Enhancement of Peripheral Nerve Regeneration with Controlled Release of Glial Cell Line-Derived Neurotrophic Factor (GDNF)

Enhancement of peripheral nerve regeneration with controlled release of glial cell line-derived neurotrophic factor (GDNF)

by Kasra Tajdaran B.Sc., University of Toronto, 2013

A thesis submitted in conformity with the requirements for the degree of master of applied science

Institute for Biomaterials and Biomedical Engineering

University of Toronto

Enhancement of peripheral nerve regeneration with controlled release of glial cell line-derived neurotrophic factor (GDNF) Kasra Tajdaran

Master of Applied Science (MASc)

Institute of Biomaterials and Biomedical Engineering

University of Toronto

2015

Abstract

Nerve injuries cause severe disability. The present investigational drug delivery strategies for enhancing peripheral nerve regeneration after nerve transection are not yet clinically translatable due to lack of efficiency or biocompatibility. We developed a local delivery system using drug-loaded poly(lactic-co-glycolic acid) (PLGA) microspheres (MS) embedded in a fibrin gel. This drug delivery system (DDS) could be applied at the nerve injury site to deliver exogenous glial cell line-derived neurotrophic factor (GDNF) to the regenerating axons. We used our developed DDS to enhance nerve regeneration in clinically applicable models of severe nerve injuries, including cases with delayed nerve repair and with large nerve defects.

Declaration of co-authorship

The original scientific content of the thesis is comprised of two articles that are submitted to peer-reviewed internationally recognized journals. In both cases these contributions were primarily the work of Kasra Tajdaran. The contributions of the co-authors are declared in the following sections in conformity with the requirements for the degree of

Master’s of Applied Science.

iii

Acknowledgement

I would like to express my sincere gratitude to my supervisors, Dr. Gregory Borschel and

Dr. Tessa Gordon for giving me an exceptional opportunity during my graduate studies.

Thank you for your support and mentorship during the last two years. Thank you for teaching me how to better practice scientific approach and critical thinking. Your advice and constructive criticism has taught me how to become a scientist and remain dedicated to my research. I would like to acknowledge my committee members, Dr. Mike Salter and Dr. Molly Shoichet for contributing to this project. Dr. Salter, thank you for your insightful comments and guidance during our meetings. Dr. Shoichet, since my first undergraduate biochemistry course, you have always inspired me and made me passionate about my research. In addition, thank you for allowing to me attend your weekly team meetings anytime I was facing challenges in my project.

I would also like to thank my best friends for the past two years and lab members at the

Borschel lab. Cecilia Alvarez-Veronesi, Joseph Catapano, Mike Willand, Mike Hendry

Steve Kemp, Cameron Chiang and Jennifer Zhang, thank you for your warm welcome and kindness from the first day we met. Cecilia, without your help and teaching me how to do everything just perfectly, I would not have been able to complete my project. I wish you the best of luck in your medical school quest. Joseph, Mike Willand, and Mike

Hendry, thank you for always making me excited to come to the lab. I am very happy that

I had a chance to know you guys and have friends like you who I can always look up to.

Jennifer, your kindness and tremendous help during the past two years has helped me to overcome some of the toughest and busiest periods of my project. You have always helped me out with anything that I’ve asked for and I cannot thank you enough for it.

Cameron Chiang and David Scholl, thank you for helping me with cryo-cutting throughout my project and thanks for always bringing a positive energy to the lab.

I would like to thank members of the Shoichet lab for letting me use their equipment and lab space and providing me with guidance through out my project. Thank you, Mike

Cook for your helpful suggestions on designing the in vitro cell viability assay. Thank you, Anup Tuladhar for helping me with working with the mass spectrometry and lyophilizer devices. Thank you Jackie Obermeyer and Ying Fang Cheng, for helping me with the neurite extension assays.

Finally, I want to thank my family. My parents have always taught me what it means to be hardworking. You have always encouraged me to find my passion in life. Thank you for always providing me with the best advice and support.

Table of contents Abstract ...... ii

Declaration of co-authorship ...... iii

Acknowledgement ...... iv

List of figures ...... viii

List of tables...... xii

1 Introduction ...... 1 1.1 Overview ...... 1 1.2 Anatomy of peripheral nervous system...... 4 1.3 Injuries to peripheral nervous system ...... 10 1.4 Tissue response to peripheral nerve injuries ...... 12 1.4.1 Peripheral nerve regeneration after chronic axotomy ...... 14 1.4.2 Peripheral nerve regeneration after chronic denervation ...... 16 1.5 Current approaches for treating nerve injuries ...... 17 1.5.1 Acellular allografts ...... 18 1.6 Neurotrophic factors support ...... 23 1.7 Local neurotrophic factor delivery to the injured peripheral nerve ...... 28 1.8 Hydrogels in drug delivery applications ...... 30 1.9 Use of polymeric microspheres as a drug vehicle ...... 31 1.9.1 Poly(esters) ...... 32 1.10 Methods of PLGA microsphere synthesis for protein delivery ...... 34 1.11 Summary and research goal ...... 35 1.11.1 Project objective and hypothesis ...... 35 1.11.2 Specific aims ...... 35 1.11.3 Scope of Thesis ...... 36 2 An engineered biocompatible drug delivery system enhances nerve regeneration after delayed repair ...... 37 2.1 Abstract ...... 38 2.2 Introduction ...... 39 2.3 Materials and Methods ...... 41 2.3.1 GDNF encapsulation in PLGA microsphere ...... 41 2.3.2 GDNF microsphere characterization ...... 42 2.3.3 GDNF DDS composite construction and in vitro release ...... 42 2.3.4 Cell seeding and culture ...... 43 2.3.5 Cell viability ...... 44 2.3.6 Experimental animals ...... 44 2.3.7 Experimental design ...... 44

2.3.8 Operative procedure ...... 46 2.3.9 Retrograde labeling of motor neurons (ventral horn cells) & sensory neurons (dorsal root ganglia) ...... 47 2.3.10 Statistical analysis ...... 48 2.4 Results ...... 49 2.4.1 In vitro microsphere characterization and GDNF release from DDS ...... 49 2.4.2 In vitro toxicity assay for fibrin gel based DDS ...... 51 2.4.3 In vivo retrograde labeling of neurons following nerve repair ...... 53 2.5 Discussion ...... 55 2.6 Conclusions ...... 58 2.7 Acknowledgements ...... 59 3 A glial cell line derived neurotrophic factor delivery system enhances nerve regeneration across acellular nerve allografts ...... 59 3.1 Abstract ...... 60 3.2 Introduction ...... 61 3.3 Materials and Methods ...... 63 3.3.1 GDNF encapsulation in PLGA microsphere ...... 64 3.3.2 GDNF microsphere characterization ...... 65 3.3.3 GDNF DDS composite construction and in vitro release ...... 65 3.3.4 Acellular nerve allograft preparation ...... 66 3.3.5 Experimental animals ...... 66 3.3.6 Experimental design ...... 67 3.3.7 Operative procedure ...... 68 3.3.8 Retrograde labeling of motor neurons (ventral horn cells) & sensory neurons (dorsal root ganglia) ...... 70 3.3.9 Histology & morphometric evaluation of nerves ...... 70 3.3.10 Statistical analysis ...... 71 3.4 Results ...... 72 3.4.1 In vitro microsphere characterization and GDNF release from DDS ...... 72 3.4.2 In vivo retrograde labeling of neurons following nerve repair ...... 74 3.4.3 Axon morphology through nerve graft ...... 77 3.4.4 In vivo nerve histology and morphometric measures of regeneration ...... 78 3.5 Discussion ...... 81 3.6 Conclusion ...... 84 3.7 Acknowledgements ...... 85 4 Conclusions and Future Work ...... 86

5 Bibliography...... 91

vii

List of figures

Figure 1-1. The peripheral nervous system has somatic and autonomic divisions. The somatic division carries information from the skin to the brain and from the brain to muscles. The autonomic division regulates involuntary functions, including activity of the heart and smooth muscles in the gut and glands. Figure adapted from (Kandel, 2013)...... 5

Figure 1-2. Peripheral nerve connective tissue anatomy. A peripheral nerve trunk consists of the outer layer, the epineurium, binding fascicles together. Each fascicle is enveloped by the perineurium and contains the endoneurium, which surrounds myelinated and unmyelinated axons...... 6

Figure 1-3. Myelinated axon anatomy. The myelinated axon is surrounded by a series of Schwann cells (SC)s. During development several Schwann cells are positioned along the length of a single axon. Each cell forms a myelin sheath approximately 1 mm long between two nodes of Ranvier. The sheath is formed as the inner tongue of the SC turns around the axon several times, wrapping the axon in layers of membrane...... 8

Figure 1-4. The sequential opening of voltage-gated Na+ and K+ channels generates the action potential. The shape of the action potential and the underlying conductance changes can be calculated from the properties of the voltage-gated Na+ and K+ channels. Figure adapted from (Hodgkin & Huxley, 1952)...... 9

Figure 1-5. Responses to axotomy in the peripheral nervous system (PNS). In the PNS, support cells aid neuronal regeneration. Proliferating Schwann cells (SC)s, macrophages, and monocytes work together to remove myelin debris, release neurotrophins, and lead axons toward their synaptic targets, resulting in restored neuronal function. Figure adapted from (Schmidt & Leach, 2003)...... 14

Figure 1-6. Chronic axotomy and chronic denervation. Chronic axotomy occurs in the proximal stump when neurons are no longer in contact with their target end-organs. Chronic denervation occurs in the distal nerve stump whereas Schwann cells are no longer connected with neuronal cell bodies and viable axons...... 15

Figure 1-7. GDNF receptor system for RET activation. Dimeric GDNF binds to the GDNF family receptor α1 (GFRα1) and brings together two tyrosine kinase RET molecules. The co- receptors are joined and initiate autophosphorylation and downstream signaling, which promotes cell survival, cell proliferation and neurite outgrowth...... 27

viii

Figure 2-1. In vivo experimental design. Experimental groups consisted of rats receiving fibrin gels loaded with GDNF MS (I), or free GDNF (II) at common peroneal (CP) nerve repair site. In 2 additional experimental groups, extra layers of fibrin were applied around the CP nerve repair site before application of the GDNF or empty MS loaded fibrin gel (III, IV). Control groups of rats received fibrin gels with empty MS (V) and no treatment after delayed nerve repair (VI). Animals with immediate nerve transection and coaptation served as the positive control group (VII)...... 45

Figure 2-2. Surgical procedures performed on rats. Two months prior to the repair, the CP nerve was transected (A) and sutured back to the surrounding muscle to prevent regeneration (B). After the initial injury, the nerve ends were repaired with epineural suture repair followed by placement of the drug delivery system (C). The experimental groups were implanted with GDNF- containing microspheres, free GDNF or empty microspheres in fibrin gel at the repair site. In order to investigate the possibility of toxicity associated with placement of the DDS at the nerve injury site, in two additional experimental groups the repair site was first surrounded with an extra layer of fibrin gel and then with the drug delivery system containing GDNF MS or empty MS. Four weeks following nerve repair, the CP nerve was harvested and labeled with retrograde dye 15 mm distally from the repair site (D). CP: common peroneal nerve ...... 48

Figure 2-3. GDNF encapsulation within polymeric microspheres extended the release time course from fibrin gel. (A) GDNF cumulative mass release from fibrin loaded with free GDNF () and GDNF microspheres (). Within the first 8 days, all free GDNF in the fibrin gel was released. Encapsulation of GDNF in MS slowed the release down to 15 days. Data are normalized to the successfully encapsulated GDNF in microspheres and free GDNF in the fibrin gel. (B) The daily mass release profile of GDNF from fibrin gel showed a more sustained release of GDNF from the MS loaded fibrin gel compared to the free GDNF loaded fibrin gel. GDNF content in the release samples was determined using ELISA. (Mean ± standard deviation, n=3 per release study) ...... 50

Figure 2-4. The developed fibrin gel based drug delivery system was biocompatible. A toxicity assay was performed by differentiating PC-12 cells to neural cell type and incubating them in 24-well plates for 48 hours with RPMI 1640 media containing: (A) alcohol as the negative control, (B) no additional substance, and the media released from empty PLGA microspheres in vitro at day 14 (C) and day 28 (D). A cell viability assay was performed in which dead cells were stained red and live cells were stained green. The cell viability for each group was calculated by finding the ratio of live cells to the total number of live and dead cells in each well. PC-12 cells incubated in vitro with the released media samples from the drug delivery system had similar viability to the control cells cultured with normal media (E) demonstrating that the drug delivery system was not toxic. Data (n = 3) represent the mean ± standard deviation. Scale bar: 200 µm...... 52

Figure 2-5. The fibrin gel based drug delivery system did not impair nerve regeneration. Common peroneal (CP) nerve regeneration at 15 mm distal from the 2 months delayed repair site was analyzed by retrograde labeling of the CP neurons 4 weeks post repair. To analyze the number of motor and sensory neurons, ventral horn cells in the spinal cord (A; 50 μm sections, all

sections counted, correction factor = 0.6 (Abercrombie, 1946)) and sensory neurons in the dorsal root ganglia (B; 20-µm sections, every fifth section counted) were counted. The empty MS treated group and the group receiving the drug delivery system with an initial extra layer of fibrin gel around the repair site had similar nerve regeneration compared to the no treatment group. Placement of the drug delivery system around the suture site did not diminish nerve regeneration. Normal uninjured values ± standard errors are represented by the dashed line. Data represent the mean ± standard error of the mean. MS: Microsphere...... 54

Figure 2-6. In vivo local GDNF release from the drug delivery system enhanced nerve regeneration. Retrograde labeling of common peroneal (CP) neurons regenerating their axons was performed 15 mm distal from the repair site 4 weeks following experimental treatment. The GDNF microsphere (MS) drug delivery system promoted significantly more motor (A) and sensory (B) neuron regeneration, similar to that of the immediate repair group, than the groups in which fibrin gels around the CP nerve were loaded with free GDNF or empty MS. Addition of an extra layer of fibrin around the suture site impaired the effectiveness of the drug delivery system due to introducing a diffusion barrier. Normal uninjured values ± standard errors are represented by the dashed line. Data represent the mean ± standard error of the mean. *p < 0.05. MS: Microsphere...... 55

Figure 3-1. In vivo experimental design. Experimental groups consisted of grafts receiving fibrin gels loaded with 2-week release formulation GDNF MS (I), 4-week release formulation GDNF MS (II) at both suture sites. Another experimental group received 2-week release formulation GDNF MS at the proximal suture site and 4-week release formulation GDNF MS at the distal site (III). Control groups received fibrin gel with empty MS (IV), and no treatment after nerve allograft implantation (V). Animals receiving isografts served as the positive control group (VI). Each group contained six wild type rats and two Thy-1 GFP rats, which were included for visualization of axonal regeneration within the implanted nerve graft...... 68

Figure 3-2. Surgical procedures performed on rats. Prior to nerve repair the common peroneal nerve was transected (A) and a 5 mm nerve gap was created (B). The nerve gap was bridged with a 10 mm nerve allograft followed by placement of the drug delivery system (C). Eight weeks following nerve repair, nerve was harvested and labeled with retrograde dye 10 mm distally from the nerve graft implantation site (D)...... 69

Figure 3-3. In vitro release of GDNF from fibrin gels loaded with microspheres. (A) Cumulative mass release of GDNF from “2 week release” formulation of DDS (), and “4 week release” formulation of DDS (). The GDNF encapsulated within the microspheres were completely released during 15 days from the “2-week release” formulation. The period of GDNF release was extended to 28 days with the 4-week release formation of DDS. The data were normalized to the successfully encapsulated GDNF in microspheres. (B) The daily mass release profile of GDNF from both formulations of microspheres confirms sustained release over 15 days for the 2-week release formulation of DDS and over 28 days for the 4-week release formulation of DDS. GDNF content in the release samples was determined using ELISA. (Mean ± standard deviation, n=3 per release study) ...... 73

Figure 3-4. In vivo GDNF release from microspheres embedded in fibril gels made the acellular nerve allografts as effective as the isografts in supporting nerve regeneration. To assess motor and sensory neuron regeneration, retrograde labeling of neurons was performed 10 mm distally from the distal repair site 8 weeks following experimental treatment. The numbers of the fluorescently labeled motoneurons were counted in the spinal cord’s ventral horn sections (A; 50 μm sections, all sections counted, correction factor = 0.6 (Abercrombie, 1946)) and sensory neurons were counted in the dorsal root ganglia sections (B; 20 µm sections, every fifth section counted). The experimental groups receiving fibrin gels loaded with MS containing GDNF had comparable nerve regeneration to the isograft group and showed significantly higher motor (A) and sensory (B) neurons regeneration through nerve allografts compared with the empty MS and no treatment control groups. The period of GDNF release from the drug delivery system did not influence the extent of nerve regeneration. The control groups receiving no treatment and fibrin gels with empty MS had similar number of regenerated neurons, indicating the drug delivery system did not diminish nerve growth. Data represent the mean ± standard error of the mean. Normal uninjured values ± standard error are represented by the dashed line. *p < 0.05. MS: Microsphere ...... 76

Figure 3-5. Axon density within the acellular nerve allograft increased after treatment with fibrin gels loaded with GDNF microspheres. Representative segments of (A) acellular nerve graft with no drug delivery system treatment, (B) with GDNF delivery system treatment, and (C) isograft, 8 weeks post implantation were obtained from the Thy-1 GFP rats (nerves are green). Longitudinal nerve graft sections (30 µm each) indicated that GDNF treatment using the microspheres enhanced allografts’ axons alignments and increased the axon density, to the same extent as the isografts. Scale bar: 300 µm...... 78

Figure 3-6. Treatment of acellular nerve allografts with GDNF delivery system increased myelinated axon regeneration and number of fibers with larger diameter. Light micrographs of nerve cross sections were analyzed in (A) no treatment group, (B) empty microspheres treated group, (C) GDNF MS treated groups, and (D) isograft treated group. Fiber frequency distribution (E) revealed similar fiber distribution for the GDNF MS and the isograft treated groups. There was a shift to the larger diameter nerve fibers (4 µm - 6 µm) for the GDNF MS and isograft groups compared with the no treatment and empty MS treated groups, which had more of the smaller fibers (2 µm - 4 µm). Histomorphometric analysis of the nerve cross-sections indicated significantly higher number of myelinated axons (E) in GDNF MS and isograft treated groups compared with the no treatment group. No groups exhibited significant differences in myelin thickness (F), fiber diameter (G), and G-ratio (H), but all were below the values of normal uninjured nerves (demonstrated by the horizontal dashed lines). Data represent the mean ± standard error of the mean. *p < 0.05. Scale bars: 10 µm. MS: Microsphere ...... 80

List of tables

Table 1-1. Neural responses to neurotrophic factors...... 24

Table 3-1. PLGA description used in drug delivery system synthesis...... 64

xii

1 Introduction

1.1 Overview

The goal of this work was to evaluate the effects of utilizing biomaterials to locally delivery neuroregenerative molecules in order to enhance nerve regeneration after peripheral nervous system (PNS) injury. Despite the substantial improvement of microsurgical techniques for nerve repair, recovery after nerve injury in the PNS is usually incomplete (Lundborg & Rosen, 2007; Lundborg, 2000; Terenghi & Wiberg,

2003). Delays in treatment further compromise recovery because of the diminished capability of neurons to regenerate their axons to their end-organs (Fu & Gordon, 1995a,

1995b) and reduced neurotrophic factor availability from the progressively denervated

Schwann cells in the growth pathway of the distal nerve stump ( ke et al., 2002; Boyd and Gordon, 2003b). The condition is exacerbated after placement of grafts to bridge nerve gaps, a common practice in human nerve repair. Recently, surgeons have been using acellular nerve allografts (ANAs) to bridge nerve gaps in humans, but these grafts lack neurotrophic factors and therefore do not support regeneration to the same extent as autografts. Given that regenerating nerve fibers preferentially elongate toward sources of neurotrophic factors (Cao & Shoichet, 2001; Perez et al., 1997), replenishing the ANAs with neurotrophic factors would likely enhance nerve gap regeneration. Previously, local application of exogenous neurotrophic factors, including glial cell line-derived neurotrophic factor (GDNF) was shown to enhance axon regeneration after delayed

repair (Boyd & Gordon, 2003a, 2003b; Wood, Gordon, Kemp, et al., 2013; Wood,

Gordon, Kim, et al., 2013).

In this work, we investigated the regenerative effect of sustained and controlled release of

GDNF to an injured nerve in the rat hindlimb after delayed nerve repair and after immediate nerve repair when an ANA placed between the transected nerve stumps. In these studies, we designed a local drug delivery system (DDS) for GDNF controlled release to the nerve injury site using drug-loaded poly(lactic-co-glycolic acid) (PLGA) microspheres (MS) embedded in a fibrin gel. This DDS served to localize the MS around the nerve repair site and ANAs while allowing sustained GDNF release.

The study presented in Chapter 2 evaluated the effects of GDNF local delivery from the

DDS on nerve regeneration after delayed nerve repair. A rat model was used in which neurons were chronically axotomized and the Schwann cells in the distal nerve stump were chronically denervated. Transected stumps of the common peroneal (CP) nerve were each ligated and sutured to surrounding muscle to prevent nerve regeneration. Two months later, the CP nerve stumps were coapted and the GDNF DDS was placed around the repair site in the experimental groups. Four weeks after nerve repair, motor and sensory neuron regeneration was analyzed using a retrograde labeling technique to enumerate those neurons that regenerated their axons into the distal nerve stump. In addition to analyzing the effectiveness of the DDS in vivo, the biocompatibility of the system was evaluated using an in vitro cell viability assay. In this assay, viability of cells cultured with the released media obtained from the DDS in vitro was compared with that of the cells cultured with normal media.

The analysis of the effect of GDNF controlled release to implanted ANAs between freshly coapted rat CP nerve stumps, using the designed DDS, constituted Chapter 3. Rats in the experimental groups received GDNF DDS at both suture sites of the acellular nerve graft. Motor and sensory neuron regeneration was assessed 8 weeks after immediate nerve graft implantation using the same retrograde labeling technique as that used in the first study. In addition, nerve morphology and maturity was assessed with histology and morphometric evaluation of nerve sections.

This introduction chapter provides the background for the experimental studies presented in Chapters 2 and 3. Following, a brief overview of PNS injuries, tissue responses to the nerve injuries and the current treatment and surgical techniques for nerve repair are presented. This information provides the context for consideration of the different types of nerve grafts that have been used including the development of the acellular nerve graft.

As a prelude to our experimental studies that provided exogenous GDNF within an acellular nerve graft, the role of neurotrophic factors in nerve regeneration is also reviewed.

1.2 Anatomy of peripheral nervous system

PNS has somatic and autonomic divisions (Figure 1–1). The somatic division includes motor and sensory neurons. Sensory neurons receive information from the skin, muscles, and joints and they supply the central nervous system (CNS) with a continuous stream of information about both the external and the internal environment of the body. The cell bodies of the sensory neurons lie in the dorsal root ganglia and cranial ganglia. Receptors associated with these cells provide information about muscle and limb position and about touch, pressure and heat at the body surface. During movement, sensory receptors in the muscles and joints are crucial to shaping coherent movement of the body. Motoneurons control the movement of skeletal muscles. The cell bodies of the motoneurons are located in the spinal cord and brainstem. The motoneurons are commonly referred to as the final common pathway of the CNS and their role is to transfer integrated information from the

CNS into movement by their appropriate activation of skeletal muscles in the periphery.

The autonomic division of the peripheral nervous system mediates visceral sensation as well as motor control of the viscera, vascular system, and exocrine glands. It consists of the sympathetic, parasympathetic, and enteric systems. The sympathetic system participates in the body’s response to stress, whereas the parasympathetic system acts to conserve body resources and restore homeostasis.

A peripheral nerve trunk that contains bundles of the sensory and/or motor axons is supported by an outer sheath of loose fibrocollagenous tissue and longitudinal blood vessels within the epineurium that surrounds the entire nerve (Figure 1-2). The internal

epineurium surrounds group of fascicles within the nerve. The fascicles are formed by group of axons surrounded by the perineurium, consisting of many layers of flattened cells, (i.e. fibroblasts), and collagen. The cells within the perineurium have tight junctions between them where the blood-nerve barrier exists (Kim et al., 2006; Weerasuriya &

Mizisin, 2011). The perineurium and epineurium provide tensile strength and it is these layers that are capable of holding sutures (Fox & Mackinnon, 2011; Webber &

Zochodne, 2010). Finally, endoneurium surrounds individual unmyelinated and myelinated axons and their Schwann cells (SCs) sheaths within each fascicle and is composed predominantly of oriented collagen fibers.

The SCs form the myelin sheath around peripheral axons (Figure 1-3), which contains a fatty substance called myelin composed of many different kinds of myelin proteins. SCs also play important role in organizing the formation of the connective tissue sheaths surrounding peripheral nerves in axon regeneration following damage in maturity.

The myelin sheath increases the velocity of action potential conduction by decreasing the capacitance of the SCs (Goldman & Albus, 1968). Ions cannot enter or exit a neuron where the axonal membrane is covered with myelin. Consequently, there is no membrane potential depolarization in the regions where the myelin wraps around the axonal plasma membrane. There are periodic gaps in the myelin sheath called the nodes of Ranvier, which have a high concentration of voltage-gated sodium and potassium channels

(Dugandgija-novakovic, 1995). The nodes of Ranvier facilitate the salutatory conduction in myelinated axons by allowing action potential to jump from node to node (Hodler,

1952).

Figure 1-3. Myelinated axon anatomy. The myelinated axon is surrounded by a series of Schwann cells (SC)s. During development, several SCs are positioned along the length of a single axon. Each cell forms a myelin sheath approximately 1 mm long between two nodes of Ranvier. The sheath is formed as the inner tongue of the SC turns around the axon several times, wrapping the axon in layers of membrane.

The resting axon membrane potential is approximately between -60 mV to -70 mV depending on the species, with the interior of the cell negatively charged with respect to the exterior of the cell. There is a sodium gradient with high sodium concentrations outside of the cell and a potassium gradient with high potassium concentrations inside the cell.

The neural cell body (soma) receives signals from other cells via synaptic contacts on the dendrites, which are slender projections from the soma. Specifically within the motoneurons, the decision by a soma to fire an action potential is determined by the summation of the effects of all the synaptic inputs impinging on the neuron, both excitatory and inhibitory. The action potential is an “all-or-nothing” event (Eyzaguirre &

Kuffler, 1955). It is initiated when the postsynaptic membrane reaches the threshold of

depolarization (about -50 mV) required to open voltage-gated sodium channels. When the channels open, sodium ions flow into the neuron down its concentration gradient, depolarizing that section of the membrane to about +35 mV before inactivating. Sodium voltage gated channels remain inactivated until the membrane potential nears resting values again. Voltage-gated potassium channels also open in response to membrane depolarization; however, they open more slowly than the voltage-gated sodium channels.

Potassium ions exit the neurons down a concentration gradient, the membrane potential returning to negative values, actually overshooting the resting potential to about -90 mV, when they close. At this point the always-functioning potassium leak channels and the

Na+/K+ ATPase pump bring the membrane back to resting potential by pumping potassium into the cell and sodium out of the cell (Figure 1-4).

1.3 Injuries to peripheral nervous system

The majority of peripheral nerve injuries, which include approximately 2.8% of all trauma cases, rarely recover completely (Ashley et al., 2007; Belkas et al, 2004; Noble et al, 1998). Unfortunately, depending on the location and type of injuries, PNS injuries affect patients with lifelong debilitating motor and sensory impairments (Jarvis & Boyce-

Rustay, 2009; Lundborg, 2003; Millesi, 1981). The most common nerve injuries are caused by lacerations and penetrating wounds such as tumor excision, gunshot wounds.

In addition, chronic compression and complete transection due to automotive and sport accidents are common causes of nerve injuries (Kline & Hudson, 1995; Kline, 1990;

Sunderland, 1978; Waldram, 2003).

The first clinical classification scheme of nerve injury was introduced by Seddon in 1943

(Seddon, 1943). This classification includes three categories of nerve fiber injury and the categories are based on the sparing or loss of nerve continuity. Neurapraxia is a nerve compression injury without damage to the axon. Consequently, Wallerian degeneration is not associated with this injury. Axonotmesis occurs after nerve transection with no disruption of the endoneurial tube, perineurium and epineurium. Complete nerve transection with compromised continuity of the epineurium is neurotmesis, which requires surgical reunion of proximal and distal nerve stumps to encourage regeneration into the disrupted endoneurial tubes.

In these three categories of nerve injuries, recovery depends on remyelination, axon regeneration through the original endoneurial pathways without surgical reunion of nerve stumps, or surgical reunion of proximal and distal nerve stumps to encourage regeneration into the disrupted endoneurial tubes, respectively. Intact endoneurial tubes

contain the denervated SCs that guide the regenerating axons back to their former targets.

However, the disruption of these tubes after nerve transection severely compromises the direction taken by the regenerating axons with random reinnervation of former endoneurial tubes and denervated targets (Brushart & Mesulam, 1980; Haftek & Thomas,

1968; Thomas et al, 1987; Young, 1949).

In 1978, Sunderland introduced a second system of nerve injury classification, which extends the Seddon categories to five different degrees of injury severity (Sunderland,

1978). The first- and second-degrees of nerve injury are similar to Seddon’s neurapraxia and axonotmesis, respectively. The third- to fifth degree injuries represent different histological based grades of injury. In the third-degree injuries, endoneurium is disrupted, while perineurium and epineurium remain intact. The forth-degree injuries result in an intact epineurium, but all the internal neural structures are disrupted. Similar to neurotmesis, the fifth degree nerve injuries are followed after complete nerve trunk transection.

Recovery from the first- and second-degree injuries is variable, ranging from poor to complete, depending on the degree of intrafascicular fibrosis. Third- and fourth-degree injuries are associated with spontaneous recovery, although rarely, and usually require surgical intervention to excise the injured area and surgical repair to improve prognosis.

Recovery after the fifth-degree injury is only possible with surgical intervention and treatment.

1.4 Tissue response to peripheral nerve injuries

After nerve transection, the distal portion of the nerve undergoes degeneration due to disconnection from the metabolic resources of the neuronal cell bodies and an increase in protease activity (Figure 1-5). As a result, within the first few days following transection, the axon cytoskeleton and cell membrane start to break down (Hirata & Kawabuchi,

2002; Lopez-Vales, 2008; Vial, 1958). This is the beginning of the Wallerian degeneration process (Waller, 1850). After destruction of the axoplasmic microtubules and neurofilaments, the SCs surrounding the axons in the distal stump shed their myelin lipids and start their phagocytosis activity of clearing myelin axonal debris (Perry, 1995).

In a few days macrophages from the blood vessels surrounding the distal nerve stump enter through the permeabilized blood-nerve barrier into the nerve stump where they make the major contribution to the phagocytosis process during 15-30 days in rat

(Avellino et al., 1995; Martini et al., 2008; Perry et al., 1995). Because the myelin protein debris contains inhibitory molecules for axonal growth (David et al., 1995), the axonal and myelin debris clearance by SCs and macrophages is crucial for successful peripheral nerve regeneration. In addition to clearing myelin debris, SCs and macrophages are known to produce cytokines which play a role in axonal growth (Gordon & Fu, 1997).

Axonal debris is cleared in the distal nerve stump and from the degenerating proximal nerve stump just distal to the first node of Ranvier (Reviewed by Gordon, 2015;

Lundborg, 2004). The SCs undergo cell division and align themselves along the endoneurium as the Bands of Bungner where they guide regenerating axons from the proximal nerve stump (Chan et al,, 2014; Gordon, 2015).

Following axon injury the neuronal cell body undergoes chromatolysis, the morphological changes in the cell that reflect the change in gene expression, now commonly referred to, in motoneurons, as a switch from a signaling mode to a growth mode with up-regulation in genes associated with neurotrophic factors, cytoskeletal proteins and neuropeptides ( Gordon, 1983, 2015). Chromatolysis is characterized by the swelling of the neuronal cell body, shift of the nucleus to the periphery, and disorganization of the basophilic granules (Nissl bodies) The proximal stump of the axotomized neurons undergo atrophy with reduced conduction velocity, recovering only when target contacts are remade by axons that regenerate from the proximal nerve stump

(Gordon & Stein, 1982).

After the membrane is sealed, axon sprouts are emitted from the proximal nerve stump, many of which are aborted (Cajal, 1928; Morris et al., 1972a, 1972b). The axonal outgrowth across a site of nerve transection and microsurgical repair occurs over a protracted period of up to 10 days in a rat (Brushart et al., 2002). Once the axons enter into the distal nerve stump, axonal regeneration proceeds at a rate of about 1-4 mm/day

(Jacobsen & Guth, 1965).

1.4.1 Peripheral nerve regeneration after chronic axotomy

Axotomy is defined as disconnection of the proximal portion of the injured nerves from their target end-organs (Figure 1-6). This condition becomes chronic if there is a delay in the repair of the transected nerve or there is a large nerve gap between proximal and distal nerve stumps. Studies had shown that chronic axotomy is an important factor that has detrimental effects on axonal regeneration and leads to poor functional recovery (Fu

& Gordon, 1995b). Fu and Gordon (1995b) evaluated the effects of progressively long periods of axotomy independent from prolonged denervation of the distal nerve stump.

Using a cross-suture paradigm (Holmes & Young, 1942), in which a proximal nerve

stump was axotomized for various periods of time prior to being cross-sutured to a freshly cut distal nerve stump. After 12 months of chronic axotomy, only 33% of motoneurons were able to successfully regenerate their axons and reinnervate their target muscles. A possible reason for poor nerve regeneration after chronic axotomy may be due to the prolonged deprivation of neurotrophic factors. Boyd and Gordon (2002) demonstrated that local delivery of growth factors to the nerve repair site after chronic axotomy increases the number of motoneurons that regenerated their axons.

Figure 1-6. Chronic axotomy and chronic denervation. Chronic axotomy occurs in the proximal stump when neurons are no longer in contact with their target end-organs. Chronic denervation occurs in the distal nerve stump when the Schwann cells are no longer connected with neuronal cell bodies and viable axons.

1.4.2 Peripheral nerve regeneration after chronic denervation

Prolonged denervation occurs in the distal nerve stump where the distal axon and the surrounding SCs are no longer in contact with neural cell bodies (Figure 1-6). This condition occurs when regeneration is prevented in the distal nerve stump or when the distances over which axons regenerate are considerable. Similar to chronic axotomy, chronic denervation has been associated with negative effects on axonal regeneration (Fu

& Gordon, 1995a). In a study, Fu and Gordon (1995a) investigated the isolated effect of chronic denervation by cross-suturing a freshly transected proximal stump to a progressively denervated distal nerve stump. When denervation was prolonged for about a year, axon regeneration and successful target muscle reinnervation occurred for less than 15% of neurons. The poor nerve regeneration after chronic denervation is related to the deterioration of the growth environment within the distal nerve stump. Such deterioration has been related to SCs atrophy, collagenization and disruption of the basal lamina (Sulaiman & Gordon, 2009; Sunderland & Bradley, 1950; You et al., 1997).

Following nerve injury, SCs gene expression changes to allow them switch to a growth- supportive phenotype (Gordon & Fu, 1997). In such phenotype, there is a transient up- regulation of laminin, neural cell adhesion molecule (N-CAM), and their receptors such as integrin’s β1-subunit (Siironen etl al., 1995). In addition, there is an increase in growth factors and neurotrophins such as GDNF (Höke et al., 2002) and the p75 neurotrophin receptors (You et al., 1997). However, over relatively short periods of time the increase in the gene expression returns to the normal levels (Brushart et al., 2013; Höke et al.,

2006, 2002). Hence, in the case of chronic denervation, the permissive growth environment of the distal nerve stump deteriorates within months.

1.5 Current approaches for treating nerve injuries

The current surgical intervention for peripheral nerve injuries typically consists of microsurgical suture of the aligned proximal and distal stumps. Even though it is possible to use this technique to repair small defects in the nerve, this approach is not suitable for longer nerve gaps, such as following sever nerve trauma or large nerve segment removal during a cancer operation. This is due to the introduction of tension into the nerve cable that is deleterious to nerve regeneration (Millesi, 1981; Sunderland et al., 2004; Terzis et al., 1975). Therefore, an autologous nerve graft (autograft) is frequently used if there is a lack of nerve tissue while performing direct nerve repair (Millesi et al., 1972). During this procedure, the patient’s healthy donor nerves are used as the autograft to bridge the nerve gap. An expandable sensory nerve such as the sural nerve that provides sensation to the lateral skin of the ankle is used for transfer to the surgical site. The advantage of autograft, in addition to it being non-immunogenic and biocompatible, is its capacity to provide a support structure for axon regeneration, cell adhesion and migration. However, the challenge is the limited amount of donor nerve available. In addition, the removal of the nerve for autografting can lead to patients’ dissatisfaction due to loss of sensory or formation of painful neuroma at the donor site. The need for multiple surgical sites can increase the risk of infection (Ehretsman et al., 1999; Mackinnon et al., 1988). The process of nerve harvesting increases the required operation time, which is economically disadvantageous (Macario, 2010).

Therefore, there is a need for substitutes to autografts. Bioengineering strategies in the

PNS have focused on simple and effective methods of developing alternative treatments

to the nerve autografts for repairing large nerve defects, while improving recovery rates and functional outcomes.

1.5.1 Acellular allografts

Non-autologous tissue and extracellular matrix (ECM)-based materials from allogenic and xenogeneic tissues obtained from cadavers and animals are gaining attention due to their large supplies and ease of use that does not require harvest from the patients

(Szynkaruk et al., 2013). However, because these tissues can induce strong immune responses, they must be used in conjunction with permanent systemic immunosuppressants (Evans et al., 1994; Ide et al., 1990; Mackinnon et al., 2001). Many efforts are being made to remove cells and antigens from the allograft tissues. These methods focus on removing the immunogenic cells, while maintaining the ECM components that are essentially similar among species. The obtained biological acellular tissues with their highly organized ECM structure can provide a scaffold component for supporting axonal growth. Currently, both acellular muscle and nerve tissues have been investigated as viable candidates for providing scaffolds.

1.5.1.1 Acellular muscle tissue

The ECM present within the muscle tissues is similar to that of the peripheral nerve tissue. The muscle’s basal lamina arrangement mimics the endoneurial tubes contained in the nerve tissues and it provides collagen type IV, fibronectin and laminin to promote axonal growth (Hall, 1997; Meek & Coert, 2002). Glasby et al. (1986) investigated nerve regeneration through processed muscles grafts in rats (Glasby et al., 1986a, 1986b).

These acellular allografts were prepared by freezing of autogenous muscles followed by thawing in distilled water. Regeneration and myelination of axons were found within the

muscles grafts with coaxial basement membrane tubes along the proximal and distal stumps of the nerve. Nerve regeneration was delayed when the graft’s basement membrane were perpendicular to the axis of the nerve fibers. The authors concluded that the rate of nerve regeneration in the processed muscle allografts depends on the availability of the basement membrane tubes. Even though, the acellular muscle tissue provides promising results for supporting nerve regeneration, the muscle tissue lacks the ability to constrain the regenerating axons within the provided scaffold. Thus, neuroma formation has been observed while using these processed muscle tissues for supporting nerve regeneration (Meek & Coert, 2002; Meek et al., 2001).

1.5.1.2 Acellular nerve tissue

Acellular nerve grafts are conceptually appealing because their physical, chemical, and mechanical properties are similar to those of a nerve autograft. These grafts lack SCs, a vital component to nerve regeneration (Gordon & Fu, 1997; Szynkaruk et al., 2013;

Wood et al., 2011), but they do not induce an immune response (Evans et al., 1994; Ide et al., 1990; Mackinnon et al., 1987). Numerous methods have been used to remove cells and antigens from nerves, as reviewed recently by Szynkaruk et al. (2013). Some of these techniques as described below include cold preservation, repeated freeze-thaw cycles and decellularization with detergents.

Cold preserved nerve allografts

Cold preservation of nerve allografts mainly developed by Mackinnon and colleagues was based on the cold storage of nerve tissue at 37oC or 5oC for various amount of time

(Evans et al., 1995; Evans et al., 1998; Hare et al., 1993; Levi et al., 1994; Moore et al.,

2011). The solution used for cold storage was initially developed at the University of

Wisconsin for renal and liver transplants (Evans et al., 1998). During this process, the cells present on the nerve grafts were not removed while they were killed only at long periods of storage of >4 weeks (Evans et al., 1998). Cold preservation at 5oC sustained the structure of the basal lamina tubes better than preservation at 37oC. However, the 5oC storage resulted in greater immunological activity shown by higher lymphocyte infiltration compared to the 37oC storage method (Evans et al., 1995, 1998,1999; Fox et al., 2005). Cold preservation was intended to provide a low cost method of decellularization while maintaining the required nerve ECM; however, its nerve regeneration capacity was inferior to other decellularization methods. In addition, the cold preserved nerve grafts implantation requires permanent systemic immunosuppression to allow axon regeneration (Grand et al., 2002; Strasberg et al.,

1996).

Freezing and Freeze-thaw techniques

The acellular allografts prepared with the freezing techniques were developed to eliminate the antigen-bearing cells while preserving the extracellular matrix integrity.

Cells within the resulting allografts were killed but not removed from the grafts. There have been two primary methods for the freezing techniques: deep freezing to -70oC

(Schröder & Seiffert, 1970; Singh, 1976; Zalewski & Gulati, 1982) and repeated freeze- thaw cycles during which grafts are repeatedly frozen and thawed five times, freezing for

3 min to -40oC and thawing for 5 min to 2oC (Gulati & Cole, 1994; Gulati, 1988; Ide et al., 1983; Osawa et al., 1987). The freezing techniques preserve the basal lamina as shown by the laminin staining of the processed allografts (Hudson, Zawko, et al., 2004).

There is evidence of using freeze-killed nerve grafts in mouse, rats, rabbits, dogs and

monkeys (Ide et al., 1990). It is difficult to draw conclusions based on these studies, because they used different graft lengths and animals. However, these studies provided important understanding regarding the role of the basal lamina in axon regeneration

(Hudson, Zawko, et al., 2004). Longer processed allografts in higher order species, such as monkeys or rabbits supports better regeneration compared to the lower species, like rats or mice (Szynkaruk et al., 2013). This may be due to the more robust basal lamina tubes in the higher order species, which support axon regeneration better. Possibly the maintenance of the basal lamina within the processed nerve allografts provides the required substrate for ingrowth of SCs and axons.

The freeze-killed nerve grafts had sufficiently low immunological reactions, as the grafts were not rejected in the performed studies. In addition, low lymphocyte infiltration was observed within the grafts after implantation (Gulati, 1998; Hall, 1986; Ide et al., 1983;

Zalewski & Gulati, 1982). The freeze-killed nerve allografts provided better outcomes compared to the cold presentation method; however, they remained inferior to the detergent method, which was developed later (Hudson, Zawko, et al., 2004). In addition, there is a distinct limit to length of axonal growth within the grafts. In rats, axons from transected nerves can regenerate into freeze-killed nerve grafts for a maximum distance of 2 cm (Danielsen et al., 1995; Gulati, 1988; Krekoski, Neubauer, Graham, & Muir,

2002; Nadim et al., 1990)

Chemical detergents

The method of producing cell-free ECM from peripheral nerve with chemical detergents was developed over time and pioneered by Johnson and colleagues (1982). Initially, an extraction-procedure involving sodium deoxycholate and Triton X-100 was used to

eliminate cellular content from human nerves. This protocol was also adapted for processing rat nerve grafts by Sondell (1998). The implanted nerve grafts supported axon regeneration in vivo and they were well tolerated within the peripheral nervous system.

Hudson and colleagues optimized the extraction process by introducing three new detergents including Triton X-200, sulfobetaine-16, and sulfobetaine-10 (Hudson, Liu, &

Schmidt, 2004). The acellular grafts produced using this protocol showed superior regeneration potential, axon density and functional recovery compared to the grafts extracted with Sondell protocol (Hudson, Zawko et al., 2004; Johnson et al., 1982).

Moreover, the grafts prepared with the Hudson protocol resulted in better maintenance of the basal lamina than the grafts processed by Sondell (Hudson, Zawko, et al., 2004).

Neubauer and colleagues investigated the enhancement of nerve regeneration through

ANAs by removing inhibitory chondroitin sulfate proteoglycan (CSPG) (Krekoski,

Neubauer et al., 2001; Neubauer et al., 2007). This inhibitory factor binds laminin and blocks its growth-promoting activity (Muir et al,, 1989; Zuo et al., 1998). Previous studies have shown that removing the inhibitory CSPGs by treating nerve grafts with chondroitinase ABC (ChABC) markedly enhances nerve regeneration (Krekoski,

Neubauer et al., 2001; Neubauer et al., 2007). The present commercially available ANAs have been developed based on the udson’s extraction method and Neubauer’s enzymatic degradation of CSPGs. The combination of these two methods led to AxoGen

Inc.’s clinically available nerve allografts.

Despite the advancement of the cell extraction process and enhancement of the ANAs to better support nerve regeneration, these acellular nerve grafts have not been as successful in restoring functional recovery to the level of that seen with autologous nerve grafts,

especially with increasing gap distances (Nadim et al., 1990). Such limitation of the

ANAs can be related to their lack of specific structural and functional characteristics of normal peripheral nerve, such as viable SCs. These cells have the ability to promote nerve regeneration by producing numerous neurotrophic factors, cytokines, and cell adhesion molecules required for nerve regeneration (Bunge, 1994). Therefore, the optimized tissue engineered ANA would provide a non-immunogenic biological substrate required for axonal growth, while delivering the molecular signals necessary for enhancing nerve regeneration (Wang et al., 2008). To this end, the primary experimental approach in this thesis has been set to supplementing ANAs with neurotrophic factors to promote nerve regeneration.

1.6 Neurotrophic factors support

Neurotrophic factors influence neuronal activity by promoting development and maturation during embryonic life. In addition, the presence of neurotrophic factors plays an important role during peripheral nerve regeneration. The basic neurotrophic factor concept is defined by the hypothesis that the trophic proteins are synthesized in the target tissues, delivered to the neuronal soma via retrograde transport where they exert a trophic and survival effect (Lindsay, 1996; Oppenheim, 1991; Purves, 1986).The various neural responses to different neurotrophic factors are summarized in Table 1-1.

Table 1-1. Neural responses to neurotrophic factors.

Neural response promoted Neurotrophic factors

Motoneuron survival BDNF, NT-3, NT-4/5, CNTF, GDNF

Motoneuron outgrowth BDNF, NT-3, NT-4/5, CNTF, GDNF

Sensory neuron survival NGF, NT-4/5, GDNF

Sensory neuron outgrowth NGF, BDNF, NT-3

Spinal cord regeneration NGF, NT-3, CNTF, FGFs

Peripheral nerve regeneration NGF, NT-3, NT-4/5, CNTF, GDNF, FGFs

Sensory nerve growth across the PNS- NGF, NT-3, GDNF, FGFs CNS transition zone

Abbreviations: Brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3), neurotrophin-4/5 (NT-4/5), ciliary neurotrophic factor (CNTF), glial cell line-derived growth factor (GDNF), nerve growth factor (NGF), acidic and basic fibroblast growth factors (FGFs). Adapted from (Giorgio Terenghi, 1999).

One family of neurotrophic factors important for promoting peripheral nerve regeneration, the neurotrophins, include nerve growth factor (NGF), brain derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3), and neurotrophin-4/5 (NT-4/5).

These proteins all share a low-affinity receptor p75 (Chao et al., 1986), to which they bind with equal affinity. The p75 receptor interact with high affinity tyrosine kinase (trk) receptors (Barbacid, 1994; Chao & Hempstead, 1995). Other important factors, outside of the neurotrophin family are ciliary neurotrophic factor (CNTF), glial cell line-derived neurotrophic factor (GDNF), and acidic and basic fibroblast growth factor (aFGF, bFGF).

In one third of primary sensory neurons instead of trk receptors family expression, these neurons express tyrosine kinase RET (Takaku, 2013). RET is a component of the receptor for GDNF. As a member of transforming growth factor β (TGF-β) superfamily

(Lin et al., 1993), GDNF forms homodimers during its physiological active form (Saarma

& Sariola, 1999). The dimerization is a result of a cysteine knot, formed by three cysteines, which is a characteristic of the TGF-β proteins (Eigenbrot & Gerber, 1997).

Another component of the receptor for GDNF is GFRα-1 (GDNF family receptor α-1).

GDNF exerts its effect and activates its intercellular signaling cascade by mainly binding to the GFRα-1 and RET co-receptors, which in turn activate the sarcoma kinase SRC protein that consequently promotes neural cell survival, cell proliferation and neurite outgrowth (Jing et al., 1996; Takaku, 2013, Figure 1-7).

GDNF is considered one of the most potent neurotrophic factors. For instance, GDNF has been shown to have better motoneuron survival outcomes in neonatal rats as compared to

BDNF and CNTF (Henderson et al., 1994; Yan et al., 1995). In addition, unlike BDNF,

GDNF administration at high doses does not induce inhibitory effects on nerve regeneration (Boyd & Gordon, 2003a; Boyd & Gordon, 2002). Due to binding of BDNF to the low-affinity p75 receptor, high doses of BDNF can causes inhibitory effects, which is revisable by preventing interaction of p75 and BDNF as shown by Boyd and Gordon

(2002).

During loss of axonal contact, in case of nerve injury, denervated SCs in the distal stump of injured peripheral nerves increase their expression of neurotrophic factors (Boyd &

Gordon 2003b). Specifically, at the distal nerve stump, GDNF and GFRα-1 expression increases while levels of RET remains constant (Höke et al., 2002). On the other hand, axotomized neural cell bodies in the spinal cord up-regulate their GFRα-1 and RET receptor expression (Boyd & Gordon 2003b). The high level RET expression in the growth cones emitted from the proximal nerve stump can mediate the GDNF nerve

regeneration effect by interacting with the GFRα-1 and GDNF expressed by the SCs at this distal nerve stump. There is also evidence to suggest that the expressed GDNF potentiates axonal regeneration by increasing level of neuronal CAMS (N-CAM) expression (Friedlander et al. 1986; Seilheimer & Schachner, 1987).

Figure 1-7. GDNF receptor system for RET activation. Dimeric GDNF binds to the GDNF family receptor α1 (GFRα1) and brings together two tyrosine kinase RET molecules. The co-receptors are joined and initiate autophosphorylation and downstream signaling, which promotes cell survival, cell proliferation and neurite outgrowth.

However, the increase in endogenous GDNF level after nerve injury is transient and limited to periods of about 4 to 8 weeks, after which there is a down-regulation of the

GDNF expression (Höke et al., 2006, 2002). Such decrease in the GDNF availability was associated with the diminished capability of neurons to regenerate their axons after chronic axotomy. Throughout various investigations, it has been shown that GDNF administration following chronic axotomy enhances nerve regeneration (Barras et al.,

2009; Barras et al., 2002; Boyd & Gordon, 2003a; Wood et al., 2012; Wood, Gordon,

Kim, et al., 2013).

In this thesis, we focused on administration of GDNF during the decline of endogenous

GDNF after chronic axotomy and large nerve gap repair. However, since continuous systemic administration of GDNF is impractical due to the inadequate bioactivity

(Poduslo & Curran, 1996) and potential side effects and toxicities (Apfel et al., 2000), we designed an alternative strategy for local and sustained release of GDNF to nerve injury site.

1.7 Local neurotrophic factor delivery to the injured peripheral nerve

As discussed previously, systemic administration of neurotrophic factors such as GDNF for promoting nerve regeneration is not a preferred method of drug delivery due to the associated systemic toxicity (Apfel et al., 2000) and non-specific delivery of the neurotrophic agents to tissues other than the injured nerve. Therefore, local delivery is more suitable for clinical purposes.

The main requirement of implantable local delivery systems is maintaining the biological activity of the drug of choice throughout the period of release, specifically for therapeutic

proteins such as GDNF that have short half-lives and are prone to denaturation (Ejstrup et al., 2010). Therefore, various drug reservoirs have been designed to allow release of bioactive proteins by providing controlled release to the targeted tissues. Amongst a variety of techniques, osmotic pumps (Boyd & Gordon, 2003a; Lewin et al., 1997) and silicone reservoirs (Santos, Rodrigo, Hontanilla, & Bilbao, 1998) have been successfully used experimentally. However, these methods are often associated with drawbacks, including device failure and higher potentials for inflammation (Aprili et al., 2009) and even nerve damage due to fibrotic foreign body response in the adjacent injured nerve due to the non-degradable components (Guilhem et al., 2009). In addition, the osmotic pump and catheter systems require a secondary surgery for removal with potential complications (Maysinger & Morinville, 1997).

Another example of local delivery devices that use polymeric matrices is the nerve guidance conduit (NGC). This tubular conduit has been primarily used for the guidance of axons through nerve gaps by isolating the nerve regenerative environment (Pfister et al., 2007). Ethylene vinyl acetate (EVA) (Bloch et al., 2001; Fine et al., 2002), fibronectin (Sterne, Brown, Green, & Terenghi, 1997) and chitosan (Kim et al., 2008) are examples of polymeric nerve guidance channel materials. In addition, the NGCs made of these materials have been successfully modified to deliver growth factors to the regenerating peripheral nerve. By incorporating growth factors into the conduit wall, the

NGC itself becomes a delivery device (Kokaiet al., 2011; Moore et al., 2010; Wood et al.,

2009). Even though, NGCs delivering neurotrophic factors provide many advantages in repairing large nerve gaps, they have not become a standard for clinical use due to their limitations in supporting regeneration of long nerve gaps. In addition, many non-

biodegradable NGCs were not as effective as other drug delivery devices because of their lack of compatibility. There are reports of implanted channels becoming disconnected due to scratching (Barras et al., 2002) and of failing to support nerve regeneration to the same extent as biological iso- and autografts (Moore et al. 2010; Barras et al. 2002).

Therefore, biodegradable polymers and hydrogel matrices, as described below, are effective delivery methods that overcome these challenges.

1.8 Hydrogels in drug delivery applications

Over the past few decades, advances in hydrogel technologies have spurred development in many biomedical applications. Since the establishment of the first synthetic hydrogels by Wichterle and Lim in 1960 (Wichterle et al., 1960), the growth of hydrogel technologies has advanced many fields ranging from pharmaceuticals (Kashyap, Kumar,

& Kumar, 2005) to biomedical implants (Corkhill et al., 1989). Hydrogels provide an ideal scaffold for tissue engineering and drug delivery systems as they are typically biocompatible, biodegradable, and have tunable mechanical properties to match that of the ECM. Moreover, hydrogels can act as a suitable drug delivery system since they lack strong hydrophobic interaction that can disturb the structure of protein molecules (Lin &

Metters, 2006). Naturally derived hydrogels, such as fibrin, collagen, chitosan, hyaluronic acid (HA), and cellulose derivatives such as methylcellulose and hydroxypropyl methylcellulose have been used extensively (Biondi etl al. 2008) along with chemically synthesized hydrogels, like polyethylene glycol (PEG) and polyvinyl alcohol (PVA)

(Peppas et al., 2000). These polymeric hydrogels have received much attention over the past decades as suitable drug delivery systems for the CNS and PNS (Biondi et al., 2008).

Fibrin gel produced from cross-linked fibrinogen and thrombin with factor XIIIa, forms naturally in the wound healing process and has been used as a tissue sealant (Amrani et al., 2001; Dunn & Goa, 1999) and a growth factor delivery vehicle for nerve repair

(Sakiyama-Elbert & Hubbell, 2000; Wood, Gordon, Kim, et al., 2013; Yin et al., 2001).

Like some of the other naturally derived polymers, fibrin supports cell adhesion and growth and its physical properties can be tuned by the fibrinogen/thrombin formulation in the design of a material well tolerated within physiological environment (Willerth et al.,

2006).

However, for drug delivery purposes natural hydrogels such as fibrin alone cannot sustain delivery of therapeutic proteins over a 7-day period (Wood, Gordon, Kim, et al., 2013) because the majority of the hydrogel structure constitutes water, and lacks the ability to impose a significant barrier against diffusion of embedded therapeutic agents. This condition is exacerbated in the case of using hydrophilic proteins such as GDNF, since they can rapidly diffuse out of the hydrogel network. Therefore, another drug delivery vehicle is required to be incorporated into the hydrogel based drug delivery systems to allow a more sustained drug release.

1.9 Use of polymeric microspheres as a drug vehicle

Due to the inability of fibrin gel to sustain protein delivery for more than several days, another strategy must be developed to achieve sustained release profiles for therapeutic proteins. One method is to use a composite system of polymeric microspheres in the fibrin gel to control the time course of therapeutic proteins delivery. In such a system, fibrin gel localizes the factors delivered at the site of the nerve injury and the microspheres sustain the release.

Polymeric microspheres were first developed in the 1970s when they were initially devised as carriers for vaccines and anticancer drugs (Ravi Kumar, 2000). Since then a large number of polymers have been investigated as candidates for synthesizing these colloidal particles for different drug delivery applications. These polymers are categorized based on their chemical structures that characterize their degradation type and period. The degradation of polymeric microspheres is dependent on the hydrophobicity of the polymer. Hydrophilic polymer chains will absorb water at a faster rate than chain hydrolysis, and consequently the microspheres will degrade by bulk erosion. Alternatively, hydrophobic polymers chains hydrolyze faster than they take up water, thus the synthesized microspheres will degrade through surface erosion. For the purpose of this thesis, we focus on hydrophilic classes of polymers.

1.9.1 Poly(esters)

Poly(esters) are bulk-eroding hydrophilic polymers in which the rate of water uptake exceeds the rate of hydrolytic chain cleavage. Many poly(esters) are biodegradable and can be used to encapsulate a variety of ingredients for biological applications, thus making them one of the largest classes of polymers used in the controlled delivery systems. Examples include radioactive imaging and therapy (Mumper & Jay, 1992), protein and small molecule delivery (Kim & Park, 1999), gene and viral delivery (Li &

Huang, 2004), and cell delivery (Chun et al., 2004).

Among poly(esters), block copolymers of poly(lactic-co-glycolic acid), PLGA, can be considered as one of the most widely used polymers for drug delivery applications, because PLGA is one of the few biodegradable polymers having FDA approval for biological use (Chun et al., 2004; Mundargi et al., 2008). PLGA offers a wide range of

tunable characteristics such as intrinsic viscosity and rate of degradation (Mundargi et al.,

2008) that may be used to encapsulate drugs with a wide range of properties and achieve a variety of release profiles for drugs. Studies have shown that by varying the relative ratios, the chain length of the copolymer blocks, lactic acid (LA) and glycolic acid (GA), the overall molecular weight of the copolymer, and the terminal functional groups on

PLGA, the period of encapsulated drug release can be adjusted from one or two weeks to several months (Pollauf et al., 2005). In addition, the resulting PLGA spheres can range in size from about 200 nm to approximately 600 μm (Cohen-Sela et al., 2009; Moritaet al., 2000), thus allowing PLGA microspheres to be used in various delivery system including the injectable forms.

However, the PLGA containing drug delivery system has its own limitations. The sonication and aqueous/organic interface formed during the emulsion process leads to extensive protein denaturing (Wischke & Schwendeman, 2008). Proteins are amphiphilic, and when they adsorb to the aqueous/organic interface, they assume the structural conformation with the lowest energy. This causes proteins to unfold and lose bioactivity.

Denatured proteins are insoluble in aqueous solutions and consequently, there is incomplete protein release and reduced biological activity. Moreover, during the degradation process of PLGA through hydrolysis of ester bonds, acidic by-products are generated. While these products have not shown toxicity in vitro and in vivo, they contribute to the denaturing of encapsulated proteins (Kang & Singh, 2001).

1.10 Methods of PLGA microsphere synthesis for protein delivery

Two methods for preparing PLGA microspheres have been developed which are the salting out and the emulsion solvent evaporation techniques (Soppimath et al., 2001;

Wischke & Schwendeman, 2008).

The salting-out method involves creation of a viscous gel by dissolving a water-soluble polymer in a high salt-concentration aqueous environment. Then the viscous gel is added to a water-miscible organic solvent containing the therapeutic proteins required to be encapsulated. The salt in the gel phase is able to produce the salting-out of the organic solvent and engender the formation of polymeric microspheres (Soppimath et al., 2001).

The disadvantage of this salting-out method is the presence of organic solvent that can lead to denaturation of the proteins and even toxicity.

In the solvent evaporation technique, which is the method used in the studies in this thesis, microspheres are formed using double emulsion process. This technique involves dissolving therapeutic protein in an inner aqueous phase, which is then added to a solution of PLGA in an organic solvent. This suspension is sonicated to form a primary

(w/o) emulsion. That is then added to an outer aqueous phase containing surfactant and sonicated to produce the double (w/o/w) emulsion. As the organic solvent diffuses into the outer aqueous phase and evaporates, the final solid PLGA microspheres are formed

(Wischke & Schwendeman, 2008).

1.11 Summary and research goal

Addressing the deleterious effects associated with chronic nerve injuries and nerve regeneration through large nerve defects is an ongoing challenge for surgeons.

Supplementing the nerve injury sites with neurotrophic factors holds promise for enhancing nerve regeneration. Toward this goal, we have developed a local drug delivery system that is comprised of GDNF-loaded PLGA microspheres dispersed in fibrin hydrogels. The microspheres offer sustained release of the encased GDNF while the hydrogel localizes the microspheres at the site of nerve injury.

1.11.1 Project objective and hypothesis

The objective of the project is to improve peripheral nerve regeneration by adapting the mass transfer properties of the proposed drug delivery system for the controlled delivery of GDNF. The project’s hypothesis is that localized GDNF delivery using microspheres improves nerve regeneration after delayed nerve repair and through an acellular nerve graft that is placed to bridge a nerve gap.

1.11.2 Specific aims

1. To optimize PNS regeneration in a delayed nerve transection repair model in rats using

GDNF microspheres.

Rationale: A clinically applicable delivery system must deliver neurotrophic factors for sufficient time and in suitable amounts to improve axon regeneration and target reinnervation. A chronic axotomy and denervation model is used in order to simulate the clinical situations in which patients present with nerve injuries that have been occurred months prior to repair. Such an experimental model can assess the effectiveness of the drug delivery system in promoting axon regeneration and target reinnervation. The results

from this specific aim determine the extent to which GDNF release from MSs improves recovery following delayed nerve repair after chronic axotomy of neurons and chronic denervation of distal nerve stumps and end organs.

2. To determine whether delivery of GDNF from microspheres enhances PNS regeneration through an ANA.

Rationale: ANAs do not contain biologically relevant levels of neurotrophic factors, presenting an additional challenge to regenerating axons. Extended delivery of exogenous

GDNF is to be used with an ANA to compensate for the lack of neurotrophic factors in the grafts.

1.11.3 Scope of Thesis

This thesis describes the development of a fibrin gel based drug delivery system for localized and sustained release of GDNF to the injured peripheral nerve. These original contributions are divided into two chapters:

Chapter 2 – The biocompatibility of the fibrin gel based drug delivery system

containing PLGA microsphere was evaluated using in vitro PC-12 cells’ viability

assay. In addition, in order to analyze the effectiveness of the GDNF containing

delivery system in vivo, a delayed CP nerve repair model in rats was used.

Chapter 3 – The developed GDNF delivery system was used to provide

neurotrophic factor support to nerve regeneration after immediate surgical repair

of transected CP nerve through a 10 mm ANAs used to bridge a 5 mm nerve gap.

2 An engineered biocompatible drug delivery system enhances nerve regeneration after delayed repair

Kasra Tajdaran1,2, Tessa Gordon1,4,5, Mathew D. Wood6, Molly S. Shoichet2,3, Gregory H. Borschel1,2,4,5

1Division of Plastic and Reconstructive Surgery, The Hospital for Sick Children 555 University Ave, Toronto, Ontario, Canada M5G1X8; telephone: 416-813-7654 ext. 28342; fax: 416-813-6637; e-mail: [email protected] 2Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, Ontario, Canada 3Department of Chemical Engineering and Applied Chemistry, University of Toronto, Toronto, Ontario, Canada 4Division of Plastic and Reconstructive Surgery, Department of Surgery, University of Toronto, Toronto, Ontario, Canada 5Program in Neuroscience, The Hospital for Sick Children Research Institute, Toronto, Ontario, Canada 6Division of Plastic and Reconstructive Surgery, Washington University School of Medicine, St. Louis, Missouri, USA

Publication Information: Submitted to Journal of Neural Engineering

The authors have no financial disclosures to reveal. The authors have no conflict of interest relevant to the subject of the manuscript. K. Tajdaran conceived, designed, and executed the experiments and wrote the manuscript. M.D. Wood edited the manuscript. M.S. Shoichet, T. Gordon and G.H. Borschel conceived the project and edited the manuscript.

2.1 Abstract

Localized drug delivery strategies could greatly benefit patients with nerve injury and could be easy for surgeons to implement. We developed a local drug delivery system

(DDS) using drug-loaded poly(lactic-co-glycolic acid) (PLGA) microspheres (MS) embedded in a fibrin gel. In an in vitro study, we investigated the biocompatibility of this

DDS by performing a toxicity assay in which we incubated PC-12 cells with the medium released from the DDS in vitro. In an in vivo study, this DDS was applied at the nerve injury site to deliver exogenous glial cell line-derived neurotrophic factor (GDNF) to the regenerating axons after delayed nerve repair. In vitro, PC-12 cells incubated with released media samples from the DDS had similar viability to control cells cultured with normal media, demonstrating that the DDS was not toxic. In vivo, the numbers of motor and sensory neurons that regenerated their axons with empty MS treatment were the same as when there was no MS treatment. The DDS increased the numbers of regenerating motor- and sensory neurons to levels indistinguishable from those observed with immediate nerve repair. The DDS increased neuron regeneration to levels double those observed with negative control groups. This biocompatible, non-toxic, fibrin gel-based

DDS enhances outcomes following severe peripheral nerve injuries.

Keywords: nerve injury, chronic axotomy; chronic denervation, glial cell line-derived neurotrophic factor, drug delivery, regenerative medicine, biomaterials.

2.2 Introduction

Peripheral nerve injuries cause severe disability, leaving patients with pain, paralysis and loss of function (Ciaramitaro et al., 2010; Kroll, Caplan, & Posner, 1990; Lundborg,

2000). Immediately following nerve injury, the expression of neurotrophic factors, such as glial cell-line derived neurotrophic factor (GDNF), and their receptors are transiently upregulated in Schwann cells and neurons before returning to basal levels (Boyd &

Gordon, 2003a). Nerves are capable of limited regeneration when the proximal nerve stump and distal end of a severed nerve are sutured together immediately

(Kouyoumdjian, 2006; Millesi, 1985). However, the clinical outcomes often remain unsatisfactory (Lundborg & Rosen, 2007; Lundborg, 2000), especially when nerve repairs are delayed beyond 1 month (Aydin MA, Mackinnon SE, Gu XM, Kobayashi J,

2004; Lundborg & Rosen, 2007; Lundborg, 2000; G Terenghi & Wiberg, 2003). Delays in nerve repair lead to a state of chronic axotomy and chronic denervation, as neurons are no longer connected to their end-organ targets (chronic axotomy) and Schwann cells are no longer in contact with axonal processes (chronic denervation) (Fu & Gordon, 1995a,

1995b). Reduced expression of neurotrophic factors and their receptors is in part responsible for poor nerve regeneration after delayed nerve repair (Boyd & Gordon,

2003a; Höke et al., 2002; You et al., 1997).

Exogenous sources of the neurotrophic factors can enhance recovery following chronic axotomy (Boyd & Gordon, 2003a; Höke et al., 2002). Conventional systemic delivery strategies, such as intravenous infusion, are relatively simple and non-invasive, but the high doses that may be required for successful outcomes, can result in systemic toxicity.

Previously, we showed that local and sustained application of exogenous GDNF using a

silicone tube and mini-osmotic pump enhances axon regeneration after delayed nerve repair (Boyd & Gordon, 2003a, 2003b; Wood, Gordon, Kemp, et al., 2013). Even though local infusions via the osmotic pumps provide sustained delivery of the biomolecules and thus enhance efficacy, pumps can cause infections and foreign body reactions adjacent to the nerve injury site, limiting their clinical potential (Follett et al., 2004; Guilhem et al.,

2009). Moreover, use of osmotic pumps requires a secondary surgery for removal of the system, which can impose further complications to the clinical outcomes. Ideally, alternative devices that provide localized and extended drug delivery to the nerve injury site, without risk of chronic nerve compression secondary to capsular fibrosis, and are biodegradable would avoid these other problems.

Fibrin is an advantageous drug delivery system because its mechanical properties can be adjusted to match the needs of peripheral nervous system tissue. Due to its biodegradable properties, it can prevent capsular fibrosis (Sameem, Wood, & Bain, 2011). In addition, fibrin gel has been widely used as sealant by surgeons performing nerve repair (Jubran &

Widenfalk, 2003; Sameem et al., 2011; Wood, Gordon, Kemp, et al., 2013). The challenges of utilizing fibrin gel as a drug reservoir however, include maintaining a sustained release period and achieving a dose high enough to induce significant repair and functional recovery.

In previous studies, we developed a fibrin gel based drug delivery system (DDS) with extended GDNF release to the injured nerve from days to weeks (Wood, Gordon, Kemp, et al., 2013; Wood, Gordon, Kim, et al., 2013). In order to attenuate the rapid rate of

GDNF release and to obtain a more sustained release for 15 days, GDNF was encapsulated in poly(lactic-co-glycolic acid) (PLGA) microspheres (MS) and the drug

containing PLGA MS were incorporated in the fibrin gel. In this study, we analyzed the biocompatibility of the PLGA MS containing fibrin gel based DDS for the first time using in vitro experiments and further analyzed the effect of the DDS on nerve regeneration in an in vivo delayed nerve repair model of nerve injury in rats.

2.3 Materials and Methods

All chemicals were obtained from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise specified.

2.3.1 GDNF encapsulation in PLGA microsphere

Glial cell line-derived neurotrophic factor (GDNF) was encapsulated in poly(lactic-co- glycolic acid) (PLGA) microspheres (MS) using a water/oil/water double emulsion, solvent evaporation method. Briefly, an inner aqueous solution of 100 µl consisting of

250 µg GDNF (Peprotech, Rocky Hill, NJ) and 12.5 mg heparin was mixed with an oil phase consisting of 230 mg PLGA 50/50 (Wako, Japan) and 12.5 mg MgCO3 in 1 mL dichloromethane (DCM)/acetone (75%/25%). The mixture was sonicated for 45 s using a

3 mm probe sonicator (Vibra-Cell™ VCX 130; Sonics and Materials, CT, USA) at 30% power. The resulting emulsion was added to 25 mL of 2.5% aqueous poly(vinyl alcohol)

(PVA) solution containing 10% NaCl and homogenized at 6,000 rpm for 60 s. The secondary emulsion was then added to 250 mL aqueous solution of 2.5% PVA and 10%

NaCl. The mixture was stirred for 3 hours with venting to allow the hardening of the microspheres by complete evaporation of the organic solvent. The hardened microspheres were collected and washed by centrifugation, lyophilized, and stored at −20°C until use.

2.3.2 GDNF microsphere characterization

The mean diameter and size distribution of the GDNF MS were measured via static light scattering using a Malvern Mastersizer 2000 laser diffraction particle sizer (Malvern

Instruments Ltd, UK), using refractive indices of 1.33 and 1.59 for water and PLGA, respectively. Encapsulation efficiency was measured by dissolving an appropriate mass of MS in 1 mL dimethyl sulfoxide (DMSO) for 1 hour at 37oC followed by addition of 10 mL of 0.05 M NaOH with 0.5% w/v sodium dodecyl sulfate (SDS) and further shaking for 1 hour at room temperature. The GDNF content of the MS was quantified by an enzyme-linked immunosorption assay (ELISA) for human GDNF according to the manufacturer’s instructions (R&D Systems, Minneapolis, MN). The absorbance was read at 450 nm with an optical subtraction at 540 nm using a multi-well plate spectrophotometer, and sample concentrations were calculated from a standard curve of known GDNF concentrations. Drug loading was determined as the GDNF mass per mg of microspheres; encapsulation efficiency was the measured drug loading of the microspheres divided by the theoretical maximum drug loading.

2.3.3 GDNF DDS composite construction and in vitro release

Fibrin gel (80 µL total volume) consisted of equal parts of fibrinogen (75–115 mg/mL, 40

µL) and thrombin (5 IU/mL, 40 µl) obtained from a Tisseel® glue kit (Baxter Healthcare,

IL, USA); and mixed according to the manufacturer’s instructions. Fibrin gels were loaded with microspheres or GDNF by incorporating 5 mg of microspheres or 5 µg of

GDNF, respectively, into the thrombin solution before it was mixed with fibrinogen to form a gel. In vitro release of GDNF from fibrin loaded with or without MS was performed by constructing 80 µL gels in 2 mL siliconized centrifuge tubes (Fisher

Scientific). The release period was measured by incubating the fibrin gels in 1 mL of phosphate buffered saline (PBS) containing 1% bovine serum albumin (BSA) at 37oC under constant gentle agitation by vortex. The PBS was collected and ELISA assays were performed to measure GDNF quantity collected from the time course release studies.

2.3.4 Cell seeding and culture

Sterile T-25 tissue culture flasks were coated by spreading collagen solution containing 4 mg/ml bovine skin collagen as a film. Flasks were allowed to dry at room temperature in a sterile laminar flow hood. Undifferentiated rat pheochromocytoma (PC-12) cells were seeded on the flasks and were maintained in RPMI 1640 medium containing 10% horse serum and 5% fetal bovine serum supplemented with 2 mM glutamine, 100 units/ml penicillin, and 100mg/ml streptomycin at 37°C incubator with 95% air-5% CO2. PC-12 cells differentiation to neural cell type were induced 24 hours later by treating the cells with 50 ng/ml GDNF in RPMI 1640 containing 1% horse serum for 7 days.

After differentiation, PC-12 cells were washed three times with serum-free RPMI 1640 media and gently triturated with 10 ml of RPMI 1640 media for at least 10 times until all cell were detached from the flask wall. Cells were centrifuged at 500 G and suspended in

RPMI 1640 media. Differentiated PC-12 cells were plated onto collagen coated 24-well plates at density of 2x10 5 cells per well in 1ml of culture media. Each well received 500

µl of experimental samples. Wells receiving 500 µl ethanol were used as the negative control group. Wells with 500 µl release samples from the empty MS drug delivery system (DDS) at day 14 and day 28 served as the experimental groups. In the positive control group, 500 µl RPMI 1640 was added to the wells. There were 3 wells assigned to each group and the study was repeated in 3 separate trials.

2.3.5 Cell viability

After 48 hours of culturing the differentiated PC-12 cells with their corresponding media, cell viability was observed using a LIVE/DEAD® Viability/Cytotoxicity kit (Invitrogen,

Carlsbad, CA, USA). The cells were fluorescently labeled with the calcein AM and ethidium homodimer dyes according to the manufacturer’s protocol. Live cells retained a dye that produced bright green fluorescence (excitation and emission of 495 nm and 515 nm, respectively), while dead cells fluoresced a bright red (excitation of 495 nm and emission of 635 nm). Images were obtained using confocal fluorescence microscopy

(Olympus IX81). PC-12 cell viability in each group was calculated by finding the ratio of the live cells’ number to the total number of cells in each well.

2.3.6 Experimental animals

Adult female Sprague–Dawley rats (Harlan, Indianapolis, USA), each weighing 250–300 g were used in this study. All surgical procedures and perioperative care measures were performed in strict accordance with the National Institutes of Health guidelines, the

Canadian Council on Animal Care (CCAC) and were approved by the Hospital for Sick

Children’s Laboratory Animal Services Committee.

2.3.7 Experimental design

Fifty-six adult female Sprague–Dawley rats were randomized into seven groups (n = 8)

(Figure 2-1). Animals receiving no DDS treatment or fibrin gels loaded with empty MS after delayed nerve repair served as the experimental control groups. Fibrin gels loaded with free GDNF or GDNF containing MS served as the primary experimental groups.

Two additional experimental groups were added to investigate the possibility of toxicity associated with the placement of the DDS at the common peroneal (CP) nerve injury site

in the hindlimb. In these two groups, the repair site was first surrounded with an extra layer of fibrin gel and then with the DDS containing GDNF MS or empty MS. In a positive control group, the nerve was coapted immediately after transection without placement of the DDS.

2.3.8 Operative procedure

All surgical procedures were performed using aseptic techniques. Isoflurane (3%) gas anesthesia was used for animal induction followed by 2.5% isoflurane gas for maintenance. The hind leg of the rat was surgically cleaned with a betadine/alcohol rub.

The sciatic nerve was exposed through a dorsolateral–gluteal muscle splitting incision.

Wounds were irrigated with saline, dried and closed in two layers, utilizing 5-0 Vicryl™

(Ethicon, OH, USA) sutures to close the muscle layers, and 4-0 Nylon sutures to close the skin. Experimental animals were recovered in a warm environment prior to returning to the housing facility.

In the first procedure, the CP nerve was dissected free (Figure 2-2A) and transected 5 mm from the sciatic nerve trifurcation. Experimental groups receiving delayed CP nerve repair had their CP nerve stumps sutured back to surrounding muscles for 2 months

(Empty MS, GDNF in fibrin, GDNF MS, and no treatment groups and groups with extra layers of fibrin, Figure 2-2B). Animals in the positive control group (Immediate repair) did not undergo any procedures at this time.

In the second procedure, CP nerves were exposed as before and the nerve stumps were repaired using 9-0 Nylon sutures (Figure 2-2C). In the groups receiving the fibrin gel based DDS, the nerve was surrounded by two 40 µL gels (Figure 2-2C), that were formed by pipetting the fibrin mixture, prior to its setting as a gel, onto Parafilm as semirectangular drops (~ 5 mm x 5 mm). The fibrin gel drops were placed centered above and below the CP nerve repair sites and secured by gently opposing the gel drops on one another. In groups with extra layer of fibrin, the CP nerve repair site was first surrounded by two 10 µL gels formed as described previously; then the fibrin gel containing GDNF

MS or empty MS was placed around the nerve. Animals in the positive control group instead had their CP nerve transected 5 mm from the sciatic trifurcation and immediately repaired with Nylon sutures and no DDS implantation.

2.3.9 Retrograde labeling of motor neurons (ventral horn cells) & sensory neurons (dorsal root ganglia)

Four weeks after nerve repair, the surgical site was reopened under general anesthesia.

The CP nerve was transected 15 mm distally from the suture repair site and the proximal nerve stump was immediately placed in a silicone well containing 4% Fluoro-Gold™ in sterile saline for 1 hour (Figure 2-2C). The silicone wells and Fluoro-Gold solutions were removed, incisions were closed and rats were allowed to recover as described previously. Seven days following the procedure, the rats were euthanized and perfused with 0.9% NaCl saline and cold 4% paraformaldehyde in PBS. The lumbar region (L3–

L6) of the spinal cord and L4–L5 of the dorsal root ganglia (DRG) were dissected free for frozen sectioning. Axial sections of the lumbar spinal cord (50 µm) or the DRGs (20 µm) were sectioned on a cryostat (Leica, ON, Canada). The number of labeled neuronal cell bodies within the ventral horn of each spinal cord section or within every fifth DRG section was counted using a fluorescent microscope with a 10× objective (100× overall magnification; Leica). Spinal cord counts were adjusted to account for split nuclei using the methods of Abercrombie (Abercrombie, 1946).

Figure 2-2. Surgical procedures performed on rats. Two months prior to the repair, the CP nerve was transected (A) and sutured back to the surrounding muscle to prevent regeneration (B). After the initial injury, the nerve ends were repaired with epineural suture repair followed by placement of the drug delivery system (C). The experimental groups were implanted with GDNF-containing microspheres, free GDNF or empty microspheres in fibrin gel at the repair site. In order to investigate the possibility of toxicity associated with placement of the DDS at the nerve injury site, in two additional experimental groups the repair site was first surrounded with an extra layer of fibrin gel and then with the drug delivery system containing GDNF MS or empty MS. Four weeks following nerve repair, the CP nerve was harvested and labeled with retrograde dye 15 mm distally from the repair site (D). CP: common peroneal nerve

2.3.10 Statistical analysis

Statistical testing was performed in GraphPad Prism 6 to confirm the normality of the data and then differences between groups were assessed by analysis of variance

(ANOVA) with Bonferroni’s post-hoc at 95% confidence intervals.

2.4 Results

2.4.1 In vitro microsphere characterization and GDNF release from DDS

Microspheres (MS) were constructed with heparin (5% wt/wt), MgCO3 (3% wt/wt), Glial cell-line derived neurotrophic factor (GDNF; 0.05% wt/wt), and poly(lactic-co-glycolic acid) (PLGA) with molecular weight of 5,000 Da. Consistent with previous studies

(Wood, Gordon, Kemp, et al., 2013; Wood, Gordon, Kim, et al., 2013), the constructed

MS demonstrated high GDNF encapsulation of 78 ± 3%, loading of 0.72 ± 0.08 µg of

GDNF per mg of the MS and average size of 45 ± 5 µm in diameter. The dynamic release of GDNF from fibrin gel in either free form or MS encapsulated form into the phosphate buffered saline (PBS) with 1% bovine serum albumin (BSA) was analyzed over 28 days at 37oC. With the exception of the first day, GDNF release from fibrin gel loaded with

GDNF MS was slower than the fibrin gel loaded with free GDNF (p < 0.05, Figure 2-

3A). In both of these drug delivery systems (DDS), there was no initial burst release from the fibrin gel, specifically during the first 24 hours of release (Figure 2-3B). From the

GDNF MS loaded DDS, GDNF had a sustained release over 15 days with the mean of

287 ± 163 ng/day. Free GDNF in fibrin had a faster release during 8 days with mean of

383 ± 333 ng/day.

1 0 0 )

% 8 0

(

a e

l 6 0

v i

t 4 0

l u

m G D N F M S u 2 0 C F re e G D N F

0 0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 A D a y s

1 0 0 0 G D N F M S

F re e G D N F )

y 8 0 0

n (

6 0 0

e l

e 4 0 0

s a

2 0 0 M

0 0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 B D a y s

samples was determined using ELISA. (Mean ± standard deviation, n=3 per release study) 2.4.2 In vitro toxicity assay for fibrin gel based DDS

The biocompatibility of the PLGA MS loaded DDS was assessed using an in vitro cell viability assay. In this assay, rat PC-12 cells were differentiated in vitro to neural cell type and cultured with or without release media obtained from the DDS containing empty

PLGA MS. All the PC-12 cells cultured with ethanol for 48 hours in the negative control group (Figure 2-4A) were dead as indicated by their red stain after performing a

LIVE/DEAD® viability assay. Based on the obtained fluorescent confocal images, the number of live cells (stained green) and dead cells (stained red) in the positive control group cultured with normal RPMI 1640 media (Figure 2-4B) and the experimental groups (Figure 2-4C-D) were similar. Quantitatively, the cell viabilities for cells cultured in the presence of empty MS release media at day 14 and day 28 (93.1 ± 1.9% and 93.3 ±

2.0% respectively) were indistinguishable from that of the positive control group (93.4 ±

2.1%, p > 0.05).

2.4.3 In vivo retrograde labeling of neurons following nerve repair

A retrograde labeling procedure was performed 4 weeks after the common peroneal (CP) nerve repair surgery to quantify the number of neurons regenerating their axons into the denervated distal CP nerve stump and to compare the extent of motor versus sensory neuron regeneration in all the experimental and control groups. Fluoro-Gold dye was applied 15 mm distal to the nerve repair site for 1 hour during the procedure. Spinal cord and dorsal root ganglia (DRG) were harvested 7 days later and sectioned for analysis under fluorescence microscopy. Two months after delayed CP nerve repair, approximately 33% of motoneurons (compared to the uninjured normal number of 400 ±

20 motoneurons) regenerated their axons into the distal CP nerve stump in the no treatment group (Figure 2-5A). Motoneuron regeneration for the control groups receiving empty MS in their fibrin gel with or without the extra layer of fibrin (118 ± 22 and 130 ± 26, respectively) were similar to the no treatment group (130 ± 18, p > 0.05,

Figure 2-5A). The same was observed for the regeneration of the sensory neurons. After delayed repair, only approximately 36% of the sensory neurons (compared to the normal uninjured number of 1000 ± 22 sensory) regenerated in the no treatment group (Figure 2-

5B). This number was similar to that of the groups received empty MS DDS with or without the extra layer of fibrin (360 ± 87, 420 ± 158, and 391 ± 147 respectively, p >

0.05, Figure 2-5B).

Significantly lower number of motoneurons regenerated their axons in the groups in which DDS with free GDNF or GDNF MS with an extra layer of fibrin was applied to the CP nerve suture site (178 ± 20 and 107 ± 30, respectively, Figure 2-6A) as compared to the immediate nerve repair positive control group (329 ± 44, representing 82% of the

normal number of CP motoneurons present in the uninjured state, p < 0.05). In contrast, significantly more motoneurons regenerated their axons after delayed nerve repair when the MS were loaded with GDNF (280 ± 47, approximately 71% of the normal number of motoneurons), such motoneurons regeneration was comparable to the immediate repair case (p > 0.05, Figure 2-6A). The increase in motoneuron regeneration was not observed when an extra layer of fibrin was applied to the suture site, probably due to creating a diffusion barrier (see discussion).

The same efficacy of the GDNF MS was demonstrated for sensory neurons: as for the motoneurons regeneration, the group receiving GDNF MS loaded DDS was the only group with statistically similar sensory neuron regeneration (771 ± 87) as compared to the immediate repair group (p > 0.05, Figure 2-6B). The sensory neurons regeneration in the groups with DDS loaded with free GDNF or GDNF MS with extra layer of fibrin

(575 ± 75, and 422 ± 124 respectively, Figure 2-6B) was, like the motoneurons, significantly lower compared to the immediate repair group (919 ± 111, p < 0.05).

e 1 0 0 0 l

4 0 0 s

a 3 0 0

y d

e 5 0 0

e a

n 2 0 0

o M

1 0 0 S

e R

R 0 0 N o E m p ty E m p ty M S N o E m p ty E m p ty M S B tre a tm e n t M S E x tra F i b ri n A tre a tm e n t M S E x tra F i b ri n

cells in the spinal cord (A; 50 μm sections, all sections counted, correction factor = 0.6 (Abercrombie, 1946)) and sensory neurons in the dorsal root ganglia (B; 20-µm sections, every fifth section counted) were counted. The empty MS treated group and the group receiving the drug delivery system with an initial extra layer of fibrin gel around the repair site had similar nerve regeneration compared to the no treatment group. Placement of the drug delivery system around the suture site did not diminish nerve regeneration. Normal uninjured values ± standard errors are represented by the dashed line. Data represent the mean ± standard error of the mean. MS: Microsphere.

* *

5 0 0 * 1 2 0 0 *

l l s 1 0 0 0

e 4 0 0

a o

L 8 0 0

3 0 0

e 6 0 0

o a

2 0 0 a

o s

g 4 0 0

e t

1 0 0 t

S e

e 2 0 0

R R

0 0

i i

A M

M i

I E E Figure 2-6. In vivo local GDNF release from the drug delivery system enhanced nerve regeneration. Retrograde labeling of common peroneal (CP) neurons regenerating their axons was performed 15 mm distal from the repair site 4 weeks following experimental treatment. The GDNF microsphere (MS) drug delivery system promoted significantly more motor (A) and sensory (B) neuron regeneration, similar to that of the immediate repair group, than the groups in which fibrin gels around the CP nerve were loaded with free GDNF or empty MS. Addition of an extra layer of fibrin around the suture site impaired the effectiveness of the drug delivery system due to introducing a diffusion barrier. Normal uninjured values ± standard errors are represented by the dashed line. Data represent the mean ± standard error of the mean. *p < 0.05. MS: Microsphere.

2.5 Discussion

The fibrin gel based drug delivery system (DDS) containing drug loaded poly(lactic-co- glycolic acid) (PLGA) microspheres (MS) is a surgically implantable hydrogel designed for localized and sustained release of drugs and therapeutic proteins such as glial cell line-derived neurotrophic factor (GDNF) to the injured peripheral nerve (Tajdaran, et al.,

2015; Wood, Gordon, Kemp, et al., 2013; Wood, Gordon, Kim, et al., 2013). In this

study, the free GDNF in a solubilized form contained within fibrin gel was released rapidly (~7 days) in vitro (Figure 2-3A). In order to extend the GDNF release time course from the fibrin gel, we incorporated GDNF loaded PLGA MS within the gel, which increased the drug release period to 15 days. In both formulations of the DDS, there was no initial GDNF burst release from the fibrin gel, which is likely due to the high concentration of fibrinogen within the fibrin gels (~50mg/mL), as the decreased porosity of the fibrin network would limit diffusion (Blomback et al., 1984; Carr &

Hardin, 1987). However, the daily amount of mass release from the fibrin gel containing free GDNF was more variable compared to the GDNF MS containing DDS (Figure 2-

3B). The more sustained release of GDNF from the PLGA MS containing DDS may be related to the formation of electrostatic interactions between GDNF, fibrinogen and

PLGA MS within the DDS composite.

Fibrin gel is a natural biocompatible hydrogel that has been extensively used clinically as nerve sealant during peripheral nerve repair in patients (Jubran & Widenfalk, 2003;

Sameem et al., 2011). In addition fibrin gel has previously been used to successfully deliver growth factors including nerve growth factor (NGF) and GDNF to regenerating axons following nerve transection injuries in rats (Jubran & Widenfalk, 2003). Polymeric

MS developed to provide controlled release of hydrophilic and hydrophobic biomolecules, have been widely investigated (Caicco et al., 2013; Garbayo et al., 2009;

Pe et al., 2000; Sinha & Trehan, 2003; Soppimath et al., 2001). Most of the MSs used for drug delivery purposes are composed of PLGA, a biodegradable polymer approved for clinical use (Sinha & Trehan, 2003). In our engineering design, we combined the two clinically used biomaterials to obtain a biocompatible DDS. Here for the first time, we

investigated the biocompatibility of this fabricated DDS with both in vitro and in vivo experiments. PC-12 cells incubated in vitro with released media samples from the empty

PLGA MS containing DDS, had similar viability compared to control cells cultured with normal media, demonstrating that the drug delivery system was not toxic (Figure 2-4).

Consistent with this, the placement of the DDS at the site of the common peritoneal (CP) nerve transection and surgical repair resulted in the numbers of motor and sensory neurons regenerating their axons in vivo being the same in the groups of no treatment, empty MS, and empty MS with an extra layer of fibrin (Figure 2-5). Importantly, the placement of the delivery system adjacent to the suture site did not attenuate nerve regeneration. Thus, the DDS was biocompatible and well tolerated within the environment around the peripheral nerves. The numbers of regenerating neurons increased when free GDNF was administered in fibrin gel, but this increase was not significant. Moreover, addition of an extra layer of fibrin around the suture site impaired the effectiveness of the drug delivery system due to introducing a diffusion barrier. In contrast, GDNF administered within the microspheres increased the numbers significantly, almost to the numbers after immediate nerve repair (Figure 2-6). In the simplest manner, the influence of the diffusion barrier after addition of an extra layer of fibrin around the nerve repair site could be described by Fick’s Law for diffusion from a plane according to the one dimensional, unidirectional, thin film approximation for non- swelling samples at short times (Brazel & Peppas, 2000):

(1)

Where Mt was the cumulative mass of drug detected at time, t, divided by the total mass released M∞. D was the drug diffusivity and l was the sample thickness. By increasing the thickness of the fibrin gel, l, and maintaining M∞ constant, at any give time, Mt should decrease as well. Therefore, in consistent with the in vivo nerve regeneration data, by addition of an extra layer of fibrin, the DDS should become less effective in delivering

GDNF when an extra layer of fibrin was added.

Our study established that the biomaterial developed by combining fibrin gel and PLGA

MS can provide a biocompatible DDS for sustained and controlled release of GDNF to the nerve injury site. Second, this study demonstrated the potential clinical value of the developed DDS. Consistent with our previous work, the release of GDNF from the DDS improved motor and sensory neurons regeneration after delayed CP nerve repair in rats.

2.6 Conclusions

The goal of this work was to evaluate the biocompatibility of a fibrin gel based drug delivery system (DDS) containing GDNF loaded MS. The release media from the DDS containing empty MS did not affect PC-12 cells’ viability in vitro. The placement of the empty MS DDS around the nerve suture site after delayed repair in a rat hind limb model did not diminish axon regeneration. In addition, we confirmed that implantation of the fibrin gels with GDNF MS at the nerve repair site improved motor and sensory neuron regeneration. We believe such engineered biomaterial offers a potential treatment of chronic peripheral nerve injuries.

2.7 Acknowledgements

We are grateful to Jennifer Zhang and David Scholl for their help on spinal cord tissue analysis. We thank the Collaborative Health Research Projects (NSERC Partnered) program for financial support.

3 A glial cell line derived neurotrophic factor delivery system enhances nerve regeneration across acellular nerve allografts

Kasra Tajdaran*1,2, Tessa Gordon1,4,5, Mathew D. Wood6, Molly S. Shoichet2,3, Gregory H. Borschel1,2,4,5

Publication Information: Submitted to Biomaterials

Acellular nerve allografts (ANA) are used clinically to bridge nerve gaps but these grafts, lacking Schwann cells and therapeutic levels of neurotrophic factors, do not support regeneration to the same extent as autografts. Here we investigated a local drug delivery system (DDS) for glial cell line-derived neurotrophic factor (GDNF) controlled release to implanted ANAs in rats using drug-loaded polymeric microspheres (MS) embedded in a fibrin gel. In a rat hindlimb nerve gap model, a 10 mm ANA was used to bridge a 5 mm common peroneal (CP) nerve gap. Experimental groups received DDS treatment at both suture sites of the allografts releasing GDNF for either 2 weeks or 4 weeks. In negative control groups, rats received no treatment or empty DDS. Rats receiving nerve isografts served as the positive control group. The numbers of motor and sensory neurons that regenerated their axons in all the groups with GDNF MS and isograft treatment were indistinguishable and significantly higher as compared to the negative control groups.

Nerve histology distal to the nerve graft demonstrated increased axon counts and a shift to larger fiber diameters due to GDNF MS treatment. The sustained delivery of GDNF to the implanted ANA achieved in this study demonstrates the promise of this DDS for the management of severe nerve injuries in which allografts are placed.

Keywords: nerve injury, acellular nerve allografts, glial cell line-derived neurotrophic factor, drug delivery, fibrinogen, poly(lactic-co-glycolic) acid, regenerative medicine, biomaterials.

3.2 Introduction

Despite substantial improvements in microsurgical techniques, patients with peripheral nerve injuries rarely recover fully (Ciaramitaro et al., 2010; Yegiyants et al., 2010).

Direct end-to-end repair of the transected peripheral nerve supports limited recovery following injury (Aydin et al., 2004; Lundborg & Rosen, 2007; Lundborg, 2000).

However, in many clinical situations, there is not enough nerve tissue to allow a tension free reconstruction (Millesi, 1981; Terzis et al., 1975). In these cases, the current surgical standard consists of using an autograft, in which a nerve graft from the same patient is used to bridge the nerve gap. Although autografts provide tension-free repair, they require a second operative site which necessitates additional operative time, a permanent scar, donor sensory loss, and could result in persistent postoperative pain (Szynkaruk et al., 2013). Moreover, due to the limitation in the available length, nerve autografts may not be feasible in cases where extensive reconstruction is required (Szynkaruk et al.,

2013).

An alternative to autografting is the use of processed nerve allografts, or acellular nerve allografts (ANA)s (Whitlock et al., 2009). ANAs retain the scaffold of nerve tissue but are made to be non-immunogenic to the recipient by a variety of processing methods,

such as repeated freeze–thaw cycles; cold preservation; and decellularization with detergents (Evans et al., 1998; A. Gulati & Cole, 1990; Szynkaruk et al., 2013). Thus,

ANAs provide a biological substrate for nerve regeneration without the requirement of immunosuppression. However, they have non-therapeutic levels of neurotrophic factors especially compared to normal denervated nerve stumps in which several growth factors are upregulated after injury (Boyd & Gordon, 2003b). Given that regenerating nerve fibers preferentially elongate toward sources of neurotrophic factors (Cao & Shoichet,

2001; Perez et al., 1997), replenishing the ANAs with key neurotrophic factors should enhance nerve gap regeneration. ANAs have been used clinically in patients for several years (Brooks et al., 2012), and we questioned whether the ability of these allografts to support nerve regeneration could be improved by supplementation with key neurotrophic factors lacking in the commercially available ANA.

Delivery of neurotrophic factors holds promise in enhancing outcome following nerve injury (Mohtaram et al., 2013). Neurotrophic factors, such as brain-derived neurotrophic factor, nerve growth factor, and glial cell line-derived neurotrophic factor (GDNF), which are essential for peripheral nervous system development, have been shown to promote axon regeneration and enhance functional recovery (Boyd & Gordon, 2003a, 2003b;

Jubran & Widenfalk, 2003). However, the challenge for achieving a clinically suitable application for GDNF is its localized and sustained release to the nerve injury site (Boyd

& Gordon, 2003a, 2003b). Current investigational methods of GDNF local delivery include viral transfected Schwann cells (Eggers et al., 2008; Wu-Fienberg et al., 2014), and catheter/mini-osmotic pump systems (Guilhem et al., 2009). While viral transduction of primary cells generates local release, regulation of GDNF release is difficult to manage

and can result in excess and toxic GDNF release. In addition, clinical translation may be a significant regulatory challenge and these methods are not currently approved for clinical use. Osmotic pump delivery systems, despite providing sustained and localized release, can hinder recovery due to risk of infection and even nerve compression secondary to capsular fibrosis (Lundborg, 2000). A sustained and tunable delivery from a biodegradable and biocompatible system is therefore preferred to effectively delivery

GDNF to the injured nerve.

Previously, we developed a microsphere-based biodegradable drug delivery system

(DDS) supporting sustained release to the injured nerve over periods of days to weeks

(Wood, Gordon, Kemp, et al., 2013; Wood, Gordon, Kim, et al., 2013). This DDS, consisting of fibrin gel containing GDNF microspheres, significantly improves axon regeneration and functional recovery after delayed nerve repair (Wood, Gordon, Kemp, et al., 2013; Wood, Gordon, Kim, et al., 2013). In this study, we combined the DDS composite system with the rat analogue of the clinically-used nerve allograft to determine the extent to which this new hybrid DDS-ANA biomaterial supported nerve regeneration.

3.3 Materials and Methods

All chemicals were obtained from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise specified.

3.3.1 GDNF encapsulation in PLGA microsphere

Glial cell line-derived neurotrophic factor (GDNF) was encapsulated in poly(lactic-co-

glycolic acid) (PLGA) microsphere (MS) using a water/oil/water double emulsion,

solvent evaporation method. Briefly, an inner aqueous solution of 100 µl consisting of

250 µg GDNF (Peprotech, Rocky Hill, NJ) and 12.5 mg heparin was mixed with 230 mg

PLGA 50/50 (Wako, Japan and Lactel Absorbable Polymers, Cupertino, CA, Table 3-1)

and 12.5 mg MgCO3 in 1 mL dichloromethane (DCM)/acetone (75%/25%). The mixture

was sonicated for 45 s using a 3 mm probe sonicator (Vibra-Cell™ VCX 130; Sonics and

Materials, CT, USA) at 30% power. The resulting emulsion was added to 25 mL of 2.5%

aqueous poly(vinyl alcohol) (PVA) solution containing 10% NaCl and homogenized at

6,000 rpm for 60 s. The secondary emulsion was then added to 250 mL aqueous solution

of 2.5% PVA and 10% NaCl. The mixture was stirred for 3 hours with venting to allow

the hardening of the microspheres by complete evaporation of the organic solvent. The

hardened microspheres were collected and washed by centrifugation, lyophilized, and

stored at −20 °C until use.

Table 3-1. PLGA description used in drug delivery system synthesis.

Formulation name PLGA inherent PLGA average GDNF initial Encapsulation viscosities (dL/g) molecular weight (Da) loading (%wt/wt efficiency in microspheres) 2-week release 0.088–0.102 5,000 250 mg (0.05%) 78 ± 3% formulation 4-week release 0.15–0.25 6,700 250 mg (0.05%) 78 ± 3% formulation

3.3.2 GDNF microsphere characterization

Microsphere mean diameter and size distribution were measured via static light scattering using a Malvern Mastersizer 2000 laser diffraction particle sizer (Malvern Instruments

Ltd, UK), using refractive indices of 1.33 and 1.59 for water and PLGA, respectively.

Encapsulation efficiency was measured by dissolving an appropriate mass of microspheres in 1 mL dimethyl sulfoxide (DMSO) for 1 hour at 37oC followed by addition of 10 mL of 0.05 M NaOH with 0.5% w/v sodium dodecyl sulfate (SDS) and further shaking for 1 hour at room temperature. The amount of GDNF was quantified by an enzyme-linked immunosorption assay (ELISA) for human GDNF according to the manufacturer’s instructions (R&D Systems, Minneapolis, MN). The absorbance was read at 450 nm with an optical subtraction at 540 nm using a multi-well plate spectrophotometer, and sample concentrations were calculated from a standard curve of known GDNF concentrations. Drug loading was determined as the GDNF mass per mg of microspheres; encapsulation efficiency was the measured drug loading of the microspheres divided by the theoretical maximum drug loading.

3.3.3 GDNF DDS composite construction and in vitro release

Fibrin gel (80 µL total volume) was constructed by mixing equal parts fibrinogen (75–

115 mg/mL, 40 µL) and thrombin (5 IU/mL, 40 µl) obtained from a Tisseel® glue kit

(Baxter Healthcare, IL, USA), and then re-suspended according to the manufacturer’s instructions. Fibrin gels were loaded with microspheres by incorporating 5 mg of microspheres into the thrombin solution before it was mixed with fibrinogen to form a gel. In vitro release of GDNF from fibrin loaded with microspheres was assessed by

using 80 µL gels in 2 mL siliconized centrifuge tubes (Fisher Scientific). The time course of release was measured by incubating the fibrin gels in 1 mL of phosphate buffered saline (PBS) containing 1% bovine serum albumin (BSA) at 37oC under constant gentle agitation by vortex. The PBS was collected and ELISA assays were performed to measure GDNF quantity collected from the time course release studies.

3.3.4 Acellular nerve allograft preparation

The processed common peroneal nerve grafts from rats were provided by AxoGen, Inc

(Alachua, FL, USA). Briefly, the harvested rat nerve tissue from donor rat’s common peroneal nerve was decellularized using the Hudson et al. protocol (Hudson, Liu, &

Schmidt, 2004; Hudson et al., 2004; Whitlock et al., 2009). The processed tissue was then depleted of chondroitin sulfate proteoglycans with chondroitinase ABC based on the method developed by Neubauer et al. (2004). The processed grafts were sterilized with gamma irradiation and frozen at -80oC. The grafts were stored at -80oC until implantation.

3.3.5 Experimental animals

Canadian Council on Animal Care (CCAC) and were approved by the Hospital for Sick

Children’s Laboratory Animal Services Committee.

3.3.6 Experimental design

Forty-eight adult female Sprague–Dawley rats were randomized into six groups (n = 8)

(Figure 3-1). Animals receiving no drug delivery system (DDS) treatment or fibrin gels loaded with empty MSs served as experimental control groups. Fibrin gels loaded with microspheres releasing GDNF in vitro for ~2 or ~4 weeks placed at both suture sites of the nerve graft served as the primary experimental groups. Another experimental group received MSs with 2-week release formulation at the proximal suture sties and microspheres with 4-week release formulation at the distal site. Rats receiving nerve isografts (i.e., grafts taken from immunologically equivalent littermates) served as the positive control group. In additional, in each group, two Thy-1 transgenic rats that expressed green fluorescent protein (GFP) in their axons (Moore et al., 2012) were included for qualitative visualization of the regenerating axons within the implanted nerve grafts.

3.3.7 Operative procedure

The sciatic nerve was exposed through a dorsolateral–gluteal muscle splitting incision.

Wounds were irrigated with saline, dried and closed in two layers, utilizing 5-0 Vicryl™

(Ethicon, OH, USA) sutures to close the muscle layers, and 4-0 Nylon sutures to close the

skin. Experimental animals were recovered in a warm environment prior to returning to the housing facility.

During the procedure, the common peroneal (CP) nerve was dissected free (Figure 3-2A) and a 5 mm nerve gap was created approximately 5 mm distal from the sciatic trifurcation (Figure 3-2B). A 10 mm segment of a nerve graft was used to bridge the nerve gap using 9-0 Nylon sutures (Figure 3-2C). In the groups receiving the fibrin gel based DDS, the proximal and distal nerve suture sites were surrounded by two 40 µL gels

(Figure 3-2C), formed by pipetting the fibrin mixture, before setting as a gel, onto

Parafilm as semirectangular drops (~ 5mm x 5 mm). The gel drops were placed centered above and below the repair sites and secured by gently opposing the gel drops on one another.

3.3.8 Retrograde labeling of motor neurons (ventral horn cells) & sensory neurons (dorsal root ganglia)

Eight weeks after nerve graft implantation, the surgical site was reopened under general anesthesia. The CP nerve was transected 10 mm distally from the distal repair site and the proximal nerve stump was immediately placed in a silicone well containing 4% Fluoro-

Gold™ in sterile saline for 1 h (Figure 3-2C). At the same time, the distal stump was harvested for histology (described below). The silicone wells and Fluoro-Gold solutions were removed, incisions were closed and rats were allowed to recover as described previously. Seven days following the procedure, the rats were euthanized and perfused with 0.9% NaCl saline and cold 4% paraformaldehyde in PBS. The lumbar region (L3–

L6) of the spinal cord and L4–L5 of the dorsal root ganglia (DRG) were dissected free for frozen sectioning. Axial sections of the lumbar spinal cord (50 µm) or DRG (20 µm) were sectioned on a cryostat (Leica, ON, Canada). The number of labeled cell bodies within the ventral horn of each spinal cord section or within every fifth DRG section was counted using a fluorescent microscope with a 10× objective (100× overall magnification; Leica). Spinal cord counts were adjusted to account for split nuclei using the methods of Abercrombie (Abercrombie, 1946).

3.3.9 Histology & morphometric evaluation of nerves

At the time of the retrograde labeling surgery, the nerve tissue taken 10 mm distally from the nerve graft distal suture site was collected (Figure 3-2D), fixed in 2% glutaraldehyde, post-fixed with 1% osmium tetroxide, ethanol dehydrated and embedded in Araldite® 502 (Polyscience, Inc., PA, USA). Thin (0.6 µm) sections were made from

the tissue using a LKB II ultramicrotome (LKB-Produckter AB, Sweden) and then stained with 1% toluidine blue for examination by light microscopy. The slides were evaluated for overall nerve architecture and quality of regenerated fibers. At 1000× overall magnification, the entire nerve cross-section was captured to count regenerated axons, measure myelin thickness and fiber diameter, and calculate G-ratio using a semi- automated MATLAB program (More et al. 2011).

In Thy-1 GFP rats, the nerve graft was harvested nine weeks after implantation at the time of spinal cord tissue dissection and fixed in cold 4% paraformaldehyde in PBS.

Longitudinal sections of the nerve graft were cut at 30 µm on a cryostat. To qualitatively analyze the axon morphology and density with the graft, the obtained sections were imaged using a fluorescent microscope with a 10× objective (100× overall magnification;

Leica).

3.3.10 Statistical analysis

Statistical testing was performed in GraphPad Prism 6 to confirm the normality of the data and then differences between groups were assessed by analysis of variance

(ANOVA) with Bonferroni’s post-hoc at 95% confidence intervals.

3.4 Results

3.4.1 In vitro microsphere characterization and GDNF release from DDS

Microspheres (MS) constructed with poly(lactic-co-glycolic acid) (PLGA), heparin and

MgCO3 demonstrated a glial cell line-derived neurotrophic factor (GDNF) encapsulation efficiency of 78 ± 3% and GDNF loading of 0.72 ± 0.08 µg per mg of microspheres. The microspheres had diameters of 45 ± 5 µm. The molecular weight of the PLGA used for the synthesis of MS, was adjusted to modulate the release kinetics of GDNF from the drug delivery system (DDS) containing the MS, into the phosphate buffered saline (PBS) with 1% bovine serum albumin (BSA) at 37oC. Consistent with previous studies(Wood,

Gordon, Kemp, et al., 2013), the dynamic release of GDNF was 15 days from MS synthesized with PLGA that had an inherent viscosity of 0.088-0.102 dL/g and a molecular weight of 5,000 Da (Figure 3-3A).

This type of DDS was called the “2-week release” formulation of DDS due to the 15-day

GDNF release period from the system. By increasing the inherent viscosity and the molecular weight of PLGA to 0.15-0.25 dL/g and 5,000 Da respectively, the encapsulated

GDNF release period from the MS was extended to 28 days (Figure 3-3A). Thus, this type of DDS was called the “4-week release” formulation of DDS. There was no initial

GDNF burst release from both DDS during the first 24 hours (Figure 3-3B). The daily mass release indicated that there was a sustained release of GDNF over 15 days with the

mean of 287 ± 25 ng/day for the 2-week release formulation of DDS and over 28 days with the mean of 220 ± 30 ng/day for the 4-week release formulation of DDS.

1 0 0 )

% 8 0

(

a e

l 6 0

v i

t 4 0

u m

u 2 0 2 W R e le a s e F o rm u la tio n C 4 W R e le a s e F o rm u la tio n

0 0 5 1 0 1 5 2 0 2 5 3 0 A D a y s

8 0 0 2 W R e le a s e F o rm u la tio n

) 4 W R e le a s e F o rm u la tio n y

a 6 0 0

(

e s

a 4 0 0

s 2 0 0

a M

0 0 5 1 0 1 5 2 0 2 5 3 0 B D a y s

formulations of microspheres confirms sustained release over 15 days for the 2-week release formulation of DDS and over 28 days for the 4-week release formulation of DDS. GDNF content in the release samples was determined using ELISA. (Mean ± standard deviation, n=3 per release study)

3.4.2 In vivo retrograde labeling of neurons following nerve repair

In order to quantify the number of common peroneal (CP) neurons regenerating their axons through the nerve graft and to compare the extend of motor versus sensory regeneration, retrograde labeling was performed 8 weeks after nerve graft implantation

(Figure 3-2D): Fluoro-Gold was applied to the regenerated axons 10 mm distally from the distal suture site of the nerve graft. The empty MS and no treatment control groups had similar number of neurons that regenerated their axons (p > 0.05, Figure 3-4), indicating fibrin gel with MS did not inhibit nerve regeneration. Following nerve gap surgical repair, approximately 50% of the CP motoneurons (compared to the uninjured normal number of 400 ± 20 motoneurons) regenerated their axons within the 8 weeks under conditions of no treatment (200 ± 420) or with microspheres that were empty (247

± 50). There was a significant increase in these numbers when GDNF was included within the microspheres (Figure 3-4A). Indeed, the acellular nerve allografts (ANAs) with GDNF MS treatment were statistically indistinguishable from the isograft (Figure

3-4A) with 397 ± 46 and 443 ± 48 motoneurons regenerating their axons, respectively (p

> 0.05), except for the group that received the 2-week release formulation of DDS at the proximal suture site and the 4-week release formulation of DDS at the distal site (365 ±

47; p < 0.05).

Sensory neuron regeneration mirrored that of the motoneurons. After the nerve graft implantation, the control groups with no treatment and empty MS containing 566 ± 74

and 605 ± 140 retrogradely labeled sensory neurons had only approximately 60% sensory neuron regeneration (compared to the normal uninjured number of 1000 ± 22 sensory,

Figure 3-4B). These numbers were significantly increased in the isograft and all the

GDNF MS treated groups (p < 0.05). The numbers of sensory neurons regenerating their axons through the ANA with GDNF DDS treatment were statistically indistinguishable from the numbers through the isograft (1151 ± 282 and 1032 ± 98, respectively; p > 0.05;

Figure 3-4B).

Overall, the numbers of motor and sensory neurons that regenerated their axons through the implanted ANAs were significantly higher in all the primary experimental groups treated with GDNF MS compared with the control groups receiving empty MS in their fibrin gels or with no treatment after the ANA implantation. Despite a decrease in the neurons’ mean number for the group that received DDS with 2 weeks GDNF release at the proximal suture site and 4 weeks GDNF release at the distal site, all the experimental groups with GDNF MS treatment demonstrated similar success of nerve regeneration through the allografts. These findings indicate that the total amount of GDNF available over time played a significant role in promoting the growth of the regenerating axons through the ANA. Thus, for the histomorphometric analysis, the GDNF MS treated group with the 2-week release formulation DDS was used as the representative experimental group.

6 0 0 *

5 0 0

l e

s 4 0 0

l e

e 3 0 0

g o

M 2 0 0

e R

1 0 0

0 N o E m p ty 2 W + 2 W 2 W + 4 W 4 W + 4 W I S O A tre a tm e n t M S G D N F M S

2 0 0 0 *

d 1 5 0 0

e 1 0 0 0

t S

e 5 0 0 R

0 N o E m p ty 2 W + 2 W 2 W + 4 W 4 W + 4 W I S O B tre a tm e n t M S G D N F M S

experimental groups receiving fibrin gels loaded with MS containing GDNF had comparable nerve regeneration to the isograft group and showed significantly higher motor (A) and sensory (B) neurons regeneration through nerve allografts compared with the empty MS and no treatment control groups. The period of GDNF release from the drug delivery system did not influence the extent of nerve regeneration. The control groups receiving no treatment and fibrin gels with empty MS had similar number of regenerated neurons, indicating the drug delivery system did not diminish nerve growth. Data represent the mean ± standard error of the mean. Normal uninjured values ± standard error are represented by the dashed line. *p < 0.05. MS: Microsphere

3.4.3 Axon morphology through nerve graft

The implanted nerve grafts were harvested from the Thy-1 GFP rats at the time of spinal cord tissue dissection to perform morphometric analysis on axon regeneration through the nerve graft. A qualitative analysis of the 30 µm longitudinal nerve graft sections revealed the axon alignment and axon density within the nerve graft (Figure 3-5A-C). There was a uniform axon distribution within the isografts and GDNF MS treated ANAs (Figure 3-

5B-C). GDNF treatment using the microspheres (Figure 3-5B) did not influence the axon alignment within ANAs and they were similar to that of the isografts and no treatment groups (Figure 3-5). An autofluorescence in all fluorescent channels was observed within the acellular nerve allografts regardless of the DDS type treatment. This could be related to the byproducts of the grafts extracellular matrix remodeling during axon regeneration through the nerve grafts.

3.4.4 In vivo nerve histology and morphometric measures of regeneration

Eight weeks following nerve repair with graft implantation, nerve samples were harvested at 10 mm distal from the nerve graft for histology analysis by light microscopy.

Qualitative analysis of nerve samples revealed similar nerve morphology for the groups receiving empty MS and no treatment (Figure 3-6A-B). All the GDNF MS groups had similar nerve morphology to the group receiving isografts (Figure 3-6C-D) with significantly higher myelinated axons present in the nerve cross sections compared with the empty MS and no treatment control groups (Figure 3-6A-B). In all the groups, regardless of the treatment, axons were uniformly distributed throughout the nerve.

Quantitative fiber frequency distribution analysis on the entire nerve cross section revealed that the GDNF MS and isograft treated groups had similar fiber distribution

(Figure 3-6E). These two groups demonstrated greater numbers of larger diameter fibers

(4 µm - 6 µm) and fewer numbers of smaller diameter fibers (2 µm - 4 µm) compared with the no treatment and empty MS treated groups (Figure 3-6E). The strong shift toward larger diameter fibers with GDNF MS treatment (p < 0.05) is another evidence that the GDNF local release to the allografts enhanced axonal regeneration.

Quantitative histomorphometric analysis of the entire nerve cross sections confirmed the significant increase in the number of myelinated axons in the GDNF MS treated groups compared with the no treatment control group (3173 ± 204 and 2028 ± 524, respectively; p < 0.05; Figure 3-6F). The number of myelinated axons that regenerated in the GDNF

MS treated groups was similar to that of the isografts treated group with 2028 ± 695 axons (p > 0.05). As a measure of nerve maturity, nerve fiber diameter, and myelin thickness was determined. There were no significant differences in myelin thickness and fiber diameter in all the groups (Figure 3-6G-H). The G-ratio, calculated as the ratio of the axon diameter to the total fiber diameter, was statistically equivalent for all groups

(Figure 3-6I).

uninjured nerves (demonstrated by the horizontal dashed lines). Data represent the mean ± standard error of the mean. *p < 0.05. Scale bars: 10 µm. MS: Microsphere

3.5 Discussion

Acellular nerve allografts (ANA) have the potential to support nerve regeneration by providing a cell free scaffold and maintaining much of the internal structural and molecular composition of the normal nerve extracellular matrix. While these allografts do not induce immunogenic responses after implantation, they lack Schwann cells and therapeutic levels of neurotrophic factors important for peripheral nerve regeneration.

Therefore, the addition of neurotrophic factor support may lead to a clinically superior

ANA (Bunge, 1994; Jesuraj et al., 2014; Wang et al., 2008).

In this study, we sought to determine the effect of localized and sustained release of glial cell line-derived neurotrophic factor (GDNF) on axon regeneration through ANAs.

Natural hydrogels, such as fibrin, have been commonly used to obtain controlled neurotrophic factor delivery to peripheral nerves (Sameem et al., 2011). In order to avoid the high burst release associated with hydrogels due to simple diffusion and non-covalent interactions as well as to prolong drug release (Yeo & Park, 2004), GDNF was incorporated within the fibrin gel in poly(lactic-co-glycolic acid) (PLGA) microspheres

(MS). Based on the PLGA molecular weight and inherent viscosity, the PLGA/fibrin gel drug delivery system (DDS) was designed to provide in vitro GDNF release for up to either 15 days (the “2-week release” formulation) or 28 days (the “4-week release” formulation). The in vitro release profile showed a sustained release without an initial burst within the first day, suggesting the formation of electrostatic or hydrophobic interactions between GDNF, fibrinogen and PLGA microspheres within the fibrin gel.

Both the 2-week release formulation and 4-week release formulation of DDS were able to deliver GDNF at levels of ≥ 100 ng/day for the entire period of release for the 2-week release formulation of DDS, and for at least 23 days for the 4-week release formulation of

DDS (Figure 3-3B). The ability to deliver more than 100 ng/day of GDNF to the nerve repair site is necessary for improved motor nerve regeneration (Boyd & Gordon, 2003a).

Previously, we have shown GDNF release for the periods of 2 weeks and 4 weeks is effective in enhancing nerve regeneration in chronic axotomy models (Wood, Gordon,

Kemp, et al., 2013). In this study, a combination of the 2-week release formulation of

DDS and 4-week release formulation of DDS were placed at two suture sites of implanted ANAs. In two experimental groups, the DDS provided GDNF release up to either 2 weeks or 4 weeks at both suture sites of the implanted grafts. In an additional experimental group, in order to provide constant GDNF concentration available to the axons’ growth cone through the graft and at both suture sites, the 2-week release formulation of DDS was placed at the proximal suture site and the 4-week release formulation of DDS was placed at the distal suture site.

Our findings demonstrated that the GDNF delivered locally to the implanted ANA from a biodegradable DDS could enhance significant axon growth in a clinically relevant model.

Implantation of the GDNF MS-containing DDS, regardless of the MS formulation, significantly improved motor and sensory neuron regeneration compared to the control groups without GDNF (Figure 3-4). Even though the numbers of motoneurons that regenerated their axons in the experimental group that received the combination of 2- week release formulation of DDS and 4-week release formulation of DDS did not match the isograft treated group, the GDNF MS-treated groups were not statistically different

from one another or the isograft treated group. The nerve histomorphometric findings paralleled the counts of the motor and sensory neurons that regenerated their axons. With the GDNF MS treatment, axon density (Figure 3-5), the total number of myelinated axons (Figure 3-6) and the frequency of the larger diameter fibers (Figure 3-7) matched the isograft group, which was statistically distinguishable from the control groups without

GDNF treatment. Nerve fiber diameter is a measure of nerve maturity and quality

(Aitken, Sharman, & Young, 1947; Aitken, 1949); more fibers with larger diameter indicates better functional recovery compared with smaller diameter fibers (Fraher &

Dockery, 1998). In the future, functional studies, such as assessment of muscle force, electrodiagnostic studies and behavioral analysis will allow us to assess the functional benefits of treating the ANAs with GDNF MS containing DDS (Fraher & Dockery,

1998).

The engineered biomaterial in this study was designed to function as a biocompatible drug delivery system for both the injured nerve and also the acellularized nerve allograft.

Because the numbers of motor and sensory neurons that regenerated their axons and the numbers of regenerated axons were similar in the empty MS treated and the no treatment groups, we conclude that the placement of the drug delivery system around the graft suture sites did not diminish nerve regeneration. The biocompatible DDS was engineered with clinical ease of use in mind. Surgeons could readily use such a system at the time of nerve reconstruction to enhance nerve regeneration. Fabricating the DDS entirely from biodegradable polymers had eliminated the need of a secondary surgery for the system removal. The degradation of the DDS in vivo prevented foreign body formation and

chronic nerve compression over time; therefore, the system is likely to be of great clinical value for nerve repair.

Our group has previously shown that the PLGA/fibrin gel composite provides a biodegradable and biocompatible DDS with the potential to significantly enhance nerve regeneration and functional outcome after a delayed nerve repair model (Wood, Gordon,

Kemp, et al., 2013). Importantly, we build on previous knowledge and show, for the first time, the local and controlled release of a neurotrophic factor to the acellular nerve allograft suture sites from this fibrin gel based DDS and the consequent axonal growth benefits.

3.6 Conclusion

In this study, a polymeric biocompatible drug delivery system (DDS) was investigated for sustained and controlled release of glial cell line-derived neurotrophic factor (GDNF) to the implanted acellular nerve allograft (ANA) for bridging a clinically relevant gap model of peripheral nerve injury. Based on the degradation rate of the DDS, GDNF was released locally in vitro over periods of 2 weeks or 4 weeks. Implantation of the DDS in vivo around the suture sites of ANAs did not induce any adverse side effects on nerve regeneration. Importantly, GDNF local administration from the DDS enhanced nerve regeneration and made the ANAs as effective as isografts in supporting nerve regeneration. The combination of the allograft biomaterial and the GDNF MS delivery system in this study has the potential to provide an “off the shelf” alternative in the current management of severe nerve injuries.

3.7 Acknowledgements

We are grateful to Jennifer Zhang and David Scholl for their help on spinal cord tissue analysis. We thank the Collaborative Health Research Projects (NSERC Partnered) program for financial support.

4 Conclusions and Future Work

Complete functional recovery from peripheral nerve injury and repair is rarely achieved despite the considerable advances in microsurgical techniques. In most clinical cases, delays in nerve regeneration and target organ reinnervation due to either delayed nerve repair or presence of large nerve defects following trauma can significantly diminish the capability of neurons to regenerate their axons.

Currently, in clinical practice, large nerve gaps are repaired by autologous nerve grafting.

This however, requires the sacrifice of a healthy nerve, usually the sural nerve, with permanent sensory impairment. Another problem is the shortage of autografts obtainable, should large defects like plexus injuries needed to be repaired. A variety of strategies have therefore been used in attempt to develop alternative repair methods. Current research is focused on developing improved scaffolds that can be used to physically guide regeneration of nerves across the nerve gaps. Past research in this area has focused either on existing natural or synthetic materials; however, none of the materials studied to date, such as the nerve conduit or the acellular nerve allograft (ANA) have matched or exceeded the performance of the nerve autograft. To this end, we have focused on the combination of materials and desired biomolecules to create new composite materials that can actively stimulate nerve regeneration.

We have developed a local drug delivery system (DDS) for controlled and sustained release of glial cell line-derived neurotrophic factor (GDNF) to the injured peripheral nerve. This is a fibrin gel based system inside which drug-loaded poly(lactic-co-glycolic acid) (PLGA) microspheres (MS) is embedded. Such engineered DDS was used to

provide neurotrophic factor support to chronically axotomized common peroneal (CP) nerve in rats following delayed nerve repair. In addition, the developed DDS was combined with a 10 mm ANA to bridge a CP nerve gap immediately after transection of the nerve. Axonal regeneration and nerve maturation were evaluated in both experiments with retrograde labeling techniques, histology and morphometric analysis of nerve sections.

The application of the GDNF loaded DDS following delayed nerve repair was effective in preventing the negative effects of chronic nerve injury. More specifically, the work presented in Chapter 2 demonstrated that only following the controlled and sustained release of GDNF from the DDS-containing MS to the chronically injured CP nerve, motor and sensory neurons regeneration was comparable to that of the ideal case of immediate nerve repair. The DDS containing free form of GDNF without MS was not effective in enhancing nerve regeneration. Furthermore, the engineered DDS was biocompatible as shown by the in vivo and in vitro experiments. The nerve regeneration after placing DDS containing empty MS with or without an extra layer of fibrin gel around the nerve injury site was similar to that of the no treatment group. Hence, the

DDS biomaterial was not deleterious to nerve regeneration. In addition, PC-12 cells cultured with the release media obtained from the DDS at different time periods demonstrated the same viability as when the cells were cultured with normal media.

Thus, the delivery system did not cause neuronal death, being biocompatible and non- toxic in vitro. The conclusion of this chapter is that the GDNF treatment using the proposed biocompatible DDS significantly promotes excellent nerve regeneration following chronic nerve injury without any detectable, unwanted toxic effects.

In Chapter 3, we addressed the limitation of ANAs in supporting nerve regeneration. We were able to modify the DDS containing MS to control the GDNF release during either 2 weeks or 4 weeks by adjusting the molecular weight of PLGA used for synthesis of the

MS. In a nerve gap model in rats, the CP nerve was transected and the nerve stumps was immediately coapted via a 10 mm ANA. The GDNF delivery system containing MS was applied around both suture sites of the interposed ANAs. Under these conditions, the number of motor and sensory neurons that regenerated was as high as the number of neurons that regenerated their axons through implanted nerve isografts. In addition, histology and morphometric analysis indicated increase in axon density within the ANAs and better nerve maturity with GDNF MS treatment.

The enhancement of nerve regeneration and maturity was observed within all the GDNF treated groups regardless of whether the period of drug release was 2 or 4 weeks. Hence it is the total amount of GDNF available within the DDS that is important in enhancing nerve regeneration rather than the differences between release periods – 2 weeks vs. 4 weeks. The implications of this chapter are that combining non-immunogenic biomaterials such as ANAs, which provide the appropriate scaffold for axonal growth, with a biocompatible delivery system for biomolecules stimulating nerve regeneration provide significant improvement in nerve regeneration and overcome the limitations associated with ANAs.

The work presented in this thesis made significant contributions to the areas of controlled drug delivery, development of ANAs and treatment of severe nerve injuries. More importantly, the combination of biomaterials developed in this study has the potential to be translated for clinical use and benefit patients following nerve injuries with an “off the

shelf” alternative. In addition, such system can even be adjusted to supplement ANAs with various neuro-regenerative factors.

The enhanced nerve regeneration in response to GDNF delivery via the DDS placed at the surgical repair site of chronically injured nerves and of implanted ANAs between transected nerve stumps is likely to impact the functional and behavioral recovery after such severe nerve injuries. It will be important to investigate target reinnervation and behavioral recovery by analyses of muscle electromyographical signals and contractile forces developed in reinnervated muscles. Further analysis can be done by recording compound action potentials in sensory nerves and by evaluation of sensory function.

By showing that the proposed GDNF-containing DDS can make a 10 mm ANA as effective as an isograft in supporting nerve regeneration, the next step would be to assess the maximum length of the ANA that can allow complete nerve regeneration with the neurotrophic factor support from the current DDS. Once the maximum ANA length is determined, the GDNF mass release properties of the delivery system can be customized to stimulate axonal regeneration through longer ANAs. An interesting future aspect for this investigation would be to identify the cell type migration within the ANAs and analyze any differences observed in the state of the migrated cells and the extent of cell migration through the grafts by increasing the length of the ANAs. In addition, with using the longer nerve grafts, molecular analysis needs to be performed to assess the penetration of GDNF within the ANAs and ensure that sufficient GDNF is available in vivo within the grafts to stimulate regeneration. In order to ensure better penetration of

GDNF within the nerve grafts, placement of the DDS in different locations along the nerve grafts should be explored. For instance, in addition to the implantation of the DDS

around the proximal and distal suture sites, the delivery system could be placed around the middle section of the grafts. In such case, further analysis is necessary to detect possible axonal entrapment within the grafts and to prevent potential formation of neuroma.

There is a clear need to develop therapies that can promote recovery after severe peripheral nerve injuries. I believe that results from this research could lead to such a therapy and help patients who are suffering from disability caused by nerve injuries.

5 Bibliography

Abercrombie, M. (1946). Estimation of nuclear population from microtome sections. Anat. Rec., 94, 239–247.

Aitken, J. (1949). The effect of peripheral connexions on the maturation of regenerating nerve fibres. J Anat, 83(Pt 1), 32–43.

Aitken, J., Sharman, M., & Young, J. (1947). Maturation of regenerating nerve fibres with various peripheral connexions. 1947;81 (Pt 1):1-22.2. J Anat, 81 (Pt1), 1–22.2.

Amrani, D. L., Diorio, J. P., & Delmotte, Y. (2001). Wound healing. Role of commercial fibrin sealants. Annals of the New York Academy of Sciences, 936, 566–79. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11460516

Apfel, S. C., Schwartz, S., Adornato, B. T., Freeman, R., Biton, V., Rendell, M., … Dyck, P. J. (2000). Efficacy and safety of recombinant human nerve growth factor in patients with diabetic polyneuropathy: A randomized controlled trial. rhNGF Clinical Investigator Group. JAMA, 284(17), 2215–21. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11056593

Aprili, D., Bandschapp, O., Rochlitz, C., Urwyler, A., & Ruppen, W. (2009). Serious complications associated with external intrathecal catheters used in cancer pain patients: a systematic review and meta-analysis. Anesthesiology, 111(6), 1346–55. doi:10.1097/ALN.0b013e3181bfab9a

Ashley, W. W., Baty, J. D., Hollander, T., Noetzel, M. J., Park, T. S., & Material, C. (2007). Long-term motor outcome analysis using a motor score composite following surgical brachial plexus repair. Journal of Neurosurgery: Pediatrics, 106(4), 276– 281. doi:10.3171/ped.2007.106.4.276

Avellino, A. M. M., Hart, D., Dailey, A. T., MacKinnon, M., Ellegala, D., & Kliot, M. (1995). Differential macrophage responses in the peripheral and central nervous system during Wallerian degeneration of axons. Neurology, Experimental, 136(2), 183–98. doi:10.1006/exnr.1995.1095

Aydin MA, Mackinnon SE, Gu XM, Kobayashi J, K. W. J. (2004). Force deficits in skeletal muscle after delayed reinnervation. Surg., 113(6), 1712–1718.

Barbacid, M. (1994). The Trk family of neurotrophin receptors. Journal of Neurobiology, 25(11), 1386–403. doi:10.1002/neu.480251107

Barras, F. M., Kuntzer, T., Zurn, A. D., & Pasche, P. (2009). Local delivery of glial cell line-derived neurotrophic factor improves facial nerve regeneration after late repair. The Laryngoscope, 119(5), 846–55. doi:10.1002/lary.20169

Barras, F. M., Pasche, P., Bouche, N., Aebischer, P., & Zurn, A. D. (2002). Glial cell line-derived neurotrophic factor released by synthetic guidance channels promotes facial nerve regeneration in the rat. Journal of Neuroscience Research, 70(6), 746– 55. doi:10.1002/jnr.10434

Belkas, J. S., Shoichet, M. S., & Midha, R. (2004). Peripheral nerve regeneration through guidance tubes. Neurological Research, 26(2), 151–60. doi:10.1179/016164104225013798

Biondi, M., Ungaro, F., Quaglia, F., & Netti, P. A. (2008). Controlled drug delivery in tissue engineering. Advanced Drug Delivery Reviews, 60(2), 229–42. doi:10.1016/j.addr.2007.08.038

Bloch, J., Fine, E. G., Bouche, N., Zurn, A. D., & Aebischer, P. (2001). Nerve growth factor- and neurotrophin-3-releasing guidance channels promote regeneration of the transected rat dorsal root. Experimental Neurology, 172(2), 425–32. doi:10.1006/exnr.2001.7778

Blomback, B., Okada, M., Forslind, B., & Larsson, U. (1984). Fibrin gels as biological filters and interfaces. Biorheology, 21(1-2), 93–104.

Boyd, J. G., & Gordon, T. (2002). A dose-dependent facilitation and inhibition of peripheral nerve regeneration by brain-derived neurotrophic factor. European Journal of Neuroscience, 15(4), 613–626. doi:10.1046/j.1460-9568.2002.01891.x

Boyd, J., & Gordon, T. (2003a). GDNF and BDNF sustain the axonal regeneration of chronically axotomized motoneurons in vivo. Experimental Neurology, 183, 610– 619.

Boyd, J., & Gordon, T. (2003b). Neurotrophic factors and their receptors in axonal regeneration and functional recovery after peripheral nerve injury. Molecular Neurobiology, 27(3), 277–324.

Brazel, C. S., & Peppas, N. a. (2000). Modeling of drug release from Swellable polymers. European Journal of Pharmaceutics and Biopharmaceutics, 49(1), 47–58. doi:10.1016/S0939-6411(99)00058-2

Brooks, D. N., Weber, R. V, Chao, J. D., Rinker, B. D., Zoldos, J., Robichaux, M. R., … Buncke, G. M. (2012). Processed nerve allografts for peripheral nerve reconstruction: a multicenter study of utilization and outcomes in sensory, mixed, and motor nerve reconstructions. Microsurgery, 32(1), 1–14. doi:10.1002/micr.20975

Brushart, T. M., Aspalter, M., Griffin, J. W., Redett, R., Hameed, ., Zhou, C., … ke, A. (2013). Schwann cell phenotype is regulated by axon modality and central-

peripheral location, and persists in vitro. Experimental Neurology, 247, 272–81. doi:10.1016/j.expneurol.2013.05.007

Brushart, T. M., Hoffman, P. N., Royall, R. M., Murinson, B. B., Witzel, C., & Gordon, T. (2002). Electrical stimulation promotes motoneuron regeneration without increasing its speed or conditioning the neuron. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience, 22(15), 6631–8. doi:20026683

Brushart, T. M., & Mesulam, M. M. (1980). Alteration in connections between muscle and anterior horn motoneurons after peripheral nerve repair. Science, 208, 603–605.

Bunge, R. (1994). The role of the Schwann cell in trophic support and regeneration. Journal of Neurology, 242(1), S19.

Caicco, M. J., Cooke, M. J., Wang, Y., Tuladhar, A., Morshead, C. M., & Shoichet, M. S. (2013). A hydrogel composite system for sustained epi-cortical delivery of Cyclosporin A to the brain for treatment of stroke. Journal of Controlled Release : Official Journal of the Controlled Release Society, 166(3), 197–202. doi:10.1016/j.jconrel.2013.01.002

Cajal, S. R. (1928). Degeneration and Regeneration of the Nervous System. New York: Oxford University Press.

Cao, X., & Shoichet, M. (2001). Defining the concentration gradient of nerve growth factor for guided neurite outgrowth. Neurosci, 103(3), 831–840.

Carr, E., & Hardin, C. L. (1987). Fibrin has larger pores when formed the presence of erythrocytes in, (2), 1069–1073.

Chan, K. M., Gordon, T., Zochodne, D. W., & Power, H. a. (2014). Improving peripheral nerve regeneration: from molecular mechanisms to potential therapeutic targets. Experimental Neurology, 261, 826–35. doi:10.1016/j.expneurol.2014.09.006

Chao, M. V, Bothwell, M. A., Ross, A. H., Koprowski, H., Lanahan, A. A., Buck, C. R., & Sehgal, A. (1986). Gene transfer and molecular cloning of the human NGF receptor. Science (New York, N.Y.), 232(4749), 518–21. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/3008331

Chao, M. V, & Hempstead, B. L. (1995). p75 and Trk: a two-receptor system. Trends in Neurosciences, 18(7), 321–6. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/7571013

Chun, K. W., Yoo, H. S., Yoon, J. J., & Park, T. G. (2004). Biodegradable PLGA microcarriers for injectable delivery of chondrocytes: effect of surface modification on cell attachment and function. Biotechnology Progress, 20(6), 1797–801. doi:10.1021/bp0496981

Ciaramitaro, P., Mondelli, M., Logullo, F., Grimaldi, S., Battiston, B., Sard, A., & Cocito, D. (2010). Traumatic peripheral nerve injuries: epidemiological findings, neuropathic pain and quality of life in 158 patients. Journal of the Peripheral Nervous System : JPNS, 15(2), 120–127.

Cohen-Sela, E., Chorny, M., Koroukhov, N., Danenberg, H. D., & Golomb, G. (2009). A new double emulsion solvent diffusion technique for encapsulating hydrophilic molecules in PLGA nanoparticles. Journal of Controlled Release : Official Journal of the Controlled Release Society, 133(2), 90–5. doi:10.1016/j.jconrel.2008.09.073

Corkhill, P. H., Hamilton, C. J., & Tighe, B. J. (1989). Synthetic hydrogels. VI. Hydrogel composites as wound dressings and implant materials. Biomaterials, 10(1), 3–10. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2653447

Danielsen, N., Kerns, J. M., Holmquist, B., Zhao, Q., Lundborg, G., & Kanje, M. (1995). Predegeneration enhances regeneration into acellular nerve grafts. Brain Research, 681(1-2), 105–8. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/7552266

David, S., Braun, P. E., Jackson, D. L., Kottis, V., & McKerracher, L. (1995). Laminin overrides the inhibitory effects of peripheral nervous system and central nervous system myelin-derived inhibitors of neurite growth. Journal of Neuroscience Research, 42(4), 594–602. doi:10.1002/jnr.490420417

Dugandgija-novakovic, S., Rock, S., & Koszowski, G. (1995). Clustering of Na + Channels Remyelinating Axons and Node of Ranvier Formation in. The Journal of Neuroscience, 15(1), 492–503.

Dunn, C. J., & Goa, K. L. (1999). Fibrin sealant: a review of its use in surgery and endoscopy. Drugs, 58(5), 863–86. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10595866

Eggers, R., Hendriks, W., Tannemaat, M., van Heerikhuize, J., Pool, C., & Carlstedt, T. (2008). Neuroregenerative effects of lentiviral vector-mediated GDNF expression in reimplanted ventral roots. Mol Cell Neurosci, 39(1), 105–117.

Ehretsman, R. L., Novak, C. B., & Mackinnon, S. E. (1999). Subjective recovery of nerve graft donor site. Ann Plast Surg, 43, 606.

Eigenbrot, C., & Gerber, N. (1997). X-ray structure of glial cell-derived neurotrophic factor at 1.9 A resolution and implications for receptor binding. Nature Structural Biology, 4(6), 435–8. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9187648

Ejstrup, R., Kiilgaard, J. F., Tucker, B. A., Klassen, H. J., Young, M. J., & La Cour, M. (2010). Pharmacokinetics of intravitreal glial cell line-derived neurotrophic factor: experimental studies in pigs. Experimental Eye Research, 91(6), 890–5. doi:10.1016/j.exer.2010.09.016

Evans, P. J., Mackinnon, S. E., Best, T. J., Wade, J. A., Awerbuck, D. C., Makino, A. P., … Midha, R. (1995). Regeneration across preserved peripheral nerve grafts. Muscle & Nerve, 18(10), 1128–38. doi:10.1002/mus.880181009

Evans, P. J., MacKinnon, S. E., Midha, R., Wade, J. A., Hunter, D. A., Nakao, Y., & Hare, G. M. (1999). Regeneration across cold preserved peripheral nerve allografts. Microsurgery, 19(3), 115–27. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10231120

Evans, P. J., Midha, R., & Mackinnon, S. E. (1994). The peripheral nerve allograft: a comprehensive review of regeneration and neuroimmunology. Prog Neurobiol, 43, 187.

Evans, P., Mackinnon, S., Levi, A., Wade, J., Hunter, D., & Nakao, Y. (1998). Cold preserved nerve allografts: changes in basement membrane, viability, immunogenicity, and regeneration. Muscle Nerve, 21, 1507–1522.

Eyzaguirre, C., & Kuffler, S. W. (1955). Processes of excitation in the dendrites and in the soma of single isolated sensory nerve cells of the lobster and crayfish. The Journal of General Physiology, 39(1), 87–117.

Fine, E. G., Decosterd, I., Papaloïzos, M., Zurn, A. D., & Aebischer, P. (2002). GDNF and NGF released by synthetic guidance channels support sciatic nerve regeneration across a long gap. The European Journal of Neuroscience, 15(4), 589–601. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11886440

Follett, K. A., Boortz-Marx, R. L., Drake, J. M., DuPen, S., Schneider, S., Turner, M., & Coffey, R. (2004). Prevention and management of intrathecal drug delivery and spinal cord stimulation system infections. Anesthesiology, 100, 1582–1594.

Fox, I. K., Jaramillo, A., Hunter, D. A., Rickman, S. R., Mohanakumar, T., & Mackinnon, S. E. (2005). Prolonged cold-preservation of nerve allografts. Muscle & Nerve, 31(1), 59–69. doi:10.1002/mus.20231

Fox, I. K., & Mackinnon, S. E. (2011). Adult peripheral nerve disorders: nerve entrapment, repair, transfer, and brachial plexus disorders. Plast Reconstr Surg, 127(105e). doi:10.1097/PRS.0b013e31820cf556.Adult

Fraher, J., & Dockery, P. (1998). A strong myelin thickness-axon size correlation emerges in developing nerves despite independent growth of both parameters. J Anat, 193(Pt 2), 195–201.

Friedlander, D. R., Grumet, M., & Edelman, G. M. (1986). Nerve growth factor enhances expression of neuron-glia cell adhesion molecule in PC12 cells. The Journal of Cell Biology, 102(2), 413–9. Retrieved from

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2114093&tool=pmcentr ez&rendertype=abstract

Fu, S., & Gordon, T. (1995a). Contributing factors to poor functional recovery after delayed nerve repair: Prolonged axotomy. J Neurosci, 15, 3886–3895.

Fu, S., & Gordon, T. (1995b). Contributing factors to poor functional recovery after delayed nerve repair: prolonged denervation. J Neurosci, 15, 3876–3885.

Garbayo, E., Montero-Menei, C., Ansorena, E., Lanciego, J., Aymerich, M., & Blanco- Prieto, M. (2009). Effective GDNF brain delivery using microspheres—A promising strategy for Parkinson’s disease. J Control Release, 135(2), 119–126.

Glasby, M. A., Gschmeissner, S. G., Hitchcock, R. J., & Huang, C. L. (1986a). The dependence of nerve regeneration through muscle grafts in the rat on the availability and orientation of basement membrane. Journal of Neurocytology, 15(4), 497–510. Retrieved from http://link.springer.com/article/10.1007/BF01611732

Glasby, M. A., Gschmeissner, S., Hitchcock, R. J., & Huang, C. L. (1986b). Regeneration of the sciatic nerve in rats. The effect of muscle basement membrane. The Journal of Bone and Joint Surgery. British Volume, 68(5), 829–33. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/3782256

Goldman, L., & Albus, J. S. (1968). Computation of impulse conduction in myelinated fibers; theoretical basis of the velocity-diameter relation. Biophysical Journal, 8(5), 596–607.

Gordon, T. (1983). Dependence of peripheral nerves on their target organs. In G. Burnstock, R. O’Brien, & G. Vrbove (Eds.), Somatic and autonomic nerve-muscle interactions. (pp. 289–316). Amsterdam: Elsevier.

Gordon, T. (2015). The biology, limits, and promotion of peripheral nerve regeneration in rats and humans. In R. Shane Tubbs, E. Rizk, J. Robert, M. M. Shoja, M. Loukas, & N. Barbaro (Eds.), Nerves and Nerve Injuries: Vol 2: Pain, Treatment, Injury, Disease and Future Directions (pp. 993–1019). Academic Press.

Gordon, T., & FU, S. (1997). The cellular and molecular basis of peripheral nerve regeneration. Mol Neurobiol, 14, 67–116.

Gordon, T., & Stein, R. B. (1982). Time course and extent of recovery in reinnervated motor units of cat triceps surae muscles. The Journal of Physiology, 323, 307–23. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1250358&tool=pmcentr ez&rendertype=abstract

Grand, A. G., Myckatyn, T. M., Mackinnon, S. E., & Hunter, D. A. (2002). Axonal regeneration after cold preservation of nerve allografts and immunosuppression with tacrolimus in mice. Journal of Neurosurgery, 96(5), 924–32. doi:10.3171/jns.2002.96.5.0924

Guilhem, I., Balkau, B., Lecordier, F., Malecot, J., Elbadii, S., & Leguerrier, A. (2009). Insulin pump failures are still frequent: a prospective study over 6 years from 2001 to 2007. Diabetologia, 52(12), 2662–2664.

Gulati, A., & Cole, G. (1990). Nerve graft immunogenicity as a factor determining axonal regeneration in the rat. J Neurosurg, 72, 114–122.

Gulati, A. K. (1988). Evaluation of acellular and cellular nerve grafts in repair of rat peripheral nerve. Journal of Neurosurgery, 68(1), 117–23. doi:10.3171/jns.1988.68.1.0117

Gulati, A. K. (1998). Immune response and neurotrophic factor interactions in peripheral nerve transplants. Acta Haematologica, 99(3), 171–4. doi:40832

Gulati, A. K., & Cole, G. P. (1994). Immunogenicity and regenerative potential of acellular nerve allografts to repair peripheral nerve in rats and rabbits. Acta Neurochirurgica, 126(2-4), 158–64. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8042549

Haftek, J., & Thomas, P. K. (1968). Electron-microscope observations on the effects of localized crush injuries on the connective tissues of peripheral nerve. J. Anat., 103, 233–243.

Hall, S. (1997). Axonal regeneration through acellular muscle grafts. J Anat., 190, 57–71.

Hall, S. M. (1986). Regeneration in cellular and acellular autografts in the peripheral nervous system. Neuropathology and Applied Neurobiology, 12(1), 27–46. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/3703154

are, G. M., Evans, P. J., Mackinnon, S. E., Nakao, Y., Midha, R., Wade, J. A., … ay, J. B. (1993). Effect of cold preservation on lymphocyte migration into peripheral nerve allografts in sheep. Transplantation, 56(1), 154–62. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8333038

Henderson, C. E., Phillips, H. S., Pollock, R. A., Davies, A. M., Lemeulle, C., Armanini, M., … Rosenthal, A. (1994). GDNF: a potent survival factor for motoneurons present in peripheral nerve and muscle. Science (New York, N.Y.), 266(5187), 1062– 4. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/7973664

Hirata, K., & Kawabuchi, M. (2002). Myelin phagocytosis by macrophages and nonmacrophages during Wallerian degeneration. Microsc Res Tech, 57, 541.

Hodgkin, A., & Huxley, A. (1952). A quantitative description of membrane current and its application to conduction and excitation in nerve. J Physiol, 117, 500–544.

Hodler, J., Stampfli, R., & Tasaki, I. (1952). Role of Potential Wave Spreading Along Myelinated Fiber in Excitation and Conduction. American Journal of Physiology, 170, 375–389.

Höke, A., Gordon, T., Zochodne, D. W., & Sulaiman, O. A. R. (2002). A decline in glial cell-line-derived neurotrophic factor expression is associated with impaired regeneration after long-term Schwann cell denervation. Experimental Neurology, 173(1), 77–85. doi:10.1006/exnr.2001.7826

ke, A., Redett, R., ameed, ., Jari, R., Zhou, C., Li, Z. B., … Brushart, T. M. (2006). Schwann cells express motor and sensory phenotypes that regulate axon regeneration. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience, 26(38), 9646–55. doi:10.1523/JNEUROSCI.1620-06.2006

Holmes, W. Z., & Young, J. (1942). Nerve regeneration after immediate and delayed suture. J. Anat., 77, 63–I06.

Hudson, T. W., Liu, S. Y., & Schmidt, C. E. (2004). Engineering an improved acellular nerve graft via optimized chemical processing. Tissue Engineering, 10(9-10), 1346– 58. doi:10.1089/ten.2004.10.1346

Hudson, T. W., Zawko, S., Deister, C., Lundy, S., Hu, C. Y., Lee, K., & Schmidt, C. E. (2004). Optimized acellular nerve graft is immunologically tolerated and supports regeneration. Tissue Engineering, 10(11-12), 1641–51. doi:10.1089/ten.2004.10.1641

Ide, C., Osawa, T., & Tohyama, K. (1990). Nerve regeneration through allogeneic nerve grafts, with special reference to the role of the Schwann cell basal lamina. Prog Neurobiol, 1, 34.

Ide, C., Tohyama, K., Yokota, R., Nitatori, T., & Onodera, S. (1983). Schwann cell basal lamina and nerve regeneration. Brain Research, 288(1-2), 61–75. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/6661636

Jacobsen, S., & Guth, L. (1965). An electrophysiological study of the early stages of peripheral nerve regeneration. Exp. Neurol., 11, 48–60.

Jarvis, M. F., & Boyce-Rustay, J. M. (2009). Neuropathic pain: models and mechanisms. Curr Pharm Des, 15, 1711.

Jesuraj, N. J., Santosa, K. B., Macewan, M. R., Moore, A. M., Kasukurthi, R., Ray, W. Z., … Sakiyama-Elbert, S. E. (2014). Schwann cells seeded in acellular nerve grafts

improve functional recovery. Muscle & Nerve, 49(2), 267–76. doi:10.1002/mus.23885

Jing, S., Wen, D., Yu, Y., Holst, P. L., Luo, Y., Fang, M., … Fox, G. M. (1996). GDNF- induced activation of the ret protein tyrosine kinase is mediated by GDNFR-alpha, a novel receptor for GDNF. Cell, 85(7), 1113–24. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8674117

Johnson, P. C., Duhamel, R. C., Meezan, E., & Brendel, K. (1982). Preparation of cell- free extracellular matrix from human peripheral nerve. Muscle & Nerve, 5(4), 335– 44. doi:10.1002/mus.880050410

Jubran, M., & Widenfalk, J. (2003). Repair of peripheral nerve transections with fibrin sealant containing neurotrophic factors. Experimental Neurology, 181(2), 204– 212.

Kandel, E. R., Schwartz, J. H., Jessell, T. M., Siegelbaum, S. A., & Hudspeth, A. J. (2013). Principles of Nueral Sciene (Fifth Eidi.). New York: McGraw Hill.

Kang, F., & Singh, J. (2001). Effect of additives on the release of a model protein from PLGA microspheres. AAPS PharmSciTech, 2(4), 30. doi:10.1208/pt020430

Kashyap, N., Kumar, N., & Kumar, M. N. V. R. (2005). Hydrogels for pharmaceutical and biomedical applications. Critical Reviews in Therapeutic Drug Carrier Systems, 22(2), 107–49. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/15862110

Kim, H. K., & Park, T. G. (1999). Microencapsulation of human growth hormone within biodegradable polyester microspheres: protein aggregation stability and incomplete release mechanism. Biotechnology and Bioengineering, 65(6), 659–67. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10550772

Kim, H., Tator, C. H., & Shoichet, M. S. (2008). Design of protein-releasing chitosan channels. Biotechnology Progress, 24(4), 932–7. doi:10.1021/bp070352a

Kim, J.-H., Kim, J.-H., Park, J.-A., Lee, S.-W., Kim, W.-J., Yu, Y.-S., & Kim, K.-W. (2006). Blood-neural Barrier: Intercellular Communication at Glio-Vascular Interface. Journal of Biochemistry and Molecular Biology, 39(4), 339–345. doi:10.5483/BMBRep.2006.39.4.339

Kline, D. G. (1990). Surgical repair of peripheral nerve injury. Muscle & Nerve, 13(9), 843–852.

Kline, D. G., & Hudson, A. R. (1995). Nerve Injuries: Operative Results for Major Nerve Injuries, Entrapments and Tumors. Philadelphia: W.B. Saunders.

Kokai, L., Bourbeau, D., Weber, D., McAtee, J., & Marra, K. (2011). Sustained growth factor delivery promotes axonal regeneration in long gap peripheral nerve repair. Tissue Eng Part A, 17(9–10), 1263–1275.

Kouyoumdjian, J. (2006). Peripheral nerve injuries: a retrospective survey of 456 cases. Muscle Nerve, 34(6), 785–788.

Krekoski, C. A., Neubauer, D., Zuo, J., & Muir, D. (2001). Axonal regeneration into acellular nerve grafts is enhanced by degradation of chondroitin sulfate proteoglycan. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience, 21(16), 6206–13. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11487643

Kroll, D., Caplan, R., & Posner, K. (1990). Nerve injury associated with anesthesia. Anesthesiology, 73(2), 202–207.

Levi, A. D., Evans, P. J., Mackinnon, S. E., & Bunge, R. P. (1994). Cold storage of peripheral nerves: an in vitro assay of cell viability and function. Glia, 10(2), 121– 31. doi:10.1002/glia.440100206

Lewin, S. L., Utley, D. S., Cheng, E. T., Verity, A. N., & Terris, D. J. (1997). Simultaneous treatment with BDNF and CNTF after peripheral nerve transection and repair enhances rate of functional recovery compared with BDNF treatment alone. The Laryngoscope, 107(7), 992–9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9217144

Li, Z., & Huang, L. (2004). Sustained delivery and expression of plasmid DNA based on biodegradable polyester, poly(D,L-lactide-co-4-hydroxy-L-proline). Journal of Controlled Release : Official Journal of the Controlled Release Society, 98(3), 437– 46. doi:10.1016/j.jconrel.2004.05.013

Lin, C.-C., & Metters, A. T. (2006). Hydrogels in controlled release formulations: network design and mathematical modeling. Advanced Drug Delivery Reviews, 58(12-13), 1379–408. doi:10.1016/j.addr.2006.09.004

Lin, L. F., Doherty, D. H., Lile, J. D., Bektesh, S., & Collins, F. (1993). GDNF: a glial cell line-derived neurotrophic factor for midbrain dopaminergic neurons. Science (New York, N.Y.), 260(5111), 1130–2. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8493557

Lindsay, R. M. (1996). Role of neurotrophins and trk receptors in the development and maintenance of sensory neurons: an overview. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 351(1338), 365–73. doi:10.1098/rstb.1996.0030

100

Lopez-Vales, R., Navarro, X., Shimizu, T., Baskakis, C., & Ko- kotos, G., Constantinou- Kokotou, V. (2008). Intracellular phospholipase A(2) group IVA and group VIA play important roles in Wallerian degeneration and axon regeneration after peripheral nerve injury. Brain, 131(10), 2620,.

Lundborg, G. (2000). A 25-year perspective of peripheral nerve surgery: evolving neuroscientific concepts and clinical significance. J Hand Surg, 25(3), 391–414.

Lundborg, G. (2003). Nerve injury and repair - a challenge to the plastic brain. J Peripher Nerv Syst, 8, 209–226.

Lundborg, G. (2004). Nerve Injury and Repair. New York: Churchill Livingstone.

Lundborg, G., & Rosen, B. (2007). Hand function after nerve repair. Acta Physiol. (Oxf.), 189(2), 207–217.

Macario, A. (2010). What does one minute of operating room time cost? J Clin Anesth, 22, 233.

Mackinnon, S. E., & Dellon, A. L. (1988). Surgery of the Peripheral Nerve. New York: Thieme Medical Publishers Incorporated.

Mackinnon, S. E., Doolabh, V. B., Novak, C. B., & Trulock, E. P. (2001). Clinical outcome following nerve allograft transplantation. Plast Reconstr Surg, 107, 1419.

Mackinnon, S. E., Hudson, A. R., Bain, J. R., Falk, R. E., & Hunter, D. A. (1987). The peripheral nerve allograft: an assessment of regeneration in the immunosuppressed host. Plastic and Reconstructive Surgery, 79(3), 436–46. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/3823218

Martini, R., Fischer, S., López-Vales, R., & David, S. (2008). Interactions between Schwann cells and macrophages in injury and inherited demyelinating disease. Glia, 56(14), 1566–77. doi:10.1002/glia.20766

Maysinger, D., & Morinville, A. (1997). Drug delivery to the nervous system. Trends in Biotechnology, 15(10), 410–8. doi:10.1016/S0167-7799(97)01095-0

Meek, M., & Coert, J. (2002). Clinical use of nerve conduits in peripheral nerve repair: review of the literature. J Reconstr Microsurg., 18(2), 97–109.

Meek, M. F., Robinson, P. H., Stokroos, I., Blaauw, E. H., Kors, G., & den Dunnen, W. F. (2001). Electronmicroscopical evaluation of short-term nerve regeneration through a thin-walled biodegradable poly(DLLA-epsilon-CL) nerve guide filled with modified denatured muscle tissue. Biomaterials, 22(10), 1177–85. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11352097

101

Millesi, H. (1981). Reappraisal of nerve repair. The Surgical Clinics of North America, 61(2), 321.

Millesi, H. (1985). Peripheral nerve repair: terminology, questions, and facts. J. Reconstr. Microsurg, 2(1), 21–31.

Millesi, H., Meissl, G., & Berger, A. (1972). The interfascicular nerve grafting of the median and ulnar nerves. The Journal of Bone and Joint Surgery, 54(4), 727–750.

Mohtaram, N. K., Montgomery, A., & Willerth, S. M. (2013). Biomaterial-based drug delivery systems for the controlled release of neurotrophic factors. Biomedical Materials (Bristol, England), 8(2), 022001. doi:10.1088/1748-6041/8/2/022001

Moore, A. M., Borschel, G. H., Santosa, K. a, Flagg, E. R., Tong, A. Y., Kasukurthi, R., … Mackinnon, S. E. (2012). A transgenic rat expressing green fluorescent protein (GFP) in peripheral nerves provides a new hindlimb model for the study of nerve injury and regeneration. Journal of Neuroscience Methods, 204(1), 19–27. doi:10.1016/j.jneumeth.2011.10.011

Moore, A. M., MacEwan, M., Santosa, K. B., Chenard, K. E., Ray, W. Z., Hunter, D. A., … Johnson, P. J. (2011). Acellular nerve allografts in peripheral nerve regeneration: a comparative study. Muscle & Nerve, 44(2), 221–34. doi:10.1002/mus.22033

Moore, A. M., Wood, M. D., Chenard, K., Hunter, D. a, Mackinnon, S. E., Sakiyama- Elbert, S. E., & Borschel, G. H. (2010). Controlled delivery of glial cell line-derived neurotrophic factor enhances motor nerve regeneration. The Journal of Hand Surgery, 35(12), 2008–17. doi:10.1016/j.jhsa.2010.08.016

More, H. L., Chen, J., Gibson, E., Donelan, J. M., & Beg, M. F. (2011). A semi- automated method for identifying and measuring myelinated nerve fibers in scanning electron microscope images. Journal of Neuroscience Methods, 201(1), 149–58. doi:10.1016/j.jneumeth.2011.07.026

Morita, T., Sakamura, Y., Horikiri, Y., Suzuki, T., & Yoshino, H. (2000). Protein encapsulation into biodegradable microspheres by a novel S/O/W emulsion method using poly(ethylene glycol) as a protein micronization adjuvant. Journal of Controlled Release : Official Journal of the Controlled Release Society, 69(3), 435– 44. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11102683

Morris, J. H., Hudson, A. R., & Weddell, G. (1972a). A study of degeneration and regeneration in the divided rat sciatic nerve based on electron microscopy. I. The traumatic degeneration of myelin in the proximal stump of the divided nerve. Z Zellforsch Mikrosk Anat, 124, 76.

102

Morris, J. H., Hudson, A. R., & Weddell, G. (1972b). A study of degeneration and regeneration in the divided rat sciatic nerve based on electron microscopy. II. The development of the “‘regenerating unit’”. Z Zellforsch Mikrosk Anat, 124, 103.

Muir, D., Engvall, E., Varon, S., & Manthorpe, M. (1989). Schwannoma cell-derived inhibitor of the neurite-promoting activity of laminin. The Journal of Cell Biology, 109(5), 2353–62. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2115873&tool=pmcentr ez&rendertype=abstract

Mumper, R. J., & Jay, M. (1992). Biodegradable radiotherapeutic polyester microspheres: optimization and in-vitro/in-vivo evaluation. Journal of Controlled Release, 18(3), 193–203. doi:10.1016/0168-3659(92)90165-N

Mundargi, R. C., Babu, V. R., Rangaswamy, V., Patel, P., & Aminabhavi, T. M. (2008). Nano/micro technologies for delivering macromolecular therapeutics using poly(D,L-lactide-co-glycolide) and its derivatives. Journal of Controlled Release : Official Journal of the Controlled Release Society, 125(3), 193–209. doi:10.1016/j.jconrel.2007.09.013

Nadim, W., Anderson, P. N., & Turmaine, M. (1990). The role of Schwann cells and basal lamina tubes in the regeneration of axons through long lengths of freeze-killed nerve grafts. Neuropathology and Applied Neurobiology, 16(5), 411–21. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2263316

Neubauer, D., Graham, J. B., & Muir, D. (2007). Chondroitinase treatment increases the effective length of acellular nerve grafts. Experimental Neurology, 207(1), 163–70. doi:10.1016/j.expneurol.2007.06.006

Neubauer, D., Graham, J., & Muir, D. (2004). Chondroitinase treatment increases the effective length of acellular nerve grafts. Exp Neurol, 207, 163–170.

Noble, J., Munro, C., Prasad, V., & Midha, R. (1998). Analysis of Upper and Lower Extremity Peripheral Nerve Injuries in a Population of Patients with Multiple Injuries. The Journal of Trauma: Injury, Infection, and Critical Care, 45(1), 116– 122.

Oppenheim, R. W. (1991). Cell death during development of the nervous system. Annual Review of Neuroscience, 14, 453–501. doi:10.1146/annurev.ne.14.030191.002321

Osawa, T., Ide, C., & Tohyama, K. (1987). Nerve regeneration through cryo-treated xenogeneic nerve grafts. Archivum Histologicum Japonicum = Nihon Soshikigaku Kiroku, 50(2), 193–208. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/3662750

103

Pe, J., Menei, P., Morel, O., Montero-menei, C. N., & Benoit, J. (2000). Intraseptal implantation of NGF-releasing microspheres promote the survival of axotomized cholinergic neurons, 21, 2097–2101.

Peppas, N. A., Bures, P., Leobandung, W., & Ichikawa, H. (2000). Hydrogels in pharmaceutical formulations. European Journal of Pharmaceutics and Biopharmaceutics : Official Journal of Arbeitsgemeinschaft Für Pharmazeutische Verfahrenstechnik e.V, 50(1), 27–46. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10840191

Perez, N., Sosa, M., & Kuffler, D. (1997). Growth cones turn up concentration gradients of diffusible peripheral target-derived factors. Exp Neurol, 145(1), 196–202.

Perry, V. H., Tsao, J. W., Fearn, S., & Brown, M. C. (1995). Radiation-induced Reductions in Macrophaae Recruitment Have Only Slight Effects on Myelin Degeneration in Sectioned Peripheral Nerves of Mice. The European Journal of Neuroscience, 7(2), 271–280.

Pfister, L. A., Papaloïzos, M., Merkle, H. P., & Gander, B. (2007). Nerve conduits and growth factor delivery in peripheral nerve repair. Journal of the Peripheral Nervous System : JPNS, 12(2), 65–82. doi:10.1111/j.1529-8027.2007.00125.x

Poduslo, J. F., & Curran, G. L. (1996). Permeability at the blood-brain and blood-nerve barriers of the neurotrophic factors: NGF, CNTF, NT-3, BDNF. Brain Research. Molecular Brain Research, 36(2), 280–6. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8965648

Pollauf, E. J., Kim, K. K., & Pack, D. W. (2005). Small-molecule release from poly(D,L- lactide)/poly(D,L-lactide-co-glycolide) composite microparticles. Journal of Pharmaceutical Sciences, 94(9), 2013–22. doi:10.1002/jps.20408

Purves, D. (1986). The trophic theory of neural concentrations. Trends in Neurosciences, 9, 486–489. doi:10.1016/0166-2236(86)90155-4

Ravi Kumar, M. N. (2000). Nano and microparticles as controlled drug delivery devices. Journal of Pharmacy & Pharmaceutical Sciences : A Publication of the Canadian Society for Pharmaceutical Sciences, Société Canadienne Des Sciences Pharmaceutiques, 3(2), 234–58. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10994037

Reynolds, M. L., & Woolf, C. J. (1993). Reciprocal Schwann cell-axon interactions. Current Opinion in Neurobiology, 3(5), 683–93. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8260817

Saarma, M., & Sariola, H. (1999). Other neurotrophic factors: glial cell line-derived neurotrophic factor (GDNF). Microscopy Research and Technique, 45(4), 292–302.

104

Sakiyama-Elbert, S. E., & Hubbell, J. A. (2000). Development of fibrin derivatives for controlled release of heparin-binding growth factors. Journal of Controlled Release : Official Journal of the Controlled Release Society, 65(3), 389–402. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10699297

Sameem, M., Wood, T., & Bain, J. (2011). A systematic review on the use of fibrin glue for peripheral nerve repair. Plast Reconstr Surg, 127(6), 2381–2390.

Santos, X., Rodrigo, J., Hontanilla, B., & Bilbao, G. (1998). Evaluation of peripheral nerve regeneration by nerve growth factor locally administered with a novel system. Journal of Neuroscience Methods, 85(1), 119–27. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9874148

Schmidt, C. E., & Leach, J. B. (2003). Neural tissue engineering: strategies for repair and regeneration. Annual Review of Biomedical Engineering, 5, 293–347. doi:10.1146/annurev.bioeng.5.011303.120731

Schröder, J. M., & Seiffert, K. E. (1970). [The fine structure of the neuromatous neurotization of nerve transplants]. Virchows Archiv. B: Cell Pathology, 5(3), 219– 35. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/4987363

Seddon, H. J. (1943). Three Types of Nerve Injury. Brain, 66(4), 238–283.

Seilheimer, B., & Schachner, M. (1987). Regulation of neural cell adhesion molecule expression on cultured mouse Schwann cells by nerve growth factor. The EMBO Journal, 6(6), 1611–6. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=553532&tool=pmcentre z&rendertype=abstract

Siironen, J., Vuorinen, V., Taskinen, H. S., & Röyttä, M. (1995). Axonal regeneration into chronically denervated distal stump 1. Electron microscope studies. Acta Neuropathot, 89, 209–218.

Singh, R. (1976). Experience with allografts in the surgery of peripheral nerves (experimental study). Acta Neurochirurgica, 34(1-4), 195–201. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/961479

Sinha, V. R., & Trehan, A. (2003). B iodegradable microspheres for protein delivery, 90, 261–280.

Sondell, M., Lundborg, G., & Kanje, M. (1998). Regeneration of the rat sciatic nerve into allografts made acellular through chemical extraction. Brain Research, 795(1-2), 44–54. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9622591

105

Soppimath, K. S., Aminabhavi, T. M., Kulkarni, A. R., & Rudzinski, W. E. (2001). Biodegradable polymeric nanoparticles as drug delivery devices. Journal of Controlled Release, 70(1-2), 1–20. doi:10.1016/S0168-3659(00)00339-4

Sterne, G. D., Brown, R. A., Green, C. J., & Terenghi, G. (1997). Neurotrophin-3 delivered locally via fibronectin mats enhances peripheral nerve regeneration. The European Journal of Neuroscience, 9(7), 1388–96. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9240396

Strasberg, S. R., Hertl, M. C., Mackinnon, S. E., Lee, C. K., Watanabe, O., Tarasidis, G., … Wong, P. Y. (1996). Peripheral nerve allograft preservation improves regeneration and decreases systemic cyclosporin A requirements. Experimental Neurology, 139(2), 306–16. doi:10.1006/exnr.1996.0104

Sulaiman, O., & Gordon, T. (2009). Role of chronic Schwann cell denervation in poor functional recovery after nerve injuries and experimental strategies to combat it. Neurosurgery, 65(4 Suppl, A105–14.

Sunderland, I. R. P., Brenner, M. J., Singham, J., Rickman, S. R., Hunter, D. A., & Mackinnon, S. E. (2004). Effect of Tension on Nerve Regeneration in Rat Sciatic Nerve Transection Model. Annals of Plastic Surgery, 53(4). Retrieved from http://journals.lww.com/annalsplasticsurgery/Fulltext/2004/10000/Effect_of_Tensio n_on_Nerve_Regeneration_in_Rat.16.aspx

Sunderland, S. (1978). Nerve and Nerve Injuries. Edinburgh: Livingstone,.

Sunderland, S., & Bradley, K. C. (1950). Endoneuriat tube shrinkage in the distal segment of a severed nerve. J. Comp. NeuroI., 93, 411–420.

Szynkaruk, M., Kemp, S. W. P., Wood, M. D., Gordon, T., & Borschel, G. H. (2013). Experimental and clinical evidence for use of decellularized nerve allografts in peripheral nerve gap reconstruction. Tissue Engineering. Part B, Reviews, 19(1), 83–96. doi:10.1089/ten.TEB.2012.0275

Tajdaran, K., Shoichet, M. S., Gordon, T., & Borschel, G. H. (2015). A novel polymeric drug delivery system for localized and sustained release of tacrolimus (FK506). Biotechnology and Bioengineering. doi:10.1002/bit.25598

Takaku, S. (2013). GDNF promotes neurite outgrowth and upregulates galectin-1 through the RET/PI3K signaling in cultured adult rat dorsal root ganglion neurons. Neurochemistry International, 63(3), 330–339.

Terenghi, G. (1999). Peripheral nerve regeneration and neurotrophic factors. Journal of Anatomy, 194(1), 1–14. doi:10.1046/j.1469-7580.1999.19410001.x

106

Terenghi, G., & Wiberg, M. (2003). Will it be possible to produce peripheral nerves? Surg Technol Int, 11, 303–10.

Terzis, J., Faibisoff, B., & Williams, B. (1975). The nerve gap: suture under tension vs. graft. Plast Reconstr Surg, 56, 166.

Thomas, C. K., Stein, R. B., Gordon, T., Lee, R. G., & Elleker, M. G. (1987). Patterns of reinnervation and motor unit recruitment in human hand muscles after complete ulnar and median nerve section and resuture. J. Neurol. Neurosurg. Psychiatry, 50(3), 259–268.

Vial, J. D. (1958). The early changes in the axoplasm during wallerian degeneration. J Biophys Biochem Cytol, 4, 551.

Waldram, M. (2003). Peripheral nerve injuries. Trauma, 5(2), 79–96.

Waller, A. (1850). Experiments on the section of the glossopharyngeal and hypoglossal nerves of the frog, and observations of the alterations produced thereby in the structure of their primitive fibres. Phil Trans Royal Soc London, 140, 423.

Wang, D., Liu, X.-L., Zhu, J.-K., Jiang, L., u, J., Zhang, Y., … Yi, J.-H. (2008). Bridging small-gap peripheral nerve defects using acellular nerve allograft implanted with autologous bone marrow stromal cells in primates. Brain Research, 1188, 44–53. doi:10.1016/j.brainres.2007.09.098

Webber, C., & Zochodne, D. (2010). The nerve regenerative microenvironment: early behavior and partnership of axons and Schwann cells. Experimental Neurology, 223(1), 51–9. doi:10.1016/j.expneurol.2009.05.037

Weerasuriya, A., & Mizisin, A. P. (2011). The blood-nerve barrier: structure and functional significance. Methods Mol Biol, 686(149).

Whitlock, E. L., Tuffaha, S. ., Luciano, J. P., Yan, Y., unter, D. a, Magill, C. K., … Borschel, G. H. (2009). Processed allografts and type I collagen conduits for repair of peripheral nerve gaps. Muscle & Nerve, 39(6), 787–99. doi:10.1002/mus.21220

Wichterle, O., & LÍM, D. (1960). Hydrophilic Gels for Biological Use. Nature, 185, 117– 118.

Willerth, S. M., Arendas, K. J., Gottlieb, D. I., & Sakiyama-Elbert, S. E. (2006). Optimization of fibrin scaffolds for differentiation of murine embryonic stem cells into neural lineage cells. Biomaterials, 27(36), 5990–6003. doi:10.1016/j.biomaterials.2006.07.036

107

Wischke, C., & Schwendeman, S. P. (2008). Principles of encapsulating hydrophobic drugs in PLA/PLGA microparticles. International Journal of Pharmaceutics, 364(2), 298–327. doi:10.1016/j.ijpharm.2008.04.042

Wood, M. D., Borschel, G. H., & Sakiyama-Elbert, S. E. (2009). Controlled release of glial-derived neurotrophic factor from fibrin matrices containing an affinity-based delivery system. Journal of Biomedical Materials Research. Part A, 89(4), 909–18. doi:10.1002/jbm.a.32043

Wood, M. D., Gordon, T., Kemp, S. W. P., Liu, E. H., Kim, H., Shoichet, M. S., & Borschel, G. H. (2013). Functional motor recovery is improved due to local placement of GDNF microspheres after delayed nerve repair. Biotechnology and Bioengineering, 110(5), 1272–81. doi:10.1002/bit.24800

Wood, M. D., Gordon, T., Kim, ., Szynkaruk, M., Phua, P., Lafontaine, C., … Borschel, G. H. (2013). Fibrin gels containing GDNF microspheres increase axonal regeneration after delayed peripheral nerve repair. Regenerative Medicine, 8(1), 27– 37. doi:10.2217/rme.12.105

Wood, M. D., Kemp, S. W. P., Weber, C., Borschel, G. H., & Gordon, T. (2011). Outcome measures of peripheral nerve regeneration. Annals of Anatomy, 193(4), 321–33. doi:10.1016/j.aanat.2011.04.008

Wood, M. D., Kim, H., Bilbily, A., Kemp, S. W. P., Lafontaine, C., Gordon, T., … Borschel, G. H. (2012). GDNF released from microspheres enhances nerve regeneration after delayed repair. Muscle & Nerve, 46(1), 122–4. doi:10.1002/mus.23295

Wu-Fienberg, Y., Moore, A. M., Marquardt, L. M., Newton, P., Johnson, P. J., Mackinnon, S. E., … Wood, M. D. (2014). Viral transduction of primary Schwann cells using a Cre-lox system to regulate GDNF expression. Biotechnology and Bioengineering, 111(9), 1886–94. doi:10.1002/bit.25247

Yan, Q., Matheson, C., & Lopez, O. T. (1995). In vivo neurotrophic effects of GDNF on neonatal and adult facial motor neurons. Nature, 373(6512), 341–4. doi:10.1038/373341a0

Yegiyants, S., Dayicioglu, D., Kardashian, G., & Panthaki, Z. (2010). Traumatic peripheral nerve injury: A wartime review. J Craniofac Surg, 21(4), 998– 1001.

Yeo, Y., & Park, K. (2004). Control of encapsulation efficiency and initial burst in polymeric microparticle systems. Archives of Pharmacal Research, 27(1), 1–12. doi:10.1007/BF02980037

108

Yin, Q., Kemp, G. J., Yu, L. G., Wagstaff, S. C., & Frostick, S. P. (2001). Neurotrophin- 4 delivered by fibrin glue promotes peripheral nerve regeneration. Muscle & Nerve, 24(3), 345–51. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11353418

You, S., Petrov, T., Chung, P., & Gordon, T. (1997). The expression of the low affinity nerve growth factor receptor in long-term denervated Schwann cells. Glia, 20, 87– 100.

Young, J. Z. (1949). Narrowing of nerve fibres at the nodes of Ranvier. J. Anat., 83(55).

Zalewski, A. A., & Gulati, A. K. (1982). Evaluation of histocompatibility as a factor in the repair of nerve with a frozen nerve allograft. Journal of Neurosurgery, 56(4), 550–4. doi:10.3171/jns.1982.56.4.0550

Zuo, J., Hernandez, Y. J., & Muir, D. (1998). Chondroitin sulfate proteoglycan with neurite-inhibiting activity is up-regulated following peripheral nerve injury. Journal of Neurobiology, 34(1), 41–54. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9469617

109