelatin-Based Conjugate Incorporating EGF

180

findings in the complete resection model for 8wt% Gtn-HPA gel with EGF and/or SDF-1α

experimental groups (Chapter 4, figure 4.7 & 4.8).

Figure 6.16 Histomorphometric analysis results using both Method 1 (Total area method) and Method 2 (standardized ROI method)

Method 1 - (A) Delivering growth factors with the Gtn-HPA gel did not affect the amount of the gel present while (B) EGF presence in the Gtn-HPA recruited significantly greater number of cells immediately surrounding the gel in the injury cavity four weeks post injection for both 8% and 12% Gtn-HPA gel formulations except GE12 group

Method 2 - (C) Delivering growth factors with the Gtn-HPA gel did not affect the extent of the gel present while (D) EGF delivered with the recruited significantly greater number of cells immediately surrounding the gel in the injury cavity four weeks post injection for both 8% and 12% Gtn-HPA gel formulations

*p<0.05, **p<0.01, *** p < 0.001, One-way ANOVA followed by Tukey post-hoc test; Bars = Mean ± SEM

Page 181: An Injectable Gelatin-Based Conjugate Incorporating EGF

181

6.4.9. Immunohistochemistry evaluation of the injury site: H&E staining provided

general morphology of the injury site. Next step was to identify the cell population that migrated

into the injury site and surrounding the gel islands. Sections were stained with various primary

antibody markers, results from which are reported below.

6.4.9.1. GFAP – astrocytes: GFAP+ cells were present in the injury site and at the

interface between the unaffected tissue and the injury site (Figure 6.17). In all the groups, the

density of astrocytes was high near the interface between the unaffected tissue and the injury

site, while it decreased towards the proximal and distal direction of the unaffected tissue. In

contrast to interface, the astrocytes were sparsely present in the injury site. Moreover, astrocytes

were rarely present in the cystic area of the injury site (Figure 6.17). In terms of morphology,

GFAP+ cells that had migrated into the injury site had long thin cells with their astrocytic

processes. However, GFAP+ cells away from the injury site had normal protoplasmic morphology

(Figure 6.18A-C). In order to compare the number of astrocytes, GFAP+ cells were quantified at

two different locations – in the injury site and at the interface between the unaffected tissue and

Interface

Gtn-HPA + EGF

A

B

C

Figure 6.17 Representative GFAP stained sections of (A) Control [N] (B) EGF in suspension [E] (C) Gtn-HPA-injected group [GE12 here] showing the interface, injury site and presence of gel in the injury site, Scale bars = 1000 µm

Page 182: An Injectable Gelatin-Based Conjugate Incorporating EGF

182

the injury site (Figure 6.17BC). Quantitative analysis of the area covered by GFAP+ cells in the

injury site indicated no significant difference among the treatment groups. However, the

quantitative analysis of the area covered by GFAP+ cells at the border of the unaffected tissue

and the injury site showed significant decrease in the presence of astrocytes in the gel-injected

groups (i.e. G8, GE8, GES8, GE12 and GE12T) compared to the control [N] and EGF in

suspension [E] group as shown in figure 6.18D (One-Way ANOVA, Tukey post-hoc test, p < 0.05).

Furthermore, migrated cells immediately surrounding the gel islands did not stain positive for the

GFAP indicating that the cells recruited by EGF to the gel islands were not astrocytes (Figure

6.19).

A: Away from injury B: Interface C: Injury site

Figure 6.18 Representative micrographs showing the GFAP+ cells’ morphology (A) in the unaffected tissue away from the injury site, (B) at the interface of the injury site and unaffected tissue and (C) in the injury site. (D) Quantitative analysis showed significantly reduced number of astrocytes at the interface outlining the injury site in the gel-injected groups [G8, GE8, GES8, GE12 and GE12] than the control [N] and EGF in suspension group [E] (One-way ANOVA, Tukey post-hoc test, *p < 0.05). However, percent area covered by GFAP+ cells was not significant in the injury site. Scale bars in A-C = 200 µm and Bars = Mean ± SEM

D

Page 183: An Injectable Gelatin-Based Conjugate Incorporating EGF

183

6.4.9.2. Iba-1 - resident and/or activated microglia: Iba-1+ cells were present

throughout the injury site across all the groups. They were sparsely present in the unaffected

tissue of the spinal cord in its typical morphology – small ramified cells with processes (Figure

6.20A). Interestingly, similar morphology of Iba1+ cells was observed in the spared tissue in the

right hemisphere of the cord (Figure 6.20C). Iba1+ cells in the injury site had different morphology

– these cells were enlarged spherical cells without any processes (Figure 6.20D). This enlarged

morphology suggests that these cells are activated microglia [21]. Interestingly, Iba1+ cells at the

border of the injury site showed both types of morphology of Iba1+ cells (Figure 6.20B). Moreover,

spared tissue closer to the injury site also had presence of enlarged Iba1+ cells. Such dichotomy

in the morphology of the Iba1+ cells in the unaffected tissue and the injury site was consistent

across all the groups including the controls. Quantitative analysis of the Iba1+ cells in the injury

site showed that the percent area occupied by Iba1+ activated microglia was similar across all

the groups – N, E, GE8, GES8, GE12, and GE12T (Figure 6.20E). This indicates that Gtn-HPA

was well tolerated and had similar microglia response to the controls. There was very sparse

presence of Iba1+ cells immediately around the Gtn-HPA corroborating the finding that Gtn-HPA

was well tolerated in the injury site.

Figure 6.19 The cells that migrated immediately surrounding the gel block stained negative for GFAP indicating that these cells are not astrocytes (left) GFAP stained section showing the presence of the gel marked by white arrow and the cellular migration around the gel block marked by red asterisks. (right) The same animal H&E showing the presence of the gel and cellular migration for comparison with the GFAP stained section. Scale bar = 1000 µm

* *

*

*

* *

* *

Gtn-HPA+EGF Gtn-HPA+EGF

Page 184: An Injectable Gelatin-Based Conjugate Incorporating EGF

184

Figure 6.20 Micrographs showing the presence and morphology of Iba1+ cells in the (A) unaffected tissue away from the injury site, (B) in the interface bordering the injury site, (C) spared tissue in the right hemisphere of the cord and (D) in the injury site, Scale bars = 200 µm (E) Quantitative analysis of Iba1+ cells showed that percent area covered by the microglia/macrophages in the injury site is similar across all the groups suggesting that implantation of Gtn-HPA was well tolerated . Bars represent Mean ± SEM

A: unaffected B: interface

C: spared tissue D: injury site

E

Page 185: An Injectable Gelatin-Based Conjugate Incorporating EGF

185

6.4.9.3. CD68 – Macrophages/monocytes: CD68 is widely considered to be the

marker for activated microphages/monocytes in response to tissue damage. This immune

reaction is not well detected by Iba1 [22]. Therefore, sections were stained with CD68-Ed1

antibody to study the immune response of macrophages after implantation of the Gtn-HPA matrix.

CD68+ cells were predominantly present in the injury site in the typical spheroidal shape with

enlarged nucleus without any processes in all the groups – N, E, G8, GE8, GES8, GE12 and

GE12T (Figure 6.21A). In the unaffected tissue away from the injury, there was sparse presence

of CD68 while the density of CD68 decreased away from the injury site in both rostral and caudal

direction. Consistent among all the Gtn-HPA injected groups, there were only few CD68+ cells in

the numerous cells that migrated immediately surrounding the Gtn-HPA gel islands (Figure

6.21B). This indicates that the migrated cells immediately surrounding the Gtn-HPA gel are not

macrophages/monocytes. Moreover, the observations are consistent with the findings of Iba1

staining that the implantation of the Gtn-HPA in the hemi-resection injury site did not elicit severe

immune reaction. As shown in figure 6.22, quantitative analysis of the percent area covered by

CD68+ cells did not reveal any statistical difference among all the groups – N, E, GE8, GES8,

GE12 and GE12T confirming the consistent finding that Gtn-HPA is a biocompatible matrix that

does not elicit severe immune response and is well tolerated (One-way ANOVA, p > 0.05).

Figure 6.21 (A) Representative micrograph showing CD68+ macrophages (red) in the injury site (B) Micrograph showing that very few cells were CD68+ in the migrated cells immediately around the Gtn-HPA gel in GE12T group. Moreover, very few CD68+ cells do not show the typical spheroidal morphology of macrophages. (C) DAPI (blue) staining of the same region showing numerous cells migrated around the Gtn-HPA gel for comparison. White dotted line marks the interface between the Gtn-HPA gel and surrounding tissue and scale bars = 200 µm

Gtn-HPA

+ EGF Gtn-HPA

+ EGF

A B C

Page 186: An Injectable Gelatin-Based Conjugate Incorporating EGF

186

6.4.9.4. NeuN – mature neurons: NeuN+ staining was observed in the neuronal cell

bodies in the viable parenchyma away from the injury site in all the groups. Strong NeuN signal

was observed in the nucleus of the cell body of the neuron while the cytoplasm showed granular

NeuN signal. The staining was positive in the gray matter band of the unaffected tissue as

expected. The density of NeuN+ cells in the viable parenchyma was greater than near the

interface between the injury site and unaffected tissue. No NeuN signal was noted in the injury

site or spared tissue. These observations were consistent across all the treatment groups – N, E,

G8, GE8, GES8, GE12 and GE12T. Figure 6.23 shows representational micrographs of the NeuN

staining at different locations in the cord for GE12 group. Given that NeuN was negative for the

entire injury site across all the groups, no quantification of the NeuN+ cells was performed to

compare among all the treatment groups. Moreover, migrated cells immediately surrounding the

Gtn-HPA were negative for NeuN for G8, GE8, GES8, GE12 and GE12T group indicating that

EGF and/or SDF-1α had no effect on the presence of mature neurons four weeks post injection

of the treatment. It would be of importance to compare these findings at longer evaluation time

after the Gtn-HPA matrix injection into the hemi-resection defect.

Figure 6.22 Quantitative analysis of the % area covered by CD68+ cells did not reveal any statistical difference among all the groups – N, E, G8, GE8, GES8, GE12 and GE12T supporting that Gtn-HPA is a biocompatible matrix and is well tolerated by the spinal cord tissue after four weeks post Gtn-HPA implantation Bars represent Mean ± SEM (One-way ANOVA, p > 0.05)

Page 187: An Injectable Gelatin-Based Conjugate Incorporating EGF

187

6.4.9.5. MMP2 - Gtn-HPA gel degradation: In the evaluation of the Gtn-HPA with EGF

and/or SDF-1α in the complete resection SCI model, we did a preliminary investigation into the

mode of Gtn-HPA degradation. 5 sections from G8 group and 5 from GE8, GE12 and GE12T

groups were stained with MMP2. Consistent with the complete resection model results, strong

MMP2 signal was observed for both Gtn-HPA gel only group [G8] and Gtn-HPA with EGF groups

[GE8, GE12, and GE12] only at the border of the Gtn-HPA gel (Figure 6.24). Quantitative analysis

of the percent area of MMP2 at the border showed significantly higher MMP2 presence in the

Gtn-HPA with EGF groups (Figure 6.24) indicating that the high cellular density of the migrated

cells released more MMP2 responsible for the degradation of the Gtn-HPA gel in vivo.

A: unaffected B: near the interface

C: injury site D: migrated cells around Gtn-HPA

Figure 6.23 Representational micrographs from GE12 group showing the NeuN staining in (A) unaffected viable parenchyma away from the injury in the gray matter of the cord, (B) near the interface between the injury site and surrounding unaffected tissue, (C) in the region where Gtn-HPA is not present and (D) around the Gtn-HPA gel island. White dotted line shows the border between the Gtn-HPA gel and migrated cells surrounding the gel. Scale bars = 200 µm

Gtn-HPA + EGF

Page 188: An Injectable Gelatin-Based Conjugate Incorporating EGF

188

A: Gtn-HPA gel only [G8]

B: Gtn-HPA gel + EGF [GE12]

C

Figure 6.24 Representative image of the Gtn-HPA gel islands with MMP2 (red) and DAPI (blue) in the injury site of (A) Gtn-HPA gel only [G8] group and in (B) representational image of Gtn-HPA + EGF [GE groups]. Quantitative analysis shows significantly greater MMP2 presence at the border of the Gtn-HPA islands in GE groups compared to the G8 group (2-tailed t-test, **p<0.01). Scale bars in A-B = 200 µm and Bars in C represent Mean ± SEM.

Page 189: An Injectable Gelatin-Based Conjugate Incorporating EGF

189

6.4.9.6. EGF – Gtn-HPA incorporated EGF: In order to evaluate the presence of EGF in

the Gtn-HPA gels four weeks post implantation of Gtn-HPA matrices in the hemi-resection defect,

6 sections from Gtn-HPA gel only group [G8] and 6 sections from Gtn-HPA + EGF groups (i.e.

GE8, GE12, and GE12T) were stained with anti-EGF antibody. All 6 sections stained positive for

EGF that co-localized with Gtn-HPA gel islands for Gtn-HPA gel groups (G8, GE12 and GE12T).

However, Gtn-HPA gel in the Gtn-HPA gel only group [G8] stained negative for the EGF,

confirming the presence of the EGF in the gel delivered with EGF four weeks post-injection (Figure

6.25). Fisher’s exact test showed significant association between the outcome of EGF staining

between both groups – G and GE groups (Table 6.8, p < 0.05). Therefore, EGF staining could be

used as reliable technique to stain the Gtn-HPA gel in the EGF-delivered groups and EGF is

locally delivered by Gtn-HPA gel four weeks post implantation in the hemi-resection defect.

Treatment Group + EGF staining - EGF staining

Gtn-HPA gel only [G8] 0 6

Gtn-HPA gel + EGF [GE8, GE12 and GE12T] 6 0

Table 6.8 Two-tailed Fisher’s exact test reports statistically significant association (p < 0.05) between the treatment groups (rows) and EGF staining outcomes (columns) four weeks post Gtn-HPA injection in the defect site. The finding suggests that EGF is being locally delivered after four weeks in the injury site.

EGF DAPI EGF/DAPI

G

tn-H

PA

ge

l +

EG

F

G

tn-H

PA

ge

l o

nly

Gtn-HPA

Gtn-HPA + EGF

Figure 6.25 Representative EGF stained micrographs showing the strong signal of EGF (green) and DAPI (blue) in the Gtn-HPA gel + EGF [GE12T] group while Gtn-HPA gel stained negative in the Gtn-HPA gel only group [G8]. Scale bars = 200 µm

Page 190: An Injectable Gelatin-Based Conjugate Incorporating EGF

190

6.4.9.7. Nestin- neural stem cells: Nestin is a type VI intermediate filament protein

found in neural stem cells (NSCs). The groups in which EGF and/or SDF-1α was locally delivered

by Gtn-HPA gel (i.e. GE8, GES8, GE12 and GE12T), histological analysis showed significantly

greater migration of cells around the Gtn-HPA gel islands in the injury site. In these four groups,

Nestin stained positive in the cells migrated surrounding the Gtn-HPA gel islands (Figure 6.26D-

G). In the injury site where Gtn-HPA is not present, few Nestin+ cells are scattered. For the Gtn-

HPA gel only group [G8], a small number of Nestin+ cells are present throughout the injury site

and around the Gtn-HPA islands. Qualitatively, however, the density of Nestin+ cells around the

Gtn-HPA islands seem to be much higher in the GE8, GES8, GE12 and GE12T groups than the

G8 group (Figure 6.26C). Control group [N] and EGF in suspension group [E] did not show any

noteworthy presence of Nestin+ cells in the injury site (Figure 6.26AB). In terms of the

morphology, Nestin staining showed elongated cells oriented perpendicular to the border of the

Gtn-HPA gel indicating the radial migratory pattern of the cells. A high magnification image of an

area of interest shows the radially aligned Nestin+ cells at the border of the Gtn-HPA islands

(Figure 6.26H). Quantitative analysis of the percent area covered by Nestin+ cells in the injury

site around the Gtn-HPA islands showed that the groups in which EGF and/or SDF-1α was locally

delivered by Gtn-HPA gel (i.e. GE8, GES8, GE12 and GE12T) showed significantly greater

presence of NSCs immediately surrounding the Gtn-HPA gel islands and into the gel compared

to the G8, E and N groups (One-Way ANOVA, Tukey post-hoc test, p < 0.05, figure 6.28). This

finding supports the hypothesis that EGF can recruit NSCs into the gel-filled lesion. However,

there was no statistical difference between GE8 and GES8 groups indicating that SDF-1α did not

enhance the recruitment of NSCs to the injury site. There was no statistical significance between

the Nestin+ cells between the control [N], EGF in suspension [E] and Gtn-HPA gel only [G8] group

indicating that the Gtn-HPA gel alone is not able to recruit NSCs to the injury site. The gelatin

content (8% vs 12%) had no effect on number of NSCs recruited to the injury site as there were

no differences between GE8 and GE12 groups (Figure 6.28).

Page 191: An Injectable Gelatin-Based Conjugate Incorporating EGF

191

Figure 6.26 (A-G) Micrographs showing the Nestin+ (red) cells in the injury site in all the experimental groups. Cells that migrated around the Gtn-HPA islands (shown by white asterisk) in the groups in which EGF was administered (GE8, GES8, GE12 and GE12T) show positive Nestin signal indicating that EGF recruited neural r stem cells (NSCs) to the injury site. (H) A high-magnification image (20x) shows the radial migratory pattern of these cells that are aligned perpendicular to the border of the Gtn-HPA islands. Scale bars = 200 µm

A: N B: E

C: G8 D: GE8

E: GES8 F: GE12

G: GE12T H: GE12-20x

* *

* *

*

*

*

*

*

*

Page 192: An Injectable Gelatin-Based Conjugate Incorporating EGF

192

In order to confirm the identity of the migrated cells into the Gtn-HPA gel in the EGF-

delivered groups (i.e. GE8, GES8, GE12 and GE12T) as NSCs, representative samples from

EGF-delivered groups were stained with Vimentin, GFAP and Musashi-1. Nestin/GFAP double

staining showed that Nestin+ cells around the Gtn-HPA islands in the stained negative for GFAP

indicating these are NSCs and not astrocytes (Figure 6.27). Vimentin, another confirmatory

marker for NSCs, was double-stained with GFAP. The images showed that the migrated cells

immediately around the Gtn-HPA islands were Vimentin+ cells and stained negative for GFAP

Figure 6.28 Quantitative analysis of the percent area covered by Nestin+ cells in the injury site around the Gtn-HPA islands showed that the groups in which EGF was locally delivered by Gtn-HPA gel (i.e. GE8, GES8, GE12 and GE12T) showed significantly greater presence of NSCs immediately surrounding the Gtn-HPA gel islands compared to the G8, E and N groups. This finding supports that Gtn-HPA alone is not able to recruit endogenous reparative cells while EGF is necessary to recruit endogenous NSCs to the injury site. (One-way ANOVA, **** p < 0.001, Bars represent Mean ± SEM)

GFAP Nestin GFAP/Nestin/DAPI

Figure 6.27 GFAP (green) and Nestin (red) signal did not co-localize in the cells surrounding the Gtn-HPA islands in the GE12T group (representational) suggesting that migrated cells are NSCs and not astrocytes. DAPI (blue) shows the presence of the cells around the Gtn-HPA island (marked by white asterisk). Scale bars = 200 µm

* * *

Page 193: An Injectable Gelatin-Based Conjugate Incorporating EGF

193

confirming the presence of NSCs (Figure 6.29). Musashi-1 is yet another RNA-binding protein

biomarker to study the presence of the CNS progenitor cells including the neural stem cells [23].

Few representative sections stained with Mushashi-1 showed Musashi-1+ signal in the cells

surrounding the Gtn-HPA islands confirming the presence of NSCs (Figure 6.30). The number of

Musashi-1+ cells seemed to be less than Nestin and Vimentin possibly indicating that Musashi-1

stains for a subset of NSCs. In addition, Nestin was double-stained with EGF marker to visualize

the presence of NSCs in the groups that delivered EGF by Gtn-HPA. We observed positive

staining of EGF in the Gtn-HPA gel and Nestin in the cells immediately surrounding the Gtn-HPA

(Figure 6.31). These observations were significant in establishing the identity of the migrated cells

in the EGF-delivered Gtn-HPA groups as neural stem cells (NSCs).

GFAP Vimentin GFAP/Vimentin/DAPI

* * *

Figure 6.29 Migrated cells around the Gtn-HPA islands are positive for Vimentin. GFAP (green) and Vimentin (red) signal did not co-localize in the cells surrounding the Gtn-HPA islands in the GE12 group (representational) suggesting that migrated cells are NSCs and not astrocytes. DAPI (blue) shows the presence of the cells around the Gtn-HPA island (marked by white asterisk). Scale bars = 200 µm

Figure 6.30 Representative micrographs of Musashi-1 (green) and DAPI (blue) showing positive Musashi-1 signal in the cells surrounding the Gtn-HPA gel in GE12 group further supporting that the migrated cells are NSCs. White-dotted line shows the border of the Gtn-HPA gel. Scale bars = 200 µm

Musashi-1 DAPI Musashi-1/DAPI

Page 194: An Injectable Gelatin-Based Conjugate Incorporating EGF

194

6.4.9.8. vWF – endothelial cells: vWF was difficult biomarker to stain the frozen spinal

cord sections. The recommended antigen retrieval buffer (High pH 9 buffer) at 97°C for 20 min

detached the spinal cord sections from the slide. Few sections that remained attached partially

showed positive vWF signal for the presence of endothelial cells distributed in the injury site

(Figure 6.32). CD31 and CD34 staining was negative for the preliminary sections that were tried.

EGF Nestin EGF/Nestin

Figure 6.31 Double-staining of EGF (green) and Nestin (red) showing Nestin+ cells immediately surrounding Gtn-HPA gel containing EGF four weeks post injection in the hemi-resection defect.

Figure 6.32 Representative micrographs showing vWF staining in the injury site indicating the presence of endothelial cells in their typical morphology. Scale bars = 200 µm

Page 195: An Injectable Gelatin-Based Conjugate Incorporating EGF

195

6.5. Discussion

In this chapter, we implanted two different Gtn-HPA gel formulations with EGF and/or

SDF-1α into the 2-mm hemi-resection injury induced in a T8 2-mm left lateral hemi-resection

surgical excision SCI model with female Lewis rats. First formulation tested was an 8% Gtn-HPA

matrix delivering the growth factors locally in the injury site. The second formulation investigated

had increased gelatin content to 12% Gtn-HPA. We evaluated the effects of implanting the Gtn-

HPA with or without the growth factors in recruiting of the endogenous cells to the injury site, in

inducing a reparative response of the damaged tissue and in aiding improved functional recovery

compared to the control group that received no injection. We evaluated these responses four

weeks after inducing the SCI and injecting the Gtn-HPA gel with or without growth factors into the

injury site. Since traumatic or contusion type of SCI model do not completely sever the cord,

surgical models such as complete or partial resection are utilized to study the healing response

and neuronal regeneration in the field of tissue engineering [24]. Moreover, lateral hemi-resection

models are widely used and considered to be ideal for studying axonal regeneration [12]. Hemi-

resection produces a standardized defect of preferred size and confirms the complete break in

the neuronal circuitry at the location of the cord. In addition, since the resection is made only one

side of the spinal cord, the contralateral side can serve as a control. The most important

advantage in the context of our work is that hemi-resection model creates a reproducible defect

in which biomaterial of a same volume can be injected along with growth-promoting factors such

as EGF and SDF-1α. In addition, we get a visual confirmation of the injection of the Gtn-HPA gel

into the hemi-resection injury site. In addition, hemi-resection models of various lengths have

been implemented to examine the ‘bridging gap’ therapies such as scaffolds, guidance channels

and pre-fabricated devices [12, 25, 26]. We studied a total of seven groups namely – control [N],

EGF in suspension [E], 8wt% Gtn-HPA gel only [G8], 8wt% Gtn-HPA gel with EGF [GE8], 8wt%

Gtn-HPA gel with EGF and SDF-1α [GES8], 12wt% Gtn-HPA gel with EGF [GE12] and 12wt%

Gtn-HPA gel with EGF with dural replacement ‘tucked-in’ around the cord [GE12T]. There were

several assessment parameters we employed to test our hypotheses mentioned in the ‘Overall

goal and hypotheses’ section of this chapter.

6.5.1. Functional evaluation: SCI results in loss of function in many systems of the body

drastically affecting the quality of life of the patients. Therefore, improvement in the function is the

most critical aspect of pre-clinical and clinical studies that aim to treat SCI. We studied bladder

function and motor function of the rats for four weeks post injection of the Gtn-HPA gels.

Page 196: An Injectable Gelatin-Based Conjugate Incorporating EGF

196

6.5.1.1. Bladder function – urine retention volume: Post-void residual volume (PVR)

is a clinical metric that is studied to gain insights into the bladder function for patients with SCI. It

is well-established that decrease in the PVR correlates with improvement in the bladder function

of the patient [27]. 3 parameters were studied to evaluate the recovery of bladder function after

inducing a lateral hemi-resection injury – urine retention volume four after four weeks, number of

animals that regained partial bladder function, and time it took to recover partial urinary function.

Urine retention volume (URV) was qualitatively recorded as small, medium or large at every

manual emptying of the bladder twice in a day during the entire duration of the experiment. For

rats that showed partial recovery, the URV decreased from large to medium or large to small

indicating that they recovered partial voiding ability. For example, 7/12 rats in the 12% Gtn-HPA

+ EGF injected groups (i.e. GE12 and GE12T) had had small URV compared to 0/6 animals in

the control group [N]. Similarly, 4/6 animals in the control group had large URV compared to 0/12

in the GE12 and GE12T groups. Chi-square analysis showed significant association between the

treatment groups and URV after four weeks suggesting that the implantation of Gtn-HPA with

factors led to greater voiding ability in the rats (Table 6.5). Chi-square analysis showed significant

association between the treatment groups and the number of animals recovering partial bladder

function. Although not statistically significant, the trend suggested that treatment groups rats

recovered partial voluntary voiding ability earlier than the than the control group. However, no

statistical significance was observed (Table 6.4). These findings suggest that the treatment

facilitated in improving the bladder function after SCI. The mechanism of the partial bladder

function recovery was either the extent of spared tissue (correlation not observed) or re-wiring of

the neural pathways [28]. Injection of Gtn-HPA with EGF seem to affect the amount of spared

tissue and thereby providing viable parenchyma for rewiring of the neural circuitry (micturition

pathway) to improve bladder function. Future studies should perform a focused study where urine

retention volume is quantitatively measured using catheterization of the bladder or other

established methods to study URV in pre-clinical models of SCI [19]. Longer survival evaluation

times would add to the understanding of the bladder recovery function as well.

6.5.1.2. Motor function – BBB scale: Beattie, Basso, and Bresnahan (BBB) scale is a

validated, standardized and well-accepted behavioral scale that is used to assess the functional

integrity of rats’ hindlimbs post SCIs, usually with the underlying assumption that the cardiac and

respiratory functions are largely unaffected. BBB involves animals walking in an open field with

their hindlimb movement carefully observed. Various voluntary motor functions such as joint

movements, weight bearing, fore/hind limb coordination, paw placement, and the directions of the

tail were judged based on the 21-point BBB scale in a randomized and blinded manner [15, 29].

Page 197: An Injectable Gelatin-Based Conjugate Incorporating EGF

197

Three analyses were undertaken to gain insights into the motor recovery function post

implantation of the Gtn-HPA matrices with or without EGF and/or SDF-1α – BBB scores over four

weeks, improvement in the BBB scores from first week to fourth week and the number of animals

showing the extent of improvement - poor (0-4 points), good (4-8 points) and better improvement

(8+ points).

Immediately after the 2 mm hemi-resection injury, animals showed complete paralysis in

both left and the right hindlimb. This is consistent with the observations of the previous hemi-

resection studies [30]. Paralysis in both hindlimbs is probably due to the presence of corticospinal

tracts in both hemispheres of the cord. Corticospinal tract (CST) is predominantly responsible for

the voluntary motor function with corticobulbar tract. Lesion in the CST tract sever 90% fibers on

the ipsilateral side while severs 10% on the contralateral side causing deficits in both ipsilateral

and the contralateral hindlimb of the lesion [31]. Despite the hemi-resection injury, most animals

showed some level of gain in the functional score on the BBB scale over four weeks (Figure 6.5).

Modest recovery was observed in the initial week across treatment groups which is attributed to

recovery from the spinal shock and ‘re-wiring’ to facilitate neuronal signal transmission [32].

However, after four weeks, EGF-delivered with Gtn-HPA groups (i.e. GE8, GES8, GE12 and

GE12T) showed significantly better recovery in the voluntary motor ability compared to the control

(N) and EGF in suspension group (E) at four weeks (Figure 6.6). After four weeks, animals in the

N, E and G8 groups showed modest recovery improving the average BBB score to 4-6 points,

animals in G8 group improved to average of 7 points while EGF-delivered groups (i.e. GE8, GES8,

GE12 and GE12T) showed improvement to 8-11 points on the BBB scale in both left and right

hindlimbs. The extent of the recovery in absolute BBB scores observed four weeks post hemi-

resection SCI in our study is consistent with the results of the similar studies [33-35]. Moreover,

there were no differences between G8, N and E groups. This indicates that Gtn-HPA gel alone is

not able to promote better functional recovery but a growth factor such as EGF is necessary to

be delivered locally to the injury site.

In order to account for the differences in the baseline scores of the animals, each rat was

treated as its own control to compare the differences in the scores at 1 (acute) and 4 (chronic)

weeks post injury. It is important to consider that since the BBB scale is not linear meaning that

the improvement from a score of 0 to 4 is different than the improvement in the score from 4 to 8

even though both would show net improvement by four points. Interestingly, the number of

animals that improved by more than 8 points were greater in the 8% or 12% Gtn-HPA delivering

EGF groups compared to the control and 8% Gtn-HPA gel only groups. Chi-square analysis

Page 198: An Injectable Gelatin-Based Conjugate Incorporating EGF

198

showed statistically significant association between the treatment group and extent of

improvement in the BBB scores by four weeks for the left hindlimb but not for the right hindlimb

(Table 6.6). Similarly, EGF-delivered Gtn-HPA groups showed greater rate of improvement in the

BBB scores in three weeks compared to the control and EGF in suspension group (Figure 6.7).

However, it is difficult to pinpoint the exact underlying mechanism due to which more animals

showed better improvement in the EGF-delivered Gtn-HPA gel injected groups compared to the

controls and EGF in suspension, given highly sophisticated nature of the neural tissue response

after SCI. However, few mechanisms are proposed in the field.

The field recognizes that the mechanisms of motor recovery in the rats with or without

specific interventions are poorly understood [36]. Various explanations of recovery include the

greater sparing of the axons, intrinsic changes in the circuitry termed as neuroplasticity, and

stimulation of the lower motor neurons not directly affected by the injury to the upper motor neuron

[37, 38]. Still, it is well known that animals that underwent lateral hemi-resection spinal cord

injuries could show notable improvement in the motor function recovery on the ipsilateral hindlimb

as observed in our study [39]. Most probable explanation is the neuroprotection of the tissue by

mitigating further secondary damage caused by the injury due to implanted Gtn-HPA matrices

with EGF. This reasoning could explain the greater net improvement in BBB scores in the Gtn-

HPA injected groups with EGF compared to the N and E reported in this study. Similar

phenomenon has been observed in various studies where the implantation of a biomaterial or

exogeneous cells has shown neuroprotective effects in preventing further degradation of the

unaffected spinal cord tissue and thereby showing better functional recovery in biomaterial or

exogenous cells treated groups [9, 35, 40, 41]. Several mechanisms of the recovery point

specifically to the extent of spared tissue present in the injury site after SCI. Spared tissue

(unaffected tissue) on the other hemisphere can contribute to re-organizing of the wiring to re-

establish the connection with spinal circuits such as central pattern generators (CPGs) nearby

the injury site [42, 43]. In addition, pre-existing spared cross-over CST neurons in the spared

contralateral side could account for the ability to use the limbs on the same side as the lesion as

reported in primates [44]. Another mechanism suggests that reorganization of the local spinal

circuitry from the spared tissue and input signal from sensory afferents could be sufficient to

generate movement of the limbs via CPGs [43]. Interestingly, our spared tissue evaluation

indicates statistically significant positive correlation between the percent spared tissue and

functional BBB score after four weeks (Figure 6.9) supporting the mentioned mechanisms

hypothesizing that greater spared tissue leads to greater functional recovery. EGF delivery with

the Gtn-HPA matrix probably promoted better preservation of the secondary damage and

Page 199: An Injectable Gelatin-Based Conjugate Incorporating EGF

199

therefore, we believe that EGF delivery with 8% or 12% Gtn-HPA gel, contributing to the

recruitment of endogenous cells, re-organizes the tissue response in the injury site leading to

greater presence of spared tissue in the treatment groups leading to greater functional recovery.

SDF-1α, as seen in all the three metrics of the functional evaluation, did not confer any

added advantage in the functional recovery compared to EGF. Moreover, there was no difference

between the 8% and 12% Gtn-HPA groups in functional recovery declaring that increased gelatin

content did not interfere with motor recovery mechanisms. Additionally, there were no differences

between the control and EGF in suspension group indicating that EGF without Gtn-HPA is not

able to localize in the injury site for longer time to exhibit any beneficial effect in the functional

recovery. In comparison with the complete resection model evaluation presented in Chapter 4,

rats showed greater motor function recovery explained by the difference in the severity of the

lesion. No spared tissue was present in the complete transection rats accounting for very minimal

recovery compared to the recovery observed in the hemi-resection rats four weeks post

implantation of the 8% Gtn-HPA gel. Future studies should validate the functional evaluation

findings presented in this chapter with other behavioral tests and electrophysiological

measurements.

6.5.2. Visualizing Gtn-HPA using MRI: spinal columns from all seven hemi-resection

groups after animal sacrifice were imaged with T1-weighted inversion recovery (T1IR) with the

protocols developed in a systematic approach described in Chapter 5. T1IR technique was able

to reliably distinguish the Gtn-HPA from the surrounding tissue. The location of the Gtn-HPA was

at the injury site as expected mostly in the left hemisphere of the cord. The analysis, performed

in randomized and blinded manner, showed significantly greater presence of the Gtn-HPA in Gtn-

HPA injected groups than the control and EGF in suspension groups indicating that the technique

is sensitive to visualizing only Gtn-HPA in the injury site in the spinal cord. Moreover, average

Gtn-HPA gel volume across all the Gtn-HPA implanted groups was found to be ~ 5.67 mm3

suggesting that out of 20 mm3 volume of the Gtn-HPA injected, 75% of the Gtn-HPA gel had

degraded by four weeks (figure 6.9). However, there were no differences in the Gtn-HPA volume

in Gtn-HPA injected groups, either in 8% Gtn-HPA gel or 12% Gtn-HPA gel. This finding suggests

that the qualitative degradation rate of the 8% and 12% Gtn-HPA gel in vivo is similar. Although

the protocol reliably distinguished the Gtn-HPA gel from the surrounding tissue, better

optimization of the protocol would need to be done before it is replicated for longitudinal in vivo

visualization of the Gtn-HPA.

Page 200: An Injectable Gelatin-Based Conjugate Incorporating EGF

200

6.5.3. Histological evaluation of the injury site: Histological evaluation was performed

four weeks after inducing SCI in coronal sections. Based on the morphology of the tissue from

Day 0 and Day 1, it was evident that the surgery had indeed created a hemi-resection defect on

the left hemisphere of the cord (Figure 6.14). Day 0 injury site showed intact contralateral tissue

with Gtn-HPA contiguous block filling in the cavity on the left hemisphere where hemi-section

injury was induced. There was separation between the Gtn-HPA gel and the surrounding tissue

because there is reported to be retraction of the cord after a surgical excision. Similar morphology

of the tissue and Gtn-HPA gel was observed on Day 1 as well. The separation between the Gtn-

HPA gel and surrounding tissue could be as a result of retraction of the cord and some acute

inflammatory degradation of the end of the cord. However, Gtn-HPA had already formed islands

by Day 7 with notable infiltration of the cells in the injury site. By day 7, some mechanical breakup

of the gel due to the animal movement and additional acute inflammatory degradation of the gel

could contribute to the residual islands of gel. Moreover, across all the Gtn-HPA injected groups,

Gtn-HPA was present in most animals four weeks post implantation in the injury site in form of

Gtn-HPA islands consistent with the findings in Chapter 4 (Figures 6.12 and 6.13). It is important

to note that our study shows for the first time 8wt% and 12wt% Gtn-HPA gel is present in the

injury site after four weeks. Presence of Gtn-HPA four weeks after implantation is a significant

improvement over the previous Collagen-genipin gel formulation. Most or all of the Collagen-

genipin gel had degraded completely by as early as two weeks in the similar 3 mm hemi-resection

SCI model [45]. It is essential for the biomaterial to persist for longer duration to facilitate the

infiltration of cells into the defect, better cell-substrate signaling, remodel the underlying ECM and

deliver signaling molecules over long period of time [46]. As far as unaffected viable parenchyma

is concerned, the histological sections displayed typical characteristics of central gray matter in

both hemispheres in both the caudal and rostral sides of the injury site. The length of the injury

site was more than 2 mm supporting the well-known phenomenon that the contralateral tissue in

the hemi-resection model suffers secondary damage further progressing the injury [47, 48].

Although no treatment was applied to the control group [N], the injury site was filled two

kinds of morphology – ‘empty spaces’ and dense collagen fibrous tissue (Figure 6.11A). Similar

morphological findings were observed in the EGF in suspension group (Figure 6.13). The empty

spaces probably represent the cystic cavitations, which are formation of fluid-filled cysts in

response to the injury. These cavities are usually filled with macrophages and the growth-

inhibiting molecules deposited in response to degradation of myelin [49]. These cavitations are

consistent in the chronology of the pathophysiological response because these cysts are

developed in the subacute and intermediate phase of the response – which coincides with the

Page 201: An Injectable Gelatin-Based Conjugate Incorporating EGF

201

four-week evaluation of our study [50, 51]. We did observe minimal spared tissue in both N and

E groups (17.5 and 15.5% respectively) suggesting that typical pathophysiological process after

SCI would lead to greater secondary damage compared to replacing the cavity contents with

growth-promoting materials. The extent of spared tissue present is consistent with the findings in

similar hemi-resection models in rats [52]. In addition, ferritin deposits were observed in both the

control and EGF in suspension group indicating the remains of the blood clot that was formed due

to the bleeding either from cord parenchyma or from surrounding muscles after creating the

surgical injury. It is possible that the blood clot provided a temporary matrix to the infiltrating

inflammatory cells especially because the dura membrane was compromised due to the surgical

nature of the model. Both the morphologies – cysts and dense fibrous collagen tissue is inhibitory

to the cellular migration into the cavity site for any potential reparative response. Therefore, goal

of this study to replace the growth-inhibitory environment with that of a biomaterial that would be

permissive of the cellular infiltration without eliciting major inflammatory response. Therefore, our

proposed therapy involves injecting a Gtn-HPA matrix in the defect site with chemoattractants,

EGF and SDF-1α to recruit endogenous neural stem cells and glial cells to the injury site.

There were two types of Gtn-HPA gel formulations implanted in the cavity created by left

lateral 2-mm hemi-resection in the cord – 8wt% Gtn-HPA and 12% Gtn-HPA gel. The gross

histological findings were similar in both 8% and 12% Gtn-HPA gels. After four weeks, Gtn-HPA

gel covered the predominant area in the injury site in all the treatment groups (i.e. G8, GE8, GES8,

GE12 and GE12T). The presence of the Gtn-HPA matrix in most animals in all the treatment

groups was assuring that the Gtn-HPA could serve as a reliable scaffold to promote regenerative

response after SCI. However, the number of animals that showed the presence of Gtn-HPA varied

across the groups. Only 2/6 of the GES8 groups showed presence of the Gtn-HPA gel while 4/6

animals showed Gtn-HPA in G8 and GE8 groups. We believe there are few potential explanations

that account for the absence of the Gtn-HPA gel in animals that were injected with Gtn-HPA matrix

in this study. First, bleeding in the cavity site during the injection could cause incomplete gelation

or more porous Gtn-HPA gel leading to a less stiff Gtn-HPA gel than anticipated. This would result

in faster degradation of the gel in vivo. Second, the animals were quick to gain consciousness

after the surgery and did movements that could lead to dislodging of the Gtn-HPA gel from the

cavity because the dural replacement collagen membrane was placed on the cord and not sutured

to the cord. Third, some animals could show more severe inflammatory response due to which

the Gtn-HPA gel degraded quickly. These potential explanations were considered in designing a

second generation Gtn-HPA gel and in better containing the gel in the cavity site. We made two

major improvements in the second generation Gtn-HPA formulation – we increased the gelatin

Page 202: An Injectable Gelatin-Based Conjugate Incorporating EGF

202

content from 8 to 12% to reduce the degradation rate of the Gtn-HPA gel [GE12 group] and

modified the placement of dural replacement by ‘tucking-in’ the membrane around the cord only

to prevent the Gtn-HPA gel from dislodging itself from the injury cavity [GE12T group]. ‘Tucking-

in’ the dural replacement around helped contain the Gtn-HPA in the injury site as all 6/6 animals

showed the Gtn-HPA present four weeks after implantation. Future studies in our laboratory would

employ this ‘tucking-in’ technique due to the improved results reported in this chapter.

Another interesting feature of the morphology of the Gtn-HPA gel was the shape of the

Gtn-HPA gel after four weeks in the Gtn-HPA implanted groups. It was predominantly present in

the injury site in form of multiple ‘islands’ of varying sizes (Figure 6.11B-D and Figure 6.15A) as

opposed to a continuous block (Figure 6.15B). These islands could mimic the small cavitary

lesions, which are common clinical manifestation of the SCI due to various disease progressions

[53]. We believe there could be several mechanisms that could leads to the formation of ‘gel

islands. First, heterogeneity in the modulus of the Gtn-HPA gel during gelation could contribute

to mechanical differences in the gel. These differences could result in varying rates of the gel

degradation leading to the break in the continuous matrix resulting in formation of gel islands over

time. Second, the number and the types of cells present around the Gtn-HPA during acute phase

of the pathophysiological response could affect the extent of cellular migration into the Gtn-HPA

matrix. For example, macrophages are shown to occupy the injury site as early as a week after

the SCI and release various inflammatory cytokines and ECM degrading molecules, especially

Matrix metalloproteinases (MMPs) [54, 55]. MMP2 is a zinc-based gelatinase and we observed

strong MMP2 signal at the border of the Gtn-HPA gel islands in this study (Figure 6.24). Therefore,

it is possible that activated macrophages released significant amount of MMP2 at the border of

the continuous Gtn-HPA block resulting in formation of gel islands during the four weeks. Third,

animal movement immediately after the surgery and within initial weeks could mechanically break

the Gtn-HPA into physically separate entities leading to the formation of gel islands. Histological

evaluation as early as 7 days post implantation also showed these islands form (Figure 6.14C).

Another feature important to discuss is the ability of the Gtn-HPA to establish a continuous

physical interface with the surrounding tissue to re-establish the lost framework due to the injury.

We observed that Gtn-HPA alone group [G8] was not able to establish an interwoven physical

structure with the surrounding tissue (Figure 6.12B). However, GE8, GES8, GE12 and GE12

groups displayed a well-defined continuous interface with the surrounding tissue in the injury site

(Figure 6.12 CD, Figure 6.15CD). Similar observations regarding the interface in the injury site

were made in the complete section model in Chapter 4. They are in accordance with the findings

Page 203: An Injectable Gelatin-Based Conjugate Incorporating EGF

203

in various studies reporting that matrices without any form of the molecular cues failed at

establishing a well-defined continuous framework with the surrounding tissue in tissue repair

approach [56, 57]. Therefore, signaling molecules such as EGF and SDF-1α are critical to include

in the injection of Gtn-HPA matrices.

One of the hypotheses of the study was that EGF and SDF-1α would promote the

recruitment of different endogenous cells types such as the neural stem cells and other glial cells

to the injury site. Qualitatively, greater number of cells had migrated to surround the Gtn-HPA gel

islands in all GE8, GES8, GE12 and GE12T groups (Figure 6.11 and Figure 6.13) compared to

the Gtn-HPA gel only [G8] group (Figure 6.10B and Figure 6.12B). Similar results were reported

in the complete resection model in Chapter 4 confirming the ability of growth factors EGF and

SDF-1α in chemotactically recruiting endogenous cells to the injury site. Moreover, this

phenomenon highlighted in our study is in accordance with the conclusion that both 8% and 12%

Gtn-HPA gel is permissive of the cellular migration both in vitro and in vivo brain ICH model as

reported based on previous work in our laboratory [20] . Increased stiffness, such as in 12% Gtn-

HPA gel, could hinder the cellular migration into the injury site [58]. However, cellular migration

observed in the 12% Gtn-HPA seems to indicate that the stiffness of the 12% Gtn-HPA gel is

within the range of stiffness of materials that cells could migrated through. It would be interesting

to see if there are quantitative differences in the gel area and migration area between the groups

with 8% and 12% Gtn-HPA gels. Quantitative analysis of the Gtn-HPA gel area and cellular

migration immediately surrounding the Gtn-HPA gel islands showed that EGF and SDF-1α were

able to recruit significantly greater number of cells to the Gtn-HPA gel islands compared to the

Gtn-HPA gel only group (Figure 6.16). However, cellular migration area was similar for 8% and

12% Gtn-HPA gel suggesting that gelatin content within the range of 8-12% has no effect on the

cellular migration. We validated this result with two different quantification methods measuring

two parameters – gel area and cellular migration area immediately surrounding the gel. The first

method accounted for the all the gel islands and cellular migration immediately around all of them

present throughout the injury site. The second method measured both the parameters in randomly

chosen images in the injury site. Statistical analyses of these parameters in both the methods

showed significantly greater recruitment of cells by EGF and SDF-1α compared to the control and

Gtn-HPA only group (Figure 6.16BD). Therefore, we deduce that Gtn-HPA gel alone would not

be able to recruit cells to the injury site and therefore, signaling molecules, i.e. EGF and SDF-1α,

are essential in recruiting endogenous cells to the injury site after SCI. This is a critical result

supporting our central hypothesis of the study. Of note, GES8 group did not show significantly

greater cellular migration compared to the GE8 groups indicating that presence of SDF-1α did not

Page 204: An Injectable Gelatin-Based Conjugate Incorporating EGF

204

enhance the recruitment of cells to the injury site – similar to the observation in the complete

resection model discussed in Chapter 4. As far as the Gtn-HPA gel area is concerned, there were

no differences among the Gtn-HPA injected groups in both the methods (Figure 6.16AC)

suggesting that introduction of the growth factor has no effect on the area of the gel present in

the injury site. Moreover, if gel area is any measure of the Gtn-HPA volume in the injury site, MRI

measurement of Gtn-HPA volume also showed no differences in the Gtn-HPA gel volumes

between the Gtn-HPA injected groups (i.e. G8, GE8, GES8, GE12 and GE12T) as shown in figure

6.10. In addition, there were no differences in the Gtn-HPA volume in groups with varying gelatin

content in the Gtn-HPA - either in 8% Gtn-HPA gel or 12% Gtn-HPA gel. This finding suggests

that the qualitative degradation rate of the 8% and 12% Gtn-HPA gel in vivo is similar.

Subsequent to the promising histological observations and conclusions, our goal was to identify

the cells that migrated around the gel islands and in the injury site.

6.5.4. Immunohistochemistry evaluation of the injury site: Several biomarkers were

stained to identify the presence and the extent of various cell types and important molecules in

the context of this study. Specifically, GFAP (astrocytes), Iba1 (microglia/activated microglia),

CD68 (macrophages), NeuN (mature neurons), Nestin (neural progenitor stem cells), Vimentin

(neural progenitor stem cells), Musashi-1 (neural progenitor stem cells), vWF (endothelial cells),

EGF and MMP2 (Gtn-HPA gel degradation) were stained to obtain a comprehensive composition

of the injury site post Gtn-HPA implantation with or without EGF and/or SDF-1α.

Astrocytes have been shown to play a multi-functional role in the context of repair and

regeneration after SCI. The multifaceted astrocytic response contributes to neuroprotection of the

spinal cord while inhibiting axonal regeneration [59]. Astrocytes are typically categorized into three

types in the CNS – naïve astrocytes, reactive astrocytes and scar-forming astrocytes [60]. Naïve

astrocytes help in maintaining the neuronal parenchyma and the blood-brain barrier while the

reactive astrocytes assist in limiting the spread of inflammatory cells in response to injury

promoting remodeling of the ECM and repair of the tissue [61]. These reactive astrocytes convert

to scar-forming astrocytes based on the environmental cues. They form the scar by adhering to

each other and releasing molecules like CSPG to produce a penetrating glial scar that impedes

axonal regeneration after SCI [62, 63]. We found the presence of GFAP+ cells in the unaffected

tissue and in the injury site but predominantly at the interface between the unaffected tissue and

the injury site. Quantification of GFAP+ cells showed no statistical differences between the groups

in the injury site (Figure 6.18D). Moreover, the cellular migration immediately around the Gtn-HPA

gel islands stained negative for the GFAP indicating that those cells are probably not reactive

Page 205: An Injectable Gelatin-Based Conjugate Incorporating EGF

205

astrocytes because GFAP is considered to be the hallmark biomarker for reactive astrocytes

(Figure 6.19) [64]. However, quantification of the GFAP+ cells at the interface between the

unaffected tissue and the injury site showed significant reduction in the GFAP+ cells or astrocytes

at the interface in the Gtn-HPA injected groups (i.e. G8, GE8, GES8, GE12 and GE12T) compared

to the control group and EGF in suspension group indicating that implantation of Gtn-HPA matrix,

with or without the growth factors, into the cavity downregulates the astrocytic response near the

injury site (Figure 6.18D) . This phenomenon probably either delays the process of glial scar

formation or reduces the extent of glial scar formation in the Gtn-HPA injected groups. Similar

findings of reduced GFAP immunodensity was reported by Alluin et al at the interface of the injury

site [65]. Interestingly, one hypothesis proposed by Macaya in his doctoral work stated that the

elongated astrocytic processes towards the center of the defect are growth-promoting while the

astrocytes bordering the injury site with dense processes are inhibitory [45]. Similar morphologies

were observed in our evaluation at the interface and in the injury site. Our results show that

although the treatment groups do not increase the presence of the growth promoting astrocytes

in the injury site, the treatment reduces the number of the inhibitory astrocytes at the border.

Similar morphology remodeling has been characterized in the previous study by Sun and Jakobs

supporting the hypothesis proposed by Macaya [66]. Results in our evaluation provide critical

evidence to suggest that the treatment reduces the glial astrogliosis leading to the beneficial repair

and migration into the injury site.

Post-injury, the resident microglia migrate to the injury site while the monocytes

differentiate into macrophages in the acute phase of the pathophysiological process after SCI.

Since they are phagocytic in nature, they contribute to the debris clearing in the injury site to

mitigate the secondary damage [21, 67]. Macrophages have been classified into two phenotypes

– M1 and M2. M1 is shown to be promoting the release of pro-inflammatory factors while M2

contributes in wound healing and repair. Interestingly, M1 have been shown to have detrimental

effect in the SCI but M2 macrophages have shown to promote regenerative response after SCI

[68]. Iba1 is a microglia/macrophage specific marker that was used to assess the inflammatory

response to the Gtn-HPA gel while CD68 was stained to assess the immune reaction to the injury.

It is reported that immune reaction to tissue damage by macrophages is not well detected by Iba1

necessitating the staining of CD68 [22].The morphology of the Iba1+ cells varied depending upon

the location on the spinal cord section (Figure 6.20A-D). Unaffected tissue had typical ramified

phenotype cells with its processes indicating that these are resident microglia. Injury site

predominantly had enlarged macrophages indicating that these were the activated microglia or

macrophages responding to the injury and contributing the clearing the debris from the myelin

Page 206: An Injectable Gelatin-Based Conjugate Incorporating EGF

206

degradation and other necrotic tissue. Spared tissue and the interface between the unaffected

tissue and the injury site had combination of both phenotypes possibly indicating that the resident

microglia are being recruited to the injury site in response to the cytokines released in the injury

site. This observation is consistent with the mechanism of the activation of microglia after SCI

[69]. Quantitative analysis did not show statistical differences among the number of

microglia/macrophages present in the injury site across all the groups including the control (Figure

6.20E). This result was validated using the CD68 staining across all the seven groups.

Morphology of the CD68+ cells was spheroidal indicative of activated macrophages phenotype.

There were very few CD68+ signal in the migrated cells around the Gtn-HPA islands in GE8,

GES8 and GE12 and GE12T groups but the morphology of the cells was not spheroidal possibly

suggesting that these were not macrophages. Even if they are truly macrophages, their number

was negligible compared to the numerous migrated cells. Similar to Iba1, analysis revealed no

differences between the groups in the injury site suggesting that Gtn-HPA does not elicit a severe

immune response greater than that observed in the control group. These consistent findings with

Iba1 and CD68 indicates that injection of Gtn-HPA, with or without the factors, did not elicit a

severe inflammatory response providing an in vivo evidence that Gtn-HPA is a biocompatible

biomaterial. Gtn-HPA and its constituents were well tolerated by the rats in the study. This is an

important finding in vivo for the Gtn-HPA material injected for the first time in the SCI models.

NeuN is a very specific biomarker for nuclear protein expressed in mature neurons. It also

has been used to study the differentiation of stem cells into neurons [70, 71]. Representative

sections from all the treatment groups were stained with NeuN to assess whether any neuronal

lineage was promoted by the potential regenerative effects of the migrated cells in the injury site.

Across all the treatment groups, we found that there were no NeuN+ cells in the injury site while

we were able to visualize characteristic neuronal cell body stained with NeuN in the unaffected

tissue and at the interface of the unaffected tissue and the injury site. Qualitatively, the number of

NeuN+ cells present in the unaffected tissue were higher than at the interface (Figure 6.23). No

NeuN+ cells were found in the cells migrated immediately surrounding the Gtn-HPA islands

indicating that Nestin+ NSCs did not have the cues required for differentiation into mature

neurons. NeuN staining is consistent with the NeuN images from complete resection model in

Chapter 4. We did not anticipate the presence of NeuN+ cells in the injury four weeks after Gtn-

HPA implantation. However, future studies should quantitatively investigate any differences

among the groups for the presence of neuronal differentiation in the injury site and near the

interface for longer duration post Gtn-HPA matrix injection.

Page 207: An Injectable Gelatin-Based Conjugate Incorporating EGF

207

Gtn-HPA is a gelatin-based biomaterial and has excellent tunable properties to control the

gelation time and stiffness by varying the concentration of its constituents [13]. In our study, Gtn-

HPA was used in an in vivo application for the SCI for the first time. Therefore, it was important

to comment on the potential mode of degradation of the Gtn-HPA in vivo. Although Gtn-HPA was

injected in the brain ICH model, it’s in vivo degradation behavior was not studied previously [20].

Matrix metalloproteinases (MMPs) were reported to be involved into the cellular migration of

neural progenitor cells [72]. Preliminary MMP2 staining in the complete resection models showed

strong signal specifically at the border of the Gtn-HPA gel island in the Gtn-HPA injected groups.

Similar qualitative observations were noted in hemi-resection model as well. Since the MMP2

signal was strong only at the border of the gel island, we hypothesize that MMP2 is responsible

for the degradation of Gtn-HPA gel to promote cellular infiltration into the Gtn-HPA matrix.

Remarkably, MMP2 is a gelatinase A enzyme and therefore could be the agent to degrade gelatin-

based Gtn-HPA biomaterial in vivo. Upon quantification, migrated cells around the Gtn-HPA

islands in EGF-delivered with Gtn-HPA groups (i.e. GE8, GES8, GE12 and GE12T) released

significantly greater MMP2 compared to the Gtn-HPA gel only group [G8] (Figure 6.24). This could

explain the reason behind observing Gtn-HPA gel only in 2/6 animals in GES8 group. One

potential hypothesis is that presence of both chemoattractants led to greater cellular migration,

thereby increasing the MMP2 release near the Gtn-HPA islands resulting in faster degradation of

the Gtn-HPA. In addition to MMP2, EGF was the other protein stained with few representative

sections to confirm the presence of EGF in the GE8, GES8, GE12 and GE12T groups. The

stained micrographs confirmed the presence of EGF exclusively in the Gtn-HPA gel island as

reported in preliminary images in Chapter 4 (Figure 6.25). However, the Gtn-HPA gel island in

the Gtn-HPA gel only group [G8] stained negative for the EGF serving as a negative control

(Figure 6.25). Fisher’s exact test confirmed the association between the treatment group and EGF

presence in the Gtn-HPA four weeks post implantation. All EGF-delivered groups positively

stained for EGF while all Gtn-HPA gel only groups stained negative for EGF. These results are

reassuring that Gtn-HPA matrix can deliver EGF locally to the injury site at least over four weeks

and agree with the in vitro release of EGF from Gtn-HPA reported in Chapter 3. Furthermore, it

could also provide a reliable method to locate the presence of the Gtn-HPA gel in groups where

EGF was delivered with the Gtn-HPA matrix. Future studies should be designed to assess the

bioactivity of the EGF in vivo. For example, EGF delivered by the Gtn-HPA gel could be pre-

labeled with a fluorophore prior to injection and could be longitudinally followed over four or more

weeks and investigate if the signal is limited to the Gtn-HPA gel location. This experiment could

offer interesting insight into the degradation mechanism and EGF release mechanism in vivo.

Page 208: An Injectable Gelatin-Based Conjugate Incorporating EGF

208

Due to difficulties with vWF staining, we were not able to stain sections from all the groups.

However, few representative rats showed the presence of vWF+ endothelial cells in the injury site

indicating that Gtn-HPA presence could promote angiogenesis in the injury site (Figure 6.32).

The identity of the cells migrated immediately surrounding the gel islands was determined

after sections were stained with Nestin to evaluate the presence of neural stem cellsin the injury

site. Nestin is an intermediate type VI protein and is considered as a biomarker for undifferentiated

CNS cells prior to committing to a specific cell lineage, considered to be the neural stem cells

(NSCs) [73]. It is reported that NSCs are found in a niche at the poles of the spinal cord ependymal

layer near the central canal. Similar niche is observed in the SVZ region in the brain [74]. These

niches stain positive for Nestin and negative for GFAP confirming the presence of NSCs [75].

There is a panel of biomarkers that can be used to further confirm the cells to be NSCs and

include Vimentin, Mushashi-1, and Sox 2 (R&D systems, SC205). In our study, cells migrated

around gel islands stained positive for Nestin, Vimentin and Mushashi-1 for EGF-delivered Gtn-

HPA groups (i.e. GE8, GES8, GE12 and GE12T) (Figures 6.26, 6.29 and 6.30). Positive staining

for all three markers in the migrated cells around Gtn-HPA further confirmed the identity of these

cells to be NSCs. In addition, double staining of GFAP/Nestin showed that these cells are not

astrocytes and similar conclusion was drawn after double staining of Vimentin/GFAP (Figure 6.27

and 6.29). Furthermore, EGF/Nestin were double-stained to confirm that NSCs migrate in

response to the presence of the EGF in the Gtn-HPA islands (Figure 6.31). We did not encounter

any Nestin+ cells in the injury site for the control or EGF in suspension group. There were some

cells that stained Nestin positive surrounding the gel islands in the Gtn-HPA gel only [G8].

Quantitative analysis of the Nestin+ cells showed significantly greater number of NSCs that have

migrated to the gel islands in GE8, GES8, GE12 and GE12T groups compared to the N, E and

G8 groups (Figure 6.28). These results show that Gtn-HPA gel alone is not capable of recruiting

important cells such NSCs while EGF and SDF-1α were able to recruit the endogenous NSCs

from the spinal cord ependymal layer niche near the central canal. Furthermore, there was no

difference between the GE8 and GES8 group indicating that SDF-1α either does not play a role

in NSCs recruitment or does not enhance the recruitment of NSCs to the injury site – finding also

observed in complete resection model in Chapter 4. This is consistent with the proposed

hypothesis that NSCs are highly responsive to EGF [76]. Hemann et al reported that intrathecal

administration of EGF resulted in significantly greater ependymal cell proliferation in the central

canal immediately rostral and caudal to the lesion compared to controls and in greater white

matter sparing [77]. Similar behavior of the NSCs was observed in our study in response to EGF

four weeks post Gtn-HPA with EGF injection. The migration of NSCs could explain the functional

Page 209: An Injectable Gelatin-Based Conjugate Incorporating EGF

209

recovery seen in EGF-delivered Gtn-HPA groups for both 8% and 12% Gtn-HPA gels. EGF could

promote better preservation of the spared tissue and reorganization of the injury site leading to

significantly better functional recovery in the rats with hemi-resection injury. Future studies should

include longer duration of evaluation and it would be interesting to find out the identity of the

migrated cells and assess if they differentiate into neural or glial fate to promote regenerative

response and improve functional recovery after the SCI.

In summation, a 2 mm hemi-resection of the spinal cord tissue at T8 level created a

standardized cavity in which 8wt% and 12% Gtn-HPA was injected with EGF and/or SDF-1α. Gtn-

HPA gel persisted in the injury site four weeks post injection. ‘Tucking-in’ the dural replacement

around the spinal cord after Gtn-HPA gel helped better contain the gel in the cavity created by

the hemi-resection in these rats. In the groups that delivered EGF and/or SDF-1α, rats showed

significantly better functional recovery in both urinary bladder function and motor function. Both

8% and 12% Gtn-HPA gels were permissive of cellular migration, making Gtn-HPA a promising

biomaterial for future studies. Injection of the Gtn-HPA matrix decreased the presence of growth-

inhibitory astrocytes at the border of the injury site thereby reducing the extent of glial scar

formation. Gtn-HPA is well tolerated and did not elicit a severe inflammatory response as

suggested by both Iba1 and CD68 biomarkers. Injecting the 8wt% Gtn-HPA matrix with EGF alone

or in combination with SDF-1α promoted recruitment of endogenous neural stem cells into the gel

islands in the injury site confirmed by positive staining for Nestin, Vimentin and Musashi-1.

However, SDF-1α did not enhance the recruitment response. Future studies should be planned

to evaluate the 12% Gtn-HPA with EGF in a contusion model of SCI to evaluate the potential

benefit of EGF delivered via Gtn-HPA to promote reparative healing response and improved

functional recovery as reported in this 2-mm hemi-resection model. The encouraging results in

this study would lay the foundation for further testing in different species of rats and in other SCI

species models such as minipigs and rabbits for eventual clinical application to improve the life

style of SCI patients – the goal that SCI researchers share throughout the world.

Page 210: An Injectable Gelatin-Based Conjugate Incorporating EGF

210

6.6. References

1. Spinal Cord Injury. 2018 [cited 2018; Available from: https://www.hopkinsmedicine.org/healthlibrary/conditions/physical_medicine_and_rehabilitation/spinal_cord_injury_85,P01180.

2. Mahdi, S.-A., Animal Models in Traumatic Spinal Cord Injury in Topics in Paraplegia, R.-M. Vafa, Editor.

3. Ahuja, C.S. and M. Fehlings, Concise Review: Bridging the Gap: Novel Neuroregenerative and Neuroprotective Strategies in Spinal Cord Injury. Stem Cells Transl Med, 2016. 5(7): p. 914-24.

4. Li, J. and G. Lepski, Cell Transplantation for Spinal Cord Injury: A Systematic Review. BioMed Research International, 2013. 2013: p. 32.

5. Haider, H. and M. Ashraf, Strategies to promote donor cell survival: combining preconditioning approach with stem cell transplantation. J Mol Cell Cardiol, 2008. 45(4): p. 554-66.

6. Hagg, T. and M. Oudega, Degenerative and spontaneous regenerative processes after SCI. J Neurotrauma, 2006. 23(3-4): p. 264-80.

7. Donnelly, D.J. and P.G. Popovich, Inflammation and its role in neuroprotection, axonal regeneration and functional recovery after SCI. Exp Neurol, 2008. 209(2): p. 378-88.

8. Hodgetts, S.I. and A.R. Harvey, Neurotrophic Factors Used to Treat Spinal Cord Injury. Vitam Horm, 2017. 104: p. 405-457.

9. Karimi-Abdolrezaee, S., et al., Synergistic Effects of Transplanted Adult Neural Stem/Progenitor Cells, Chondroitinase, and Growth Factors Promote Functional Repair and Plasticity of the Chronically Injured Spinal Cord. The Journal of Neuroscience, 2010. 30(5): p. 1657-1676.

10. Facts and Figures at a Glance, in SCI Data Sheet. 2018, University of Alabama at Birmingham: National Spinal Cord Injury Statistical Center. p. 2.

11. Musker, P. and G. Musker, Pneumocephalus and Brown-Sequard syndrome caused by a stab wound to the back. Emergency Medicine Australasia, 2011. 23(2): p. 217-219.

12. Xiaofei Wang, X.-M.X., Spinal Cord Lateral Hemisection and Implantation of Guidance Channels, in Animal Models of Acute Neurological Injuries, Z.C.X. J. Chen, X.-M. Xu, J.H. Zhang, Editor. 2009.

13. Wang, L.S., et al., Injectable biodegradable hydrogels with tunable mechanical properties for the stimulation of neurogenesic differentiation of human mesenchymal stem cells in 3D culture. Biomaterials, 2010. 31(6): p. 1148-57.

14. Cholas, R., H.P. Hsu, and M. Spector, Collagen scaffolds incorporating select therapeutic agents to facilitate a reparative response in a standardized hemiresection defect in the rat spinal cord. Tissue Eng Part A, 2012. 18(19-20): p. 2158-72.

15. Basso, D.M., M.S. Beattie, and J.C. Bresnahan, A sensitive and reliable locomotor rating scale for open field testing in rats. J Neurotrauma, 1995. 12(1): p. 1-21.

16. Shatil, A.S., K.M. Matsuda, and C.R. Figley, A Method for Whole Brain Ex Vivo Magnetic Resonance Imaging with Minimal Susceptibility Artifacts. Frontiers in neurology, 2016. 7: p. 208-208.

17. Taweel, W.A. and R. Seyam, Neurogenic bladder in SCI patients. Research and reports in urology, 2015. 7: p. 85-99.

18. Kelly, C.E., Evaluation of voiding dysfunction and measurement of bladder volume. Reviews in urology, 2004. 6 Suppl 1(Suppl 1): p. S32-S37.

19. Soler, R., et al., Bladder dysfunction in a new mutant mouse model with increased superoxide--lack of nitric oxide? The Journal of urology, 2010. 183(2): p. 780-785.

Page 211: An Injectable Gelatin-Based Conjugate Incorporating EGF

211

20. Lim, T.C., Injectable Chemokine-Releasing Gelatin Matrices for Enhancing Endogenous Regenerative Responses in the Injured Rat Brain, in Health, Science and Technology. 2014, Massachusetts Institute of Technology: Cambridge, MA.

21. Zhou, X., X. He, and Y. Ren, Function of microglia and macrophages in secondary damage after SCI. Neural regeneration research, 2014. 9(20): p. 1787-1795.

22. Hendrickx, D.A.E., et al., Staining of HLA-DR, Iba1 and CD68 in human microglia reveals partially overlapping expression depending on cellular morphology and pathology. J Neuroimmunol, 2017. 309: p. 12-22.

23. Kaneko, Y., et al., Musashi1: an evolutionally conserved marker for CNS progenitor cells including neural stem cells. Dev Neurosci, 2000. 22(1-2): p. 139-53.

24. Kwon, B.K., T.R. Oxland, and W. Tetzlaff, Animal models used in spinal cord regeneration research. Spine (Phila Pa 1976), 2002. 27(14): p. 1504-10.

25. Lai, B.Q., et al., Graft of a tissue-engineered neural scaffold serves as a promising strategy to restore myelination after rat spinal cord transection. Stem Cells Dev, 2014. 23(8): p. 910-21.

26. Ajay, B., et al., Mechanically engineered hydrogel scaffolds for axonal growth and angiogenesis after transplantation in SCI. Journal of Neurosurgery: Spine, 2004. 1(3): p. 322-329.

27. Linsenmeyer, T.A., Bladder Management for Adults with Spinal Cord Injury: A clinical practice guideline for health-care providers., in Consortium for Spinal Cord Medicine. 2006, Paralyzed Veterans of America. p. 61.

28. Benarroch, E.E., Neural control of the bladder. Recent advances and neurologic implications, 2010. 75(20): p. 1839-1846.

29. Scheff, S.W., D.A. Saucier, and M.E. Cain, A statistical method for analyzing rating scale data: the BBB locomotor score. J Neurotrauma, 2002. 19(10): p. 1251-60.

30. Vipin, A., et al., Natural Progression of Spinal Cord Transection Injury and Reorganization of Neural Pathways. J Neurotrauma, 2016. 33(24): p. 2191-2201.

31. Nolte, J., The Human Brain: An Introduction to Its Functional Anatomy. 5th ed. 2001: Elsevier Health Sciences. 650.

32. Batista, C.M., et al., Behavioral Improvement and Regulation of Molecules Related to Neuroplasticity in Ischemic Rat Spinal Cord Treated with PEDF. Neural Plasticity, 2014. 2014: p. 16.

33. Hwang, D., et al., Transplantation of human neural stem cells transduced with Olig2 transcription factor improves locomotor recovery and enhances myelination in the white matter of rat spinal cord following contusive injury. BMC Neuroscience, 2009. 10(1): p. 117.

34. Arvanian, V.L., et al., Chronic spinal hemisection in rats induces a progressive decline in transmission in uninjured fibers to motoneurons. Experimental neurology, 2009. 216(2): p. 471-480.

35. Kim, Y.C., et al., Transplantation of Mesenchymal Stem Cells for Acute Spinal Cord Injury in Rats: Comparative Study between Intralesional Injection and Scaffold Based Transplantation. J Korean Med Sci, 2016. 31(9): p. 1373-82.

36. Rasmussen, R. and E.M. Carlsen, Spontaneous Functional Recovery from Incomplete Spinal Cord Injury. J Neurosci, 2016. 36(33): p. 8535-7.

37. Muir, G.D. and A.A. Webb, Mini-review: assessment of behavioural recovery following SCI in rats. Eur J Neurosci, 2000. 12(9): p. 3079-86.

38. Maier, I.C. and M.E. Schwab, Sprouting, regeneration and circuit formation in the injured spinal cord: factors and activity. Philos Trans R Soc Lond B Biol Sci, 2006. 361(1473): p. 1611-34.

39. Rossignol, S. and A. Frigon, Recovery of locomotion after SCI: some facts and mechanisms. Annu Rev Neurosci, 2011. 34: p. 413-40.

Page 212: An Injectable Gelatin-Based Conjugate Incorporating EGF

212

40. Kushchayev, S.V., et al., Hyaluronic acid scaffold has a neuroprotective effect in hemisection SCI. J Neurosurg Spine, 2016. 25(1): p. 114-24.

41. Faccendini, A., et al., Nanofiber Scaffolds as Drug Delivery Systems to Bridge Spinal Cord Injury. Pharmaceuticals (Basel), 2017. 10(3).

42. Hilton, B.J., et al., Re-Establishment of Cortical Motor Output Maps and Spontaneous Functional Recovery via Spared Dorsolaterally Projecting Corticospinal Neurons after Dorsal Column Spinal Cord Injury in Adult Mice. J Neurosci, 2016. 36(14): p. 4080-92.

43. Barriere, G., et al., Prominent role of the spinal central pattern generator in the recovery of locomotion after partial spinal cord injuries. J Neurosci, 2008. 28(15): p. 3976-87.

44. Darian-Smith, C., Primary afferent terminal sprouting after a cervical dorsal rootlet section in the macaque monkey. J Comp Neurol, 2004. 470(2): p. 134-50.

45. Macaya, D., Biomaterials-Tissue Interaction of an Injectable Collagen-Genipin Gel in a Rodent Hemi-Resection Model of Spinal Cord Injury in Health, Science and Technology. 2013, Massachusetts Institute of Technology: Massachusetts Institute of Technology. p. 210.

46. Kumar, P., Future biomaterials for enhanced cell-substrate communication in SCI intervention. Future science OA, 2017. 4(2): p. FSO268-FSO268.

47. Kim, C.Y., PEG-assisted reconstruction of the cervical spinal cord in rats: effects on motor conduction at 1 h. Spinal cord, 2016. 54(10): p. 910-912.

48. Wang, Y., et al., Effective improvement of the neuroprotective activity after SCI by synergistic effect of glucocorticoid with biodegradable amphipathic nanomicelles. Drug Deliv, 2017. 24(1): p. 391-401.

49. van Niekerk, E.A., et al., Molecular and Cellular Mechanisms of Axonal Regeneration After Spinal Cord Injury. Molecular & cellular proteomics : MCP, 2016. 15(2): p. 394-408.

50. Fleming, J.C., et al., The cellular inflammatory response in human spinal cords after injury. Brain, 2006. 129(Pt 12): p. 3249-69.

51. Ahuja, C.S., et al., Traumatic SCI. Nat Rev Dis Primers, 2017. 3: p. 17018. 52. Xu, Z.X., et al., Acellular Spinal Cord Scaffold Implantation Promotes Vascular Remodeling with

Sustained Delivery of VEGF in a Rat Spinal Cord Hemisection Model. Curr Neurovasc Res, 2017. 14(3): p. 274-289.

53. Sundarakumar, D.K., et al., Evaluation of Focal Cervical Spinal Cord Lesions in Multiple Sclerosis: Comparison of White Matter–Suppressed T1 Inversion Recovery Sequence versus Conventional STIR and Proton Density–Weighted Turbo Spin-Echo Sequences. American Journal of Neuroradiology, 2016. 37(8): p. 1561-1566.

54. Campbell, E.J., et al., Neutral proteinases of human mononuclear phagocytes. Cellular differentiation markedly alters cell phenotype for serine proteinases, metalloproteinases, and tissue inhibitor of metalloproteinases. J Immunol, 1991. 146(4): p. 1286-93.

55. Elkington, P.T., J.A. Green, and J.S. Friedland, Analysis of matrix metalloproteinase secretion by macrophages. Methods Mol Biol, 2009. 531: p. 253-65.

56. Deguchi, K., et al., Implantation of a New Porous Gelatin–Siloxane Hybrid into a Brain Lesion as a Potential Scaffold for Tissue Regeneration. Journal of Cerebral Blood Flow & Metabolism, 2006. 26(10): p. 1263-1273.

57. Mooney, D.J. and H. Vandenburgh, Cell Delivery Mechanisms for Tissue Repair. Cell Stem Cell, 2008. 2(3): p. 205-213.

58. Lautscham, L.A., et al., Migration in Confined 3D Environments Is Determined by a Combination of Adhesiveness, Nuclear Volume, Contractility, and Cell Stiffness. Biophysical journal, 2015. 109(5): p. 900-913.

59. White, R.E., D.M. McTigue, and L.B. Jakeman, Regional heterogeneity in astrocyte responses following contusive SCI in mice. J Comp Neurol, 2010. 518(8): p. 1370-90.

Page 213: An Injectable Gelatin-Based Conjugate Incorporating EGF

213

60. Okada, S., et al., Astrocyte reactivity and astrogliosis after SCI. Neuroscience Research, 2018. 126: p. 39-43.

61. Herrmann, J.E., et al., STAT3 is a critical regulator of astrogliosis and scar formation after SCI. J Neurosci, 2008. 28(28): p. 7231-43.

62. Okada, S., et al., Conditional ablation of Stat3 or Socs3 discloses a dual role for reactive astrocytes after SCI. Nature Medicine, 2006. 12: p. 829.

63. Nathan, F.M. and S. Li, Environmental cues determine the fate of astrocytes after SCI. Neural regeneration research, 2017. 12(12): p. 1964-1970.

64. Sofroniew, M.V. and H.V. Vinters, Astrocytes: biology and pathology. Acta neuropathologica, 2010. 119(1): p. 7-35.

65. Alluin, O., et al., Examination of the combined effects of chondroitinase ABC, growth factors and locomotor training following compressive SCI on neuroanatomical plasticity and kinematics. PLoS One, 2014. 9(10): p. e111072.

66. Sun, D. and T.C. Jakobs, Structural remodeling of astrocytes in the injured CNS. The Neuroscientist : a review journal bringing neurobiology, neurology and psychiatry, 2012. 18(6): p. 567-588.

67. Wang, X., et al., Macrophages in SCI: phenotypic and functional change from exposure to myelin debris. Glia, 2015. 63(4): p. 635-51.

68. Kigerl, K.A., et al., Identification of two distinct macrophage subsets with divergent effects causing either neurotoxicity or regeneration in the injured mouse spinal cord. J Neurosci, 2009. 29(43): p. 13435-44.

69. Jin, X. and T. Yamashita, Microglia in central nervous system repair after injury. The Journal of Biochemistry, 2016. 159(5): p. 491-496.

70. Verdiev, B.I., et al., Molecular genetic and immunophenotypical analysis of Pax6 transcription factor and neural differentiation markers in human fetal neocortex and retina in vivo and in vitro. Bull Exp Biol Med, 2009. 148(4): p. 697-704.

71. Gusel'nikova, V.V. and D.E. Korzhevskiy, NeuN As a Neuronal Nuclear Antigen and Neuron Differentiation Marker. Acta naturae, 2015. 7(2): p. 42-47.

72. Lim, T.C., et al., Chemotactic recruitment of adult neural progenitor cells into multifunctional hydrogels providing sustained SDF-1alpha release and compatible structural support. Faseb j, 2013. 27(3): p. 1023-33.

73. Suzuki, S., et al., The neural stem/progenitor cell marker nestin is expressed in proliferative endothelial cells, but not in mature vasculature. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society, 2010. 58(8): p. 721-730.

74. Hamilton, L.K., et al., Cellular organization of the central canal ependymal zone, a niche of latent neural stem cells in the adult mammalian spinal cord. Neuroscience, 2009. 164(3): p. 1044-56.

75. Namiki, J., et al., Nestin protein is phosphorylated in adult neural stem/progenitor cells and not endothelial progenitor cells. Stem Cells Int, 2012. 2012: p. 430138.

76. Lillien, L. and H. Raphael, BMP and FGF regulate the development of EGF-responsive neural progenitor cells. Development, 2000. 127(22): p. 4993-5005.

77. Jimenez Hamann, M.C., C.H. Tator, and M.S. Shoichet, Injectable intrathecal delivery system for localized administration of EGF and FGF-2 to the injured rat spinal cord. Exp Neurol, 2005. 194(1): p. 106-19.

Page 214: An Injectable Gelatin-Based Conjugate Incorporating EGF

214

Chapter 7:

Conclusions

Page 215: An Injectable Gelatin-Based Conjugate Incorporating EGF

215

The work presented in this thesis dealt with developing and evaluating an injectable

gelatin-based matrix (Gtn-HPA) incorporating EGF and SDF-1α in promoting tissue repair and

functional recovery after spinal cord injury. Specifically, Gtn-HPA was formulated to be injected

into a 2-mm cavity created by complete and hemi-resection SCI models. Following are the

conclusions which can be drawn from the data generated in this thesis:

1. EGF can be directly incorporated into a Gtn-HPA gel for its sustained release. In a neutral

buffer in vitro, 30% of the growth factor remains in the gel after 4 weeks. In vivo,

immunohistochemistry confirms that, after 4 weeks, EGF is present in gel remaining in the SCI

defect. By extension, SDF-1α, which has a molecular weight (10.6 kDa) larger than EGF (6.2

kDa) would likely display comparable release kinetics. These results are important in ruling out

the necessity of implementing secondary nano- and micro-meter particles as release vehicles for

EGF, and SDF-1α. Moreover, Gtn-HPA gel was persistent in the injury site after 4 weeks – a

significant improvement from the previously designed collagen-genipin gel for SCI.

2. Injection of Gtn-HPA/EGF immediately after SCI (a 2 mm long hemi-resection defect) provides

a notable neuroprotective effect resulting in a statistically meaningful improvement in behavioral

outcome at 4 weeks. However, injection of Gtn-HPA/EGF immediately after SCI (a 2 mm complete

resection defect) did not result in a statistically improved functional outcome at 4 weeks, likely

due to the extreme nature of the injury.

3. Gtn-HPA gel incorporating EGF accommodates the migrating of endogenous NSCs, clearly

demonstrated at 4 weeks post-injury. The results show that the gel is necessary, but not sufficient,

to enable the gel-filled defect to be infiltrated by cells with the potential to regenerate the neural

circuitry. EGF delivered without Gtn-HPA did not result in migration of endogenous NSCs

indicating the importance of delivery via Gtn-HPA. These results were consistent in both 2 mm

complete resection and 2-mm hemi-resection SCI models. Moreover, SDF-1α did not confer any

additional benefit of migration of endogenous NSCs into the Gtn-HPA gel.

4. An MRI protocol can be implemented to enable the visualization of the Gtn-HPA gel and its

distinction from spinal cord tissues and CSF. Specifically, T1-weighted inversion recovery (T1IR)

sequence with inversion time of 450 ms resulted in visualization of the Gtn-HPA gel.

5. Injection of Gtn-HPA, alone or incorporating EGF, resulted in reduced presence of reactive

astrocytes at the border outlining the injury area indicating delayed or reduced formation of glial

scar. However, injection of Gtn-HPA, alone or incorporating EGF, does not affect the number of

reactive astrocytes present in the injury site.

Page 216: An Injectable Gelatin-Based Conjugate Incorporating EGF

216

6. Gtn-HPA was found to be biocompatible because it did not elicit any severe inflammatory

response based on macrophage marker immunohistochemistry. Moreover, Gtn-HPA was present

in the injury site in form of ‘islands.

7. We found no differences between the 8% and 12% Gtn-HPA across various histological and

functional outcomes indicating that the 12% Gtn-HPA is also permissive of the cellular migration

into the Gtn-HPA gel.

Page 217: An Injectable Gelatin-Based Conjugate Incorporating EGF

217

Chapter 8:

Limitations and Future directions

Page 218: An Injectable Gelatin-Based Conjugate Incorporating EGF

218

8.1. Limitations of the thesis: Despite the extensive assessment performed to characterize

and evaluate Gtn-HPA/EGF/SDF-1α both in vitro and in vivo for spinal cord injury repair, there

were several limitations of this research work.

In vitro characterization:

• Although preliminary mechanical testing work was done specifically with 8% and 12% Gtn-

HPA to be used in our formulation, the technical difficulties with the shape and

homogeneity of the Gtn-HPA did not allow for an extensive assessment of the mechanical

properties of the gel to more accurate comparison with the stiffness of the spinal cord.

Moreover, the experiments characterized the stiffness modulus of the pre-formed Gtn-

HPA gels, the gelation rate profile for our formulations was not conducted due to the

evaporation issues encountered during preliminary trials. A proper mold for the fabrication

should be prepared for comparative analysis of 8% and 12% Gtn-HPA gel.

• EGF release profile from the 8% and 12% Gtn-HPA was studied for four weeks. However,

similar study was not planned with SDF-1α because SDF-1α was not incorporated alone

in the in vivo group. Moreover, we were not able to perform immunohistochemical

evaluation of the Gtn-HPA gel remaining after the four-week study to confirm the presence

of EGF primarily because of technical considerations in sectioning Gtn-HPA. Based on

the experience with the Gtn-HPA in prior work in our laboratory, it was difficult to

cryoprotect the Gtn-HPA gel to avoid artifacts and crystal formation during fast-freezing of

the Gtn-HPA.

• All the in vitro assessments were performed under conditions not encountered by the Gtn-

HPA gel in vivo such as Gtn-HPA degrading enzymes, blood during gelation, mechanical

forces experienced in the cavity during the movement of the animal, and breakage of the

Gtn-HPA gel due to perturbations. These factors could affect the gelation rate, gelation

time, final stiffness and the homogeneity in the Gtn-HPA gel in vivo.

• The bioactivity and encapsulation efficiency of EGF and SDF-1α was not assessed after

being encapsulated in the 8% and 12% Gtn-HPA. Although the in vivo results provide an

indirect mechanism to study the biological effects of EGF, a standardized in vitro assay

was not performed.

• While we found the presence of EGF in the Gtn-HPA gel in vivo after four weeks by

immunohistochemical staining, no comment could be made regarding the bioactivity of the

EGF in vivo.

Page 219: An Injectable Gelatin-Based Conjugate Incorporating EGF

219

In vivo evaluation in SCI model:

• Although standardized surgical excision models employed in this evaluation were ideal to

create a reproducible standardized defect, to visually confirm the injection of the Gtn-HPA

gel in the cavity, and to study regenerative responses in the cavity, contusion models are

more relevant clinical models studied for spinal cord injury. One of the disadvantages of

contusion models, however, is the variability of the defect, with respect to size and

location. Typically, contusion models do not breach the dura and therefore exhibit reduced

inflammatory response. Moreover, the defects produced by contusion would be fully

contained within the cord tissue constraining the gel which would be injected.

Nonetheless, standardized surgical excision models provide a proof of principle of the

biocompatibility of Gtn-HPA and the chemotactic effect of EGF on the migration of

endogenous neural progenitor stem cells.

• Isoflurane was used as an anesthetic during the surgery while sodium pentobarbital was

used in previous studies. Animals are quick to gain consciousness due to the use of

isoflurane disrupting the gelation of Gtn-HPA in vivo by their movements post-operation.

Moreover, we believe that isoflurane causes greater bleeding in the musculature during

the surgery affecting the in-situ gelation of Gtn-HPA.

• The injection of the gel was timed to be immediately after the creation of the surgical defect

in order to enable direct visualization of the gelation process and to provide assurance

that the gel filled the defect. One of the disadvantages of the procedure was that the

retraction of the cut ends of the spinal cord, while slight, pulled the cord away from the gel,

creating a microscopic separation. Another disadvantage was that the acute inflammatory

response may have led to degradation of the ends of the cord further contributing to a

separation between the gel and the cord ends.

• The unconstrained nature of the hemi-resection defect enabled even the compromised

movement of the animals to dislodge and mechanically breakup the gel. While “tucking

in” the autologous fascia, which served as replacement for the excised dura, may have

helped to constrain the gel, there was still fragmentation of the gel.

• Although Gelfoam aided in achieving complete hemostasis in the injury site before

injection of Gtn-HPA, few animals showed bleeding in the cavity site during injection. Other

approaches to control the bleeding should be considered for future studies.

• Fabrication step of the Gtn-HPA was not ideal given the subsequent addition of H2O2

after HRP. In an ideal case, a two-syringe model should be employed for simultaneous

delivery and mixing of HRP and H2O2.

Page 220: An Injectable Gelatin-Based Conjugate Incorporating EGF

220

• Since the Gtn-HPA was injected using a pipette directly into the cavity, the injection rate

of Gtn-HPA was not controlled due to the cumbersome nature of an automatic injector.

• We performed two types of functional evaluation – open field locomotor test and bladder

function. However, these functions do not address the other aspects after spinal cord

injury such as sensory function, pain, electrophysiological deficits. Although the scoring

for BBB scale was done in a blinded and randomized manner, a greater number of

evaluators should score the animals for hindlimb function.

• Our results present the findings from the acute injection of Gtn-HPA because the Gtn-HPA

gel was injected immediately after the injury – a scenario not replicable in the clinic.

• We evaluated the effects of injection of Gtn-HPA/EGF/SDF-1α for four weeks post

injection. Therefore, we are able to assess the beginning aspects of the tissue remodeling

and regeneration. In addition, although preliminary histological was performed on Day 0,

1 and 7 post injection, a comprehensive study with immunohistochemical staining was not

performed at early time points.

• Considering that Gtn-HPA does not embed well with paraffin-embedding procedure, we

used the fast-freezing method to prepare serial sections of the spinal cord. However, the

thickness of the sections was limited to 30 µm as opposed to 6 µm achieved with paraffin-

sectioning protocol. In addition, despite cryoprotection of the spinal cord samples with

sucrose solution, there were artifacts introduced in the sections due to crystal formation in

the tissue resulting in incomplete picture of the injury site after four weeks.

• Histomorphometry analysis was performed on one section per animal from the middle of

cord. Although that was representative of the injury site for the animal, volumetric analysis

could be performed in future studies to gain more insight into the injury site.

• Immunohistochemical analysis was restricted to one section per animal. New methods

such as CLARITY could give a volumetric analysis of the injury sites in terms of various

cells and biomarkers.

• MRI-based protocol was reliable in distinguishing Gtn-HPA from the surrounding tissue.

However, the protocol was not sensitive to detect edema, hemorrhage and other

pathophysiological features. Moreover, MRI examination was performed using fixed spinal

columns ex vivo. Future studies should employ longitudinal in vivo assessment of the

location and the volume of the Gtn-HPA over the duration of the experiment.

Page 221: An Injectable Gelatin-Based Conjugate Incorporating EGF

221

8.2. Future directions: The research work presented in this thesis commends further

investigation of injectable Gtn-HPA incorporating EGF for tissue repair and regeneration after

spinal cord injury. The future directions are based on the findings and the limitations of this work.

In vitro characterization:

• More comprehensive characterization of the mechanical properties of the Gtn-HPA should

be done to evaluate the effect that injecting into cavity filled with blood or varying injection

rates has on the gelation profile of Gtn-HPA. Moreover, reliable protocol with proper

dimensions of the Gtn-HPA should be devised to study the stiffness of the Gtn-HPA gel

using compression testing and frequency sweep testing.

• 3D migration assays with two different formulations of Gtn-HPA - 8% and 12% - can be

performed with various cells such as neural progenitor stem cells, astrocytes, and

macrophages. In addition, the ability of Gtn-HPA/EGF in recruiting oligodendrocytes would

be interesting to evaluate in future experiments. In addition, the migration assay could

study the dose-dependent migration of these cells into the core.

• Detailed release profile kinetics studied with both EGF and SDF-1α can give critical

information about the behavior of the Gtn-HPA in sustained release of these factors. In

addition, embedding procedure for Gtn-HPA should be optimized so that

immunohistochemical assessment of the presence of EGF and SDF-1α can be performed

after four weeks.

• Other injectable biomaterials such as HA-Tyr and collagen should be evaluated both in

vitro and in vivo in comparison with Gtn-HPA. Moreover, other mitogenic factors such as

FGF-2, BDNF, PRP and NGF should be studied in combination with EGF if the

combination can enhance the response seen in these experiments.

In vivo evaluation:

• Contusion model of the spinal cord should be employed to validate that Gtn-HPA/EGF can

induce the migration of neural progenitor stem cells into the injury site. Since contusion is

a more relevant clinical model, it would be necessary to reproduce the data in higher

animal contusion models such as minipigs and monkeys from a translation point of view.

• Although the treatment groups showed promising results after 28 days, longer evaluation

more than four weeks (i.e. 8 weeks or longer) should be studied to gain insight into the

chronic effects of the proposed intervention.

Page 222: An Injectable Gelatin-Based Conjugate Incorporating EGF

222

• BBB scale is a well-established scale to measure functional recovery after spinal cord

injury. Different functional evaluation tests such as electrophysiological studies and

retrograde tracing should be incorporated in future to get a comprehensive view of

functional recovery. Quantitative analysis of urine retention volumes using catheterization

could also be considered.

• This study provided a comprehensive study of experimental groups to evaluate the effect

of Gtn-HPA/EGF/SDF-1α in a rodent spinal cord injury model. Few groups not evaluated

– SDF-1α in suspension, SDF-1α alone in the Gtn-HPA and 12% Gtn-HPA gel alone

should be evaluated in future studies for comparison.

• Since immediate injection of the treatment is not feasible in the clinicals setting, a modified

model to inject Gtn-HPA/EGF 7-14 days after inducing SCI should be considered.

• This work studied the serial coronals sections of the injury site in which axonal

regeneration is difficult to detect. Axial sections of the cord should be simultaneously

studied to get a comprehensive view of the injury site.

• MRI evaluation was able to distinguish Gtn-HPA in a reliable manner. Further optimization

of the ex vivo protocol may be performed to identify edema, hemorrhage and cavitation.

Furthermore, we believe that T1IR method in conjunction with T2-weighted image would

provide the most information about the injury site. In addition, future studies may compare

the histological or functional findings with the MRI metrics. Therefore, the protocol may be

optimized for in vivo imaging of the rats so that the degradation characteristics of the Gtn-

HPA gel can be studied in a longitudinal manner in future. It would be important to consider

the effect of cardiac and respiratory movements of the rat on the MRI images during

acquisition.

• In future, one may consider greater panel of biomarkers to study various cell types such

as oligodendrocytes and Schwann cells shown to promote regeneration in the defect area.

Moreover, ECM proteins in the injury site should be comprehensively studied using

histological and immunohistochemical methods.

• Finally, future formulation of the Gtn-HPA may plan to include cells delivered in the injury

site in a contusion or surgical excision models. These cells should be chosen such that

their recruitment would be difficult to achieve using chemotactic factors. This could ensure

repopulation of the spinal cord defect with all the major cell constituents present in the

healthy tissue.

Page 223: An Injectable Gelatin-Based Conjugate Incorporating EGF

223

Appendix:

1. Table outlining the total number of animals evaluated in thesis work:

Defect Experimental group Duration Study size (n)

3 mm left hemi-resection

8% Gel [G8] 2 weeks 2

8% Gel [G8] 4 weeks 2

8% Gel + SDF-1a [GS8] 2 weeks 1

8% Gel + SDF-1a [GS8] 4 weeks 2

8% Gel + SDF-1a in PCN [GS8P] 2 weeks 2

8% Gel + SDF-1a in PCN [GS8P] 4 weeks 2

1 mm left hemi-resection

Control [N] 4 weeks 6

8% Gel + EGF [GE8] 4 weeks 6

2 mm full-resection Control [N] 4 weeks 5

8% Gel [G8] 4 weeks 6

8% Gel + EGF [GE8] 4 weeks 3

8% Gel +EGF +SDF-1a [GES8] 4 weeks 5

2 mm left hemi-resection

Control [N] 4 weeks 6

EGF in PBS suspension [E] 4 weeks 5

8% Gel [G8] 4 weeks 6

8% Gel + EGF [GE8] 4 weeks 6

8% Gel +EGF + SDF-1a [GES8] 4 weeks 6

12% Gel + EGF [GE12] Day 0 1

12% Gel + EGF [GE12] Day 1 3

12% Gel + EGF [GE12] Day 7 3

12% Gel + EGF [GE12] 4 weeks 6

12% Gel + EGF - tucked-in dura [GE12T]

4 weeks 6

Total 90

Page 224: An Injectable Gelatin-Based Conjugate Incorporating EGF

224

2. Supplies to ensure before starting animal study

ARF (Diane/Sylvia):

- Oxygen cylinders

- Isoflurane

- Nembutal

- Heparin

- Lactated ringers

- Meloxicam

- Baytril/Cephazolin

- Heating pads & empty cages

- 4 face masks

- Surgical gloves

- Weighing scale

- Syringes (3 ml, 1 ml)

- Suture clips

- Scalpels

- Clip removers

- Razor for shaving rat’s hairs

Dr. Hsu

- Number clips for ears

- Gelfoam

Surgery and preparation:

- EGF (aliquots)

- SDF-1α (aliquots)

- PBS

- Gtn-HPA

- calibrated pipettes, tips

- Filters

- HRP (aliquots)

- H2O2 (aliquots)

Page 225: An Injectable Gelatin-Based Conjugate Incorporating EGF

225

3. Rat Spinal Cord Surgery Preparation and Post-Operative Care Checklist

Pre-surgical Set Up- Day++ before:

Confirm availability for the dates with Dr. Hsu

Order animals through BVARI (email [email protected] and cc Dr. Spector, Diane Ghera from ARF and Nickey Ortega for approval), procedures must wait 48 hrs after arrival

Set up room adjacent to O.R. a. Get 4 (short) empty plastic rat cages from the 3rd floor. b. Place cages on heat pad. c. Place a small cloth (blanket) in each cage. d. Set up oxygen and tubing with face masks for 4 rats.

Autoclave 1.5 ml tubes, water (if needed for reconstitution)

Place box with syringes and gloves into procedure/animal care room.

Make stock solutions of Meloxicam and Cephazolin. Keep refrigerated. a. Meloxicam (1mg/kg): 5 mg/ml bottle, 0.04 ml for 200 g rat, 0.05 ml for 250 g rat b. Cephazolin (35mg/kg): Add 9 ml saline to 1 g cephazolin, 0.07 ml (for 200 g rat, 0.09

for 250 g rat) Pre-surgical Morning of:

Turn on heating pads, make sure water level is OK, put on bag of lactate ringers

Take care of any animals already operated on (Bladder, Meloxicam & Baytril)

Have EGF and SDF-1a (as needed) ready on ice – do all the calculations before)

Fabricate gel and hold on ice Pre-surgical Immediate:

Weigh and record weight of rat

Take the rat to Dr. Hsu and set the isoflurane at 4% isoflurane. Once the rat is unconscious, keep isoflurane level at 2.5% to maintain anesthesia)

Intra-operatively:

Once the rat is under anesthesia (2.5%), shave the back, especially the surgery region

Get ready with the gel, PBS, any growth factors for the injection

Confirm with Dr. Hsu that it is 2 mm left hemi-resection defect

Inject 20 ul (2 mm left hemi-resection defect) of the gel into the injury site – calculations

As Dr. Hsu is finishing up, weight the rat and bring the second rat in. Post-Operative Care:

1. Take rat from the O.R. and place in heated cage in adjacent room. 2. Inject rat with 3 ml of warm Lactated Ringer’s Solution 3. Inject rat (1 mg/kg) Meloxicam - 5 mg/ml bottle, 0.04 ml for 200 g rat, 0.05 ml for 250 g rat 4. Inject rat (35 mg/kg) with 0.07 ml (for 200 g rat, 0.09 for 250 g rat) Cefazolin 5. For the first hour after surgery, monitor rat every 10 minutes for breathing 6. Thereafter, monitor rats every 45 minutes

Once the rat regains consciousness, express the bladder and place in cage with wood-chip bedding, a long nozzle water bottle, and 2-3 pieces of dry food on the bottom for easy access to the rat. Fill out postoperative card!!!

7. Return in the evening to express bladder and administer Cephazolin as per dosage guidelines (Ensure light/dark timings in the room with Diane Ghera or Sylvia Smith)

Daily Care thereafter:

1. 2x Daily: a. Bladder expression, b. Cephazolin – twice daily for 1 week 2. 1x Daily: Meloxicam - once every 22-24 hours for 4 days 3. 1x Weekly: a. Weigh b. videotape movement

Keep track of everything on the VABHS record sheet for each rat

Page 226: An Injectable Gelatin-Based Conjugate Incorporating EGF

226

4. Functional evaluation – Open field locomotor test and 1-week clip removal protocol

BBB:

- Bring the animals (maximum 4 at a time) on the second floor wrapped in the gray cloth

cover in my box located in the rat room on 3rd floor.

- In the pre-OR room, have 2 green chux tightly on the table using tape.

- Weigh the animal and record in the VABHS record sheet

- Gently take the rat and place her in the middle of the table and record video for 3 min

- Put rat back in the cage and repeat for others

- Take notes of the hindlimbs movement with BBB scale

- Randomize the videos for randomized and blind rating

Clip removal (after 1 week – reserve time with Diane Ghera / Sylvia Smith)

Items needed: Clip remover tool, tweezers, Vetbond (if needed)

- After BBB, bring the cart with rats in the surgery room 216A.

- Turn on the oxygen at 1.5-2 L/min

- As you are ready to bring in the rat, turn the isoflourane to 4%

- With the rat glove in the right hand, bring the rat and put her into the nose cone. At 4%, it

will roughly take 1-2 minute for the rat to anesthetize

- Pinch the tail to make sure rat is anesthetized. Once anesthetized, turn the isoflourane

to 2%

- Using the clip remover tool, take out all the clips. Sometimes, due to blood and/or other

material, it will take some force to take out the clips.

- If you notice any open area due to clip removal, use vetbond to close it.

- Once all the clips are removed, turn off isoflurane and put the rat back in the cage.

- Repeat the procedure for each rat.

- Once all are done, turn off the oxygen and isoflurane. MAKE SURE both are turned off

- Clean up and throw away the clips

- Bring the rats back to the rat room on 3rd floor wrapped with gray cloth cover and put

them back in their original place.

- Make sure the rats are awake in the cage!

Page 227: An Injectable Gelatin-Based Conjugate Incorporating EGF

227

5. Animal sacrifice - transcardial perfusion of female Lewis rats

Items needed:

• Medium dissecting scissors

• Small dissecting scissors, scalpel

• 18 gage needles

• 4% paraformaldehyde solution (PFA)

• PBS

• Heparin solution

• Ice bucket

• Perfusion pump

• Specimen containers – 50 ml Eppendorf tube

• Red biohazard bags

• 1 ml syringes with 25 gage/ 5/8 inch long needle (for administering IP Nembutal)

• Nembutal

• 100ml measuring cylinder

• Metal pan with grill

• Two glass beakers

• Rat glove

• Box of gloves

• Paper towels, paraffin

Prepare:

-120 ml of heparin solution at 20 units/ml PBS, put on ice (Add 2.4 ml Heparin in 120 ml PBS

located on the shelf opposite the chemical hood – labeled with my name)

-100 ml paraformaldehyde on ice

-50 ml Eppendorf tube with 4% PFA – label it with animal number and the surgery date

-Once both are prepared in glass beakers, put them tilted at an angle in the ice bucket

-The ‘in’ tube should be in the heparinized saline beaker: use tape to fix the tube in the solution

to avoid air bubbles.

-The ‘out’ tube should be fixed with 18 Gage needle (pink cover) wrapped tightly with paraffin to

ensure no leakage

Procedure:

1. Flush air from pump tubing with heparinized saline solution (remove any air bubbles) 2. Anesthetize rat with 150 mg/kg pentobarbital (Nembutal) IP injection (200 g → 0.6 ml,

201-225 g → 0.65 ml, 225g → 0.675 ml, 225-250 g → 0.7 ml, 250 g → 0.75 ml) 3. Using medium dissecting scissors cut skin across abdomen 4. Using medium dissecting scissors cut skin along midline of chest 5. Using medium dissecting scissors cut sternum along midline of chest 6. Make lateral cut from bottom of sternum to the left side of thorax 7. Expose the heart

Page 228: An Injectable Gelatin-Based Conjugate Incorporating EGF

228

8. Insert needle into left ventricle and please always be on the lookout for the needle in the ventricle. Due to animal movement the needle could come out of the ventricle so hold the needle inside during the perfusion

9. Turn on pump at medium setting and speed 3 10. Quickly cut right atrium using small dissecting scissors 11. Increase pump speed to 7 12. Just as the heparin solution is about to finish, turn off the pump and switch to PFA

solution (be careful not to introduce any air bubbles into the line), then restart the pump at speed 7.

13. Once the rat is completely perfused with PFA, discard the needle in the sharps container.

14. Flip the rat body and dissect out the skin layer to expose the collagen fascia cover. Any sized scalpel would work well for this.

15. Make cuts along the ribs on the sides. Top cut should be around the neck and the bottom cut should be around the ‘white’ fascia region to ensure the defect site is part of the resected spinal column

16. Put the column in 4% PFA in 50 ml Eppendorf tube and store it in 4C cold room on the first floor.

17. Discard the rat carcass in the red biohazard bag. Please discard all the blood, saline and PFA in the container under the chemical hood. For thick blood, use paper towels to clean the metal pan and the grill. Those paper towels can go into the red biohazard bag.

18. Clean all the tools with plenty of water in the sink. Air dry the metal pan, grill and the white board while the tolls (scissors, etc) can be cleaned and dried with paper towels.

19. Put the pump back into its box! 20. Please discard the red biohazard bag in the animal facility carcass freezer room #R2-

205 (room is near 2nd floor entrance in animal facility) 21. Keep the dirty cage in the wash room on the 3rd floor animal facility. Please make sure to

keep the VABHS white sheet and the animal cards with you (the red and the pink card).

Page 229: An Injectable Gelatin-Based Conjugate Incorporating EGF

229

6. Fast freezing spinal cords and embedding procedure

Supplies to ensure:

• Medium and small dissecting scissor, Bone ronguers

• Scalpel/sharp blade

• Dry ice

• Aluminum foil & mold

• OCT Tissue-Tek compound

• Liquid nitrogen

• Isopentane (2-methylbutane)

• Metal bowl and tongs

• PBS

• Surgical gloves & markers

Spinal cord harvest from spinal column:

• Get liquid nitrogen, dry ice, isopentane

• Make Aluminum foil molds for tissue embedding

• Carefully take out extra tissue (muscle and bone) from the spinal column

• Using small dissecting scissors, slowly and carefully isolate the cord

• Make cuts at 45 angle and then proceed carefully not to puncture it/damage it

• Do not let the tissue dry out so dip in PBS as needed

• Once the cord is isolated, rinse it in PBS and then put it in the Al foil mold (marked with the sample #) with OCT compound

o Ensure that the cord is covered by OCT on all surfaces o Ensure that the cord is stable in the mold – cord at the bottom with floor of the

mold o Ensure there are no air bubbles o Ensure the cord is nicely placed in the middle of the mold

Tissue Embedding using isopentane and liquid nitrogen

• Fill isopentane in the metal bowl

• Fill Styrofoam box with liquid nitrogen (take all necessary precautions and wear protective gear)

• Lower the isopentane metal bowl into the liquid nitrogen

• Let it cool till the bottom of the bowl is opaque (white)

• Remove the bowl using tongs and then submerge the mold with the cord in isopentane for fast freezing

o If the isopentane turns completely clear, it is not cool enough. Cool it in liquid nitrogen again

o Lower one mold at a time in isopentane. Do not try to embed 2 spinal cord tissues in isopentane

• Let the OCT freeze in isopentane completely. You should see outside to inside freezing. Freezing of the entire block should take less than a minute

• Place the frozen block in dry ice for some time to ensure all the vapors either from isopentane or liquid nitrogen are gone

• Cover the block with Al foil marked with sample #

• Store the tissue blocks at -80 C (cryosection to produce serial 25-30 µm thick sections with CT ~ -17-19 C and OT = -20 C)

Page 230: An Injectable Gelatin-Based Conjugate Incorporating EGF

230

7. Typical immunohistochemistry staining protocol

Antibody Company Dilution Antigen retrieval

2 abs Dilution

Iba-1 (Rb) Wako 1:400 pH6 buffer 488 Rb 1:300

CD68 (Ms) Biorad 1:200 pH6 buffer Cy3 Red 1:300 Procedure:

1. Warm the tissue sections at 550C in slide warmer for 3 hours.

2. Allow the slides to cool at room temperature (RT) for 15 minutes.

3. Immersed the slides in PBS at RT for 30-40 min.

4. Antigen retrieval step - citrate buffer pH 9 (200 ml diH2O + 1.875 ml unmasking solution) or high

pH 9 buffer

a. Preheat the buffer containing the staining rack in pressure cooker at 1000C

b. Set up the temperature to the required level well in advance and check in between for

proper increase in temperature. Immerse the slides in preheated buffer in staining rack

and heat in the pressure cooker at the required temperature for 20 minutes.

c. After heat retrieval step, the slides are allowed to cool in the buffer itself for 15 minutes

at RT.

5. Immerse the slides in 0.3% TritonX-100 in PBS for 30 minutes.

6. Immerse the slides in PBS for quick dips and then for 5 minutes.

7. Fix the slides in ice cold methanol (precooled at -200C) for 10 minutes.

8. Then transfer the slides to PBS and allow to stand for 5 minutes.

9. The slides are taken out from PBS and the sides are blotted with care being careful not to wipe

off section. Mark a closed circle around tissue section with an IHC marking pen.

10. Protein block step

Block the slides with serum-free protein block for 1 hour.

11. Wash the slides as follows:

a. 0.05% Tween20 in PBS (5 min)

b. PBS (5 min)

c. PBS (5 min)

12. Dilute the primary antibodies in DAKO antibody diluent to the working dilution and add to the

slides. Slides are incubated at 40C overnight.

13. Wash the slides as follows:

a. 0.05% Tween20 in PBS (5 min)

b. PBS (10 min)

c. PBS (10 min)

14. Incubate with secondary antibodies for 2 hours (at room temperature). (Secondary antibody

constituted using DAKO antibody diluent.

15. Wash the slides as follows:

a. 0.05% Tween20 in PBS (5 min)

b. PBS (10 min)

c. PBS (10 min)

16. After 3 quick dips in diH2O, blot excess liquid and mount in DAPI containing fluorescent mounting

media (DAKO) with cover slip.

17. Store at 4 C

Page 231: An Injectable Gelatin-Based Conjugate Incorporating EGF

231

8. Quantification of Immunohistochemical biomarkers

1) Import original IHC images into ImageJ/Fiji.

2) Assign appropriate scale.

3) Convert the image to 8-bit image

4) Uniform threshold was applied to all the images for each biomarker (threshold for each

biomarker was determined using at least 15 samples images and compared with original IHC

images.

5) Post-threshold, a standardized ROI of 0.3 mm2 was created.

6) Percent area of the signal in the standardized ROI was measured.