Biopolymer-Nanotube Nerve Guidance Conduit Drug Delivery for Peripheral Nerve Regeneration Ohan S

University of Connecticut OpenCommons@UConn

Doctoral Dissertations University of Connecticut Graduate School

3-26-2019 Biopolymer-Nanotube Nerve Guidance Conduit Drug Delivery for Peripheral Nerve Regeneration Ohan S. Manoukian University of Connecticut - Storrs, [email protected]

Follow this and additional works at: https://opencommons.uconn.edu/dissertations

Recommended Citation Manoukian, Ohan S., "Biopolymer-Nanotube Nerve Guidance Conduit Drug Delivery for Peripheral Nerve Regeneration" (2019). Doctoral Dissertations. 2081. https://opencommons.uconn.edu/dissertations/2081 Biopolymer-Nanotube Nerve Guidance Conduit Drug Delivery for Peripheral Nerve Regeneration

Ohan Shant Manoukian, Ph.D.

University of Connecticut, 2019

Peripheral nerve injuries account for roughly 3% of all trauma patients with over 900,000 repair procedures annually in the U.S. Of all extremity peripheral nerve injuries, 51% require nerve repair with a transected gap. The current gold-standard autograft repair has several associated shortcomings, while engineered constructs typically offer suboptimal results only suitable for short gaps. In order to achieve enhanced peripheral nerve regeneration and greater functional recovery, we have developed a biodegradable, chitosan-based nerve guidance conduit (NGC) with aligned microchannel porosity and halloysite nanotubes for the sustained release of 4-Aminopyridine (4AP), a small molecule drug that promotes nerve conduction and neurotrophic factor release. In vitro studies characterized the polymer NGC system and confirmed the sustained release of 4AP. Human Schwann cells showed a dose-dependent elevation of critical proteins including nerve growth factor (NGF), myelin protein zero (P0), and brain derived neurotrophic factor

(BDNF). When treated with 10 µg/mL 4AP, NGF, P0, and BDNF showed 128.58±0.57%, 310.75±11.92%, and 170.94±4.43% increases over no-drug control groups, respectively. When tested in vivo using a critical- size (15 mm) rat sciatic nerve transection model, the results of functional walking track analysis, morphometric evaluations of myelin development, and histological assessments of various markers confirmed the equivalency of the drug-conduit repair to autograft controls. At 8 weeks post-operative, the sciatic functional index (SFI) of autograft (-44.89±4.73) was not statistically different from drug-conduit (-

45.12±4.45), while myelin thickness (G-ratio) evaluations also showed no statistical difference between autograft (0.61±0.10) and drug-conduit (0.65±0.06). The conduit’s aligned microchannel architecture and sustained release of 4-Aminopyridine may facilitate physical guidance of axons to distal target reinnervation, while increasing nerve conduction, and in turn neurotransmitter and neurotrophic factor release. Overall, our NGC facilitates more efficient and efficacious peripheral nerve regeneration via a drug delivery system that is feasible for clinical application, with potentially transformative implications. Biopolymer-Nanotube Nerve Guidance Conduit Drug Delivery for Peripheral Nerve Regeneration

Ohan Shant Manoukian

B.S.E., University of Connecticut, 2015

A Dissertation

Submitted in Partial Fulfillment of the

Requirements for the Degree of

Doctor of Philosophy

at the

University of Connecticut

2019

Ohan Shant Manoukian

2019

APPROVAL PAGE

Doctor of Philosophy Dissertation

Biopolymer-Nanotube Nerve Guidance Conduit Drug Delivery for Peripheral Nerve Regeneration

Presented by

Ohan Shant Manoukian, B.S.E.

Major Advisor ______

Sangamesh G. Kumbar, Ph.D.

Associate Advisor ______

Swetha Rudraiah, Ph.D.

Associate Advisor ______

Syam Nukavarapu, Ph.D.

Associate Advisor ______

Ivo Kalajzic, M.D., Ph.D.

Associate Advisor ______

Kazunori Hoshino, Ph.D.

University of Connecticut

2019

iii

DEDICATION

To my parents, Mourad and Hilda, without whose endless guidance, unconditional love, and unwavering

support and encouragement, none of this would be possible.

ACKNOWLEDGEMENTS

This work would not have materialized, and I would not be where I am today, without the guidance and support of many mentors who helped me throughout all phases of my professional and personal life. I would like to, first and foremost, thank my major advisor, Dr. Sangamesh Kumbar, who has been an excellent mentor throughout my years at the University of Connecticut. Dr. Kumbar provided me with invaluable guidance and mentorship, and helped me grow both professionally and personally. Without his support, this work would never have been realized. I would also like to acknowledge and thank my thesis committee members, Dr. Swetha Rudraiah, Dr. Syam Nukavarapu, Dr. Ivo Kalajzic, and Dr. Kazunori

Hoshino for their time, technical expertise, and valuable scientific inputs and critiques. Thank you for all the discussions and conversations which undoubtedly helped strengthen this work.

I would like to thank the faculty and staff in the Department of Biomedical Engineering for their help throughout my time as a student, in particular Dr. Ki Chon, Dr. David Kaputa, Dr. Patrick Kumavor,

Ms. Jennifer Desrosiers, and Ms. Lisa Ephraim, all of whom made my time within the department an enjoyable experience. I would like to thank Dr. Jenna Bartley, from the Department of Immunology at

UConn Health, for all of her help and support with the walking track analysis portion of my animal work, as well as her invaluable advice and encouragement, as a mentor and friend.

I would also like to extend my gratitude to all members of the lab, past and present, who provided guidance and scientific support, and also made the daily lab life enjoyable. In particular, Dr. Roshan James who helped me every step of the way in my early graduate school career, taught me everything he knew, and helped develop and strengthen my research techniques and scientific skills. Dr. Namdev Shelke, Dr.

Aja Aravamudhan, and Dr. Daisy Ramos for teaching me laboratory basics and guiding me with my early work, and in particular Dr. Ramos for all her help and friendship over the years. I would also like to thank

Dr. Michael Arul, Sama Abdulmalik, Sara Katebifar, Jonathon Nip, Naseem Sardashti, and Jiana Baker for their help throughout my thesis work and for making my time in the lab memorable.

It would be incomplete to not also acknowledge and thank the National Science Foundation (NSF)

Division of Graduate Education, Graduate Research Fellowship Program (GRFP), for awarding me an NSF

Graduate Research Fellowship in 2017, without whose financial support this work would not be possible.

Being awarded an NSF Graduate Research Fellowship (Grant No. DGE-1747453), for my past and present contributions to science and engineering, was truly a highlight of my graduate school career, as well as a great moment of pride and a humbling honor.

I would like to thank all of my friends from all circles who have stood by me and made this time a truly enjoyable and adventurous journey. To my dearest friends from the Armenian community, where I was born and raised, and where lifelong friendships have been forged, your continued respect and support is undoubtedly appreciated. To all of my friends from my amazing times at the University of Connecticut, you have all been an integral part of my professional and personal successes throughout my college years and beyond, and I thank you all for that.

Last, but certainly not least, I would like to thank my family for their unwavering support. To my parents, Mourad and Hilda, without whose unconditional love and guidance I would not be where I am today, thank you for teaching me the values of education, hard work, perseverance, integrity, and humility.

Your countless sacrifices, and continued support and encouragement, have led me to achieve my wildest dreams and I hope I have, and can continue to, make you both proud.

TABLE OF CONTENTS

APPROVAL PAGE ...... iii DEDICATION...... iv ACKNOWLEDGEMENTS ...... v CHAPTER 1: PERIPHERAL NERVE INJURY AND REGENERATION ...... 1 1.1. INTRODUCTION ...... 1 1.2. BIOLOGY OF PERIPHERAL NERVE INJURY AND REGENERATION ...... 2 1.3. AUGMENTED PERIPHERAL NERVE REGENERATION ...... 9 1.4. EXOGENOUS NEUROTROPHIC GROWTH FACTORS ...... 13 1.5. DRUG DELIVERY STRATEGIES AND CELL THERAPIES ...... 18 1.6. ANIMAL MODELS AND FUNCTIONAL ASSESSMENTS ...... 24 1.7. CLINICAL APPLICATIONS AND COMMERCIALIZED NGCs ...... 27 CHAPTER 2: ALIGNED MICROCHANNEL POLYMER-NANOTUBE COMPOSITES FOR PERIPHERAL NERVE REGENERATION: SMALL MOLECULE DRUG DELIVERY ...... 29 2.1. INTRODUCTION ...... 29 2.2. MATERIALS AND METHODS ...... 32 2.3. RESULTS ...... 40 2.4. DISCUSSION ...... 52 2.5. CONCLUSION ...... 55 CHAPTER 3: BIOPOLYMER-NANOTUBE NERVE GUIDANCE CONDUIT DRUG DELIVERY FOR PERIPHERAL NERVE REGENERATION: IN VIVO ASSESSMENT ...... 57 3.1. INTRODUCTION ...... 57 3.2. MATERIALS AND METHODS ...... 58 3.3. RESULTS ...... 63 3.4. DISCUSSION ...... 73 3.5. CONCLUSION ...... 77 CHAPTER 4: SUMMARY OF FINDINGS ...... 78 CHAPTER 5: FUTURE PERSPECTIVES ...... 83 REFERENCES ...... 86

vii

LIST OF FIGURES

Figure 1: Two approaches to bridging a peripheral nerve injury defect ...... 1 Figure 2: Stages of peripheral nerve injury, repair, and regeneration ...... 6 Figure 3: Tissue engineering scaffolds and biomaterials ...... 11 Figure 4: Nerve guidance conduits and associated attributes ...... 11 Figure 5: Biochemical pathways for NGF- & BDNF-mediated neuronal repair and regeneration .. 15 Figure 6: Unidirectional freezing apparatus setup for fabrication of aligned pores ...... 34 Figure 7: Scanning electron microscopy (SEM) images showing aligned microchannels ...... 41 Figure 8: Tensile mechanical testing of fabricated nerve guidance conduits ...... 42 Figure 9: Water absorption (swelling) ...... 43 Figure 10: X-Ray diffraction spectra ...... 44 Figure 11: Differential scanning calorimetry (DSC) spectra ...... 45 Figure 12: Porosity measurements ...... 45 Figure 13: Thermogravimetric analysis (TGA) spectra ...... 46 Figure 14: In vitro degradation ...... 47 Figure 15: Cumulative drug release and burst release characterizations ...... 48 Figure 16: Cell toxicity and viability assays ...... 50 Figure 17: Immunofluorescence protein staining and quantification ...... 52 Figure 18: Wistar rat critical-size sciatic nerve transection surgical procedure ...... 60 Figure 19: DigiGait walking track analysis setup ...... 61 Figure 20: Scanning electron microscopy (SEM) images of fabricated nerve guidance conduits ..... 64 Figure 21: Walking track analysis - Sciatic functional index ...... 65 Figure 22: Walking track analysis - Paw area ...... 66 Figure 23: Walking track analysis - Swing phase ...... 67 Figure 24: Walking track analysis - Stance phase ...... 68 Figure 25: Transmission electron microscopy (TEM) of the regenerated sciatic nerve ...... 69 Figure 26: Histological staining and quantification of myelin ...... 70 Figure 27: Immunohistochemical staining and quantification of NF-H and S100 ...... 71 Figure 28: Histological cross sections of liver tissues to assess potential HNT toxicity ...... 72 Figure 29: Photograph of 4AP drug-conduit repaired sciatic nerve ...... 73 Figure 30: Summary of key advantages of the developed nerve guidance conduit system...... 78 Figure 31: Key in vitro findings and achievements...... 80 Figure 32: Key in vivo findings and achievements...... 81 Figure 33: Future directions of the field and scientific perspectives ...... 83

viii

LIST OF TABLES

Table I: Various classes of nerve fibers and their associated properties, location, and functions ...... 3 Table II: Classifications of nerve injuries according to sunderland and seddon ...... 4 Table III: Commercially available nerve guidance conduits ...... 28 Table IV: Cumulative 4AP release data fit to various drug release models ...... 49

CHAPTER 1: PERIPHERAL NERVE INJURY AND REGENERATION

1.1. INTRODUCTION

Peripheral nerve injury (PNI) occurs in roughly 2.8% of all trauma patients and can be extremely debilitating dependent on severity and classification of injury [1]. There are 50,000 procedures performed annually, in the United States alone, to repair damaged peripheral nerves, with an associated annual cost of

7 billion USD [2]. Despite relatively low incidence, there is significant associated morbidity and peripheral nerve repair outcomes historically have been poor, particularly when treatment is delayed and/or damage is severe [3]. Axon regeneration can occur naturally, but is considered a generally slow process at a rate of

1 mm/day, with any defect size greater than 3 cm often resulting in poor functional recovery [4-6].

The severity and effects of PNI range from sensory and motor discomfort to severely debilitating impairment, dependent on the physiology of the injury. Treatment of severe PNI requires surgical intervention, most often the autograft, however nerve guidance conduits, functionalized for transportation and delivery of cells, growth factors, and drugs represent a new wave of neuroregenerative research with potentially transformative implications (Figure 1).

Figure 1: Schematic illustration showing two approaches to bridging a peripheral nerve injury defect: the commonly-used autograft and a nerve guidance conduit, which may be used to deliver cells, growth factors, and/or various drugs.

This chapter aims to serve as a comprehensive, in-depth perspective on peripheral nerve injury, repair and regeneration, with a particular focus on nerve guidance conduits and drug delivery strategies. A comprehensive background of peripheral nerve injury and repair pathology, and an in-depth look into augmented nerve regeneration, nerve guidance conduits, and drug delivery strategies provides a state-of- the-art perspective on the field.

1.2. BIOLOGY OF PERIPHERAL NERVE INJURY AND REGENERATION

1.2.1. Anatomy and Organization of Peripheral Nerves

The peripheral nervous system (PNS) includes all parts of the total nervous system excluding the brain and spinal cord, which are classified as the central nervous system (CNS). It includes the peripheral nerves, the cranial nerves, spinal nerves, and neuromuscular junctions. Within the PNS, bundles of nerve fibers, or axons, conduct sensory and motor information to and from the CNS. Nerves are composed of nervous tissue, connective tissue supporting their structure, as well as blood vessels for nutrients. The most outer surface of a nerve is surrounded by a fibrous connective tissue known as the epineurium. One level in, axons are bundled into fascicles which are surrounded by the perineurium, composed of fibroblasts and collagen. Lastly, individual axons are sheathed by the endoneurium, composed primarily of oriented collagen fibers.

All nerves connected to the spinal cord are classified as spinal nerves. In contrast, the nerves attached to the brain are classified as the cranial nerves, which are outside of the scope of this review. Spinal nerves are separated as sensory and motor and connect to the spinal cord as the dorsal nerve root and the ventral nerve root, respectively. Efferent (motor and autonomic) neurons receive signals from neurons in the CNS, while afferent (sensory) neurons receive signals from specialized cell types which are then sent to the CNS to provide sensory information or interneurons in the spinal cord for reflex responses [7]. The

PNS is composed of two main cell types: neurons and neuroglia. Neurons are the primary functional elements of the PNS consisting of a cell body (soma), dendrites which are branching extensions that receive signals, an axon which is the primary extension which communicates signals through the neuron, and lastly

2 the axon terminal, which form junctions with other neurons or muscles and release neurotransmitters. The dendrites of the neuron receive excitatory or inhibitory signals from other cells and transmit electrical signals to the neuron’s soma, which are then conducted as electrical impulses (action potentials) down the axon.

In the PNS, axons are classified as myelinated or unmyelinated based on the presence of an insulating myelin sheath surrounding the axon. Made up of proteins and fatty acids, myelin sheath is produced by Schwann cells around the axon and acts as an electrical insulator. Thus, the myelin serves to increase the axon potential conduction velocities, which is particularly important for extended axons, where unmyelinated axon conduction velocities range from 0.5-10 m/s and myelinated axons conduct up to 150 m/s (Table I) [8].

Table I: Various classes of nerve fibers and their associated properties, location, and functions.

1.2.2. Peripheral Nerve Injury

In 1942, sir Herbert Seddon became the first to classify nerve injuries according to the severity of demyelination, damage to axons, and damage to surrounding tissues (Table II) [8]. Neurapraxia is the mildest form of injury, characterized by focal demyelination with no damage to axons or surrounding tissues, typically occurring from mild compression of the nerve. Demyelination typically causes decreases in conduction velocity, however depending on severity, it can cause conduction blocks and muscle

3 weakness. The next level of injury, axonotmesis, involves focal demyelination and damage to the axons, while maintaining continuity of the surrounding nerve tissues. Lastly, neurotmesis refers to the most severe form of nerve injury where there is a full transection (segmental defect) of the axons as well as all the surrounding connective tissue layers [8]. With such an injury, there is complete discontinuity of the nerve and surgical repair is required for recovery.

To differentiate the extent of damage to surrounding connective tissues, namely the epineurium, perineurium, and endoneurium, sir Sydney Sunderland expanded on Seddon’s scheme with a five-grade

PNI classification (Table II) [9]. The first and second grades of the Sunderland classification correspond to neurapraxia and axonotmesis, respectively, as described above. Grade III, IV, and V injuries are described as damage to the axon along with the associated endoneurium, perineurium, and epineurium, respectively. With damage to the connective tissue layers surrounding the axons, there is increased likelihood of variable recovery, often incomplete, due to misdirection of regenerating axons which may be unable to innervate sensory or muscle targets. Grade IV injuries lead to localized neuroma formation resulting from internal hemorrhage and fibrous tissues blocking the distal growth of regenerative axons, whereas grade V injuries, with complete nerve discontinuity, result in end-bulb neuroma.

Table II: Classifications of nerve injuries according to Sunderland and Seddon with associated descriptions and recovery.

1.2.3. Compression Injuries

Compression injuries, different from crush injuries, fall under the Seddon class of neurapraxia, or

Sunderland’s grade I for PNI. These injuries are generally characterized by focal demyelination, localized to the site of compression, with no damage to the axon or surrounding connective tissues. Carpal tunnel syndrome, tarsal tunnel syndrome, as well as sciatica are all examples of compression injuries of nerves

[10]. Here, normal nerve morphology is still present with adequate neuromuscular junctions. However, focal demyelination is characteristic of a compression injury and is shown by an increased G-ratio, a ratio of the axon diameter to the nerve fiber (axon plus myelin sheath) diameter. Similarly, there is a marked decrease in the distance between adjacent nodes of Ranvier, known as the internodal length.

1.2.4. Crush and Transection Injuries

Crush injuries are typically the result of acute blunt-force trauma, causing traumatic compression of the nerve. This can include trauma from any blunt object, for example a bat, that does not transect the nerve [8]. The resultant crush injury can be categorized, in part or wholly, by any of the classes in both the

Seddon and Sunderland schemes. Thus, the severity of crush injuries is highly variable and dependent on the magnitude of trauma and location of injury. In contrast, transection injuries, representing grade V nerve injuries, show a complete discontinuation of the nerve where axons are transected and there is a physical gap, or segmentation, between the proximal and distal ends. Transection injuries are commonly due to lacerations from a sharp object such as a knife, blade, or glass shard, as well as gunshot wounds and explosions [11]. It is important to note that for ballistic wounds, namely gunshot and explosive fragmentations, often times there may be a combination of both crush and transection of the nerve, partially associated with shear and compressive stresses caused by the shockwave of the passing of the projectile or fragment through tissues.

1.2.5. Wallerian Degeneration

Upon PNI, Wallerian degeneration is initiated distal to the injury zone, starting with axonal breakdown, as a measure to alleviate abnormal axon regeneration (Figure 2). This intrinsic breakdown of

5 distal axons and myelin is the initiator of the degeneration and repair cascade, which does not alone include the nerve, but is a multi-cellular complex response involving multiple components [12-14]. The coordination and recruitment of macrophages, neutrophils, and adaptive cells (T and B cells) along with de-differentiating Schwann cells is vital to the successful repair and regeneration of the injured nerve [15].

These cells are recruited to the site of injury within the initial hours and days after nerve injury, with the primary goals of clearing inhibitory debris and the production of an environment supportive of nerve regeneration. Despite aggressive recruitment and initiation of axonal degeneration, detached axons remain intact for several days after the initial injury and have been shown to be capable of transmitting action potentials when stimulated.

Figure 2: The various stages of peripheral nerve injury, repair, and regeneration.

The involvement of macrophages in axonal degeneration was long thought to be limited to removal of myelin debris. However, macrophages along with Schwann cells help promote an environment supportive of axonal regeneration with the release extracellular matrix (ECM) proteins, neurotrophins, cytokines, and chemokines. Mainly originating from circulating hematogenous monocytes, the recruitment of macrophages begins 2-3 days post-injury and peaks at 7-days post-injury [16-19]. Trophic cues released by de-differentiating Schwann cells as a response to peripheral nerve injury initiate and regulate macrophage infiltration. Schwann cell released factors include Monocyte Chemoattractant Protein-1 (MCP-

1/CCL2), IL-1α, IL-1β, S100A8, S100A9 as well as ECM proteins including prominently collagen VI [20-

23]. Moreover, localized macrophages also release similar factors including MCP-1, IL-1α, and IL-1β which further recruit macrophages to the site of injury in a positive-feedback loop mechanism [22-25].

1.2.6. Endogenous Nerve Repair and Regeneration

Initially, swelling occurs at both the proximal and distal ends of the damaged nerve, resulting from futile continuation of retrograde and anterograde axonal transport [26]. This eventually leads into the breakdown of the distal stump, via the granular disintegration of the cytoskeleton [27]. This breakdown is preceded shortly by the influx of extracellular ions, namely, Ca+ and Na+, which serve to recruit nearby macrophages to the site via signals from Schwann cells. Furthermore, mRNA of brain-derived neurotrophic factor (BDNF) and glial cell line-derived neurotrophic factor (GDNF) have shown marked upregulation in the distal stump.

In contrast to the distal stump breakdown, changes in the proximal stump vary highly based on the location of the injury with relation to the neuronal body. If the injury is in close proximity to the neuronal body, apoptosis may occur [28]. Otherwise, the breakdown of the proximal stump is typically limited, only progressing to the first node of Ranvier. The primary focus of the proximal portion of the nerve is chromatolysis – the changing of its cell genetic makeup, to focus on a regeneration phenotype [8]. There is an endogenous upregulation of neuroprotectant proteins, for example heat-shock protein-27, in an effort to promote neuronal survival [29, 30]. Nerves that undergo this process have shown upregulation of protein

7 associated with growth including gap-43, tubulin, and actin, while neurofilament proteins, involved with axonal maintenance, are downregulated [31-33].

Dependent on the type and severity of nerve injury repair and recovery can occur in two primary ways: through collateral branching of intact axons or by the regeneration of injured axons [34]. In slight to moderate injuries, collateral branching begins in the first week and continues for anywhere from 3-6 months, until recovery and reinnervation. It is interesting to note that there are initially more sprouting axonal branches than actual number of nerves innervating a target-organ [35]. The branches that do not receive neurotrophic factors from the target-organ eventually undergo a process to degenerate [36, 37]. In contrast, in more severe injuries, where a vast majority of nerve axons have been affected, the primary mode of recovery is axonal regeneration [38]. The regeneration occurs in three broad phases: Wallerian degeneration, axonal regeneration, and target-organ reinnervation.

1.2.7. Schwann Cell Involvement

Typically known for their myelin-producing abilities, Schwann cells are vital, primary mediators in triggering the cascading events in Wallerian degeneration and changes in their protein expression are critical to axon regeneration [8]. The ability for Schwann cells to de-differentiate and later re-differentiate enables them to control their cellular behavior dependent on the situation. In an immediate response to PNI,

Schwann cells detached from axons in the distal nerve begin to de-differentiate and alter gene expression

[39-41]. Changes in protein and factor expression take place as soon as within 48 hours where Schwann cells cease myelin production, in turn upregulating regeneration-associated genetic markers for example,

GAP-43 and neurotrophic factors such as Nerve Growth Factor (NGF), BDNF, GDNF, etc. [42-44]. In conjunction with this increase in regenerative factor expression, Schwann cells begin to proliferate, reaching peak proliferation around 4 days post-injury. These proliferating cells align in their basal lamina tubes to form the notorious Bands of Büngner, providing a supportive basis for regenerating axons [45, 46].

In addition to promoting regeneration, Schwann cells play a key role in the removal of myelin debris, which may otherwise hinder repair, acting as a physical barrier to regenerating axons. Post-PNI, the

myelin sheath of Schwann cells remains to form small ovoids [46], which contain molecules that are

inhibitory to axonal growth such as myelin-associated glycoprotein (MAG) and oligodendrocyte-myelin

glycoprotein (OMgp) [47, 48]. Schwann cells possess the ability to degrade their own myelin and

phagocytose extracellular debris, making them one of the major phagocytic cells for the first 5 days after

PNI [49]. This Schwann-cell mediated myelin degradation is highly regulated by the phospholipase-A2

(PLA2) enzyme family, where cytosolic and secreted PLA2 hydrolyze the phospholipid phosphatidylcholine

into lysophosphatidylcholine, eliciting myelin breakdown, and arachadonic acid [12, 50]. PLA2 levels

remain upregulated in Schwann cells, as well as macrophages, for 2 weeks, correlating to the time course

of Wallerian degeneration after PNI, where the return of PLA2 to basal levels occurs approximately 3 weeks

post-PNI, associated closely with axon regeneration and remyelination [51].

Despite the efficient and effective response of Schwann cells to control and combat the effects of

PNI, their efforts that successfully promote axon regeneration are time-dependent. Schwann cells in the

distal nerve typically struggle to survive and facilitate axon regrowth, with their ability to survive declining

within 8 weeks of injury and denervation [41, 52]. Denervated Schwann cells die by apoptosis within

months after PNI, with persistent cells yielding to atrophy, and this time-sensitive loss of support is

generally considered one key factor that limits the successful regeneration of peripheral nerve axons over

long distances [44, 53-56].

1.3. AUGMENTED PERIPHERAL NERVE REGENERATION

1.3.1. Tissue Grafts for Nerve Repair

1.3.1.1. Autograft

Highly regarded as the gold-standard for the treatment of nerve lesions, the autograft consists of

harvesting a healthy segment of nerve directly from the injured patient [57]. The autograft offers a great

advantage in that the tissue has minimal chance of tissue rejection (nonimmunogenic), having been

harvested from the same patient. Furthermore, autografts provide appropriate neurotrophic factors and

viable Schwann cells for axonal regeneration and nerve repair. The size of the nerve gap and the location

of the injury often play a large role in deciding the choice of autograft [58], with the Sural nerve serving as

the most commonly used autograft [59, 60]. The first clinical experiments with autograft for nerve repair

date back to 1870 in which Philipeaux and Vulpian reported performing the first nerve autografts in dogs

[61]. Despite its successes, the autograft offers a number of considerable disadvantages, namely the

necessity of an additional surgical procedure, causing another nerve injury. This results in the loss of

function at the donor site, often termed donor site morbidity, and may lead to the formation of a painful

neuroma [14, 62].

1.3.1.2. Allograft

Nerve allografting, or the use of donor-related or cadaveric nerve allografts, is an alternative to the

traditional autograft technique for segmental nerve injuries. The allotransplantation of tissues, including

nerve allografts, require systemic immunosuppression in order to limit the risks of potential host system

rejection. The health hazards and vulnerability associated with systemic immunosuppression and morbidity

of immunomodulatory therapy severely limits the widespread use of nerve allografts. Unlike organ

transplantation, nerve allografts do not require lifelong immunosuppression. Rather, systemic

immunosuppression can be terminated within approximately 24 months, once adequate host Schwann cells

have migrated into the nerve allograft [58]. Recent advances in medicine and immunology research have

led to several techniques to limit antigenicity and the risk of rejection of allografts including irradiation,

cold preservation and freezing, and lyophilization [63-67]. The use of allografts relies upon the viability of

both host and donor Schwann cells, with the native and transplanted Schwann cells responsible for proper

remyelination and acting as the source for neurotrophic factors [68]. There remains a high research interest

in biomedical and biological fields to understand the host response to nerve allograft transplantation, in

order to elucidate mechanisms and techniques to mitigate such responses.

1.3.2. Tissue Engineering Strategies: Biomaterials and Nerve Guidance Conduits

Nerve guidance conduits (NGC) represent a multidisciplinary, multidimensional, and cutting-edge

transformative material platform to apply tissue engineering principles and properties (biodegradability,

10 biocompatibility, mechanical mimicry, etc.) to develop biomaterial-based scaffolding to aid in nerve repair and regeneration (Figure 3).

Figure 3: Tissue engineering scaffolds and biomaterials are versatile tools for the repair and regeneration of various tissues.

Often polymer-based, NGCs are designed to bridge nerve defects in order to guide the regenerating axons to the appropriate distal target – significantly improving the efficacy of repair and target reinnervation. The development of NGCs requires input and expertise in biomaterials, biomedical engineering, regenerative medicine, pharmaceuticals, neurobiology, and orthopedic surgery. Over the last decade, vast research efforts have been devoted to NGCs of various compositions, with varying novelties, functionalizations, and capabilities, including the ability to carry and deliver pharmaceutical payloads, small-molecule drugs, and growth factors (Figure 4).

Figure 4: Nerve guidance conduits are engineered tubes which are used to bridge peripheral nerve gaps, capable of promoting key attributes to facilitate quicker and more efficacious peripheral nerve repair and regeneration.

1.3.2.1. Natural Materials for Nerve Regeneration

There are several key advantages which make natural polymers particularly appealing for

biomedical, tissue engineering, and regenerative medicine applications. Most notably, natural polymers are

well-known for their biodegradability, biocompatibility, and economic availability (natural abundance with

lower associated costs). Despite their advantages, which make natural materials particularly well-suited for

nerve regeneration applications, natural materials tend to have weak mechanical properties which often

serves as a profound obstacle for feasible regenerative strategies. Overcoming such limitations requires the

modification of materials in order to enhance physiochemical properties including mechanical strength,

degradation/swelling, and stability. Perhaps the most widely used modification is crosslinking, which refers

to the chemical process by which one polymer chain is linked to another. Crosslinking often results in

higher tensile strength, greater modulus, improved fluid resistance, better crush resistance, etc. Another

common method to create more useful materials is the science of composites – blending natural polymers

with other materials to achieve suitable properties.

Among the countless number of natural materials used for fabrication of NGCs, or otherwise for

repair of PNI, chitosan [69], collagen [70], and silk fibroin [71] have garnered significant attention due to

their biodegradability, biocompatibility, and nontoxic properties. Various studies have shown optimistic

results with regards to nerve repair, showing myelinated axons, recovering nerve functionality, and

colonization of nervous tissue. Other materials of interest for nerve regeneration applications include gelatin

[72], hyaluronic acid [73], and fibrin [74].

1.3.2.2. Synthetic Materials for Nerve Regeneration

Many synthetic materials are popularly used for various tissue engineering and regenerative

medicine applications, including polycaprolactone (PCL) [75], polylactic acid (PLA) [76], polyglycolic

acid (PGA), polylactic-co-glycolic acid (PLGA) [77, 78]. Such materials possess adequate mechanical

strength and structural integrity, however they possess variable levels of biodegradation. Non-degradable

synthetic polymers have also been used for such applications, silicone being the prime example, however

12 these induce complications of foreign body response, inflammation, and secondary surgery. PCL [79], PLA

[80], PGA [81], and PLGA [82] have all been developed and tested as NGCs for nerve regeneration showing successful in vitro and in vivo results, achieving improved nerve diameter, myelination, and functional recovery.

1.4. EXOGENOUS NEUROTROPHIC GROWTH FACTORS

1.4.1. Biofunctionalization by Neurotrophic Factors

Neurotrophic factors are known to provide significant regulation for Schwann cell (SC) myelination in peripheral nerve development through various mechanisms. For each growth factor, there are different signal transduction pathways used and concentration levels necessary to encourage axonal regrowth and remyelination. The neurotrophic factors that act to enhance recovery include nerve growth factor (NGF), glial cell-line derived neurotrophic factor (GDNF), brain-derived neurotrophic factor

(BDNF), neutrohphin-3 (NT-3), and neutrohpin-4/5 (NT-4/5), which will all be examined later on in this section. The specificity for neurotrophic factor levels needed in injury repair can be seen in Richner et al.’s identification that NGF is required for the survival of nociceptive and thermoceptive dorsal root ganglion

(DRG) neurons, NT-3 is necessary for proprioceptive neurons, and BDNF and NT-4/5 are required for motor neurons survival [83]. The effects of the neurotrophic factors are harnessed through direct protein delivery, stimulating their expression or using cell therapies to deliver the proteins locally.

In this section, the mechanisms that allow these factors to encourage regeneration will be explored using experimental evidence along with the presentation of challenges that must be overcome to harness their regenerative potential. The effects of these factors will be shown through research studies involving their direct and indirect use. Later on, the biomedical engineering strategies designed to overcome these challenges for delivery will be presented and analyzed.

1.4.2. Challenges in Neurotrophic Growth Factor Delivery

Neurotrophic growth factors pose difficulty in drug delivery systems based on pharmacokinetics, pharmacodynamics and physiochemical properties, requiring significant regulation for their release. In

order to maintain their stability, growth factors must be processed in mild conditions, including ambient or

low temperatures, in the absence of organic solvents, with stabilizing additives, and in aseptic sterilization

[84]. To overcome the time-dependent limitation, studies often test stimuli that encourage the expression

levels of several neurotrophic factors. Guo et al.’s 2018 study using chitosan conduits containing

simvastatin/Pluronic F-127 hydrogel was found to promote peripheral nerve regeneration by upregulating

the expression of pleiotrophin, hepatocyte growth factor, VEGF and GDNF [85]. Another mechanism that

is commonly used to stimulate growth factor expression is stimulating neuronal activity through the use of

electrical stimulus or exercise for the upregulation of BDNF and other neurotrophic factors [86]. These

types of treatments have shown to be especially effective in pre-clinical testing. Gordon et al. demonstrated

the efficacy of electrical stimulation correlated to an acceleration of axon outgrowth and increased

motoneuron regeneration.

1.4.3. Neurotrophic Growth Factors for Peripheral Nerve Regeneration

1.4.3.1. Nerve Growth Factor

Nerve growth factor (NGF) has been known to play a vital role in developing sympathetic and

sensory nerve cells, directing growth of their regenerating axons [87]. NGF is upregulated following PNI

and functions through binding to Trk receptors, activating a regulatory pathway that induces cytoskeleton

assembly and the outgrowth of neurites (Figure 5). Neurotrophins that bind to Trk receptors activate Ras-

extracellular signal-regulated kinase (ERK) pathway leading to enhanced neurite outgrowth and axon

survival. NGF has been used in a range of PNI treatment studies, including Kuihua et al.’s 2013 study

testing NGF encapsulated in aligned SF/P(LLA-CL) nanofibers to sustain release of NGF over 60 days.

Results suggest that this sustained release mechanism using nanofibers effectively enhanced regeneration

[88]. Since NGF degrades quickly, these encapsulation methods help extend the life span for the NGF

supply significantly by preventing degradation.

While this pathway can be stimulated by using an exogenous supply of NGF, the pathway can also

be stimulated at several points by common pharmacologic agents (e.g. erythropoietin, tacrolimus,

geldanamycin, acetyl-l-carnitine, and n-acetyl-cysteine), which have all been explored as potential

treatments[89]. An example using one of these stimuli in combination with neurotrophic factors can be seen

in Labroo et al.’s study treating chicken embryos in vitro with the immunosuppressant, FK506 (tacrolimus),

along with GDNF and/or NGF. This combination therapy significantly increased neuroregenerative activity

potentially due to FK506 enhancing the growth factors’ efficacy at the injury site [90].

1.4.3.2. Brain-Derived Neurotrophic Factor

Figure 5: A simplified schematic illustrating the biochemical pathways for NGF- and BDNF-mediated neuronal repair and regeneration. Both NGF and BDNF are upregulated following PNI and function through binding to Trk receptors, activating a regulatory pathway that induces cytoskeleton assembly, axonal outgrowth, and myelination.

Brain-derived neurotrophic factor (BDNF) interacts with p75 and TrkB receptors leading to

enhanced myelination in Schwann cells, promotion of neuronal cell survival, and enhancement of neurite

outgrowth. In 2016, its mechanism was investigated by Lin et al., finding that this activity is due to BDNF

activating the JAK/STAT pathway in Schwann cells, leading to cytokine production that encourages

regeneration [91]. A significant amount of research has been focused on investigating pathway

intermediates to control stimulatory effects. The molecule netrin-1 is an important molecule for nerve

regeneration and its overexpression was found to increase the expression of BDNF and NGF, leading to

improved functional recovery and the increased presence of Schwann cells when using transplanted bone

marrow mesenchymal stem cells that produce the molecule [92]. This molecule is a potential target for

stimulating BNDF’s signaling pathway.

The levels of BDNF concentration have a direct influence on axonal regeneration, with low

exogenous BDNF doses found to enhance motor neuron axon regeneration while high doses inhibit

regeneration both initially and long-term following transection, exhibiting dose-dependent facilitation and

inhibition [83, 93]. A 2012 study identified that BDNF expressed in the neurons and Schwann cells work

together to facilitate motor axonal regeneration, potentially mediated by retrograde signaling on the TrkB

receptors of growing axons from the Schwann cell BDNF [94]. These findings led to Gordon et al.’s

investigation of exercise and electrical stimulation as a therapeutic treatment for PNIs, due to the

upregulation of BDNF and other neurotrophic factors [86]. Arbat-Plana et al.’s 2017 study found that a

regulated physical exercise regimen offers a better TrkB activation method than using pharmacological

TrkB pathway agonists, attributed to the increase in BDNF [95]. This treatment is a candidate with strong

clinical potential without the limitations associated with drug delivery.

1.4.3.3. Neurotrophin-3

Neurotrohin-3 (NT-3) has been found to have reduced concentration levels for up to two weeks

following nerve transection with the indication that Schwann cells are the factor’s primary source. In early

studies, the exogenous application of NT-3 led to maintained conduction velocity supporting NT-3 as a

molecule that promotes nerve regeneration. Because persistent denervation causes Schwann cell atrophy

and the reduction of neurotrophic factor production, an exogenous NT-3 supply offers a viable option for

stimulating functional recover because Schwann cells require NT-3 during the early myelin developed in

regeneration [83]. Wang et al.’s 2016 study identified that NT-3 bound to collagen binding domain

promoted ordered axon outgrowth and enhanced functional recovery [96]. NT-3 operates through the

activation of TrkC and evidence suggests that it also drives regeneration through GABA excitation

potentiating presynapses [97]. Based on the information presented, neurotrohpin-3 and agonists for its

pathway are often tested for regenerative potential.

1.4.3.4. Glial Cell-Derived Neurotrophic Factor

In a 2018 study by Nagai et al., it was determined that decreased glial cell-derived neurotrophic

factor (GDNF) leads to decreased levels of m2 muscarinic acetylcholine receptor (m2MAchR) and gamma-

aminobutyric acid receptor α1 (GABAARα1) in injured motoneurons. Following injury, GDNF levels

decrease from normal levels and reach their lowest point around 3-5 days later, returning to standard levels

after 2 weeks [98]. A 2016 study by Tajdaran et al. used the a drug delivery system with PLGA

microspheres to deliver exogenous GDNF, finding increased counts for motor and sensory nerves when

repaired immediately after inflicting injury and in a delayed repair [99]. Marquardt et al. used GDNF

preconditioning to overcome the negative effects that result from mismatching Schwann cells and neuron

combinations (“i.e. motor neurons/muscle nerve-derived SCs vs. motor neurons/cutaneous nerve-derived

SCs”) demonstrating the clinical potential for this treatment [100]. Exogenous supplies of GDNF

compensate for the decreased level endogenous production to promote redevelopment.

1.4.3.5. Neurotrophin-4/5

Neurotrophin-4/5 (NT-4/5) is able to enhance peripheral nerve regeneration through the signal’s

interaction with two different neurotrophin receptors, TrkB and p75NTR, like the BDNF signaling pathway

[101]. Similar to NT-3, there is a downregulation of NT-4/5 mRNA levels following peripheral nerve injury

with the levels returning around 2 weeks post injury. In rodent nerve injury models, NT-4/5 was found to

restore motor axon growth while improving the axon count, diameter and the myelin thickness [83]. English

et al. identified that axon regeneration enhancement was dependent on the expression of NT-4/5 in the

proximal nerve stump when using treadmill training as a stimulant, while changes in the distal segment did

not have the same neuroregenerative effects. Electrical stimulation and treadmill training were found to be

effective in enhancing axon regeneration, which led English et al. to agree with previous researchers in that

the stimulants likely cause the autocrine or paracrine neurotrophic signals [102]. While there have been

minimal studies testing the direct application of NT-4/5 to the injury site, these results support the

molecule’s potential for enhancing regeneration.

1.4.3.6. Vascular Endothelial Growth Factor

Vascular endothelial growth factor (VEGF) was initially discovered as an angiogenic factor based

on the cellular signaling mechanisms for VEGF-A. Despite this, VEGF-B has downstream signaling targets

that regulate both the nervous and cardiovascular systems with activity distinctly different from VEGF-A.

In the peripheral nervous system, VEGF-B has protective effects that mediate repair after injury that are

independent of its vascular effects. Evidence suggests that VEGF-B signals through PI3K in order to

elongate neurites along with an upregulation of pigment epithelium-derived factor (PEDF) with

dopaminergic pathways requiring further investigation [103]. In testing resveratrol’s effects on PNI

regeneration, resveratrol was found to stimulate p300 acetyltransferase-mediated VEGF signaling of the

ventral spinal cord for rats that experienced sciatic nerve injury and lead to improved motor function and

repair [104]. Based on these results, resveratrol has regenerative potential and should be studied

mechanistically. In a 2017 study by Di et al., an exogenous supply of VEGF-B was found to promote

recovery and trophic function through reactivating PI-3K/Akt-GSK-3β-mTOR signaling, attenuating

neuronal oxidative stress, and elevating PEDF expression [105]. These finding support the use of VEGF-B

as a potential therapeutic agent and, with the evidence from previous studies, can have its pathway

stimulated by drug molecules.

1.5. DRUG DELIVERY STRATEGIES AND CELL THERAPIES

1.5.1. Drug Delivery via NGCs

Various mechanisms can be used to incorporate small drug molecules and growth factors of interest

into the NGC matrix. In a 2013 study, Tang et al. used an NGF concentration gradient in conduits, created

using adsorption of the growth factor in a solution to form a gradient with the molecules immobilized in

the PLCA membranes. This method had results comparable to the autograft standard and superior to the

membrane with a uniform nerve growth factor concentration [106]. Recently, immunotherapy approaches

have also been explored for nerve regeneration. Mokarram et al. showed early modulation of the immune

environment at the site of nerve injury by delivering fractalkine, a chemokine protein found commonly in

18 neural cells, using a nerve guidance conduit [107]. The authors of the paper showed dramatic increases in regeneration using both histological and electrophysiological assessments to characterize the efficacy of their delivery system, potentially introducing an alternative paradigm to facilitate the endogenous repair undergone by nerve tissues after PNI. In a 2010 study, Wood et al. created an affinity-based delivery system using various neurotrophic factors and polymerizing silicone tubes with a fibrinogen solution for sustained delivery. In this study, the test group using the delivery system for GDNF produced superior or equivalent functional recovery and motor nerve regeneration results, when compared to an isograft model [108]. This affinity based method is commonly used involving heparin and heparin sulfate to protect growth factors from proteolytic cleaving while also making them available to trigger signal transduction pathways [109].

The structure of the scaffold itself can also be altered to promote sustained delivery of a given drug molecule. Hong et al. used multilayered PLGA scaffold (the first layer aligned for topographical cues and the following two layers randomly oriented) to create different rates for biodegradation and controlled release of growth factors, where the scaffold can be tailored chemically and morphologically to provide specific cues to the regenerating tissues. The in vivo study found positive results through behavioral tests, mechanical allodynia analysis and morphological studies of the tissue [110]. Madduri et al. argues that the further investigation in a clinical trial is challenged by the rapid release of these factors from the NGC matrix and the harsh and difficult conditions necessary to process these devices [109].

Small-molecule drugs and growth factors can also be suspended in matrices that fill the lumen of

NGCs (intraluminal matrix) in order to provide local release. The active material used in these matrices can be laminin, heparin, heparin sulfate, alginate, and hydrogel-forming collagen. The matrices are viscous in order to provide slow fluid exchange coupled with rapid release of growth factors. Molecular interactions with the growth factors can vary, including ionic, electrostatic, and hydrophobic interactions along with ligand-receptor binding based on the physiochemical nature of the growth factors [84]. An example of a luminal filler that mimics the extracellular matrix is the dual-component hydrogel system (cGEL) that

Kohn-Polster et al. designed and found to be superior to the use of empty conduits by providing mechano-

19 biochemical stimulation for regeneration [111]. Gnavi et al. used biocompatible and biodegradable gelatin- based hydrogel to deliver VEGF and was found to prevent the growth factor’s degradation for it to promote nerve regeneration [112]. The matrix used in the lumen must be specialized for the delivery system to optimize regenerative effects.

1.5.2. Microspheres for Drug Delivery

NGC are commonly combined with polymeric microspheres as a method to deliver growth factors and maintain dose regimens. Growth factors get loaded into the microspheres and then incorporated into the NGC walls or the lumen to degrade over time, releasing the drug to enhance regeneration. Tajdaran et al. reports that PGLA is the most widely used polymer for microsphere drug delivery, typically ranging in size from 200nm to 600µm with the release time ranging from one week to several months, based on the composition of the microspheres. They also can be used to hold drugs with a variety of properties and drug profiles to promote sustain release to an injury site [113].

In a 2017 study on the effects of timed EPO release using PLGA microspheres, Zhang et al. showed that rats with this treatment mechanism had better recovery that those undergoing intraperitoneal injections along with the positive and negative controls, seen through functional, electrophysiological and histological analysis [114]. While intraperitoneal injections are often used in these drug delivery studies, this study suggests that release using microsphere promote improved regeneration results without requiring repeat injections. Using PLGA microspheres for PNI treatment Tajdaran et al.’s 2015 study tested the sustained released of tacrolimus (FK506) as a method to reduce drugs systemic effects of immunosuppression. FK506 was added to a fibrin gel that was then encapsulated into PLGA microspheres that released FK506 over 28 days, finding enhanced neurite outgrowth and nerve regeneration without systemic immunosuppression

[115]. Using similar methods they also tested GDNF in fibrin gel incorporated into PLGA microspheres which improved axon regeneration along with fiber maturity [99]. This study demonstrates that microspheres are useful for harnessing local effects of drug molecules and limiting systemic effects.

While PLGA is the most commonly used material for microspheres because of its FDA approval,

other materials, like gelatin, can also be used as the base molecule for microspheres. To encourage the

sustained release of neurotropic factors, Matsumine et al. performed a study using basic fibroblast growth

factor (bFGF) in acidic gelatin hydrogel microsphere compared to silicone tubes infused with bFGF. The

bFGF microspheres resulted in faster regeneration and more mature nerve axons after 7 weeks, supporting

the positive influence of sustained, controlled release of stimulatory molecules [116]. To further study the

effects of this delivery method, nerve recovery should be tested varying the size of the microspheres to

optimize regenerative effects based on the surface area to volume ratio. Microsphere solutions can also be

used to fill the luminal space of NGCs to provide similar effects as when they are incorporated into the

conduit structures. Chen et al. used this method by injecting microspheres with NGF into the lumen of

conduits using a microsyringe, finding no outflow from this region into the surrounding tissues. While

testing the effect of icariin (an element of Epimedium extract) on PNI recovery, they found that both

suspended icariin and microspheres with NGF in the lumen promoted improved nerve conduction velocity

and fiber myelination [117].

1.5.3. Cell Therapies for Peripheral Nerve Regeneration

1.5.3.1. Schwann Cells

Schwann cells play a crucial role for creating an environment conducive to recovery after peripheral

nerve injury. Barton et al. explains how, when damaged axons transition from the “conducting” phenotype

to the “growing” phenotype, myelinating Schwann cells will de-differentiate to become non-myelinating

Schwann cells for approximately 6 months following injury until the re-differentiate into myelinating

phenotype. Non-myelinating Schwann cells feature down-regulation of the genes responsible for

myelination and an up-regulation of genes that lead to neurotrophic factor and adhesion molecule

production. This change in phenotype is associated with Schwann cells’ proliferation and with the overall

effects of providing growth factors locally, influencing immune response, and clearing cellular debris [118].

Common treatment methods for Schwann cell-based therapies include their direct transplant and stem cell

21 transplants that then differentiate into Schwann cell-like cells. They also have been combined with neuroprotective agents and inhibitory molecules to encourage improved regeneration [118].

Nerve guidance conduits seeded with Schwann cells have been developed as an alternative treatment to autograft nerve transplants with support given by in vivo studies. Early studies found the development of new nerve fascicles at the distal nerve stump following repair with a vein graft NGC seeded with Schwann cells producing increased myelin density and average conduction velocity [119]. Barton et al.’s study tested a co-transplant of olfactory ensheathing cells and Schwann cells to find that the combined transplant yielded better results than when treating with either of the cell types in isolation. Transplanted

Schwann cells are capable of acting as drug delivery modalities, supplying endogenous neurotrophic factors to injury sites with the capacity for the cells to migrate at the site. The secretion of neurotrophic factors and cell-adhesion molecules provide a guiding effect for regenerating axons. SCs can be genetically modified to secrete viral and non-viral vectors, which have both undergone pre-clinical in vivo studies to determine their potential. Designing clinical-grade vectors comes with challenges regarding their clinical safety, control, target specificity and resultant immune response, despite the pre-clinical model successes.

Adenovirus-based gene therapies are the most commonly tested viral vectors with extended release duration, the ability to directly or retrogradely transduce neurons, and minimal potential for insertional mutagenesis [120].

It is important to note that the clinical use of Schwann cells has been limited due to a lack of feasibility and commercial viability, supported by an FDA phase 1 clinical trial for the treatment of MS which found that there was no evidence supporting the sustained viability of transplanted Schwann cells

[118]. In an evaluation of the survival of transplanted Schwann cells in vivo, Gambhir et al. used a nucleus- restricted fluorescent label and co-administered NRG1β, a therapeutic with the potential to drive rapid

Schwann cells migration, finding improved cell viability [121]. These conflicting studies suggest that more research is necessary to determine the effectiveness of transplanting Schwann cells or if using stem cells is a valid alternative once the cells differentiate and produce neurotrophic factors.

1.5.3.2. Stem Cells

Stem cells from various origins have been used in peripheral nerve regeneration studies in vitro and

in vivo as a cell therapy treatment, including embryonic stem cells (ESC), induced pluripotent stem cells

(iPSC), adult mesenchymal stem cells (MSC), and neural stem cells (NSC). This treatment method employs

similar mechanisms to stimulate regeneration as the previously mentioned Schwann cell transplants,

because implanted stem cells are often driven to develop into Schwann cell-like cells that will provide

neurotrophic factors to the injury site [57]. This promotes the development of a positive microenvironment

that encourages the survival of neuronal cells and the formation of new nerve cells [122].

Transdifferentiation of stem cells into Schwann cell-like cells has been a field of interest for

researchers, with several techniques showing promising results. Chemical treatments incorporating a

cocktail of growth factors have shown consistent results in driving stem cells towards such Schwann cell-

like lineages [123-125]. As an alternative to chemical growth factor transdifferentiation, in a 2018 study,

Manoukian et al. experimented with in vitro electrical stimulation of human MSCs showing positive results

for morphological and phenotypical changes towards Schwann cell-like or neuronal lineages [126].

Whether stem cells are transdifferentiated prior to incorporation or not, there are several developed

techniques for their delivery in vivo. Stem cells can be suspended in a matrix or medium which is then

microinjected into the nerve endings or around the repair site. Intraneural injections can be traumatic to the

intra-neural architecture without the ability to predict the distribution of the cells once injected, making this

an unfavorable method for delivery [127]. Fibrin matrix or hydrogels are often used to suspend the stem

cells for local delivery. These suspension matrices can be loaded into conduits and are used to repair nerve

defects by injecting them either into the lumen or the matrix of the conduit [122].

In a study testing a neurotrophic factor delivery system, Meyer et al. transplanted Schwann cells

genetically modified to overexpress neurotrophic factors and used NVR-hydrogel-enriched chitosan

conduits. The NVR-Gel significantly reduced the neuroregenerative effects of the growth factors being

expressed, impairing axonal outgrowth. Despite these negative findings, the hindrance was overcome when

23 the Schwann cells were engineered to overexpress FGF-218kDa [128]. Conduits the have the cells cultured or seeded onto them do not experience the inhibition that cells experience when moving through the matrices like hydrogels. The conduits that are commercially available from natural or synthetic materials have useful substrates that allow cell adhesion although the degradation profiles for these conduits need to be considered in the cell therapy development [127]. Stem cell treatments have been found to promote regeneration in the presence and absence of luminal fillers.

1.6. ANIMAL MODELS AND FUNCTIONAL ASSESSMENTS

The functional assessment selected to determine recovery status is dependent on the selected animal model, the functions that the injured nerve controls, and the desired measurements. The sciatic nerve, the leading nerve model, innervates the portion of the musculoskeletal system used for walking as well as fine motor functions of the toes, providing distinct biomechanical evidence of nerve injury. To detect sensory recovery, nerves require stimulus to provoke the withdrawal reflex response, indicating both sensation redevelopment and muscle re-innervation. Outside of function specific testing, studies often perform tests to determine the amount of development that has occurred during recovery through histology and morphology.

Early testing included the use of larger animal breeds (non-human primates, dogs, cat, etc.) for better translational data, but there has been a significant decrease in their use for ethical and humane reasons in the United States [129]. The animal type and nerve selected for induced PNI’s limits the functional assessments available for testing. Several of the functional assessments (nerve conduction and muscle atrophy studies) can be used universally for nerve function while others are only available for specific animals due to size restrictions or are based on certain motor functions (walking gait, vibrissae movement, and finger function). Researchers tend to migrate toward rodent models (most popularly, the rat model) as the field standard producing comparable data. Despite this, there has been movement toward replacing in vivo studies with in vitro models to for animal welfare but are limited due to nerve tissue’s structural complexity [130].

1.6.1. Rat Sciatic Nerve Model

The sciatic nerve in rats is the standard test model for nerve regeneration studies due its relatively large size and accessibility, along with dramatic functional deficits that result from its injury permitting functional recovery assessment. It stretches from the sciatic notch in the pelvis until it trifurcates mid-thigh into the tibial, peroneal, and sural nerves, providing a large enough region to facilitate surgeries [130]. The sciatic and following nerves have all been used in detecting leg, foot and toe function for recovery. Changes due to defects include changes in paw placement, stepping patterns, gait phases, strength and mobility [131].

Studies using larger animal breeds tend toward us of the femoral, tibial and peroneal nerves because these animals do not have the same size restriction as rodents while producing comparable defects [129].

Standardized functional indices for the sciatic, tibial, and peroneal nerves are commonly used as well as the

Basso Beattie and Bresnahan (BBB) locomotive rating for assessing mobility.

1.6.2. Walking Track Analysis and Sciatic Functional Index (SFI)

For the detection of gross motor function in gait across animal models, walking track analyses are used extensively. Earlier methods used for walking track analysis used different colored ink to mark the rodent’s foot prints, which were used to measure the print length along with the 1-5 and 2-4 toe spread as the rodent walked on a paper path [132]. In recent history, technological advancements have made digital walking track analysis possible where software programs have been designed to analyze the forelimb- hindlimb coordination for rodents and detect movements that cannot be caught visually with static and dynamic paw parameters [37]. These tests analyze aspects of the animals’ walking pattern incorporating stride length and the area of the paw in the considered factors for determining sciatic nerve injury [133].

Computerized walking gait systems use a transparent motorized treadmill with ventral high-speed video cameras to record normal walking, which is then digitized leading to calculation of a variety of gait parameters including the Sciatic Functional Index (SFI) without the high level variability and confounding factors affecting manually recorded data [134]. This method can be translated for the tibial and peroneal nerves to produce corresponding indices [135, 136]. Arguably the “gold standard” for functional assessment

25 of sciatic nerve damage and repair, the SFI has been well established in literature for a variety of studies aimed at peripheral nerve regeneration [80, 110, 137, 138]. To calculate the SFI, the following Equation 1 must be used, where EPL is the experimental print length, NPL is the normal print length, ETS is the experimental toe spread, NTS is the normal toe spread, EIT is the experimental intermediary toe spread, and NIT is the normal intermediary toe spread [136]:

퐸푃퐿−푁푃퐿 퐸푇푆−푁푇푆 퐸퐼푇−푁퐼푇 푆퐹퐼 = −38.3 × ( ) + 109.5 × ( ) + 13.3 × ( ) − 8.8 (1) 푁푃퐿 푁푇푆 푁퐼푇

1.6.3. Basso, Beattie, and Bresnahan (BBB)

The BBB locomotor rating is a motor coordination index used for open field testing on rodents.

This system takes into account 10 behavioral categories including limb movement, trunk position, paw placement, coordination, toe clearance, predominant paw position, trunk instability, and stepping along with the general positioning of the abdomen and the tail using a standardized scoring sheet for observation.

The BBB scale spans a wide spectrum, ranging from a score of 0, consisting of no observable hindlimb movement, to a score of 21, consisting of consistent plantar stepping and coordinated limb movement.

Relative scoring within these categories makes this method reliable for comparison when done by trained examiners. Basso et al. identified the curvilinear relationship between functional motor recovery and tissue sparing for spinal cord injuries and further deduced that this technique can be used to accurately predict the spared white matter at an injury site for spinal cord injuries [130, 139]. Barbosa et al. later used this method for detecting functional recovery from a sciatic nerve injury in rats, to detect sensorimotor function. The comparison of BBB scores indicated that the hydrochloric extract treated animals experienced accelerated functional recovery [140]. The BBB does not take fine motor function of the paws into consideration like the SFI in its inclusion of toe spread and paw area, serving as a fundamental difference between the two, along with its more subjective basis of rating. Most nerve regeneration studies use a combination of the functional recovery assessments [141, 142], like Mohammadi et al.’s study of thyroid hormone on rats, confirming faster and better recovery of axons through BBB and SFI [143].

1.6.4. Nerve Conduction Velocity and Compound Motor Action Potential (CMAP)

A functional assessment measuring the compound motor action potentials (CMAP) requires the nerve to be stimulated proximal to the injury site under anesthesia using an electrode needle, recording the evoked potential with a microneedle inserted into the muscle and corresponding joint [144-147]. This study provides information on the transmission of action potentials resulting in muscle movement, indicating the return of motor neuron innervation. For detection of sensory neuron re-innervation, this study can be done with the stimulating electrodes placed distal to the injury site and the microneedle electrode proximal [130].

Ichihara et al. used these methods to record the distance from the stimulation site and measure the amplitudes of action potentials from peak-to-peak to calculate the nerve conduction velocity [144]. This method can also be used for large animal models, as seen in its use in a study of ultrasound guided plasma rich protein factor’s influence on recovery from a crush injury in an ovine (sheep) model [148].

1.7. CLINICAL APPLICATIONS AND COMMERCIALIZED NGCs

While there have been significant tissue engineering advancements over the past decade using nerve guidance conduits (NGC) and pharmaceutical agents, few of these developments have been tested clinically. The current standard-of-care for PNIs includes gold standard the autografts, FDA-approved hollow conduits, and decellularized nerve allografts [149]. Presently, there is a wide range of conduit materials and therapeutics with a variety of mechanisms that have exhibited their ability to enhance nerve regeneration and have not been tested clinically. Further in vitro and in vivo studies are needed to determine the potential side effects of tissue engineering treatments prior to their clinical debut. The sural nerve is the most common nerve used for autograft procedures but, as a sensory nerve, there is limited functional recovery developed in the treatment of mixed motor-sensory nerves [150]. Research is needed for treatment without limitations or restrictions based on the defect size or the time between the injury and repair [151].

Currently there are FDA-approved conduits for treatment of PNI composed of a wide variety of biomaterials, both natural and synthetic, with a wide range of diameters, lengths, and properties. Table III summarizes some of the popular commercially available FDA-approved conduits.

Table III: Commercially available nerve guidance conduits and repair products with associated information and properties.

CHAPTER 2: ALIGNED MICROCHANNEL POLYMER-NANOTUBE COMPOSITES FOR

PERIPHERAL NERVE REGENERATION: SMALL MOLECULE DRUG DELIVERY

This chapter is adapted from an article published in Journal of Controlled Release, Vol. 296, Manoukian, Ohan S., et al. “Aligned microchannel polymer-nanotube composites for peripheral nerve regeneration: Small molecule drug delivery.” Copyright Elsevier (2019). https://doi.org/10.1016/j.jconrel.2019.01.013.

2.1. INTRODUCTION

Peripheral nerve injury (PNI) often results in a loss of sensory, motor, and autonomic functions in the affected region. Moderate and severe PNI is often marred with very slow and frequently incomplete regeneration and can cause severe loss of sensory and motor function, where recovery is inversely proportional to the severity of damage [14, 152, 153]. Such injuries have a significant, and sometimes permanent, impact on patients and their day-to-day activities. Unlike cellular repair in other injuries, the wound healing response of the peripheral nerve does not involve mitosis and cellular proliferation [152].

In contrast, upon PNI, a calcium-mediated process known as Wallerian degeneration is initiated distal to the injury zone, starting with axonal breakdown, as a measure to alleviate abnormal axon regeneration [20].

Regeneration at the proximal stump begins within 24-48 hours, as a growth cone protrudes from the axonal stump and grows toward the target organ [154], but this endogenous repair usually does not sustain itself beyond 12 months [52].

Axonal regeneration is an extremely slow process that occurs at a rate of ~1mm/day and requiring at least 12-18 months for muscle reinnervation and initial functional recovery [155]. However, nerve gaps in the peripheral nervous system larger than 5mm often cannot regenerate naturally [156]. In the case of a critical-sized nerve defect, where the lesion extends too far away from the proximal end of the intact axon, regeneration to the distal stump is vitally slow and there is a great risk of traumatic neuroma formation, which not only renders the neuron ineffective, but is often very painful [157]. Those injuries which recover spontaneously (neurapraxic lesions) are thought to represent instances in which at least some of axons retain continuity through the lesion site [158]. If all axons are transected, then recovery depends on surgical intervention [37, 159]. Axonal and myelin sheath degradation stimulate proliferation and migration of

Schwann cells and other neuronal support cells to the site of injury, forming conduits which guide regenerating axons to their target, known as the bands of Büngner [160, 161]. In addition to acting as a conduit, Schwann cells are a source of key neurotrophic factors, which exert trophic influence to which regenerating axons preferentially respond to, stimulating axonal regeneration by upregulating neuronal cell adhesion molecules and trophic cues [162, 163].

Current treatments for PNI depend largely upon surgical intervention, most popularly the use of autografts and allografts. These are limited by donor site morbidity and sensory loss, scarring, neuroma formation, limited supply, and potential immunosuppression [164]. Typical engineering approaches focus on synthetic or biological nerve conduits, however, due to variable, suboptimal outcomes these options are not often clinically utilized or used for noncritical small diameter sensory nerves only. Transient electrical stimulation has been extensively studied and has been reported to increase the speed of functional recovery after crush PNI due to increased nerve conduction. However, critical-sized nerve defects require the use of nerve guidance conduits to bridge the gap and implementation of electrical stimulation poses a challenge in a clinical setting [165].

In an effort to mimic the effects of transient electrical stimulation, namely promotion of nerve impulse conduction via a pharmacological approach, the sustained release of a pharmacological agent that promotes impulse conduction may aid in regeneration of peripheral nerve following injury. A small- molecule drug, 4-Aminopyridine (4AP), with a molar mass of 94.11 g/mol is a voltage-gated potassium channel blocker and FDA approved in 2010 as AMPYRA® (Acorda Therapeutics, Inc.) for treatment of multiple sclerosis (MS). Due to its potassium channel-blocking abilities, it has been shown to prolong nerve action potentials and strongly promote neurotransmitter release [166-168]. Furthermore, direct administration has shown to enhance both speed and extent of functional recovery, as well as promote remyelination, following minor nerve crush injury [158]. However, the use of 4-aminopyridine has never been reported for applications in a segmental nerve defect, where there is a physical gap between proximal and distal ends of the nerve, or in any sustained-release formulations for such an application.

The ability of 4AP to block potassium channels causes delayed repolarization and allows it to prolong action potentials and amplify neurotransmitter release within neurons. This serves to potentiate electrical performance and is the key factor which contributes to 4AP’s clinical feasibility and utility as a therapeutic for MS [169], spinal cord injury [170], myasthenia gravis [171], and Lambert-Eaton syndrome

[172]. The 4AP-induced blocking of potassium channels also has an effect on Schwann cells, which are present on the neurons and which migrate to the site of injury. The reduction of voltage-gated potassium channels has been correlated to the onset of myelin formation and developmental myelinogenesis [173,

174]. As such, a pharmacological reduction of potassium channels may have similar effects, leading to substantial myelin differentiation and neurotrophin release.

A number of natural and synthetic polymers are currently employed in biomedical applications, particularly as tissue engineering scaffolds and/or drug delivery systems. Most natural polymers are particularly attractive due to their inherent biocompatibility and biodegradability. Chitosan (Cht), the partially deacetylated derivative of chitin, is one such natural polymer derived from the shell of shrimp, crab, and other crustaceans. As such, it is abundant, affordable, and easy to commercially produce.

Furthermore, several decades of polymer research have shown chitosan to be biodegradable, biocompatible, and possess antimicrobial properties [175-178]. Chitosan and its complexes have been studied for several biomedical applications including wound healing [179, 180], drug delivery [181], and implants [182]. These include applications related to nerve repair and regeneration [14, 183-185]. However, chitosan often lacks appropriate mechanical properties, particularly Young’s modulus, for many biomedical applications. Its appeal as a hydrophilic polymer encourages the formation of hydrogels, however those limit the end- application, not having sufficient structural and mechanical integrity.

As such, we investigated the use of halloysite nanotubes (HNT) to form a stronger chitosan-HNT composite structure. HNT’s have garnered major research attention over the last five years due to their nano-nature and intrinsic properties. Halloysite is a naturally-occurring aluminosilicate nanotube and has been researched for applications as reinforcements in nanocomposites, where they serve to increase

31 mechanical properties [186]. The neighboring alumina and silica layers, and their waters of hydration, curve and form multilayer tubes due to a packing disorder. They possess a hollow lumen (10-150 nm diameter) which provides excellent ability to carry, encapsulate, and transport chemical agents [186]. They can be effective in controlling the initial burst drug release typical of polymeric matrices and can prolong the sustained drug release, which remains a challenge for natural polymer systems.

In this study, we describe the fabrication, characterization, and testing of a porous chitosan sponge conduit scaffold, fortified with HNTs, with aligned micro-channel porosity as a drug delivery system and nerve guidance conduit (NGC) to bridge peripheral nerve defects, capable of guiding regenerated axons and delivering a sustained release of 4AP to the local target site.

2.2. MATERIALS AND METHODS

2.2.1. Materials

High molecular weight chitosan, halloysite nanotubes, epichlorohydrin, and FITC were purchased from Sigma-Aldrich (St. Louis, MO). Glacial acetic acid, sodium hydroxide pellets, phosphate buffered saline (PBS), LIVE/DEAD™ Viability/Cytotoxicity Kit, regenerated cellulose dialysis tubing with a

MWCO of 3,500 Da, Nunc™ Lab-Tek™ II Chamber Slides, 4% Paraformaldehyde solution, Triton™ X-

100, and NucBlue DAPI reagent were all purchased from Fisher Scientific (Fair Lawn, NJ). Normal goat serum, anti-NGF antibody (catalog #ab6199), anti-P0/MPZ antibody (catalog #ab61851), anti-BDNF antibody (catalog #ab108319), and goat anti-rabbit Texas Red (catalog #ab6719) were purchased from

Abcam (Cambridge, MA). 4-Aminopyridine (>99%) was purchased from Alomone Labs (Jerusalem,

Israel). Human Schwann cells and cell media were purchased from Sciencell Research Laboratories Inc.

(Carlsbad, CA).

2.2.2. Halloysite Nanotube Drug Loading

4AP-saturated solution was produced by dissolving 4AP in ultrapure distilled water at a concentration of 50 mg/mL at room temperature. Following dissolution, dry HNT was mixed with the 4AP- saturated solution in a weight ratio of 1:2. To encourage better dispersion and prevent precipitation,

32 ultrasonication was employed for 1 hour. Cyclic vacuum pumping in/out was employed to enhance the replacement of air in the HNT’s internal lumen with saturated liquid [187]. Vacuum was applied for 30 minutes, then the vacuum was broken, allowing 4AP solution to enter the lumen. This vacuum process was repeated three times and finally left overnight for higher drug loading. The HNT-4AP powder was separated by centrifugation (5000 rpm, 20 minutes) and washed with DI water three times to remove unloaded drug.

Post-centrifugation, HNT-4AP powder was collected and freeze-dried for 24 hours to obtain dry powder.

10 mg of dry HNT-4AP powder was weighed and evaluated for drug loading using thermogravimetric analysis [186, 187].

2.2.3. Fabrication of Composite Nerve Guidance Conduits

High molecular weight chitosan (Cht) (3.0% w/v) was dissolved in 2% (v/v) acetic acid solution at

50 °C in a water bath. The chitosan solution was stirred overnight until a homogenous solution was obtained. Air remaining in the solution is removed by vacuum pump for 24 hours. Conduits were prepared by injecting chitosan solution into custom-made molds of appropriate dimensions. The molds were subjected to 1 h of vacuum to remove air bubbles and allow equilibration at room temperature. The following unidirectional freezing was performed using a modified version of a published methodology

[188]. The molds were then placed in insulating Styrofoam containers, such that only the bottom surface was exposed. The molds, covered with Styrofoam, were then placed onto the surface of a pre-cooled stainless-steel plate submersed in a 5 cm-deep pool of liquid nitrogen. As such, this created a uni-axial thermal gradient, which resulted in unidirectional freezing of the Chitosan solution (Figure 6). Samples were allowed to freeze for 30-45 min and were then subjected to freeze-drying to create to final structure.

Random (unaligned) conduits were fabricated by simply freezing the chitosan solution overnight at -80 °C prior to freeze-drying. Similarly, samples with HNT and HNT-4AP were fabricated with the addition of

5.0% (w/w) of dry powder to the stirring chitosan solution. The freezing and lyophilization of HNT-4AP conduits was identical to that of neat chitosan conduits.

Figure 6: Simplified schematic of the unidirectional freezing setup which uses a uni-axial thermal gradient to create linear formation of ice crystals. (b) The formation of linear ice crystals is shown using the uni-axial thermal gradient setup. (c) Simplified schematic illustrating the process of incorporating 4AP drug into Halloysite Nanotubes (HNT) and subsequently mixing with chitosan in a custom mold to produce drug-loaded conduits. A photograph of the prototype is shown along with representative SEM images of the aligned, porous microstructure.

2.2.4. Crosslinking of Composite Conduits

Freeze-dried chitosan conduits were crosslinked using alkaline epichlorohydrin (ECH). Alkaline epichlorohydrin solution (0.01 mol/L ECH) was prepared in NaOH solution (0.067 mol/L) and samples were incubated in ECH solution at room temperature for 30 minutes [189, 190]. After this period, the samples were rinsed thoroughly with DI water three times to remove unreacted ECH. Samples were then vacuum dried to remove residual water.

2.2.5. Water Absorption (Swelling)

Crosslinked samples were tested for water absorption properties by soaking samples in 37 °C PBS

(pH 7.4). Pre-incubation, samples were weighed to obtain initial dry weight. At predetermined timepoints, samples were removed from PBS, excess liquid was removed by gently tapping the samples onto Kimwipe

34 paper and weighed to obtain the wet weight. The percent water absorption was calculated using the simple ratio of weights given in Equation 2, where Wt is the wet weight (g) and W0 is the dry weight (g) prior to immersion in PBS [191].

푊 −푊 푊푎푡푒푟 푎푏푠표푟푝푡푖표푛 (%) = 푡 0 ∗ 100 (2) 푊0

2.2.6. Conduit Porosity

Porosity of composite structures was calculated using a standard calculation of volume and mass,

3 given by Equation 3 [191], where Vm is total volume of samples (cm ), Wm is mass of samples (g), ρ is density of chitosan (1.342 g/cm3). For all porosity measurements, samples (n=5) were formed in cylindrical molds and measurements of height and diameter were made in order to calculate total volume.

푊푚 푉푚− 푃표푟표푠푖푡푦 (%) = 휌 ∗ 100 (3) 푉푚

2.2.7. Scanning Electron Microscopy (SEM) Imaging

Surface morphology of samples was visualized and analyzed using scanning electron microscopy

(SEM) (JEOL JSM-6335F, JEOL USA, Inc., MA, USA). Samples (n=3 for each formulation) were appropriately cut, adhered to carbon tape on SEM stub, and coated with Au/Pd using a Polaron E5100 sputtering system (Quorum Technologies, East Sussex, UK) for 3 minutes prior to imaging; This achieved a conductive layer of approximately 18 nm. ImageJ (NIH, Rockville, MD) was used to process and analyze images.

2.2.8. In Vitro Degradation

The degradation of conduit conduits (n=5) was studied in PBS (pH 7.4) and PBS+lysozyme at 37

°C. Conduits were initially weighed and immersed in PBS solution containing 4 mg/mL lysozyme at 37 °C for 12 weeks [192, 193]. Samples were collected after 1, 2, 4, 6, 8, and 12 weeks, washed three times in DI to remove ions adsorbed on the surface and freeze-dried. The freeze-dried samples were weighed and the

35 degradation (mass loss) was calculated using Equation 4 [194], where Mi is the initial weight (g) of the conduits prior to immersion and Mf is the final weight (g).

푀 −푀 푀푎푠푠 푙표푠푠 (%) = 푖 푓 ∗ 100 (4) 푀푖

2.2.9. Mechanical Testing

Crosslinked neat chitosan (0% HNT) and 5.0% HNT samples (n=6) were tested in the wet state by incubating in PBS (pH 7.4) overnight at room temperature prior to testing. Testing was performed on an

Instron 3300 Single Column Universal Testing System (Instron, Norwood, MA, USA) at a tension rate of

5 mm/min. The testing yielded both tensile strength and Young’s modulus results of the various samples.

2.2.10. Crystallinity (X-Ray Diffraction)

X-Ray diffraction (XRD) studies were performed using a Bruker D2 Phaser (Bruker AXS,

Madison, WI, USA). XRD patterns for neat chitosan, neat 4AP, HNT-4AP, and composite CHT/HNT-4AP structures were recorded. Values were recorded from 5-40° (2θ) at a scanning speed of 0.2 deg/s. Cu-Kα (λ

1.546 Ǻ) radiation at 40 kV and 30 mA was used as the X-Ray source.

2.2.11. Thermal Characterizations (DSC and TGA)

DSC thermograms were obtained for neat chitosan, neat 4AP, HNT-4AP, and composite

CHT/HNT-4AP samples using a TA Instruments DSC Q100 (TA Instruments, New Castle, DE, USA). 10 mg of each sample in an aluminum pan was heated from 0-400 °C at a rate of 10 °C/min under 50 mL/min nitrogen purge. The DSC data was analyzed using the associated TA Instrument’s Universal Analysis software package. Similarly, thermogravimetric analysis (TGA) of aforementioned samples was done using a TA Instruments TGA Q-500 (TA Instruments, New Castle, DE, USA). A 10 mg sample in a platinum pan was heated from 0-700 °C at a rate of 10 °C/min under nitrogen purge. Thermograms were recorded to capture weight loss as a function of increasing temperature.

2.2.12. Halloysite Nanotube Distribution, Alignment, and Imaging

In an effort to qualitatively determine the distribution of HNTs throughout the chitosan polymer matrix, HNTSs were loaded with FITC in a similar manner as fully described above. Briefly, FITC was dissolved in acetone (1 mg/mL) and following dissolution, dry HNT was mixed with the FITC-saturated solution in a weight ratio of 1:2. The FITC-loaded HNTs were then centrifuged, freeze-dried to obtain powder, and later applied as 5% to the chitosan solution to create a composite structure. The structures were fabricated as described above, and thin circular slices were cut and the FITC-loaded HNTs were imaged using confocal microscopy. The images were processed and analyzed for directionality using ImageJ.

2.2.13. In Vitro Human Schwann Cell Viability

In order to assess the biocompatibility of chitosan conduits, an in vitro assessment was conducted using human Schwann cells at passage two. Cell viability was tested using a technique commonly known as live/dead staining, where kits combine fluorescent reagents to yield two-color discrimination of the population of live cells (green) from the dead-cell population (red). The LIVE/DEAD™

Viability/Cytotoxicity Kit was used and samples for live/dead assay were prepared by following assay manufacturer’s instructions. Thereafter, confocal microscopy (Zeiss LSM 880, Carl Zeiss AG, Oberkochen,

Germany) was performed for imaging.

Chitosan conduits were prepared as per the aforementioned method and sterilized under ultraviolet

(UV) light for 1 hour on each side. Post-sterilization, conduits were soaked in appropriate medium for 1 hour prior to seeding 30,000 cells on top surface in untreated 24-well plates. Cell staining and imaging was performed 7 days post-seeding and confocal microscopy was used to obtain Z-stack images of conduits (10 slices, in 10 µm increments) in order to assess cell penetration within the 3D conduit and overall cell viability. Images were analyzed and maximum projection and gradient hyperstacks were created using

ImageJ (NIH, Bethesda, MD).

2.2.14. In Vitro 4AP Drug Release

For 4AP drug release studies, the following groups were prepared as per aforementioned methods:

(1) crosslinked composite: Cht+5% HNT-4AP, (2) composite: Cht+5% HNT-4AP, (3) crosslinked Cht+5%

4AP, (4) Cht+5% 4AP, and (5) HNT-4AP. All groups had a sample size of n=5. Approximately 20±2 mg of crosslinked and uncrosslinked samples were taken in regenerated cellulose dialysis tubing with a MWCO of 3,500 Da. Each bag was filled with 1 mL PBS, secured on both ends, immersed in 10 mL of PBS, and incubated in a 37 °C orbital shaker. For each formulation a sample size of n=5 was used. At various predetermined time intervals, 1 mL of the release media was collected and replaced with fresh media. The concentration of 4AP in the release media was determined by UV-Vis spectroscopy (Genesys 10S, Thermo

Fisher Scientific, Waltham, MA, USA) at a λmax of 260 nm using a standard curve previously created for

4AP in PBS [76, 176, 195].

2.2.15. Cell Dose-Response Toxicity and Proliferation

Human Schwann cells were used for all evaluations of cell dose response to 4AP treatments.

Schwann cells, purchased frozen at passage one, were expanded to and harvested at passage two, as per the manufacturer’s protocols. Cells were cultured on poly-L-lysine coated (2 µg/cm2) tissue culture polystyrene

(TCPS) well plates in manufacturer’s Schwann cell culture medium comprised of basal media, supplemented with 5% fetal bovine serum (FBS), 1% penicillin-streptomycin solution, and 1% Schwann

o cell growth supplement. Cell cultures were maintained in a humidified incubator at 37 C and 5% CO2.

For evaluation of potential toxicity effects, a dose-response study was performed with human

Schwann cells using a lactase dehydrogenase (LDH) assay (n=5). The LDH assay is a strongly sensitive method for quantifying cytotoxicity, or degree of inhibition of cell growth, resulting from interaction of a test material with cells. Human Schwann cells were treated with five, varying doses of 4AP and tested for percent LDH leakage after 24 hours in culture [196, 197]. Dose-response studies evaluating Schwann cell proliferation were performed using a standard colorimetric MTS assay. A colored formazan product is produced by dehydrogenase enzymes in metabolically active cells. An MTS assay of human schwann cells

38 was performed at day 1, 7, 14, and 21 with cells treated with 1, 5, and 10 μg/mL of 4AP, compared to an untreated control which received no drug.

2.2.16. Immunofluorescent Staining and Quantification

For immunofluorescent staining experiments, human Schwann cells, harvested at passage two, were seeded onto poly-L-lysine coated (2 µg/cm2) chamber slide wells at a concentration of 6,000 cells/cm2.

Cells were initially cultured using aforementioned Schwann cell culture medium, and media was replaced with 4AP-media solution (drug media) 24 hours post-seeding. Cells received drug media with 4AP in a concentration of 1, 5, or 10 µg/mL. These groups were compared to control groups where cells received standard Schwann cell culture media without drug. All cell media was changed every three days for the duration of the 14-day study (n=3).

At 14 days post-seeding, immunofluorescent staining was performed in order to examine the dose- response changes in Schwann cell proteins and neurotrophic factors, compared to control groups. Briefly, samples were washed using PBS (all washes performed in triplicate) to get rid of cell culture medium and were subsequently fixed for 30 minutes using 4% paraformaldehyde solution. Following fixation, samples were again washed in triplicate. Samples were permeabilized using 0.2% Triton X-100 prepared in PBS for

10 minutes then washed with PBS. Samples were blocked by incubating with 10% normal goat serum in

PBST (PBS with 0.1% Tween 20) for 30 minutes at room temperature followed by PBS wash. The primary antibodies, anti-NGF (1:200), anti-P0/MPZ (1:200), and anti-BDNF (1:1000) were all diluted to their respective ratios in the aforementioned blocking solution and added to each sample. Samples were incubated with primary antibody solution overnight at 4°C. The following day samples were washed three times with PBS followed by incubation for 1 hr with secondary antibody, Texas Red (goat anti-rabbit) diluted 1:2000 in blocking solution, and carefully protected from light. Samples were again washed with

PBS and DAPI stain was applied prior to imaging using a Zeiss LSM 510 Meta confocal laser scanning microscope (Carl Zeiss, Oberkochen, Germany) [77, 176].

The resulting images were processed, analyzed, and quantified using ImageJ. To quantify relative

NGF, P0, and BDNF fluorescence levels, a single in-focus plane was chosen for imaging. Using ImageJ, the area, integrated density, and mean gray value were measured for five independent images per sample, along with adjacent background readings. The total corrected cellular fluorescence (TCCF) = integrated density – (area of selected region × mean fluorescence of background readings), was calculated [198, 199].

This TCCF is presented as relative fluorescence of samples with arbitrary units, displayed as box-plots with

5-95% confidence intervals.

2.2.17. Statistical Analysis

All data is expressed as mean ± standard deviation (mean±s.d.). All results were first evaluated using t-tests or one-way/two-way analysis of variation (ANOVA) followed by Dunnett’s multiple comparisons test. All statistical analyses were performed with a confidence level of 95% (p<0.05) using

GraphPad Prism 7 (GraphPad Software, Inc. La Jolla, CA).

2.3. RESULTS

2.3.1. Conduit Scaffold Characterization

The aforementioned method of fabricating porous conduits resulted in highly porous, sponge-like scaffolds with specified geometries (Figure 7a). When imaged using SEM, a highly-aligned porous microstructure can be seen with microtubule-like pores aligned parallel to the longitudinal section (Figure

7b). Fenestrations can be seen interconnecting the longitudinally-aligned channels [200, 201]. As a result of the unidirectional freezing, and subsequent anisotropic growth of ice crystals, the cross-sectional images reveal elliptical forms, rather than symmetrical circles (Figure 7c). This is due to the fact that the cross- section of cellular ice crystals are not rotationally symmetric, but rather elliptical [200]. Given their elliptical geometry, measurements of pore Feret diameter, defined as the longest distance between any two points along the particle boundary, also known as the maximum size [202], yielded pores of 59.3±14.2 µm

(mean±s.d.). These findings are in agreement with previous studies which have investigated the effects of

40 topography on axon regeneration and shown that pores sized 20-60 µm along the longitudinal direction were optimal for maximum axon penetration and minimum axon misdirection [201, 203].

Figure 7: SEM images showing highly porous channel microstructure of (a) conduit cross sectioned, SB = 1 mm; (b) conduit sectioned longitudinally, SB = 100 µm; (c) conduit cross sectioned, SB = 100 µm. A highly linear pore configuration can be seen throughout the construct. (d) FITC-loaded HNTs were imaged within the chitosan matrix, showing an even distribution throughout the construct. (e) The directionality of HNTs was calculated, showing a highly aligned distribution of HNTs resulting from the alignment of the polymer matrix. (f) Human Schwann cells seeded on conduit showed alignment and proliferation in the direction of the aligned polymer matrix. (g) The directionality of seeded Schwann cells confirmed an aligned distribution of cells on the aligned microchannel conduit.

These aligned and interconnected pores allow for longitudinal and horizontal migration of Schwann cells [201], which is critical for successful regeneration of nerve tissue, as well as a large surface area to provide local and sustained delivery of pharmacological agent to the site of injury [204]. The distribution of HNTs within the polymer matrix showed even distribution of aligned HNTs (Figure 7d). The directionality of HNTs was calculated using ImageJ, revealing HNT particles aligned preferentially to

90.6±16.6° (mean±s.d.) with a goodness of fit R2 of 0.99 (Figure 7e). The direction of alignment is rather arbitrary based on image orientation, however a strong goodness of fit to a particular direction is evidence of HNTs preferentially oriented in a particular direction, in this case the direction of polymer alignment discussed previously. Similarly, when Schwann cells were cultured on conduits with aligned microchannel porosity, they show a clear favorable alignment correlating to the alignment of the polymer matrix (Figure

7f). The directionality calculations show that cells were aligned to 109.8±15.2° (mean±s.d.) with a goodness of fit R2 of 0.97 (Figure 7g).

2.3.2. Mechanical Properties

An investigation of the mechanical properties, namely Young’s modulus (modulus of tensile elasticity) and tensile strength, of conduits with and without halloysite reinforcement was conducting using uniaxial tensile testing. Significant differences (p<0.05) were found for both modulus and tensile strength with the addition of halloysite reinforcement. The Young’s modulus was calculated to be 0.13±0.02 and

0.23±0.07 MPa (mean±s.d.) for random (unaligned) conduits with 0% HNT (no halloysite) and 5% HNT, respectively. In contrast, the Young’s modulus for aligned conduits were 0.18±0.02 and 0.33±0.09 MPa

(mean±s.d.) for aligned conduits with 0% HNT and 5% HNT, respectively (Figure 8).

Figure 8: Tensile mechanical testing of fabricated nerve guidance conduits, both with aligned pores and random porosity, and with and without halloysite reinforcement (labelled 5% HNT and 0% HNT, respectively). Young’s modulus is shown to increase in composite samples with 5% HNT as compared to samples with 0% HNT. Alignment of pores was shown to have anisotropic mechanical properties, increasing the modulus, particularly for the aligned composite with 5% HNT which showed the greatest moduli. Where native peripheral nerve is generally considered to have a Young’s modulus of 0.50 MPa (red line), aligned composite conduits showed very similar moduli. *=p<0.05, **=p<0.01, and ***=p<0.001.

2.3.3. Water Absorption (Swelling)

A systematic study was done to assess the optimal crosslinking agent and duration for the fabricated conduits. Crosslinking was critical for achieving structural integrity and swelling rates. Uncrosslinked samples immersed in PBS showed >10,000% swelling within the initial 10 minutes, where polymer samples became gel and could not be measured post-10 minute mark (Figure 9a). Thermal crosslinking resulted in decrease of water absorption across all crosslinking durations, however, substantial decreases were not achieved as water absorption was still 3000-7000% (Figure 9b). Chemical crosslinking using Sulfuric Acid

(H2SO4) achieved substantial decreases in water absorption rates, with 60 minute crosslinking duration achieving ~400% absorption. However, Sulfuric Acid treatment resulted in a breakdown of the structure’s physical integrity, with substantial deterioration of the polymer samples (Figure 9c). Epichlorohydrin chemical crosslinking was able to achieve significant decreases in water absorption, particularly, when treatment duration was 30, 60, or 120 minutes. Water absorption at the 180-mark was measured to be ~400-

500%, with no structural damage (Figure 9d). As a result, ECH was chosen as the crosslinking method moving forward for all conduit fabrication and preparation.

Figure 9: Water absorption, shown as percent of dry weight, with various crosslinking methods and durations.

2.3.4. Crystallinity (X-Ray Diffraction)

X-Ray Diffraction (XRD) of 4-Aminopyridine showed several sharp peaks, namely at 14.40°,

18.80°, 24.27°, 24.91°, and 32.57°. The sharp peaks of 4AP observed represent the drug to be crystalline in nature. Pristine chitosan does not exhibit any sharp peaks, indicating its amorphous nature (Figure 10a).

Composite chitosan with 4AP showed a hollow diffused pattern, where no sharp peaks of 4-AP are observed, indicating that the crystalline nature of drug has been converted to amorphous during complexation. XRD of the HNT-4AP showed peaks due to the crystal nature of 4AP and some diffused peaks owing to the HNT (Figure 10b). This indicates that there is no complex formation between HNT and

4AP, merely a filling of the hollow HNT lumen with condensed drug. These findings confirm that the entrapped 4AP is dispersed mono-molecularly in the composite chitosan matrix.

Figure 10: X-Ray diffraction spectra of 4-Aminopyridine, HNT-4AP, pristine Chitosan, and the composite structure indicating various degrees of crystallinity.

2.3.5. Thermal Analysis

The DSC spectra of 4-aminopyridine is shown with sharp endothermic peak at 165.00 °C which represents the crystalline nature of the drug (Figure 11a). DSC spectra of HNT-4AP did not show any endothermic peak. The thermograms of composite chitosan showed a shift in the endothermic peaks from the original endothermic peaks of pristine chitosan and 4-AP, that is, from 103.07 °C to 82.88 °C and 165.00

°C to 136.59 °C, respectively, which confirms the formation of composite chitosan drug complex (Figure

11b). The shifting of peaks from the higher temperature to lower indicates that the crystalline form of the drug is converted into a more amorphous form during complex formation.

Figure 11: Differential scanning calorimetry spectra of 4-Aminopyridine, HNT-4AP, pristine Chitosan, and the composite structure indicating various endothermic transitions.

2.3.6. Porosity

As previously detailed, composite conduits yieled pores ranging in diameter from 20-60 µm with an average size of 59.3±14.2 µm (mean±s.d.). The porosity of conduit structures was evaluated as per the aforementioned Equation (1). The porosity values were calculated as follows: Pristine chitosan

96.03±0.326%, composite 96.03±0.122%, and crosslinked composite 88.66±0.494% (Figure 12).

Figure 12: Porosity measurements for conduits of various formulation. The addition of HNT resulted in negligible porosity change from pristine chitosan. However the effect of crosslinking is seen on the polymer matrix density, as porosity decreases.

2.3.7. Drug Loading Efficiency

Analysis of TGA profiles (Figure 13) indicated that 4AP degrades at 130-160 °C and the peak at

500 °C indicates the presence of halloysite. At this temperature, halloysite does not fully degrade, but undergoes a loss of its hydroxyl group [187]. TGA profile of the halloysite drug composites, HNT-4AP

(Figure 13b) shows the decomposition peaks for both 4AP and halloysite. The amount of 4AP in the composite was quantified as mass lost over the given temperature range and is expressed in percent of the total mass of the composite. TGA profiles constructed using the parameters identified from derivative curves indicated a 7.69 wt% overall drug loading [187]. It has been commonly established in literature that successful drug loading of unmodified HNTs reached 5-10 wt% of the tube weight. Taking the density of the organic drug as 1.26 and halloysite 2.53 g/cm3, one could see that this value almost coincides with the complete ﬁlling of the lumen with condensed drug [186, 205].

Figure 13: TGA spectra (a) before and (b) after loading of 4AP into HNT. The modification of HNT loaded with 4AP (b) is shown as the TGA curve shows decomposition peaks for both 4AP and halloysite. The amount of 4AP loaded into HNT was quantified as mass lost over the given temperature range corresponding to the decomposition of 4AP and is expressed as the percent of total mass of HNT-4AP. (c) Pristine chitosan and composite Cht/HNT-4AP.

2.3.8. In Vitro Degradation

The degradation of composite conduits was studied under physiologically-simulated conditions in

PBS and PBS supplemented with lysozyme. The degradation rate of chitosan by lysozyme is inversely related to the molecular weight and degree of crystallinity of the chitosan, and proportionally related to its deacetylation [206]. Lysozyme recognizes and digests N-acetyl glucosamine sequences in the chitosan molecules, thus digestibility increases with increasing degree of N-acetylation in the polymer chain [207].

Following 12 weeks of incubation in PBS and PBS+lysozyme, 83.05±1.01% and 70.99±1.27% weight of

46 samples remained (mean±s.d.), respectively (Figure 14a). Inversely, this resulted in 16.95±1.01% and

29.01±1.27% weight loss, respectively. Weight loss in PBS+lysozyme at weeks 2, 4, 6, 8, and 12 were all statistically significant compared to that of in PBS. To supplement degradation data, samples subjected to

PBS+lysozyme were collected at various timepoints and imaged using SEM (Figure 14b). Microscopy images show a consistent breakdown of the polymer microstructure, with fragmenting, cracking, and deterioration gradually increasing in relation the timepoint.

Figure 14: (a) Degradation of crosslinked composite Cht/HNT-4AP samples in PBS pH 7.4 (black) and PBS with 4mg/mL lysozyme (red) (n=3 samples/timepoint, mean±s.d.). Degradation represented by the weight (%) of polymer remaining. *=p<0.05 vs PBS, **=p<0.01 vs PBS, and ***=p<0.001 vs PBS. (b) Representative SEM images showing crosslinked composite samples at various timepoints during degradation in PBS+lysozyme.

2.3.9. In Vitro Drug Release

Drug release from a variety of samples was tested in PBS (pH 7.4) at 37 °C. Loaded HNTs showed a sustained release of 4AP within 6 h (Figure 15b). When drug-loaded HNTs were incorporated into the chitosan polymer matrix (composite), sustained drug release was shown to be extended, with 94.68±3.62% of drug being released within 168 h (7 days) (Figure 15c). This release was shown to be in stark contrast to that of the release of 4AP from chitosan polymer conduits without the incorporation of halloysite. In this case, where 4AP is simply mixed into the chitosan solution prior to fabrication, a 97.84±3.05% release was shown within 0.5 h (mean±s.d.). When this formulation, excluding halloysite, was crosslinked, a

58.19±2.79% release was shown within 7 days. Lastly, formulations of crosslinked composite conduits, where 4AP-loaded HNT was incorporated into chitosan conduits and the resultant conduit was crosslinked, showed the longest sustained release of 4AP with 30.11±2.02% release within 7 days. No significant differences in drug release properties were noted when drug-loaded chitosan conduits were exposed to UV during the process of sterilization. The structure of 4-AP released from conduits remained intact as evidenced through UV-spectroscopy and bioactivity.

Figure 15: (a) Schematic of 4AP drug loading into HNT, followed by sustained release of drug from lumen. (b) Cumulative release of 4AP from halloysite nanotubes; (c) cumulative release of 4AP from Cht (black diamond), crosslinked Cht (blue square), composite Cht/HNT-4AP (red circle), and crosslinked composite Cht/HNT-4AP (orange triangle) (n=5 samples/group, mean±s.d.). (d) Burst release of 4AP from samples, where burst release is quantified as percent cumulative drug release within the first 1 h (n=5 samples/group, mean±s.d.). **=p<0.01, ****=p<0.0001.

Figure 15d illustrates the effect of the formulations on the early release of 4AP, focusing on the burst release, quantified as the percent cumulative drug release within the first 1 h. Drug-loaded halloysite

(HNT-4AP) samples showed 88.20±6.62%, composite conduits showed 47.27±2.71%, crosslinked

Cht/4AP conduits (no halloysite) showed 28.67±0.91%, and lastly crosslinked composite conduits showed the lowest burst release with 20.03±1.57% drug release within the first 1 h (mean±s.d.). Uncrosslinked

48 samples of Cht/4AP which did not include halloysite were excluded from the burst release quantification as nearly 100% of drug was released in less than 1 h. Thus, according to the data representation, this would quantify as 100% burst release.

The drug release results from tested formulations were fit to various release kinetic models including Zero Order, First Order, Higuchi, and Korsmeyer-Peppas (K-P) (Table IV). In the K-P equation, the n value is the release exponent that indicates the mechanism of drug release. All three samples showed an n<0.5, indicating quasi-Fickian diffusion of drug from the matrix [208], suggesting partial diffusion as the drug release kinetic. All four models show an interesting trend for the diffusion rate constant, k, where the Composite formulation showed the largest k value, crosslinked (CL) Cht/4AP showed the second largest, and CL Composite showed the lowest. Thus, calculations from all four models agree that 4AP release from a CL Composite Cht/4AP-HNT matrix was slightly slower than release from a CL Cht/4AP matrix, which was in turn, slightly slower than release from a Composite Cht/4AP-HNT. These results are also in agreement with the cumulative drug release plot discussed earlier in Figure 15.

Table IV: Cumulative 4AP release data from respective samples fit to the Zero-order, First-order, Higuchi, and Korsmeyer- Peppas drug release models.

2.3.10. Human Schwann Cell Studies

In an effort to elucidate the effects of 4AP on physiologically-relevant cells, human Schwann cells were tested for dose-dependent responses to 4AP with regards to toxicity, proliferation, as well as expression of neurotrophic factors and proteins. The LDH toxicity assay, where the loss of intracellular

LDH and its release (leakage) into the culture medium is an indicator of irreversible cell death due to cell membrane damage [209], showed no statistical differences in LDH leakage between any of the 4AP treatment groups and the control group, which received no drug, indicating no cytotoxicity caused by the presence of varying doses of 4AP (Figure 16a). When analyzed for cell proliferation (Figure 16b), MTS assay of 4AP-treated Schwann cells cultured on TCPS for 1, 7, 14, and 21 days revealed conclusive dose- dependent inhibition, which is in agreement with previous literature [210].

Figure 16: (a) Lactase dehydrogenase (LDH) leakage cytotoxicity assay testing various doses of 4AP treatment on human Schwann cells in culture. ns=no significance. (b) MTS assay of human Schwann cells cultured on TCPS with no drug (control) or 1, 5, and 10 µg/mL of 4AP for up to 21 days. *=p<0.05 vs control, ***=p<0.001 vs control, and ****=p<0.0001 vs control. (c) MTS assay of human Schwann cells cultured on composite conduits for 21 days. **=p<0.01 vs day 1. (d) Human Schwann cells cultured on composite conduits and stained using live/dead cell staining. A Z-stack projection of 10 confocal microscopy images, each 10µm deep, is shown (i) and the same image is shown pseudo-colored for cell z-position using a color-depth gradient (ii).

MTS assay in conjunction with live/dead cell staining techniques were employed to study the viability, proliferation, and infiltration of Schwann cells cultured on 3D composite conduits. Over the

50 course of 21 days, MTS assay of human Schwann cells seeded onto 3D composite conduits showed increases in cell number with significant proliferation by day 14 and day 21 (Figure 16c). Cell growth and proliferation over the 21-day period confirms the biocompatibility of the chitosan-based composite conduit system, as well as the composite polymer’s ability to facilitate cell adhesion and proliferation. Given the

3D nature of the composite conduit, beyond cell adhesion and proliferation, cell infiltration into and within the 3D architecture was examined. At day 14, live/dead staining was employed to samples seeded with human Schwann cells. Z-stack images were taken at 10 µm slices and the composite images were projected and thresholded using a color-depth gradient (Figure 16d). Cell infiltration can be seen throughout the entire 100 µm projection with the majority of cells infiltrating 50-80 µm deep.

Schwann cell neurotrophic factor and protein expression was also investigated as a measure of dose-response to 4AP (Figure 17). Human Schwann cells cultured for 14 days with varying treatments of

4AP were subjected to immunofluorescent staining to assess the expression of three markers of interest: nerve growth factor (NGF), myelin protein zero (P0), and brain-derived neurotrophic factor (BDNF).

Schwann cells were shown to exhibit dose-dependent increases in the expression of all three markers, compared to control groups, which received no drug. The dose-dependent increases in NGF, P0, and BDNF expression were shown qualitatively via immunofluorescent staining, and calculated as a quantification of immunofluorescent staining, expressed as relative fluorescence (Figure 17b). NGF and P0 expression showed statistically-significant increases in both 5 and 10 µg/mL 4AP-treatment groups, while BDNF expression increases were only significant in the 10 µg/mL treatment group, when compared to no-drug control groups. When treated with 10 µg/mL 4AP, NGF, P0, and BDNF showed 128.58±0.57%,

310.75±11.92%, and 170.94±4.43% increases over no-drug control groups, respectively.

Figure 17: (a) In vitro immunofluorescent staining images showing human Schwann cells stained for NGF, P0, and BDNF (all stained red, respectively by row) in a dose response study to 4AP drug treatment. Control groups, which received media without drug, were compared to 1, 5, and 10 µg/mL drug media solutions, respectively by column. SB=10µm. (b) Quantification of immunofluorescent staining of NGF, P0, and BDNF (left to right, respectively). The intensity of staining obtained with each antibody was measured and displayed as box-plots with 5 to 95% confidence intervals (n=5 images/group). *=p<0.05 vs control, ***=p<0.001 vs control, and ****=p<0.0001 vs control.

2.4. DISCUSSION

The aforementioned fabrication method for polymeric nerve guidance conduit represents a straightforward method for achieving a tubular structure with aligned microchannel porosity for successful nerve regeneration. The use of the chitosan polymer not only grants natural biodegradability and

52 biocompatibility, but also has a broad antimicrobial spectrum to which gram-negative, gram-positive bacteria, and fungi are highly susceptible [211]. In addition to its alignment, the composite system’s incorporation of aluminosilicate halloysite nanotubes, which serve as an excellent drug-carrier given their hollow lumen, represents a highly effective method for the incorporation of small-molecule drug, in addition to mechanical benefits. The calculated HNT drug loading of 7.686% was well within the published range of 5-10 wt% of the tube weight [186, 205], confirming the complete filling of the tubes with condensed drug. Furthermore, silica-based scaffolds have been found to be effective for regeneration of soft tissues. Though the exact mechanism is unknown, it is believed that the silicate ions stimulate specific pathways responsible for the regeneration process [212].

A successful nerve guidance conduit should have a Young’s modulus similar to that of native nerve tissues to provide sufficient strength and flexibility to be implanted and sutured during implantation, and to resist in vivo physiological loading and mechanical mismatch during nerve regeneration [213]. Although variations exist, it is generally accepted that Young’s modulus of peripheral nerve, in the longitudinal direction, is in the range of 0.50 MPa [214]. In addition to its benefits as an excellent drug-release vehicle, the developed composite conduit’s inclusion of halloysite nanotubes is an effective method to achieve tunable mechanical properties. Conduits with 5% HNT were shown to have a mean Young’s modulus of

0.33±0.09 MPa, very similar to that of native nerve. The increase in Young’s modulus, with the increasing contents of HNT, are in agreement with other reports in literature [215, 216]. It is also important to note that no significant differences in tensile properties were noted when chitosan conduits were exposed to UV during the process of sterilization.

The composite, when processed via freeze-drying, resulted in a highly-porous, foam-like structure well-suited for many tissue engineering and regenerative medicine applications. Prior to freeze-drying, utilization of the unidirectional freezing technique resulted in consistent, longitudinally-aligned pores clearly visible in SEM images. The alignment of the chitosan polymer, and subsequently the halloysite nanotube fillers, resulted in mechanically anisotropic properties, preferential to the aligned longitudinal

53 orientation. This is similar to the anisotropic structure of natural nerves, but extends also to muscles, tendons, blood vessels, and ligaments, which all have anisotropic mechanical properties [217, 218].

Furthermore, anisotropic pore structures have been correlated to excellent cell penetration, where highly aligned and porous scaffolds present significant contact guidance cues which promote cell motility within the network [219] – all of which are significant advantages with regards to peripheral nerve regeneration.

This was all evident in the seeding of human Schwann cells on the aligned microchannel conduits, where directionality testing confirmed Schwann cell alignment correlating to the polymer alignment.

Longitudinally-aligned pores within the conduit may serve as axonal tracts and provide directionality to the regenerating axon, in addition to providing increased surface area as compared to traditional tubular constructs. Given the hydrogel nature of the processed material, chemical crosslinking was necessary to stabilize the structure’s integrity in aqueous environments. As an alternative to the most commonly used chitosan crosslinking agent, glutaraldehyde [220], which has been shown to be an irritant with known toxicity [221], epichlorohydrin was successfully used to crosslink the chitosan-based structure and stabilize its water absorption in aqueous conditions.

According to the pathophysiology of peripheral nerve injury and repair, the first 7-21 days offer the greatest potential for facilitated regeneration as this is the period of Wallerian degeneration, cell recruitment, and the beginning of axonal regeneration. Release of 4AP using a hydrophilic polymeric nerve grafts is challenging. Many such scaffold formulations in our preliminary work released significant amount of 4AP (>75%) in the first 24 hours (data not shown). The primary purpose of this study was to solve this release problem along with scaffold suturability and its retention under animal leg movement at the defect site. 4-Aminopyridine release from fabricated conduits was shown to be dependent on the incorporation of halloysite nanotubes as well as the structure’s crosslinking. Release of water-soluble drugs from natural polymeric matrices is often quick, while the release of small-molecule drugs is considered rapid; This was observed in samples excluding HNTs and crosslinking. 4AP-loaded HNTs showed a sustained release of drug within 6 h, in agreement with drug-release values from HNTs previously reported in literature [186,

205]. Crosslinking was shown to have a profound effect on sustained drug release, with the crosslinked composite structure containing chitosan and 4AP-loaded HNTs showed the slowest release. As such, the tunability of fabricated conduit system’s drug-loading and drug release characteristics are highlighted, with optimized conduits releasing 4AP from well over seven days.

4AP treatment of human Schwann cells resulted in significant, dose-dependent upregulation of critical trophic factors, namely nerve growth factor (NGF), myelin protein zero (P0), and brain derived neurotrophic factor (BDNF), as shown by the immunofluorescent staining images and quantification of relative fluorescence. NGF is a neurotrophic factor expressed by migrating Schwann cells in developing or regenerating peripheral nerves and has been shown to stimulate the regeneration of axons along with BDNF

[222, 223]. BDNF is a neurotrophin which has been shown to enhance peripheral nerve system myelination

[224] and be required for peripheral nerve regeneration and remyelination after injury; studies show

Schwann cells may contribute as the primary source of BDNF during regeneration [225]. Lastly, P0 is a major myelin protein critical in the development of myelin, which sheathes axons and acts as an electrical insulator, greatly speeding up action potential conduction. As such, all three factors are critically important to healthy regenerating nerves and their upregulation may be the key to more efficient and effective nerve regeneration. The MTS cell proliferation assay determined significant decreases in human Schwann cell proliferation correlated to dose-dependent 4AP treatment. This 4AP-triggered decrease in proliferation corresponds to an increase in myelin differentiation and upregulation of neurotrophins and proteins, complimenting the aforementioned immunostaining data. These results are strongly in agreement with previous studies, which report that proliferation and myelin differentiation appear to represent two incompatible developmental possibilities, both in vivo and in vitro [226, 227].

2.5. CONCLUSION

The developed polymer-nanotube composite drug delivery conduit demonstrates high potential for facilitating peripheral nerve regeneration. As an alternative to the conduction effects of electrical stimulation, we showed the design and optimization of a nerve guidance conduit with aligned

55 microchannels for the sustained release of a small-molecule drug that promotes nerve impulse conduction.

A critical challenge is the sustained release of small drug molecules using hydrophilic polymers, due to their faster diffusion rates. We successfully demonstrated the sustained release of 4-Aminopyridine as a growth factor alternative to enhance the rate of nerve regeneration. The composite conduit released one third of the encapsulated 4AP in the first 7 days and the release was shown to be quasi-Fickian diffusion.

In vitro studies confirmed the beneficial effects of 4AP, increasing expression of key proteins such as nerve growth factor, myelin protein zero and brain derived neurotrophic factor in human Schwann cells.

Future in vivo studies will investigate the effects of the designed nerve guidance conduit in combination with 4AP in a critical-sized 15-mm sciatic nerve defect in a rat model over a period of 8 weeks.

Nerve repair and regeneration will be benchmarked against appropriate controls including the nerve autograft (reversed nerve) and the neat conduit (no drug) in terms of quality of tissue healing (histology), myelin formation (microscopy), and functionality assessment via walking track analysis. We hypothesize that drug-eluting conduits with aligned microchannels will enhance the rate of nerve regeneration, superior to control conduit groups, and comparable to the autograft repair. These nerve guidance conduits and 4AP sustained delivery may serve as an attractive strategy for nerve repair and regeneration.

CHAPTER 3: BIOPOLYMER-NANOTUBE NERVE GUIDANCE CONDUIT DRUG DELIVERY

FOR PERIPHERAL NERVE REGENERATION: IN VIVO ASSESSMENT

3.1. INTRODUCTION

In our previous work we described the fabrication and characterization of a chitosan-based nerve guidance conduit scaffold doped with halloysite nanotubes (HNT), a naturally derived dioctahedral aluminosilicate clay mineral [69]. These NGCs were fabricated using a unique process for unidirectional freezing, resulting in the formation of longitudinally aligned microchannel porosity throughout the tubular construct. The advantage of such aligned porosity is simply to influence physical cues and directionality for guided axon growth and more efficacious regeneration of nerves. Further, it was used as a drug delivery vehicle for the sustained delivery of the FDA-approved small-molecule drug 4-Aminopyridine (4AP), a voltage-gated potassium channel blocker, which prolongs action potentials and amplifies neurotransmitter release within neurons and Schwann cells for greater nerve regeneration. To our knowledge, this was the first use of this drug for in a sustained-release formulation for applications in segmental nerve defects, in particular for critical-sized defects [69]. Its successes were shown in vitro as well as in a preliminary in vivo study, where the aligned microchannel NGC was implanted into a critical-sized (15 mm) rat sciatic nerve defect and showed optimistic signs of nerve healing.

In an effort to further advance the field of NGC-based nerve regeneration, in this study we describe the full in vivo investigation of the previously developed drug delivery NGC. A critical-sized, 15 mm, sciatic nerve defect was created in Wistar rats, where the defect was bridged with the 4AP drug-loaded conduits and compared to control groups, including the gold-standard autograft, as well as a non-drug- loaded conduit. The effects of repair and regeneration were evaluated using a large battery of tests and assessments including electron microscopy, various histology techniques, and walking track analysis for functional motor recovery. The results of functional walking track analysis, morphometric evaluations of myelin development, and histological assessments of various markers confirmed the equivalency of

57 implanted drug-conduits repair to autograft control treatments, with potentially transformative implications of the fields of regenerative medicine and peripheral nerve repair, regeneration, and engineering.

3.2. MATERIALS AND METHODS

3.2.1. Materials

Chitosan (high molecular weight), halloysite nanotubes, and epichlorohydrin were purchased from

Sigma-Aldrich (St. Louis, MO). Glacial acetic acid, phosphate buffered saline (PBS) were purchased from

Fisher Scientific (Fair Lawn, NJ). 4-Aminopyridine (>99%) was purchased from Alomone Labs

(Jerusalem, Israel). S100B antibody was purchased from Invitrogen (Grand Island, NY). Neurofilament-H was purchased from Cell Signaling Technology (Beverly, MA). All surgical tools were purchased from

Fine Science Tools (Foster City, CA). Neutral buffered formalin, sodium hydroxide pellets, and Vicryl® suture were purchased from Fisher Scientific Company (Suwannee, GA). Nylon monofilament suture was purchased from McKesson Medical Surgical (Cheshire, CT). Ultrapure water was used from Millipore

MilliQ system (EMD Millipore, Burlington, MA).

3.2.2. Drug Loading and Conduit Fabrication

Drug loading of 4AP into halloysite nanotubes was achieved following the previously detailed methodology. Similarly, the fabrication and unidirectional freezing of conduits follows the aforementioned methodology.

3.2.3. Scanning Electron Microscopy (SEM) Imaging

Images of the conduit surface architecture and microstructure were captured using a JEOL JSM-

6335F scanning electron microscopy (SEM) (JEOL USA, Inc., MA, USA). Samples cut and attached to

SEM stubs were coated with Au/Pd using a Polaron E5100 sputtering system (Quorum Technologies, East

Sussex, UK) for 3 minutes prior to imaging, achieving a conductive layer of approximately 18 nm.

3.2.4. Animal Subjects

Forty (40) female Wistar rats (Charles River Laboratories, Wilmington, MA) weighing approximately 200 g were randomly divided into four groups: sham (control), autograft repair, conduit repair, and drug-conduit repair. All animals were housed in pairs with free access to food and drinking water in a temperature-controlled room with a 12-hour light-dark cycle. All animals were allowed to acclimate to housing conditions for at least one week prior to surgery. All animals were maintained and cared for according to methods approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Connecticut.

3.2.5. Surgical Procedure

Animals were anesthetized by inhalation of 2.5% isoflurane carried in medical-grade oxygen to induce and maintain a surgical anesthetic plane. The right hind leg was shaved and cleaned using betadine and isopropyl alcohol. Body temperature was maintained at approximately 37 °C by placing the animal on a surgical warming pad. All surgical procedures were conducted under aseptic conditions using sterile tools.

A 30 mm incision was made parallel to the femoral axis and the right sciatic nerve was exposed through a gluteal muscle splitting incision (Figure 18). The sciatic nerve was carefully dissected free of surrounding tissues and a 15 mm nerve segment was sharply transected and removed. For conduit and drug-conduit groups, a 20 mm conduit was secured to the proximal and distal stumps using 8-0 Nylon monofilament suture (Ethicon Inc., Somerville, NJ), with approximately 2.5 mm of nerve stump inside the conduit. For autograft groups, the transected nerve segment was reversed and repaired using 8-0 Nylon monofilament sutures. The muscle and skin incisions were closed with 5-0 Vicryl® suture (Ethicon Inc., Somerville, NJ).

In sham-operated animals the sciatic nerve was surgically exposed, as mentioned above, without any injury to the nerve. Pain was managed using a multi-modal analgesic approach with injections of Buprenorphine every 12-18 h for two days post-operative and injections of Meloxicam every 24 h for three days post- operative. A saturated solution (1.3%) of aqueous picric acid (2,4,6-trinitrophenol, Sigma-Aldrich Co., St.

Louis, MO) was applied to the denervated foot immediately after surgery, and twice a week for a month,

59 to deter scratching and biting [228, 229]. At eight weeks post-operative, animals were euthanized by CO2 asphyxiation.

Figure 18: Wistar rat critical-size sciatic nerve transection surgery with repair using a composite conduit. (a) A 30 mm incision was made parallel to the femoral axis and the right sciatic nerve was exposed through a gluteal muscle splitting incision. (b) A 15 mm (critical-size) segment of nerve was measured, sharply transected, and removed creating a segmental defect. (c) Top: The transected 15 mm sciatic nerve segment. Bottom: The conduit is shown pre-implantation, measuring 20 mm long with a 5 mm diameter. (d) A 20 mm sterile conduit is sutured to proximal and distal ends of the nerve. (e, f) The gluteal muscle and skin incisions, respectively, are sutured closed and the incision site is cleaned.

3.2.6. Walking Track Analysis and Sciatic Functional Index

Walking track analysis was performed with the DigiGait (Mouse Specifics, Inc., Framingham, MA) imaging and analysis system to assess functional recovery following sciatic nerve transection and repair.

The DigiGait system uses unique ventral plane video imaging to generate computerized paw prints while the animal walks on a speed-controlled motorized treadmill belt (Figure 19a). The dynamic gait signal is corrected for optimizing signal-to-noise ratio and digitized to analyze and define a wide variety of gait

60 parameters, including sciatic functional index (SFI), paw area, and stance and stride information (Figure

19b).

Figure 19: The DigiGait automated treadmill system performs walking track analysis to quantify differences associated with injury severity. (a) A surgically-operated Wistar rat walking on the enclosed treadmill within the DigiGait system. (b) A ventral high-speed camera records walking and software digitizes the footprints in preparation for walking track analysis. (c) Differences in several animal paw measurements can differentiate between normal (left; green) and injured (right; red) paws. Measurements such as inter-toe (IT, 2-4, distance), toe spread (TS, 1-5, spread), and paw length (PL) are used to calculate values such as the sciatic functional index (SFI).

In order to assess the functional recovery of sciatic nerve damage and repair, the SFI has been well established in literature for a variety of studies aimed at peripheral nerve regeneration [80, 110, 137, 138].

The SFI is determined by comparing the geometric representation of the affected hind paw from an injured rat, and comparing it to the contralateral paw. To calculate the SFI, Equation 5 is used, where EPL is the experimental print length, NPL is the normal print length, ETS is the experimental toe spread, NTS is the normal toe spread, EIT is the experimental intermediary toe spread, and NIT is the normal intermediary toe spread (Figure 19c) [136]:

퐸푃퐿−푁푃퐿 퐸푇푆−푁푇푆 퐸퐼푇−푁퐼푇 푆퐹퐼 = −38.3 × ( ) + 109.5 × ( ) + 13.3 × ( ) − 8.8 (5) 푁푃퐿 푁푇푆 푁퐼푇

Animals were individually placed on the DigiGait treadmill several times to acclimate them with the apparatus prior to surgery. Baseline DigiGait data was collected for all animals prior to surgery.

Following surgery, animals were tested on the DigiGait treadmill every two weeks for eight weeks using an optimal speed of ~20 cm/s. Data was saved when animals walked consistently for at least 5 seconds, resulting in ~15 complete strides. Animals were then returned to their paired housing cages.

3.2.7. Transmission Electron Microscopy (TEM)

Axonal regeneration was investigated by transmission electron microscopy (TEM, HITACHI H -

7650) in the middle regions of the regenerated sciatic nerve. Samples were fixed in 2.5% glutaraldehyde in

0.1 M phosphate buffer (pH 7.4) for 4 hours at room temperature, then overnight at 4 °C. Samples were post-fixed with 1% osmium tetroxide, dehydrated, and embedded using Poly/Bed 812 embedding kit

(Polysciences Inc., Warrington, PA). Ultrathin sections of 70 nm were cut, stained with uranyl acetate and lead citrate, and then examined under a transmission electron microscope. The axon diameter, myelinated nerve fiber diameter, and thickness of myelin sheath was evaluated using ImageJ software. For each specimen, three images were taken. The G-ratio was calculated according to Equation 6, where Do is the outer diameter of the corresponding fiber and Di is the inner axon diameter:

퐷 퐺푟푎푡푖표 = 푖⁄ (6) 퐷표

3.2.8. Histological Evaluation

Post-euthanasia, the right sciatic nerve was harvested and washed with PBS before being transferred to a histological container filled with 10% neutral buffered formalin and fixed at 4 °C overnight.

Samples were then washed with PBS and transferred to 70% ethanol prior to embedding. Samples (n=3) were embedded in paraffin and the middle region of the regenerated nerve (center of the conduit or autograft) was cut to approximately 5 µm thick slices in preparation for histological assessments. Following standard protocols, hematoxylin and eosin (H&E) and Luxol Fast Blue (LFB) staining was performed, for myelin identification, on sectioned slices of sham-operated and repaired nerves. Similarly,

62 immunohistochemistry staining was performed for S100 and Neurofilament-H (NF-H, NF200) markers.

Histology slides were imaged using an Aperio CS2 high-resolution digital slide scanner (Leica Biosystems

Inc., Buffalo Grove, IL) and all quantification was performed using ImageJ software using a minimum of

5 images per group. Liver tissues were also harvested, fixed, and stained for H&E in order to assess any potential accumulation and toxicity of HNTs in the rat liver.

3.2.9. Data Presentation and Statistical Analysis

All data were analyzed using Microsoft Excel and GraphPad Prism 7 (GraphPad Software, Inc. La

Jolla, CA). Data were analyzed via one-way analysis of variation (ANOVA) with Tukey’s posthoc corrections for multiple comparisons and reported as mean ± standard deviation (SD) with the exception of gait data. Gait data, excluding SFI, are expressed as a ratio between the ipsilateral (right) hind paw and contralateral (left) hind paw times 100% and reported as mean ± standard-error-of the-mean (SEM). Gait data were analyzed via two-way repeated measures ANOVA (time x group) with Tukey’s posthoc corrections for multiple comparisons tests as appropriate. All statistical analyses were performed with a confidence level of 95% (α=0.05) using GraphPad Prism 7.

3.3. RESULTS

3.3.1. Surface Characterization of Nerve Guidance Conduits

The surface morphology of conduits with aligned microchannels was imaged and observed using

SEM. Conduits showed ultra-porous microstructures with elliptical pores with an average size of 59.3±14.2

μm in diameter (mean±SD) (Figure 20a). The elliptical geometry is an expected result of the anisotropic growth of ice crystals during the unidirectional freezing process, and is explained in our previous publication. SEM images confirm highly-aligned porosity in the form of longitudinal microchannels, with fenestrations interconnecting the channels (Figure 20b). Topographically, the microstructure and architectural environment of the conduits are within range for what has been shown to be optimal for maximum axon penetration and minimum axon misdirection [201, 203].

Figure 20: Scanning electron microscopy (SEM) images of fabricated nerve guidance conduits (NGCs) in (a) cross section (Size bar=1 mm) and (b) longitudinal (Size bar=100 μm). The highly porous microstructure is shown with consistent aligned microchannel architecture. Conduit section planes are shown in graphic overlays for each image.

3.3.2. Animal Surgeries

Drug-loaded conduits were tested against non-drug (neat) conduits and autograft repairs, in a 15 mm critical-sized Wistar rat sciatic nerve model. Autograft repair animals as well as sham-operated animals served as controls. Post-surgery, there were no obvious signs of systemic or regional inflammation and no surgical complications were reported with all animals surviving surgery.

One noteworthy complication was autotomy or automutilation by the rats. This is a well-described phenomenon of the rat sciatic nerve model where severing and resection of the sciatic nerve results in numbness of the foot and the rats often chew off the toes of their denervated foot, thus preventing meaningful walking track data [229-231]. This autotomy behavior is postulated to be associated with parathesias, similar to phantom limb pain [231], and has been historically shown to affect up to 50 and

100% of rats undergoing sciatic nerve transection, dependent on strain [232]. Unfortunately, despite the well-established literary nature of autotomy in rat sciatic nerve models, this complication is often not declared in nerve regeneration studies. For the purposes of our study all animals were intensively monitored and examined for signs of autotomy. Rats with severe wounds were excluded from the study.

3.3.3. Functional Recovery Evaluation

3.3.3.1. Sciatic Functional Index

The recovery of hindlimb motor function was assessed using walking track analysis evaluating a

number of gait parameters, in particular, sciatic functional index (SFI) biweekly over the 8-week test period.

Quantitative assessment of SFI ranges from 0 to -100, indicating normal function to complete dysfunction,

respectively. As a baseline, a pre-operative SFI value of -10.15±1.18 (mean±SEM) was measured for all

rats. There was a sharp decrease in SFI at 2 weeks post-operative for all surgical groups with values as

follows: autograft (-57.37±2.13), conduit (-58.41±0.98), and drug-conduit (-52.62±3.38) (mean±SEM)

(Figure 21). Fluctuating changes were quantified for all groups (excluding sham) at 4 weeks and 6 weeks

post-operative, with 8-week SFI values as follows: autograft (-44.89±4.73), conduit (-51.79±4.40), and

drug-conduit (-45.12±4.45) (mean±SEM). There were no statistically significant (p<0.05) differences

between any of the test groups (autograft, conduit, and drug conduit), however all test groups were

significantly different than sham at all timepoints excluding the week 0 baseline.

Figure 21: Sciatic functional index (SFI) compares geometric representations of injured hind paw to the contralateral uninjured paw, where values are scaled from ~-10 (representing nearly normal function) to ~-100 (dramatic injury). Improvements can be seen in test groups from 2-8 weeks post-operative, with autograft and drug-conduit groups having nearly identical trends. All data are shown as mean±SEM. n=4-6 per group per timepoint. *=p<0.05 vs sham.

3.3.3.2. Paw Area

There was a dramatic decrease in paw area 2 weeks post-operative, compared to baseline,

corresponding to the limited capability of toe spread of injured rats. Given the data representation for gait

testing as a ratio of ipsilateral vs contralateral, naturally a value of approximately 100% is considered

normal. Paw area values may also be affected by the rat heel, which drops with nerve injury, and may form

a longer foot print. Paw area did not show a marked change over the 8-week study, with no statistically

significant (p<0.05) differences between any of the test groups (autograft, conduit, and drug conduit)

(Figure 22). All test groups were significantly different than sham at all timepoints excluding the week 0

baseline. Paw area values 8 weeks post-operative were slightly lower than previous measurements, perhaps

due to the relatively quicker healing of the aforementioned heel drop which would further decrease the

footprint. The paw area values at 8 weeks post-operative were as follows: autograft (61.40±7.04%), conduit

(62.30±2.82%), and drug-conduit (63.48±1.55%) (mean±SEM).

Figure 22: Paw area reflects the size of the paw print during the stance phase, which decreases post-nerve injury. All data are shown as mean±SEM. Data are expressed as ratios between the ipsilateral (right) hind paw and the contralateral (left) hind paw times 100%. n=4-6 per group per timepoint. *=p<0.05 vs sham.

3.3.3.3. Swing Time

Swing time refers to the period of time the foot is off of the surface, swinging forward during the

gait cycle. Typically, swing time is increased as a result of injury, as animals are more reluctant to (or

physically cannot) put weight on, or bear load with, the affected leg. This was quantified with the sciatic

nerve transection, where 2 weeks post-operative there is an increase in swing time for all test groups (Figure

23). The swing time stabilizes around 6 weeks post-operative with 8-week swing time values as follows:

autograft (129.94±3.52%), conduit (144.31±18.66%), and drug-conduit (126.53±2.86%) (mean±SEM). At

week 6 post-operative, sham values were determined to be statistically significant (p<0.05) to conduit only.

At week 8 post-operative sham values were determined to be statistically significant (p<0.05) to autograft

and conduit only. All other values were not significant.

Figure 23: Duration of swing phase where the paw is not in contact with the glass runway, which increases post-nerve injury. All data are shown as mean±SEM. Data are expressed as ratios between the ipsilateral (right) hind paw and the contralateral (left) hind paw times 100%. n=4-6 per group per timepoint. *=p<0.05 vs sham.

3.3.3.4. Stance Time

Stance time refers to the period of time the foot is in contact with the surface. In direct contrast

with swing time, typically stance time is decreased as a result of injury. A baseline average across all groups

67 of 98.27±0.96% (mean±SEM) was recorded for stance time (Figure 24). At 2 weeks post-operative, conduit showed the lowest value (81.54±5.99%), compared to autograft and drug conduit, however the value was only significant to that of sham animals (p<0.05). At 8 weeks post-operative all test groups did not show major changes from week 2 post-operative values, with 8 weeks post-operative values as follows: autograft

(88.81±1.43%), conduit (84.59±2.50%), and drug-conduit (85.92±3.37%) (mean±SEM). Values for sham

2 weeks post-operative were determined to be statistically significant (p<0.05) to conduit only. At 8 weeks post-operative sham values were determined to be statistically significant (p<0.05) to all three test groups.

All other values were not significant.

Figure 24: Duration of the stance phase where the paw is in contact with the glass runway, which decreases post-nerve injury. All data are shown as mean±SEM. Data are expressed as ratios between the ipsilateral (right) hind paw and the contralateral (left) hind paw times 100%. n=4-6 per group per timepoint. *=p<0.05 vs sham.

3.3.4. Nerve Morphology and Myelin Assessment

Nerve regeneration was observed in all test groups 8 weeks after implantation with differing degrees of myelination (Figure 25a). Axon diameter was found to differ between groups where sham axons measured 5.87±5.46µm, autograft axons 3.42±2.89µm, conduit 5.78±8.80µm, and drug-conduit axons

7.24±8.78µm (mean±SD) (Figure 25b). More importantly, assessment of myelin thickness using G-ratio measurements concluded that, autograft had the thickest myelin at 0.61±0.10, followed by drug-conduit at

0.65±0.06, and conduit presented the thinnest myelin at 0.74±0.07 (mean±SD) (Figure 25c). These G-ratio values were compared to sham (0.57±0.08), where conduit and drug-conduit groups were significantly different (p<0.05) to sham, however autograft was not. Furthermore, autograft and drug-conduit groups were significantly different (p<0.05) to conduit values.

Figure 25: Drug-conduit implants supported axon regeneration comparable with autografts at 8 weeks post-operation. (A) Transmission electron microscopy (TEM) of cross sections of the middle region of the regenerated sciatic nerve at 8 weeks post- operation. Black arrows indicate unmyelinated C-fiber axons and white arrows indicate myelinated A-fiber axons. Images were obtained at 1000x magnification; Size bars=10 μm. (B) Scatter plot of axon diameter against G-ratio (axon diameter/fiber diameter) with logarithmic regression curves . (C) Test and control myelin sheath thicknesses (G-ratio). G-ratio is the ratio of inner axon diameter to outer fiber diameter. Data are expressed as mean±SD. n=35 measurements per group. #=p<0.05 vs sham, *=p<0.05 vs conduit.

3.3.5. Histological Evaluation

Luxol fast blue, a myelin sheath stain that stains phospholipids (the main constituent of myelin) blue, was used for the assessment of myelin presence and distribution. All groups showed myelin

69 production, as seen by the characteristic light blue staining of longitudinal sections (Figure 26a). When myelin was quantified, autograft (29.98±3.07%), conduit (26.34±3.99%), and drug-conduit (32.57±5.78%) were statistically significant (p<0.05) to sham (46.15±2.81%), however there was no significance between groups (Figure 26b). Using the same Luxol fast blue staining, the number of infiltrating cells for each group was quantified and expressed as fold change, normalized to sham values (Figure 26c). Analysis showed an increase in infiltrating immune cells for test groups, as expected. All three groups were statistically significant to sham, and autograft values (2.27±0.20) were statistically lower than conduit values (3.05±0.43). Autograft and drug-conduit values (2.81±0.55) were shown to be not significant.

Figure 26: Myelin production and number of infiltrating immune cells were similar between drug-conduit and autograft at 8 weeks post-operation. (a) Histological longitudinal sections of the middle region of the regenerated sciatic nerve stained for myelin (Luxol fast blue staining) at 8 weeks post-operation. Images were obtained at 20x magnification, where myelin is stained light blue and cell nuclei are stained violet; Size bars=200 μm. (b) Quantification of average myelin area (%) from the middle region of the regenerated nerve. (c) Number of infiltrating immune cells normalized to the sham group and expressed as fold change. Data are expressed as mean±SD, n=5 per group. #=p<0.05 vs sham, *=p<0.05 vs conduit.

In similar assessments, neurofilament-H (NF-H) and S100 were evaluated using appropriate immunohistochemical stains of longitudinal sections where the marker stains dark brown (Figure 27a).

Staining for NF-H can be seen across all groups, with a noticeably lower content in the conduit group.

Percent area values for NF-H were as follows: sham (22.49±2.16%), autograft (15.21±2.07%), conduit

(6.37±0.81%), and drug-conduit (16.05±2.54%) (Figure 27b). All three test groups were statistically significant (p<0.05) to sham, and autograft and drug-conduit values were significantly greater than conduit, but were not significant to one another. S100 staining showed a similar trend to that of NF-H (Figure 27c), where autograft (30.66±2.86%), conduit (19.95±7.45%), and drug-conduit (34.06±1.23%) values were statistically significant (p<0.05) to sham values (42.01±3.22%). Again, autograft and drug-conduit values were statistically greater than conduit values, however were not significant to one another.

Figure 27: Drug-conduit and autograft implants show similar neurogenic marker expression at 8 weeks post-operation. (a) Immunohistochemical histology assessment showing longitudinal sections of the middle region of the regenerated sciatic nerve stained for Neurofilament (NF-H) and S100 at 8 weeks post-operation. Images were obtained at 20x magnification; Size bars=200 μm. Marker proteins stain brown. (b) Quantification of average Neurofilament area (%) from the middle region of the regenerated nerve. (c) Quantification of average S100 area (%) from the middle region of the regenerated nerve. Data are expressed as mean±SD, n=5 per group. #=p<0.05 vs sham, *=p<0.05 vs conduit.

In an effort to investigate any potential toxicity of halloysite nanotubes, cross-sections of rat liver tissue were stained with hematoxylin and eosin (H&E). Figure 28 shows the histological assessment of hepatic tissues harvested from sham, autograft, conduit, and drug-conduit groups. The hepatic tissues in the control animals (sham and autograft) showed a normal hepatic architecture [233], and no differences were observed in the conduit and drug-conduit test groups, revealing that HNTs at the given dose had no remarkable adverse effects on the liver. No fatty degeneration, local hydropic degeneration, or spotted necrosis was observed in the hepatic tissue of conduit or drug-conduit groups, which may occur with HNT administration at particularly high doses, where systemic clearance and excretion is not sufficient [233,

234].

Figure 28: No apparent hepatotoxicity was observed in HNT-containing conduits or any test group. Histological cross sections (H&E stain) of liver tissue in rats 8 weeks post-operation were used to assess potential toxicity of HNT conduits. Sham and autograft groups do not contain halloysite. No fatty degeneration, local hydropic degeneration, or spotted necrosis was observed in the hepatic tissue of conduit or drug-conduit groups, which may occur with HNT administration at high doses, where systemic clearance and excretion is not sufficient. Size bar=100 μm.

3.4. DISCUSSION

The ultimate goal of all nerve repair and regeneration strategies is reestablishing continuation of the nerve channel to reinnervate the target tissue and conduct electrical impulses. The developed aligned conduit, delivering a sustained release of 4AP, showed successful nerve repair and regeneration, both histologically and functionally, showing regenerative efficacy equal to that of the gold-standard autograft

(Figure 29). The critical-size sciatic nerve transection is a severe nerve injury model which results in significant sensory and motor loss of the hindlimb. Autologous nerve grafting (autograft) is considered to be the gold-standard treatment for PNI with a large segmented gap [235]. Due to several disadvantages of autografts, engineered nerve conduits have been developed as alternatives for segmental PNI treatment.

Recent studies have shown the ability of nerve guidance conduits to improve the repair and regeneration across a nerve gap, with marked increases in functional recovery, which was also shown in our study.

Figure 29: Photograph of repaired sciatic nerve harvested at 8 weeks post-operative. The nerve is shown cut in cross-section at the central region, where the native proximal sciatic nerve can be seen in white (left), the 4AP drug-loaded conduit can be seen in the yellow/brown color, and regenerated sciatic nerve can be seen exposed through the conduit center (right).

Chitosan is a natural, biodegradable, and versatile biopolymer which has been shown to be highly biocompatible and effectively used in antibacterial, antifungal, and regenerative medicine applications. The development of chitosan scaffolds for tissue engineering and regenerative medicine applications has been extensively studied with great successes, despite mechanical fragility and poor mechanical integrity. To

73 combat such disadvantages, chitosan scaffolds can be doped with supporting compounds to allow for mechanical strengthening, creating a mechanically competent composite. Biopolymers are often doped with nanofillers which serve to increase the scaffolds’ mechanical strength, but can also enhance biological compatibility due to the surface irregularities of the scaffold pores, that arise due to the insoluble nanofillers, which promote cell adhesion. In our previous work, the doping of conduits with HNTs is detailed.

Halloysite nanotubes can have lengths ranging from 0.3-1.5 μm, with an inner lumen that has a diameter of

~40-70 nm. The neighboring alumina and silica layers, with differing waters of hydration, curve and form multilayer tubes due to a packing disorder. The innermost alumina surface is positively charged, while the outermost silicate surface is negatively charged [236]. The hollow lumen of HNTs provide excellent ability to carry, encapsulate, and transport drugs and biomacromolecules, where they can be effective in sustaining drug release while also controlling burst release characteristics of smaller, hydrophilic molecules. When mixed with cationic chitosan in a dilute acidic solution, the negatively-charged HNTs form a chitosan-HNT complex via electrostatic interactions. Furthermore, halloysite can form hydrogen bonds with the amino- and hydroxyl-groups of chitosan, further increasing the interfacial compatibility and dispersion of HNTs in chitosan [236].

Typically, after severing the sciatic nerve, the rat is unable to walk normally up on its toes. As a result, the heel drops and forms a long foot print, until reinnervation of the nerve is established and muscle function is returned [230]. Post-injury, there is also a noticeable collapse of the affected limb toe spread which is slowly restored with regeneration of the sciatic nerve. Taking advantage of such post-injury changes, the calculation of sciatic functional index (SFI) serves as a strong indicator for functional recovery of the sciatic nerve. A sharp reduction in SFI was observed post-surgery, and the SFI noticeably increased over the eight-week test period. The equivalency of autograft and drug-conduit groups (-44.89±4.73 and -

45.12±4.45, respectively) at 8-weeks post-operative is a strong positive indicator of the efficacy of the drug- conduit. Neat (no drug) conduit groups did not show any statistical differences from values of autograft and drug-conduit groups in functional assessments, which may reflect on the functional capabilities of the

74 aligned microchannel porosity to physically guide regenerating axons, improving target reinnervation.

Autograft and drug-conduit groups performed virtually identically in quantifications of paw area, swing time, and stance time, further confirming the drug-conduit’s ability to improve functional recovery post- nerve transection.

The endogenous repair of transected PNI is marked by three broad phases: Wallerian degeneration, axonal regeneration, and target-organ reinnervation. Schwann cells are vital primary mediators of the cascading events of Wallerian degeneration, including the recruitment of macrophages, along with increases in critical factors including brain-derived neurotrophic factor (BDNF) and nerve growth factor

(NGF) [42-44]. The ability of the small-molecule drug 4AP to increase BDNF and NGF release in Schwann cells in vitro was shown in a previous publication [69]. Schwann cells are also critical in the alteration of gene expression, removal of myelin, formation of Bands of Büngner, and axon regeneration and remyelination. The protein marker S100 exists only in glial cells of the central nervous system and Schwann cells of the peripheral nervous system. Many S100-positive Schwann cells were observed via immunohistochemical staining in all groups, with autograft and drug-conduit groups showing relatively equal levels of S100, significantly greater than neat (no-drug) conduit groups. The increased expression of

S100 may be a result of increased Schwann cell recruitment to the injury site due to chemotactic effects of amplified neurotransmitter release as a direct result of 4AP administration. This trend was also seen with the detection of neurofilament-H (NF-H), a neurofilament protein which is a primary component of the mature neuronal cytoskeleton, and significantly downregulated at the early period following axonal transection [237]. When neurofilament-positive axons were quantified, levels were not significantly different between autograft and drug-conduit groups, which both showed higher levels than neat (no-drug) conduit groups.

The degree of repair and regeneration achieved by the nerve conduit was further confirmed by the investigation of myelin production and thickness, both histologically and using TEM. Using a Luxol Fast

Blue stain for myelin, histological studies revealed abundant, well-myelinated axons in the mid-portion of

75 the regenerated nerve. Autograft groups showed no statistical differences compared to both neat conduit and drug-conduit groups. Similarly, in TEM morphometric studies, myelin thickness was calculated to be statistically equivalent between autograft and drug-conduit groups, which were significantly greater than neat (no-drug) conduit groups (smaller G-ratio). This is perhaps attributed to the ability of 4AP drug to upregulate the release of myelin proteins in Schwann cells, including myelin protein zero (P0), which was shown in our previous publication [69].

We conducted several extensive studies and assessments to evaluate peripheral nerve regeneration using our previously developed and characterized nerve guidance conduit. The advantages of the nerve guidance conduit were (1) the aligned microchannel porosity, which can physically guide the regenerating axons across the injury gap for better distal target reinnervation; and (2) the sustained release of small- molecule 4AP drug which can increase nerve conduction, increasing the release of synaptic neurotransmitters and critical Schwann cell neurotrophic factors including BDNF, NGF, and myelin protein zero. Studies showed varying levels of nerve regeneration across the autograft, conduit, and drug-conduit groups. All three groups showed successful increases in functional recovery without significant differences, however, the drug-conduit outperformed the neat (no drug) conduit in terms of histological and morphometric evaluations of myelin, S100, and NF-H. Overall, the drug-conduit showed equivalency to autograft repair in a majority of the functional, morphological, and histological assessments. These results indicate the success of the conduit’s aligned microchannel architecture to facilitate physical guidance and reinnervation, but also highlight the enhanced efficacy of regeneration achieved by the drug conduit, a testament to the contributions of 4AP to the repair and regeneration process. This represents a significant advance in the fields of peripheral nerve repair, regeneration, and engineering, where the developed aligned nerve guidance conduit was shown to be a clinically feasible drug delivery vehicle for small molecule drug

4AP in vivo, showing its strong ability to treat peripheral nerve injury.

3.5. CONCLUSION

The present study looked at novel nerve guidance conduits with aligned microchannel porosity delivering a sustained release of 4-Aminopyridine for peripheral nerve regeneration in a critical-size rat sciatic nerve transection model. The results of functional walking track analysis, morphometric evaluations of myelin development, and histological assessments of various markers confirmed the equivalency of implanted drug-conduits repair to autograft control treatments. The conduit’s aligned microchannel architecture may play a vital role in facilitating physical guidance and distal target reinnervation, while the sustained release of 4-Aminopyridine may increase nerve conduction, and in turn synaptic neurotransmitter release and upregulation of critical Schwann cell neurotrophic factors, which further aid in more efficient and efficacious peripheral nerve regeneration via a drug delivery system that is feasible for clinical application.

CHAPTER 4: SUMMARY OF FINDINGS

The presented project describes, in detail, engineering efforts to combat slow and ineffective peripheral nerve injury repair by investigating, and later, developing and testing a novel approach to peripheral nerve regeneration. A nerve guidance conduit was developed in order to bridge peripheral nerve defect injuries, in an effort to help promote more efficient and successful nerve repair and regeneration.

Chemically composed of chitosan, a natural, biodegradable, biocompatible, and FDA-approved polymer, the nerve guidance conduit was developed and fully characterized for physiochemical properties. In an effort to enhance regenerative cues, aligned microchannel porosity was created within the conduit structure using a unidirectional freezing technique (Figure 30). Such physical cues will help correctly guide sprouting axons to re-innervate the target distal organ more effectively. Due to chitosan’s poor mechanical integrity, halloysite nanotubes were added as filler composite material, which served to increase the mechanical strength of the conduit system, and also acted as the primary drug-delivery vehicle for the selected drug, 4-Aminopyridine. The inclusion of halloysite nanotubes was shown to be homogenous and was shown to have no significant impacts on physiological toxicity. The composite structure was shown to have adequate mechanical modulus, within comparable range of native peripheral nerve.

Figure 30: Summary of key advantages of the developed nerve guidance conduit system.

The composite nerve guidance conduit was further tested for biodegradation properties and human

Schwann cell viability. The degradation profile of the polymer system was characterized using standard

78 calculations over a 12-week period. The degradation was visually inspected by performing SEM over the course of the testing period. Cell viability on the conduit structure was analyzed using a standard MTS cell assay as well as fluorescent microscopy. The conduit showed adequate cell attachment and viability, with normal cell proliferation, and showed longitudinal alignment of cells according to the aligned nature of the polymer matrix.

4-Aminopyridine, an FDA-approved small molecule voltage-gated potassium channel blocker, was successfully loaded into halloysite nanotubes and incorporated into the composite nerve guidance conduit structure. The release of 4-Aminopyridine drug was fully characterized from a variety of conduit formulations, including from halloysite nanotubes without a polymer matrix. The drug release profiles were also fit to multiple kinetic release models which were in agreement with the findings. The final composite structure achieved sustained release of 4-Aminopyridine, showing only ~30% drug release within 7 days.

The loading of 4-Aminopyridine into halloysite nanotubes prior to inclusion into the polymer matrix showed a tremendous boost in controlling the burst release of drug typically seen with small-molecules from polymeric systems.

The effect of 4-Aminopyridine on human Schwann cells was tested in vitro in a systematic assessment of cytotoxicity, proliferation, and protein expression. Human Schwann cells were treated with a wide range of 4-Aminopyridine doses (0.001-10 µg/mL) and evaluated for cytotoxicity using a standard

LDH assay, which showed no significant differences in cell activity between groups, indicating no cytotoxicity. The proliferation of Schwann cells was investigated when treated with 1, 5, and 10 µg/mL of

4-Aminopyridine and compared to a no-drug control. Early timepoints of 1- and 7-days post-seeding showed no differences in cell proliferation, however at 14- and 21-days post-seeding, 4-Aminopyridine decreased cell proliferation when compared with the no-drug control. This decrease in proliferation was in agreement with findings in literature and was further confirmed with evaluations of protein expression.

Immunofluorescent staining was used to evaluate the expression of three key neurotrophic factors: nerve growth factor (NGF), myelin protein zero (P0), and brain-derived neurotrophic factor (BDNF). At 14 days

79 post-seeding, 4-Aminopyridine showed a dose-dependent increase in all three protein markers. These increases in neurotrophic factor expression can be directly correlated to the aforementioned decreases in proliferation, where past literature states that proliferation and myelin differentiation have been proven to be incompatible processes, unable to occur concurrently both in vitro and in vivo. A summary of key in vitro findings and achievements is shown in Figure 31.

Figure 31: Key in vitro findings and achievements.

To test the feasibility and efficacy of the drug-delivery conduit in vivo, a 15 mm (critical size) sciatic nerve defect was created in female Wistar rats, serving as a severe peripheral nerve injury model.

The nerve defect was bridged and repaired with an autograft control, a neat (no-drug) conduit, and a 4AP drug-loaded conduit. Over an 8-week test period, the animals were subjected to walking track analysis where gait parameters were measured as an analysis of functional recovery. After 8-weeks post-operative, animals were euthanized and nerve repair and regeneration was evaluated using several histological stains

80 and myelin development was evaluated using electron microscope. The results of functional walking track analysis, morphometric evaluations of myelin development, and histological assessments of various markers confirmed the equivalency of the drug-conduit repair to autograft controls. Repaired nerves show formation of thick myelin, presence of S100 and Neurofilament markers, and promising functional recovery. Although the neat (no-drug) conduit showed positive nerve regeneration, it was deficient compared to the regenerative efficacy of autograft and drug-conduits. A summary of key in vivo findings and achievements is shown in Figure 32.

Figure 32: Key in vivo findings and achievements.

Overall the nerve guidance conduit achieved peripheral nerve regeneration across a critical size nerve defect. The drug-conduit’s performance was measured to be statistically equivalent to the gold- standard autograft in a majority of the performed evaluations, including the assessment of functional recovery. The conduit’s aligned microchannel architecture may play a vital role in physically guiding axons

81 to distal target re-innervation, while the sustained release of 4-Aminopyridine may increase nerve conduction, and in turn synaptic neurotransmitter release and upregulation of critical Schwann cell neurotrophic factors. This facilitates more efficient and efficacious peripheral nerve regeneration via a drug delivery system that is feasible for clinical application, with potentially transformative implications.

CHAPTER 5: FUTURE PERSPECTIVES

Clinically, the repair of peripheral nerve injuries is positioned at a crossroads of medical disciplines, namely orthopedics and neurosurgery. However, surgeons classically trained in the nervous system and neurosurgery most commonly operate in the areas of the central nervous system. As such, most (particularly mild and moderate) peripheral nerve repair falls to trauma surgery and orthopedic surgery professionals.

Clinical investigation of endogenous repair mechanisms, neuroma (scar-tissue) prevention strategies, endplate re-innervation modalities, as well as cellular involvements and biochemical pathways are critical for future improvements to nerve repair, re-innervation, and functional recovery. The overwhelming amount of research and development focused on cellular, pharmaceutical, and bioengineering modalities to further improve peripheral nerve repair and regeneration, as well as patient outcome, is a testament to the severity of the unmet clinical need and the potentially transformative implications of any such treatment alternatives. A summary of key future directions and scientific perspectives is shown in Figure 33.

Figure 33: Future directions of the field and scientific perspectives

Ideally, a nerve guidance conduit (NGC) systems needs to be biocompatible, biodegradable, adequately flexible yet suturable, and should have a combination of physical topographic cues as well as biological agents for enhanced repair and regeneration. To date, researchers around the world have developed various strategies in combination with NGC for improved regeneration outcomes. The transplantation of Schwann cells as well as stem cells (adipose-derived, mesenchymal, neural, etc.) have

83 shown successes in combination with NGCs, but pose feasibility concerns with translation to clinic. The use of biopolymer surface modification and the delivery of growth factors (NGF, GDNF, BDNF, VEGF,

NT-3, etc.) and drugs have also been popularly researched. More recently, physical stimuli such as ultrasound and electrical stimulation have been investigated for their regenerative effects. Lastly, genetic manipulation, by means of transfection and viral vectors has been attempted, but is still considered non- translatable.

Given the complex nature of peripheral nerve injury, repair, and regeneration there have been few tissue engineering developments to currently used treatments, mainly citing clinical feasibility concerns

(design, logistical, or financial) and differing preclinical and clinical results. This highlights the unmet need for better preclinical models focused on optimized clinical feasibility and translatability, in order for tested tissue engineering approaches to successfully bridge the lab-to-clinic gap [238]. Promising technologies, such as lab-on-a-chip methods, exist and are being developed which hold significant promise for more accurate and relevant in vitro testing models [239]. The combination of pharmaceutical and molecular therapies (i.e. protein and RNA delivery) will undoubtedly be involved for enhancement of peripheral nerve repair and regeneration. In combination with novel surgical techniques, the use of new nerve guidance conduits and constructs may alleviate shortcomings of traditional autografts and allografts to facilitate.

Future directions aimed at the delivery of biomacromolecules, such as silencing RNAs (siRNA), proteins and peptides, and DNA therapies may facilitate intrinsic growth mechanisms, potentially leading to more natural, endogenous growth and repair [240]. The rapid developmental growth of 3-dimensional (3D) printing technologies also hold high promise in such fields which may revolutionize the repair of peripheral nerve injuries with patient-specific parameters for more accurate and helpful nerve guidance conduits [239].

In the future development of new nerve constructs, the use of modern techniques such as 3D- printing, laser fabrication, two-photon and single-photon stereo lithography, extrusion-based fabrication, and rapid prototyping will undoubtedly result in the fabrication of more complicated constructs with very high degrees of accuracy. These modern technologies may enable the ability to more accurately mimic

84 nerve anatomy and architecture, such as bands of Büngner, leading to more endogenous-like repair.

Furthermore, the use of modern fabrication techniques such as 3D-printing and extrusion-based fabrication may facilitate the manufacturing of structures with advanced geographical features including capillary blood vessels and other vasculature, which will become increasingly important as constructs are developed for larger defects. The ability to accurately predict such fabrication and performance using computer and mathematical modeling will continue to be a useful tool in the design of new NGCs. Lastly, more rigorous in vitro testing as well as more translatable in vivo testing, by means of large animal studies, will be required to determine the efficacy of such constructs to successfully bridge the lab-to-clinic gap.

Despite the successes of the current study, there are several limitations. Firstly, due to greater variability in functional assessments, the number of animals in each group may need to be increased to obtain more walking track data points, which may help in the identification of statistically significant improvements between groups. Second, it is possible that functional differences between groups, specifically the neat (no-drug) conduit and drug-conduit, may be evident at later time points given the differences observed in TEM and histology between these groups at the 8-week post-operative time point.

As such, extended the study to include future timepoints may yield more evident changes and improvements in functional recovery. Nevertheless, the functional assessments provided key insights into differences between the groups, and highlighted the performance equivalence between the control autograft and test drug-conduit. Third, the release of 4AP was not quantified and characterized in vivo. The quantification and optimization of 4AP released in vivo will be critical to harness the full capabilities of this drug for optimized nerve regeneration. According to the pathophysiology of peripheral nerve injury and repair, the first 7-21 days offer the greatest potential for facilitated regeneration as this is the period of Wallerian degeneration, cell recruitment, and the beginning of axonal regeneration. As such, this will be the period where 4AP will have the greatest functional impacts. The dosing of drug in the implanted conduits should be evaluated, modified, and optimized in future studies.

REFERENCES

[1] J. Noble, C. A. Munro, V. S. Prasad, and R. Midha, "Analysis of upper and lower extremity peripheral nerve injuries in a population of patients with multiple injuries," Journal of Trauma and Acute Care Surgery, vol. 45, pp. 116-122, 1998. [2] G. R. Evans, "Peripheral nerve injury: a review and approach to tissue engineered constructs," The Anatomical Record: An Official Publication of the American Association of Anatomists, vol. 263, pp. 396-404, 2001. [3] S. Yegiyants, D. Dayicioglu, G. Kardashian, and Z. J. Panthaki, "Traumatic peripheral nerve injury: a wartime review," Journal of Craniofacial Surgery, vol. 21, pp. 998-1001, 2010. [4] I.-C. Liao, H. Wan, S. Qi, C. Cui, P. Patel, W. Sun, et al., "Preclinical evaluations of acellular biological conduits for peripheral nerve regeneration," Journal of tissue engineering, vol. 4, p. 2041731413481036, 2013. [5] B. Holmquist, M. Kanje, J. M. Kerns, and N. Danielsen, "A mathematical model for regeneration rate and initial delay following surgical repair of peripheral nerves," Journal of neuroscience methods, vol. 48, pp. 27-33, 1993. [6] A. R. Nectow, K. G. Marra, and D. L. Kaplan, "Biomaterials for the development of peripheral nerve guidance conduits," Tissue Engineering Part B: Reviews, vol. 18, pp. 40-50, 2011. [7] M. T. Jobe and S. Martinez, "Canale & Beaty: Campbell's Operative Orthopaedics," ed: Elsevier Philadelphia:, 2013. [8] R. M. Menorca, T. S. Fussell, and J. C. Elfar, "Peripheral nerve trauma: mechanisms of injury and recovery," Hand clinics, vol. 29, p. 317, 2013. [9] S. Sunderland, "A classification of peripheral nerve injuries producing loss of function," Brain, vol. 74, pp. 491-516, 1951. [10] A. Chhabra, S. Ahlawat, A. Belzberg, and G. Andreseik, "Peripheral nerve injury grading simplified on MR neurography: as referenced to Seddon and Sunderland classifications," The Indian journal of radiology & imaging, vol. 24, p. 217, 2014. [11] D. Zochodne and D. Levy, "Nitric oxide in damage, disease and repair of the peripheral nervous system," Cellular and molecular biology (Noisy-le-Grand, France), vol. 51, pp. 255-267, 2005. [12] A. D. Gaudet, P. G. Popovich, and M. S. Ramer, "Wallerian degeneration: gaining perspective on inflammatory events after peripheral nerve injury," Journal of neuroinflammation, vol. 8, p. 110, 2011. [13] A. DeFrancesco-Lisowitz, J. Lindborg, J. Niemi, and R. Zigmond, "The neuroimmunology of degeneration and regeneration in the peripheral nervous system," Neuroscience, vol. 302, pp. 174-203, 2015. [14] M. Anderson, N. B. Shelke, O. S. Manoukian, X. Yu, L. D. McCullough, and S. G. Kumbar, "Peripheral nerve regeneration strategies: electrically stimulating polymer based nerve growth conduits," Critical Reviews™ in Biomedical Engineering, vol. 43, 2015. [15] L. I. Benowitz and P. G. Popovich, "Inflammation and axon regeneration," Current opinion in neurology, vol. 24, pp. 577-583, 2011. [16] M. Bendszus and G. Stoll, "Caught in the act: in vivo mapping of macrophage infiltration in nerve injury by magnetic resonance imaging," Journal of Neuroscience, vol. 23, pp. 10892- 10896, 2003. [17] M. Mueller, C. Leonhard, K. Wacker, E. B. Ringelstein, M. Okabe, W. F. Hickey, et al., "Macrophage response to peripheral nerve injury: the quantitative contribution of resident and hematogenous macrophages," Laboratory investigation, vol. 83, p. 175, 2003. [18] V. Perry, M. Brown, and S. Gordon, "The macrophage response to central and peripheral nerve injury. A possible role for macrophages in regeneration," Journal of Experimental Medicine, vol. 165, pp. 1218-1223, 1987.

[19] H. Taskinen and M. Röyttä, "The dynamics of macrophage recruitment after nerve transection," Acta neuropathologica, vol. 93, pp. 252-259, 1997. [20] P. Chen, X. Piao, and P. Bonaldo, "Role of macrophages in Wallerian degeneration and axonal regeneration after peripheral nerve injury," Acta neuropathologica, vol. 130, pp. 605-618, 2015. [21] N. Kiguchi, Y. Kobayashi, F. Saika, and S. Kishioka, "Epigenetic upregulation of CCL2 and CCL3 via histone modifications in infiltrating macrophages after peripheral nerve injury," Cytokine, vol. 64, pp. 666-672, 2013. [22] S. Shamash, F. Reichert, and S. Rotshenker, "The cytokine network of Wallerian degeneration: tumor necrosis factor-α, interleukin-1α, and interleukin-1β," Journal of Neuroscience, vol. 22, pp. 3052-3060, 2002. [23] G. K. Tofaris, P. H. Patterson, K. R. Jessen, and R. Mirsky, "Denervated Schwann cells attract macrophages by secretion of leukemia inhibitory factor (LIF) and monocyte chemoattractant protein-1 in a process regulated by interleukin-6 and LIF," Journal of Neuroscience, vol. 22, pp. 6696-6703, 2002. [24] K. Namikawa, T. Okamoto, A. Suzuki, H. Konishi, and H. Kiyama, "Pancreatitis-associated protein-III is a novel macrophage chemoattractant implicated in nerve regeneration," Journal of Neuroscience, vol. 26, pp. 7460-7467, 2006. [25] F. E. Perrin, S. Lacroix, M. Avilés-Trigueros, and S. David, "Involvement of monocyte chemoattractant protein-1, macrophage inflammatory protein-1α and interleukin-1β in Wallerian degeneration," Brain, vol. 128, pp. 854-866, 2005. [26] R. George and J. W. Griffin, "The proximo-distal spread of axonal degeneration in the dorsal columns of the rat," Journal of neurocytology, vol. 23, pp. 657-667, 1994. [27] R. Miledi and C. Slater, "On the degeneration of rat neuromuscular junctions after nerve section," The Journal of Physiology, vol. 207, pp. 507-528, 1970. [28] W. Tetzlaff and M. Bisby, "Neurofilament elongation into regenerating facial nerve axons," Neuroscience, vol. 29, pp. 659-666, 1989. [29] S. E. Lewis, R. J. Mannion, F. A. White, R. E. Coggeshall, S. Beggs, M. Costigan, et al., "A role for HSP27 in sensory neuron survival," Journal of Neuroscience, vol. 19, pp. 8945-8953, 1999. [30] S. HALL, "Mechanisms of repair after traumatic injury," in Peripheral Neuropathy (Fourth Edition), ed: Elsevier, 2005, pp. 1403-1433. [31] W. Tetzlaff, S. W. Alexander, F. D. Miller, and M. A. Bisby, "Response of facial and rubrospinal neurons to axotomy: changes in mRNA expression for cytoskeletal proteins and GAP-43," Journal of Neuroscience, vol. 11, pp. 2528-2544, 1991. [32] S. Y. Fu and T. Gordon, "The cellular and molecular basis of peripheral nerve regeneration," Molecular neurobiology, vol. 14, pp. 67-116, 1997. [33] M. Costigan, R. J. Mannion, G. Kendall, S. E. Lewis, J. A. Campagna, R. E. Coggeshall, et al., "Heat shock protein 27: developmental regulation and expression after peripheral nerve injury," Journal of neuroscience, vol. 18, pp. 5891-5900, 1998. [34] A. Aguayo, J. Peyronnard, and G. Bray, "A quantitative ultrastructural study of regeneration from isolated proximal stumps of transected unmyelinated nerves," Journal of Neuropathology & Experimental Neurology, vol. 32, pp. 256-270, 1973. [35] F. Sanders and J. Young, "The influence of peripheral connexion on the diameter of regenerating nerve fibres," Journal of Experimental Biology, vol. 22, pp. 203-212, 1946. [36] J. Griffin and P. Hoffman, "Degeneration and regeneration in the peripheral nervous system," Peripheral neuropathy, vol. 1, pp. 361-388, 1993. [37] W. W. Campbell, "Evaluation and management of peripheral nerve injury," Clinical Neurophysiology, vol. 119, pp. 1951-1965, 2008/09/01/ 2008. [38] E. Lunn, M. Brown, and V. Perry, "The pattern of axonal degeneration in the peripheral nervous system varies with different types of lesion," Neuroscience, vol. 35, pp. 157-165, 1990.

[39] A. D. Guertin, D. P. Zhang, K. S. Mak, J. A. Alberta, and H. A. Kim, "Microanatomy of axon/glial signaling during Wallerian degeneration," Journal of neuroscience, vol. 25, pp. 3478- 3487, 2005. [40] B. B. Murinson, D. R. Archer, Y. Li, and J. W. Griffin, "Degeneration of myelinated efferent fibers prompts mitosis in Remak Schwann cells of uninjured C-fiber afferents," Journal of Neuroscience, vol. 25, pp. 1179-1187, 2005. [41] H. M. Liu, L. H. Yang, and Y. J. Yang, "Schwann cell properties: 3. C-fos expression, bFGF production, phagocytosis and proliferation during Wallerian degeneration," Journal of Neuropathology & Experimental Neurology, vol. 54, pp. 487-496, 1995. [42] B. D. Trapp, P. Hauer, and G. Lemke, "Axonal regulation of myelin protein mRNA levels in actively myelinating Schwann cells," Journal of Neuroscience, vol. 8, pp. 3515-3521, 1988. [43] F. V. White, A. D. Toews, J. F. Goodrum, D. L. Novicki, T. W. Bouldin, and P. Morell, "Lipid metabolism during early stages of Wallerian degeneration in the rat sciatic nerve," Journal of neurochemistry, vol. 52, pp. 1085-1092, 1989. [44] R. Pellegrino, M. Politis, J. Ritchie, and P. Spencer, "Events in degenerating cat peripheral nerve: induction of Schwann cell S phase and its relation to nerve fibre degeneration," Journal of neurocytology, vol. 15, pp. 17-28, 1986. [45] J. W. Griffin and W. J. Thompson, "Biology and pathology of nonmyelinating Schwann cells," Glia, vol. 56, pp. 1518-1531, 2008. [46] G. Stoll, J. Griffin, C. Y. Li, and B. Trapp, "Wallerian degeneration in the peripheral nervous system: participation of both Schwann cells and macrophages in myelin degradation," Journal of neurocytology, vol. 18, pp. 671-683, 1989. [47] J. K. Huang, G. R. Phillips, A. D. Roth, L. Pedraza, W. Shan, W. Belkaid, et al., "Glial membranes at the node of Ranvier prevent neurite outgrowth," Science, vol. 310, pp. 1813-1817, 2005. [48] Z. L. Shen, F. Lassner, A. Bader, M. Becker, G. F. Walter, and A. Berger, "Cellular activity of resident macrophages during Wallerian degeneration," Microsurgery, vol. 20, pp. 255-261, 2000. [49] V. Perry, J. Tsao, S. Feam, and M. Brown, "Radiation‐induced reductions in macrophage recruitment have only slight effects on myelin degeneration in sectioned peripheral nerves of mice," European Journal of Neuroscience, vol. 7, pp. 271-280, 1995. [50] R. Martini, S. Fischer, R. López‐Vales, and S. David, "Interactions between Schwann cells and macrophages in injury and inherited demyelinating disease," Glia, vol. 56, pp. 1566-1577, 2008. [51] S. De, M. A. Trigueros, A. Kalyvas, and S. David, "Phospholipase A2 plays an important role in myelin breakdown and phagocytosis during Wallerian degeneration," Molecular and Cellular Neuroscience, vol. 24, pp. 753-765, 2003. [52] O. A. Sulaiman and T. Gordon, "Effects of short‐and long‐term Schwann cell denervation on peripheral nerve regeneration, myelination, and size," Glia, vol. 32, pp. 234-246, 2000. [53] M. E. Vargas and B. A. Barres, "Why is Wallerian degeneration in the CNS so slow?," Annu. Rev. Neurosci., vol. 30, pp. 153-179, 2007. [54] O. A. Sulaiman and T. Gordon, "Transforming growth factor‐β and forskolin attenuate the adverse effects of long‐term Schwann cell denervation on peripheral nerve regeneration in vivo," Glia, vol. 37, pp. 206-218, 2002. [55] H. Li, G. Terenghi, and S. Hall, "Effects of delayed re‐innervation on the expression of c‐erbB receptors by chronically denervated rat Schwann cells in vivo," Glia, vol. 20, pp. 333-347, 1997. [56] E. I. Dedkov, T. Y. Kostrominova, A. B. Borisov, and B. M. Carlson, "Survival of Schwann cells in chronically denervated skeletal muscles," Acta neuropathologica, vol. 103, pp. 565-574, 2002. [57] K. Belanger, T. M. Dinis, S. Taourirt, G. Vidal, D. L. Kaplan, and C. Egles, "Recent Strategies in Tissue Engineering for Guided Peripheral Nerve Regeneration," Macromolecular Bioscience, vol. 16, pp. 472-481, 2016. [58] W. Z. Ray and S. E. Mackinnon, "Management of nerve gaps: autografts, allografts, nerve transfers, and end-to-side neurorrhaphy," Experimental neurology, vol. 223, p. 77, 2010.

[59] S. E. Mackinnon, A. L. Dellon, and J. Zhu, Surgery of the peripheral nerve vol. 396: Thieme Medical Publishers New York, 1988. [60] T. Norkus, M. Norkus, and T. Ramanauskas, "Donor, recipient and nerve grafts in brachial plexus reconstruction: anatomical and technical features for facilitating the exposure," Surgical and Radiologic Anatomy, vol. 27, pp. 524-530, 2005. [61] E. S. Dellon and A. L. Dellon, "The first nerve graft, Vulpian, and the nineteenth century neural regeneration controversy," Journal of Hand Surgery, vol. 18, pp. 369-372, 1993. [62] J. W. D. Chiu, "Painful neuromas: a review of treatment modalities," Ann Plast Surg, vol. 43, 1999. [63] P. Anderson and M. Turmaine, "Peripheral nerve regeneration through grafts of living and freeze‐ dried CNS tissue," Neuropathology and applied neurobiology, vol. 12, pp. 389-399, 1986. [64] P. J. Evans, S. E. Mackinnon, T. J. Best, J. A. Wade, D. C. Awerbuck, A. P. Makino, et al., "Regeneration across preserved peripheral nerve grafts," Muscle & Nerve: Official Journal of the American Association of Electrodiagnostic Medicine, vol. 18, pp. 1128-1138, 1995. [65] G. Lawson and M. Glasby, "A comparison of immediate and delayed nerve repair using autologous freeze-thawed muscle grafts in a large animal model: the simple injury," Journal of Hand Surgery, vol. 20, pp. 663-700, 1995. [66] L. Marmor, "The repair of peripheral nerves by irradiated homografts," Clinical Orthopaedics and Related Research®, vol. 32, p. 187, 1964. [67] A. Martini, "The lyophilized homologous nerve graft for the prevention of neuroma formation (animal experiment study)," Handchirurgie, Mikrochirurgie, plastische Chirurgie: Organ der Deutschsprachigen Arbeitsgemeinschaft fur Handchirurgie: Organ der Deutschsprachigen Arbeitsgemeinschaft fur Mikrochirurgie der Peripheren Nerven und Gefasse: Organ der V.. vol. 17, pp. 266-269, 1985. [68] T. E. Trumble and F. G. Shon, "The physiology of nerve transplantation," Hand clinics, vol. 16, pp. 105-122, 2000. [69] O. S. Manoukian, M. R. Arul, S. Rudraiah, I. Kalajzic, and S. G. Kumbar, "Aligned microchannel polymer-nanotube composites for peripheral nerve regeneration: Small molecule drug delivery," Journal of Controlled Release, 2019. [70] Y. Cui, C. Lu, D. Meng, Z. Xiao, X. Hou, W. Ding, et al., "Collagen scaffolds modified with CNTF and bFGF promote facial nerve regeneration in minipigs," Biomaterials, vol. 35, pp. 7819- 7827, 2014. [71] C. Wang, Y. Jia, W. Yang, C. Zhang, K. Zhang, and Y. Chai, "Silk fibroin enhances peripheral nerve regeneration by improving vascularization within nerve conduits," Journal of Biomedical Materials Research Part A, vol. 106, pp. 2070-2077, 2018. [72] S. Gnavi, L. Di Blasio, C. Tonda‐Turo, A. Mancardi, L. Primo, G. Ciardelli, et al., "Gelatin‐based hydrogel for vascular endothelial growth factor release in peripheral nerve tissue engineering," Journal of tissue engineering and regenerative medicine, vol. 11, pp. 459-470, 2017. [73] R. C. Thomas, P. Vu, S. P. Modi, P. E. Chung, R. C. Landis, Z. Z. Khaing, et al., "Sacrificial crystal templated hyaluronic acid hydrogels as biomimetic 3D tissue scaffolds for nerve tissue regeneration," ACS Biomaterials Science & Engineering, vol. 3, pp. 1451-1459, 2017. [74] J. Du, J. Liu, S. Yao, H. Mao, J. Peng, X. Sun, et al., "Prompt peripheral nerve regeneration induced by a hierarchically aligned fibrin nanofiber hydrogel," Acta biomaterialia, vol. 55, pp. 296-309, 2017. [75] O. S. Manoukian, M. R. Arul, N. Sardashti, T. Stedman, R. James, S. Rudraiah, et al., "Biodegradable polymeric injectable implants for long‐term delivery of contraceptive drugs," Journal of applied polymer science, vol. 135, p. 46068, 2018. [76] R. James, O. S. Manoukian, and S. G. Kumbar, "Poly (lactic acid) for delivery of bioactive macromolecules," Advanced drug delivery reviews, vol. 107, pp. 277-288, 2016.

[77] P. Lee, O. S. Manoukian, G. Zhou, Y. Wang, W. Chang, X. Yu, et al., "Osteochondral scaffold combined with aligned nanofibrous scaffolds for cartilage regeneration," RSC Advances, vol. 6, pp. 72246-72255, 2016. [78] O. S. Manoukian, A. Aravamudhan, P. Lee, M. R. Arul, X. Yu, S. Rudraiah, et al., "Spiral Layer- by-Layer Micro-Nanostructured Scaffolds for Bone Tissue Engineering," ACS Biomaterials Science & Engineering, vol. 4, pp. 2181-2192, 2018. [79] M. Barbarisi, G. Marino, E. Armenia, Q. Vincenzo, F. Rosso, M. Porcelli, et al., "Use of polycaprolactone (PCL) as scaffolds for the regeneration of nerve tissue," Journal of Biomedical Materials Research Part A, vol. 103, pp. 1755-1760, 2015. [80] C. O. Goulart, F. R. P. Lopes, Z. O. Monte, S. V. Dantas Jr, A. Souto, J. T. Oliveira, et al., "Evaluation of biodegradable polymer conduits–poly (L-lactic acid)–for guiding sciatic nerve regeneration in mice," Methods, vol. 99, pp. 28-36, 2016. [81] Y. Peng, K.-Y. Li, Y.-F. Chen, X.-J. Li, S. Zhu, Z.-Y. Zhang, et al., "Beagle sciatic nerve regeneration across a 30 mm defect bridged by chitosan/PGA artificial nerve grafts," Injury, 2018. [82] H. Liu, P. Lv, Y. Zhu, H. Wu, K. Zhang, F. Xu, et al., "Salidroside promotes peripheral nerve regeneration based on tissue engineering strategy using Schwann cells and PLGA: in vitro and in vivo," Scientific reports, vol. 7, p. 39869, 2017. [83] M. Richner, M. Ulrichsen, S. L. Elmegaard, R. Dieu, L. T. Pallesen, and C. B. Vaegter, "Peripheral nerve injury modulates neurotrophin signaling in the peripheral and central nervous system," Molecular neurobiology, vol. 50, pp. 945-970, 2014. [84] L. A. Pfister, M. Papaloïzos, H. P. Merkle, and B. Gander, "Nerve conduits and growth factor delivery in peripheral nerve repair," Journal of the Peripheral Nervous System, vol. 12, pp. 65-82, 2007. [85] Q. Guo, C. Liu, B. Hai, T. Ma, W. Zhang, J. Tan, et al., "Chitosan conduits filled with simvastatin/Pluronic F-127 hydrogel promote peripheral nerve regeneration in rats," Journal of Biomedical Materials Research Part B: Applied Biomaterials, vol. 106, pp. 787-799, 2018. [86] T. Gordon and A. W. English, "Strategies to promote peripheral nerve regeneration: electrical stimulation and/or exercise," European Journal of Neuroscience, vol. 43, pp. 336-350, 2016. [87] R. Levi-Montalcini, "Nobel lecture: the nerve growth factor: thirty-five years later," ed, 2018. [88] Z. Kuihua, W. Chunyang, F. Cunyi, and M. Xiumei, "Aligned SF/P(LLA-CL)-blended nanofibers encapsulating nerve growth factor for peripheral nerve regeneration," Journal of Biomedical Materials Research Part A, vol. 102, pp. 2680-2691, 2014. [89] K. M. Chan, T. Gordon, D. W. Zochodne, and H. A. Power, "Improving peripheral nerve regeneration: From molecular mechanisms to potential therapeutic targets," Experimental Neurology, vol. 261, pp. 826-835, 2014/11/01/ 2014. [90] P. Labroo, J. Shea, H. Sant, B. Gale, and J. Agarwal, "EFFECT OF COMBINING FK506 AND NEUROTROPHINS ON NEURITE BRANCHING AND ELONGATION," Muscle & nerve, vol. 55, pp. 570-581, 12/28 05/23/nihms-submitted 07/19/pmc-release 2017. [91] G. Lin, H. Zhang, F. Sun, Z. Lu, A. Reed-Maldonado, Y.-C. Lee, et al., "Brain-derived neurotrophic factor promotes nerve regeneration by activating the JAK/STAT pathway in Schwann cells," Translational andrology and urology, vol. 5, p. 167, 2016. [92] X. Ke, Q. Li, L. Xu, Y. Zhang, D. Li, J. Ma, et al., "Netrin-1 overexpression in bone marrow mesenchymal stem cells promotes functional recovery in a rat model of peripheral nerve injury," Journal of Biomedical Research, vol. 29, pp. 380-389, 07/30. [93] J. Boyd and T. Gordon, "A dose‐dependent facilitation and inhibition of peripheral nerve regeneration by brain‐derived neurotrophic factor," European Journal of Neuroscience, vol. 15, pp. 613-626, 2002.

[94] J. C. Wilhelm, M. Xu, D. Cucoranu, S. Chmielewski, T. Holmes, K. S. Lau, et al., "Cooperative roles of BDNF expression in neurons and Schwann cells are modulated by exercise to facilitate nerve regeneration," Journal of Neuroscience, vol. 32, pp. 5002-5009, 2012. [95] A. Arbat-Plana, S. Cobianchi, M. Herrando-Grabulosa, X. Navarro, and E. Udina, "Endogenous modulation of TrkB signaling by treadmill exercise after peripheral nerve injury," Neuroscience, vol. 340, pp. 188-200, 2017. [96] H. Wang, H. Ni, S. Han, W. Xu, J. Wang, B. Yuan, et al., "The Promotion of Neural Regeneration in A Rat Facial Nerve Crush Injury Model Using Collagen-Binding NT-3," Annals of Clinical & Laboratory Science, vol. 46, pp. 578-585, 2016. [97] M. P. Kashyap, C. Roberts, M. Waseem, and P. Tyagi, "Drug Targets in Neurotrophin Signaling in the Central and Peripheral Nervous System," Molecular neurobiology, pp. 1-17, 2018. [98] Y. Nagai, T. Ishijima, and K. Nakajima, "Temporal downregulation of glial cell line-derived neurotrophic factor (GDNF) in the axotomized adult rat facial nucleus," Biomedical Research, vol. 29, pp. 3124-3128, 2018. [99] K. Tajdaran, T. Gordon, M. D. Wood, M. S. Shoichet, and G. H. Borschel, "An engineered biocompatible drug delivery system enhances nerve regeneration after delayed repair," Journal of Biomedical Materials Research Part A, vol. 104, pp. 367-376, 2016. [100] L. M. Marquardt and S. E. Sakiyama-Elbert, "GDNF preconditioning can overcome Schwann cell phenotypic memory," Experimental neurology, vol. 265, pp. 1-7, 2015. [101] A. W. English, W. Meador, and D. I. Carrasco, "Neurotrophin-4/5 is required for the early growth of regenerating axons in peripheral nerves," European Journal of Neuroscience, vol. 21, pp. 2624-2634, 2005. [102] A. W. English, D. Cucoranu, A. Mulligan, J. A. Rodriguez, and M. J. Sabatier, "Neurotrophin‐4/5 is implicated in the enhancement of axon regeneration produced by treadmill training following peripheral nerve injury," European Journal of Neuroscience, vol. 33, pp. 2265-2271, 2011. [103] B. Caballero, S. J. Sherman, and T. Falk, "Insights into the mechanisms involved in protective effects of VEGF-B in dopaminergic neurons," Parkinson’s Disease, vol. 2017, 2017. [104] Z. Ding, J. Cao, Y. Shen, Y. Zou, X. Yang, W. Zhou, et al., "Resveratrol Promotes Nerve Regeneration via Activation of p300 Acetyltransferase-Mediated VEGF Signaling in a Rat Model of Sciatic Nerve Crush Injury," Frontiers in Neuroscience, vol. 12, p. 341. [105] G. Di, X. Zhao, X. Qi, S. Zhang, L. Feng, W. Shi, et al., "VEGF-B promotes recovery of corneal innervations and trophic functions in diabetic mice," Scientific reports, vol. 7, p. 40582, 2017. [106] S. Tang, J. Zhu, Y. Xu, A. P. Xiang, M. H. Jiang, and D. Quan, "The effects of gradients of nerve growth factor immobilized PCLA scaffolds on neurite outgrowth in vitro and peripheral nerve regeneration in rats," Biomaterials, vol. 34, pp. 7086-7096, 2013. [107] N. Mokarram, K. Dymanus, A. Srinivasan, J. G. Lyon, J. Tipton, J. Chu, et al., "Immunoengineering nerve repair," Proceedings of the National Academy of Sciences, vol. 114, pp. E5077-E5084, 2017. [108] M. D. Wood, M. R. MacEwan, A. R. French, A. M. Moore, D. A. Hunter, S. E. Mackinnon, et al., "Fibrin matrices with affinity-based delivery systems and neurotrophic factors promote functional nerve regeneration," Biotechnology and Bioengineering, vol. 106, pp. 970-979, 2010. [109] S. Madduri and B. Gander, "Growth factor delivery systems and repair strategies for damaged peripheral nerves," Journal of Controlled Release, vol. 161, pp. 274-282, 2012. [110] M.-H. Hong, H. J. Hong, H. Pang, H.-J. Lee, S. Yi, and W.-G. Koh, "Controlled Release of Growth Factors from Multilayered Fibrous Scaffold for Functional Recoveries in Crushed Sciatic Nerve," ACS Biomaterials Science & Engineering, vol. 4, pp. 576-586, 2018. [111] C. Kohn-Polster, D. Bhatnagar, D. J. Woloszyn, M. Richtmyer, A. Starke, A. H. Springwald, et al., "Dual-Component Gelatinous Peptide/Reactive Oligomer Formulations as Conduit Material and Luminal Filler for Peripheral Nerve Regeneration," International journal of molecular sciences, vol. 18, p. 1104, 2017.

[112] S. Gnavi, L. Blasio, C. Tonda-Turo, A. Mancardi, L. Primo, G. Ciardelli, et al., "Gelatin-based hydrogel for vascular endothelial growth factor release in peripheral nerve tissue engineering," Journal of Tissue Engineering and Regenerative Medicine, vol. 11, pp. 459-470, 2017. [113] K. Tajdaran, K. Chan, T. Gordon, and G. H. Borschel, "Matrices, scaffolds, and carriers for protein and molecule delivery in peripheral nerve regeneration," Experimental neurology, 2018. [114] W. Zhang, Y. Gao, Y. Zhou, J. Liu, L. Zhang, A. Long, et al., "Localized and Sustained Delivery of Erythropoietin from PLGA Microspheres Promotes Functional Recovery and Nerve Regeneration in Peripheral Nerve Injury," BioMed Research International, vol. 2015, p. 478103. [115] K. Tajdaran, M. S. Shoichet, T. Gordon, and G. H. Borschel, "A novel polymeric drug delivery system for localized and sustained release of tacrolimus (FK506)," Biotechnology and Bioengineering, vol. 112, pp. 1948-1953, 2015. [116] H. Matsumine, R. Sasaki, Y. Tabata, M. Matsui, M. Yamato, T. Okano, et al., "Facial nerve regeneration using basic fibroblast growth factor-impregnated gelatin microspheres in a rat model," Journal of Tissue Engineering and Regenerative Medicine, vol. 10, pp. E559-E567, 2016. [117] B. Chen, S.-p. Niu, Z.-y. Wang, Z.-w. Wang, J.-x. Deng, P.-x. Zhang, et al., "Local administration of icariin contributes to peripheral nerve regeneration and functional recovery," Neural regeneration research, vol. 10, p. 84, 2015. [118] M. J. Barton, J. S. John, M. Clarke, A. Wright, and J. Ekberg, "The glia response after peripheral nerve injury: a comparison between Schwann cells and olfactory ensheathing cells and their uses for neural regenerative therapies," International journal of molecular sciences, vol. 18, p. 287, 2017. [119] F. Zhang, B. Blain, J. Beck, J. Zhang, Z. Chen, Z.-w. Chen, et al., "Autogenous venous graft with one-stage prepared Schwann cells as a conduit for repair of long segmental nerve defects," Journal of reconstructive microsurgery, vol. 18, pp. 295-300, 2002. [120] S. Madduri and B. Gander, "Schwann cell delivery of neurotrophic factors for peripheral nerve regeneration," Journal of the Peripheral Nervous System, vol. 15, pp. 93-103, 2010. [121] H. S. Gambhir, E. Raharjo, J. Forden, R. Kumar, C. Mishra, G. F. Guo, et al., "Improved method to track and precisely count Schwann cells post-transplantation in a peripheral nerve injury model," Journal of neuroscience methods, vol. 273, pp. 86-95, 2016. [122] L. Jiang, S. Jones, and X. Jia, "Stem cell transplantation for peripheral nerve regeneration: Current options and opportunities," International journal of molecular sciences, vol. 18, p. 94, 2017. [123] P. J. Kingham, D. F. Kalbermatten, D. Mahay, S. J. Armstrong, M. Wiberg, and G. Terenghi, "Adipose-derived stem cells differentiate into a Schwann cell phenotype and promote neurite outgrowth in vitro," Experimental neurology, vol. 207, pp. 267-274, 2007. [124] G. Keilhoff, A. Goihl, K. Langnäse, H. Fansa, and G. Wolf, "Transdifferentiation of mesenchymal stem cells into Schwann cell-like myelinating cells," European journal of cell biology, vol. 85, pp. 11-24, 2006. [125] S. Jang, H.-H. Cho, Y.-B. Cho, J.-S. Park, and H.-S. Jeong, "Functional neural differentiation of human adipose tissue-derived stem cells using bFGF and forskolin," BMC cell biology, vol. 11, p. 25, 2010. [126] O. S. Manoukian, S. Stratton, M. R. Arul, J. Moskow, N. Sardashti, X. Yu, et al., "Polymeric ionically conductive composite matrices and electrical stimulation strategies for nerve regeneration: In vitro characterization," Journal of Biomedical Materials Research Part B: Applied Biomaterials, 2018. [127] N. G. Fairbairn, A. M. Meppelink, J. Ng-Glazier, M. A. Randolph, and J. M. Winograd, "Augmenting peripheral nerve regeneration using stem cells: A review of current opinion," World journal of stem cells, vol. 7, p. 11, 2015.

[128] C. Meyer, S. Wrobel, S. Raimondo, S. Rochkind, C. Heimann, A. Shahar, et al., "Peripheral Nerve Regeneration through Hydrogel-Enriched Chitosan Conduits Containing Engineered Schwann Cells for Drug Delivery," Cell Transplantation, vol. 25, pp. 159-182, 2016. [129] R. Alvites, A. Rita Caseiro, S. Santos Pedrosa, M. Vieira Branquinho, G. Ronchi, S. Geuna, et al., "Peripheral nerve injury and axonotmesis: State of the art and recent advances," Cogent Medicine, vol. 5, p. 1466404, 2018/01/01 2018. [130] S. Geuna, "The sciatic nerve injury model in pre-clinical research," Journal of Neuroscience Methods, vol. 243, pp. 39-46, 2015/03/30/ 2015. [131] C. M. Nichols, T. M. Myckatyn, S. R. Rickman, I. K. Fox, T. Hadlock, and S. E. Mackinnon, "Choosing the correct functional assay: A comprehensive assessment of functional tests in the rat," Behavioural Brain Research, vol. 163, pp. 143-158, 2005/09/08/ 2005. [132] M. Bervar, "Video analysis of standing — an alternative footprint analysis to assess functional loss following injury to the rat sciatic nerve," Journal of Neuroscience Methods, vol. 102, pp. 109-116, 2000/10/30/ 2000. [133] J. Du, J. Liu, S. Yao, H. Mao, J. Peng, X. Sun, et al., "Prompt peripheral nerve regeneration induced by a hierarchically aligned fibrin nanofiber hydrogel," Acta Biomaterialia, vol. 55, pp. 296-309, 2017/06/01/ 2017. [134] K. E. A., S. P. K., E. P. E., M. A. V., D. A. Salvatore, S. D. J., et al., "Validity and reliability of the CatWalk system as a static and dynamic gait analysis tool for the assessment of functional nerve recovery in small animal models," Brain and Behavior, vol. 7, p. e00723, 2017. [135] G. M. Hare, P. J. Evans, S. E. Mackinnon, T. J. Best, J. R. Bain, J. P. Szalai, et al., "Walking track analysis: a long-term assessment of peripheral nerve recovery," Plastic and reconstructive surgery, vol. 89, pp. 251-258, 1992/02// 1992. [136] J. R. Bain, S. E. Mackinnon, and D. A. Hunter, "Functional evaluation of complete sciatic, peroneal, and posterior tibial nerve lesions in the rat," Plast Reconstr Surg, vol. 83, pp. 129-38, Jan 1989. [137] F. M. Pestana, R. C. Domingues, J. T. Oliveira, D. F. Durço, C. O. Goulart, H. R. Mendonça, et al., "Comparison of morphological and functional outcomes of mouse sciatic nerve repair with three biodegradable polymer conduits containing poly (lactic acid)," Neural regeneration research, vol. 13, p. 1811, 2018. [138] R. Bamba, D. C. Riley, J. S. Kim, N. L. Cardwell, A. C. Pollins, R. B. Shack, et al., "Evaluation of a nerve fusion technique with polyethylene glycol in a delayed setting after nerve injury," The Journal of hand surgery, vol. 43, pp. 82. e1-82. e7, 2018. [139] D. M. BASSO, M. S. BEATTIE, and J. C. BRESNAHAN, "A Sensitive and Reliable Locomotor Rating Scale for Open Field Testing in Rats," Journal of Neurotrauma, vol. 12, pp. 1-21, 1995. [140] R. A. Barbosa, T. L. G. M. Nunes, T. L. G. M. Nunes, A. O. d. Paixão, R. B. Neto, S. Moura, et al., "Hydroalcoholic extract of red propolis promotes functional recovery and axon repair after sciatic nerve injury in rats," Pharmaceutical Biology, vol. 54, pp. 993-1004, 2016/06/02 2016. [141] A. Golzadeh and R. Mohammadi, "Effect of local administration of platelet-derived growth factor B on functional recovery of peripheral nerve regeneration: A sciatic nerve transection model," Dental research journal, vol. 13, p. 225, 2016. [142] R. L. Wood, K. S. Karlinsey, A. D. Thompson, M. N. Rigby, G. D. Boatright, W. G. Pitt, et al., "Baseline effects of lysophosphatidylcholine and nerve growth factor in a rat model of sciatic nerve regeneration after crush injury," Neural regeneration research, vol. 13, p. 846, 2018. [143] R. Mohammadi, K. Amini, A. Yousefi, M. Abdollahi-Pirbazari, A. Belbasi, and F. Abedi, "Functional Effects of Local Administration of Thyroid Hormone Combined With Chitosan Conduit After Sciatic Nerve Transection in Rats," Journal of Oral and Maxillofacial Surgery, vol. 71, pp. 1763-1776, 2013/10/01/ 2013. [144] S. Ichihara, Y. Inada, A. Nakada, K. Endo, T. Azuma, R. Nakai, et al., "Development of New Nerve Guide Tube for Repair of Long Nerve Defects," Tissue Engineering Part C: Methods, vol. 15, pp. 387-402, 2009.

[145] J. Jin, M. Park, A. Rengarajan, Q. Zhang, S. Limburg, S. K. Joshi, et al., "Functional motor recovery after peripheral nerve repair with an aligned nanofiber tubular conduit in a rat model," Regenerative medicine, vol. 7, pp. 799-806, 2012. [146] C. Tallon, E. Rockenstein, E. Masliah, and M. H. Farah, "Increased BACE1 activity inhibits peripheral nerve regeneration after injury," Neurobiology of Disease, vol. 106, pp. 147-157, 2017/10/01/ 2017. [147] Y. Yao, Y. Cui, Y. Zhao, Z. Xiao, X. Li, S. Han, et al., "Efect of longitudinally oriented collagen conduit combined with nerve growth factor on nerve regeneration after dog sciatic nerve injury," Journal of Biomedical Materials Research Part B: Applied Biomaterials, vol. 106, pp. 2131- 2139, 2018. [148] S. Mikel, A. E., D. D., P. R., S. P., F. N., et al., "Ultrasound-guided plasma rich in growth factors injections and scaffolds hasten motor nerve functional recovery in an ovine model of nerve crush injury," Journal of Tissue Engineering and Regenerative Medicine, vol. 11, pp. 1619-1629, 2017. [149] B. S. Spearman, V. H. Desai, S. Mobini, M. D. McDermott, J. B. Graham, K. J. Otto, et al., "Tissue-Engineered Peripheral Nerve Interfaces," Advanced Functional Materials, vol. 28, p. 1701713, 2018. [150] S. Jones, H. M. Eisenberg, and X. Jia, "Advances and Future Applications of Augmented Peripheral Nerve Regeneration," International Journal of Molecular Sciences, vol. 17, p. 1494. [151] B. He, Z. Zhu, Q. Zhu, X. Zhou, C. Zheng, P. Li, et al., "Factors predicting sensory and motor recovery after the repair of upper limb peripheral nerve injuries," Neural Regeneration Research, vol. 9, pp. 661-672, 02/12/accepted 2014. [152] M. G. Burnett and E. L. Zager, "Pathophysiology of peripheral nerve injury: a brief review," Neurosurgical focus, vol. 16, pp. 1-7, 2004. [153] X. Gu, F. Ding, and D. F. Williams, "Neural tissue engineering options for peripheral nerve regeneration," Biomaterials, vol. 35, pp. 6143-6156, 2014. [154] J. W. Griffin, M. V. Hogan, A. B. Chhabra, and D. N. Deal, "Peripheral nerve repair and reconstruction," The Journal of Bone & Joint Surgery, vol. 95, pp. 2144-2151, 2013. [155] B. J. Pfister, T. Gordon, J. R. Loverde, A. S. Kochar, S. E. Mackinnon, and D. K. Cullen, "Biomedical engineering strategies for peripheral nerve repair: surgical applications, state of the art, and future challenges," Critical Reviews™ in Biomedical Engineering, vol. 39, 2011. [156] X. Jiang, S. H. Lim, H.-Q. Mao, and S. Y. Chew, "Current applications and future perspectives of artificial nerve conduits," Experimental neurology, vol. 223, pp. 86-101, 2010. [157] D. Grinsell and C. Keating, "Peripheral nerve reconstruction after injury: A review of clinical and experimental therapies," BioMed research international, vol. 2014, p. 13, 2014. [158] K. C. Tseng, H. Li, A. Clark, L. Sundem, M. Zuscik, M. Noble, et al., "4‐Aminopyridine promotes functional recovery and remyelination in acute peripheral nerve injury," EMBO molecular medicine, vol. 8, pp. 1409-1420, 2016. [159] K. L. Ljungquist, P. Martineau, and C. Allan, "Radial nerve injuries," The Journal of hand surgery, vol. 40, pp. 166-172, 2015. [160] J. L. Salzer, A. K. Williams, L. Glaser, and R. P. Bunge, "Studies of Schwann cell proliferation. II. Characterization of the stimulation and specificity of the response to a neurite membrane fraction," The Journal of cell biology, vol. 84, pp. 753-766, 1980. [161] Y.-J. Son and W. J. Thompson, "Schwann cell processes guide regeneration of peripheral axons," Neuron, vol. 14, pp. 125-132, 1995. [162] D. R. Friedlander, M. Grumet, and G. M. Edelman, "Nerve growth factor enhances expression of neuron-glia cell adhesion molecule in PC12 cells," The Journal of cell biology, vol. 102, pp. 413- 419, 1986. [163] B. Seilheimer and M. Schachner, "Regulation of neural cell adhesion molecule expression on cultured mouse Schwann cells by nerve growth factor," The EMBO Journal, vol. 6, p. 1611, 1987.

[164] A. M. Moore, M. MacEwan, K. B. Santosa, K. E. Chenard, W. Z. Ray, D. A. Hunter, et al., "Acellular nerve allografts in peripheral nerve regeneration: a comparative study," Muscle & nerve, vol. 44, pp. 221-234, 2011. [165] M. Sakuma, I. R. Minev, S. Gribi, B. Singh, C. J. Woolf, and S. P. Lacour, "Chronic electrical nerve stimulation as a therapeutic intervention for peripheral nerve repair," Bioelectron Med, vol. 2, pp. 43-48, 2015. [166] H. B. Jensen, M. Ravnborg, U. Dalgas, and E. Stenager, "4-Aminopyridine for symptomatic treatment of multiple sclerosis: a systematic review," Therapeutic advances in neurological disorders, vol. 7, pp. 97-113, 2014. [167] S. I. Judge and C. T. Bever, "Potassium channel blockers in multiple sclerosis: neuronal K v channels and effects of symptomatic treatment," Pharmacology & therapeutics, vol. 111, pp. 224- 259, 2006. [168] R. Sherratt, H. Bostock, and T. Sears, "Effects of 4-aminopyridine on normal and demyelinated mammalian nerve fibres," Nature, vol. 283, pp. 570-572, 1980. [169] F. A. Davis, D. Stefoski, and J. Rush, "Orally administered 4‐aminopyridine improves clinical signs in multiple sclerosis," Annals of neurology, vol. 27, pp. 186-192, 1990. [170] R. R. Hansebout, A. R. Blight, S. Fawcett, and K. Reddy, "4-Aminopyridine in chronic spinal cord injury: a controlled, double-blind, crossover study in eight patients," Journal of neurotrauma, vol. 10, pp. 1-18, 1993. [171] H. Lundh, O. Nilsson, and I. Rosen, "Effects of 4-aminopyridine in myasthenia gravis," Journal of Neurology, Neurosurgery & Psychiatry, vol. 42, pp. 171-175, 1979. [172] S. Agoston, T. Van Weerden, P. Westra, and A. Broekert, "Effects of 4-aminopyridine in Eaton Lambert syndrome," BJA: British Journal of Anaesthesia, vol. 50, pp. 383-385, 1978. [173] G. Wilson and S. Chiu, "Potassium channel regulation in Schwann cells during early developmental myelinogenesis," Journal of Neuroscience, vol. 10, pp. 1615-1625, 1990. [174] B. W. Connors, B. R. Ransom, D. M. Kunis, and M. J. Gutnick, "Activity-dependent K+ accumulation in the developing rat optic nerve," Science, vol. 216, pp. 1341-1343, 1982. [175] S. Islam, M. R. Bhuiyan, and M. Islam, "Chitin and chitosan: structure, properties and applications in biomedical engineering," Journal of Polymers and the Environment, vol. 25, pp. 854-866, 2017. [176] O. S. Manoukian, A. Aravamudhan, P. Lee, M. Arul, X. Yu, S. Rudraiah, et al., "Spiral Layer-by- Layer Micro-Nanostructured Scaffolds for Bone Tissue Engineering," ACS Biomaterials Science & Engineering, 2018. [177] O. S. Manoukian, J. Nip, O. Abiola, A. Gopalakrishna, and S. G. Kumbar, "Tissue-Engineered Medical Products," Biomaterials and Nanotechnology for Tissue Engineering, p. 289, 2016. [178] R. C. F. Cheung, T. B. Ng, J. H. Wong, and W. Y. Chan, "Chitosan: An update on potential biomedical and pharmaceutical applications," Marine drugs, vol. 13, pp. 5156-5186, 2015. [179] T. Dai, M. Tanaka, Y.-Y. Huang, and M. R. Hamblin, "Chitosan preparations for wounds and burns: antimicrobial and wound-healing effects," Expert review of anti-infective therapy, vol. 9, pp. 857-879, 2011. [180] O. Manoukian, A. Ahmad, C. Marin, R. James, A. Mazzocca, and S. Kumbar, "Bioactive nanofiber dressings for wound healing," in Wound Healing Biomaterials, ed: Elsevier, 2016, pp. 451-481. [181] Y. Luo and Q. Wang, "Recent development of chitosan-based polyelectrolyte complexes with natural polysaccharides for drug delivery," International journal of biological macromolecules, vol. 64, pp. 353-367, 2014. [182] K. Nawrotek, M. Tylman, K. Rudnicka, J. Balcerzak, and K. Kamiński, "Chitosan-based hydrogel implants enriched with calcium ions intended for peripheral nervous tissue regeneration," Carbohydrate polymers, vol. 136, pp. 764-771, 2016.

[183] G. Haipeng, Z. Yinghui, L. Jianchun, G. Yandao, Z. Nanming, and Z. Xiufang, "Studies on nerve cell affinity of chitosan-derived materials," Journal of biomedical materials research, vol. 52, pp. 285-295, 2000. [184] M. Patel, P. J. Vandevord, H. Matthew, B. Wu, S. Desilva, and P. H. Wooley, "Video-gait analysis of functional recovery of nerve repaired with chitosan nerve guides," Tissue engineering, vol. 12, pp. 3189-3199, 2006. [185] Y. Yuan, P. Zhang, Y. Yang, X. Wang, and X. Gu, "The interaction of Schwann cells with chitosan membranes and fibers in vitro," Biomaterials, vol. 25, pp. 4273-4278, 2004. [186] Y. Lvov, W. Wang, L. Zhang, and R. Fakhrullin, "Halloysite clay nanotubes for loading and sustained release of functional compounds," Advanced Materials, vol. 28, pp. 1227-1250, 2016. [187] R. Yendluri, D. P. Otto, M. M. De Villiers, V. Vinokurov, and Y. M. Lvov, "Application of halloysite clay nanotubes as a pharmaceutical excipient," International journal of pharmaceutics, vol. 521, pp. 267-273, 2017. [188] S. Stokols and M. H. Tuszynski, "The fabrication and characterization of linearly oriented nerve guidance scaffolds for spinal cord injury," Biomaterials, vol. 25, pp. 5839-5846, 2004. [189] A.-H. Chen, S.-C. Liu, C.-Y. Chen, and C.-Y. Chen, "Comparative adsorption of Cu (II), Zn (II), and Pb (II) ions in aqueous solution on the crosslinked chitosan with epichlorohydrin," Journal of Hazardous materials, vol. 154, pp. 184-191, 2008. [190] Y. Wei, S. Hudson, J. Mayer, and D. Kaplan, "The crosslinking of chitosan fibers," Journal of Polymer Science Part A: Polymer Chemistry, vol. 30, pp. 2187-2193, 1992. [191] W.-C. Hsieh, C.-P. Chang, and S.-M. Lin, "Morphology and characterization of 3D micro-porous structured chitosan scaffolds for tissue engineering," Colloids and Surfaces B: Biointerfaces, vol. 57, pp. 250-255, 2007. [192] Z. Zhang and H. Cui, "Biodegradability and biocompatibility study of poly (chitosan-g-lactic acid) scaffolds," Molecules, vol. 17, pp. 3243-3258, 2012. [193] G. Wang, G. Lu, Q. Ao, Y. Gong, and X. Zhang, "Preparation of cross-linked carboxymethyl chitosan for repairing sciatic nerve injury in rats," Biotechnology letters, vol. 32, p. 59, 2010. [194] J. Yang, F. Wang, and T. Tan, "Degradation behavior of hydrogel based on crosslinked poly (aspartic acid)," Journal of applied polymer science, vol. 117, pp. 178-185, 2010. [195] O. S. Manoukian, M. R. Arul, N. Sardashti, T. Stedman, R. James, S. Rudraiah, et al., "Biodegradable polymeric injectable implants for long‐term delivery of contraceptive drugs," Journal of Applied Polymer Science, vol. 135, 2018. [196] S. Vijayakumar and S. Ganesan, "In vitro cytotoxicity assay on gold nanoparticles with different stabilizing agents," Journal of Nanomaterials, vol. 2012, p. 14, 2012. [197] M. Bezuidenhout, P. Liu, S. Singh, M. Kiely, K. M. Ryan, and P. A. Kiely, "Promoting cell proliferation using water dispersible germanium nanowires," PloS one, vol. 9, p. e108006, 2014. [198] A. Burgess, S. Vigneron, E. Brioudes, J.-C. Labbé, T. Lorca, and A. Castro, "Loss of human Greatwall results in G2 arrest and multiple mitotic defects due to deregulation of the cyclin B- Cdc2/PP2A balance," Proceedings of the National Academy of Sciences, vol. 107, pp. 12564- 12569, 2010. [199] R. A. McCloy, S. Rogers, C. E. Caldon, T. Lorca, A. Castro, and A. Burgess, "Partial inhibition of Cdk1 in G2 phase overrides the SAC and decouples mitotic events," Cell cycle, vol. 13, pp. 1400-1412, 2014. [200] A. Bozkurt, G. A. Brook, S. Moellers, F. Lassner, B. Sellhaus, J. Weis, et al., "In vitro assessment of axonal growth using dorsal root ganglia explants in a novel three-dimensional collagen matrix," Tissue engineering, vol. 13, pp. 2971-2979, 2007. [201] A. Bozkurt, R. Deumens, C. Beckmann, L. O. Damink, F. Schügner, I. Heschel, et al., "In vitro cell alignment obtained with a Schwann cell enriched microstructured nerve guide with longitudinal guidance channels," Biomaterials, vol. 30, pp. 169-179, 2009. [202] Y. Hu, G. Du, and N. Chen, "A novel approach for Al2O3/epoxy composites with high strength and thermal conductivity," Composites Science and Technology, vol. 124, pp. 36-43, 2016.

[203] R. Sridharan, R. B. Reilly, and C. T. Buckley, "Decellularized grafts with axially aligned channels for peripheral nerve regeneration," Journal of the mechanical behavior of biomedical materials, vol. 41, pp. 124-135, 2015. [204] S. Y. Chew, R. Mi, A. Hoke, and K. W. Leong, "Aligned Protein–Polymer Composite Fibers Enhance Nerve Regeneration: A Potential Tissue‐Engineering Platform," Advanced functional materials, vol. 17, pp. 1288-1296, 2007. [205] Y. M. Lvov, M. M. DeVilliers, and R. F. Fakhrullin, "The application of halloysite tubule nanoclay in drug delivery," Expert opinion on drug delivery, vol. 13, pp. 977-986, 2016. [206] N. Nwe, T. Furuike, and H. Tamura, "The mechanical and biological properties of chitosan scaffolds for tissue regeneration templates are significantly enhanced by chitosan from gongronella butleri," Materials, vol. 2, pp. 374-398, 2009. [207] S.-i. Aiba, "Studies on chitosan: 4. Lysozymic hydrolysis of partially N-acetylated chitosans," International journal of biological macromolecules, vol. 14, pp. 225-228, 1992. [208] S. C. Basak, K. S. Kumar, and M. Ramalingam, "Design and release characteristics of sustained release tablet containing metformin HCl," Revista Brasileira de Ciências Farmacêuticas, vol. 44, pp. 477-483, 2008. [209] G. Fotakis and J. A. Timbrell, "In vitro cytotoxicity assays: comparison of LDH, neutral red, MTT and protein assay in hepatoma cell lines following exposure to cadmium chloride," Toxicology letters, vol. 160, pp. 171-177, 2006. [210] S. Chiu and G. Wilson, "The role of potassium channels in Schwann cell proliferation in Wallerian degeneration of explant rabbit sciatic nerves," The Journal of Physiology, vol. 408, pp. 199-222, 1989. [211] R. C. Goy, D. d. Britto, and O. B. Assis, "A review of the antimicrobial activity of chitosan," Polímeros, vol. 19, pp. 241-247, 2009. [212] Q. Wang, Smart Materials for Tissue Engineering: Fundamental Principles vol. 24: Royal Society of Chemistry, 2016. [213] A. Cooper, N. Bhattarai, and M. Zhang, "Fabrication and cellular compatibility of aligned chitosan–PCL fibers for nerve tissue regeneration," Carbohydrate Polymers, vol. 85, pp. 149-156, 2011. [214] X. Gu, F. Ding, Y. Yang, and J. Liu, "Construction of tissue engineered nerve grafts and their application in peripheral nerve regeneration," Progress in neurobiology, vol. 93, pp. 204-230, 2011. [215] T. S. Gaaz, A. B. Sulong, M. N. Akhtar, A. A. H. Kadhum, A. B. Mohamad, and A. A. Al- Amiery, "Properties and applications of polyvinyl alcohol, halloysite nanotubes and their nanocomposites," Molecules, vol. 20, pp. 22833-22847, 2015. [216] K. Fujii, A. N. Nakagaito, H. Takagi, and D. Yonekura, "Sulfuric acid treatment of halloysite nanoclay to improve the mechanical properties of PVA/halloysite transparent composite films," Composite Interfaces, vol. 21, pp. 319-327, 2014. [217] W.-J. Li, R. L. Mauck, J. A. Cooper, X. Yuan, and R. S. Tuan, "Engineering controllable anisotropy in electrospun biodegradable nanofibrous scaffolds for musculoskeletal tissue engineering," Journal of biomechanics, vol. 40, pp. 1686-1693, 2007. [218] M. R. Wrobel and H. G. Sundararaghavan, "Directed migration in neural tissue engineering," Tissue Engineering Part B: Reviews, vol. 20, pp. 93-105, 2013. [219] S. R. Caliari and B. A. Harley, "The effect of anisotropic collagen-GAG scaffolds and growth factor supplementation on tendon cell recruitment, alignment, and metabolic activity," Biomaterials, vol. 32, pp. 5330-5340, 2011. [220] M. Beppu, R. Vieira, C. Aimoli, and C. Santana, "Crosslinking of chitosan membranes using glutaraldehyde: Effect on ion permeability and water absorption," Journal of membrane science, vol. 301, pp. 126-130, 2007. [221] T. Takigawa and Y. Endo, "Effects of glutaraldehyde exposure on human health," Journal of occupational health, vol. 48, pp. 75-87, 2006.

[222] E. S. Anton, G. Weskamp, L. F. Reichardt, and W. D. Matthew, "Nerve growth factor and its low-affinity receptor promote Schwann cell migration," Proceedings of the National Academy of Sciences, vol. 91, pp. 2795-2799, 1994. [223] R. M. Lindsay, "Nerve growth factors (NGF, BDNF) enhance axonal regeneration but are not required for survival of adult sensory neurons," Journal of Neuroscience, vol. 8, pp. 2394-2405, 1988. [224] J. M. Cosgaya, J. R. Chan, and E. M. Shooter, "The neurotrophin receptor p75NTR as a positive modulator of myelination," Science, vol. 298, pp. 1245-1248, 2002. [225] J. Y. Zhang, X. G. Luo, C. J. Xian, Z. H. Liu, and X. F. Zhou, "Endogenous BDNF is required for myelination and regeneration of injured sciatic nerve in rodents," European Journal of Neuroscience, vol. 12, pp. 4171-4180, 2000. [226] R. Mirsky and K. R. Jessen, "Schwann cell development, differentiation and myelination," Current opinion in neurobiology, vol. 6, pp. 89-96, 1996. [227] K. R. Jessen and R. Mirsky, "Schwann cells: early lineage, regulation of proliferation and control of myelin formation," Current opinion in neurobiology, vol. 2, pp. 575-581, 1992. [228] M. S. Firouzi, M. Firouzi, M. H. Nabian, L. O. Zanjani, S. A. Zadegan, R. S. Kamrani, et al., "The effects of picric acid (2, 4, 6-trinitrophenol) and a bite-deterrent chemical (denatonium benzoate) on autotomy in rats after peripheral nerve lesion," Lab animal, vol. 44, p. 141, 2015. [229] R. A. Weber, W. H. Proctor, M. R. Warner, and C. N. Verheyden, "Autotomy and the sciatic functional index," Microsurgery, vol. 14, pp. 323-327, 1993. [230] J. Chang and G. Gupta, Tissue engineering for the hand: research advances and clinical applications: World Scientific, 2010. [231] C. Murray, Amputation, prosthesis use, and phantom limb pain: An interdisciplinary perspective vol. 14: Springer, 2009. [232] M. Carr, T. Best, S. Mackinnon, and P. Evans, "Strain differences in autotomy in rats undergoing sciatic nerve transection or repair," Annals of plastic surgery, vol. 28, pp. 538-544, 1992. [233] X. Wang, J. Gong, Z. Gui, T. Hu, and X. Xu, "Halloysite nanotubes‐induced Al accumulation and oxidative damage in liver of mice after 30‐day repeated oral administration," Environmental toxicology, vol. 33, pp. 623-630, 2018. [234] G. I. Fakhrullina, F. S. Akhatova, Y. M. Lvov, and R. F. Fakhrullin, "Toxicity of halloysite clay nanotubes in vivo: a Caenorhabditis elegans study," Environmental Science: Nano, vol. 2, pp. 54- 59, 2015. [235] G. Lundborg, "A 25-year perspective of peripheral nerve surgery: evolving neuroscientific concepts and clinical significance," The Journal of hand surgery, vol. 25, pp. 391-414, 2000. [236] E. A. Naumenko, I. D. Guryanov, R. Yendluri, Y. M. Lvov, and R. F. Fakhrullin, "Clay nanotube–biopolymer composite scaffolds for tissue engineering," Nanoscale, vol. 8, pp. 7257- 7271, 2016. [237] P. N. Hoffman, S. C. Pollock, and G. G. Striph, "Altered gene expression after optic nerve transection: reduced neurofilament expression as a general response to axonal injury," Experimental neurology, vol. 119, pp. 32-36, 1993. [238] S. Mobini, B. S. Spearman, C. S. Lacko, and C. E. Schmidt, "Recent advances in strategies for peripheral nerve tissue engineering," Current Opinion in Biomedical Engineering, vol. 4, pp. 134- 142, 2017. [239] B. N. Johnson, K. Z. Lancaster, I. B. Hogue, F. Meng, Y. L. Kong, L. W. Enquist, et al., "Correction: 3D printed nervous system on a chip," Lab on a Chip, vol. 16, pp. 1946-1946, 2016. [240] A. Faroni, S. A. Mobasseri, P. J. Kingham, and A. J. Reid, "Peripheral nerve regeneration: experimental strategies and future perspectives," Advanced drug delivery reviews, vol. 82, pp. 160-167, 2015.