Growth Factor Administration for Nerve, Skeletal Muscle and Cardiac Tissue Repair: a Key Tool in Regenerative Medicine

UNIVERSITA' DEGLI STUDI DI TORINO PhD Programme in Experimental Medicine and Therapy XXVIII cycle

GROWTH FACTOR ADMINISTRATION FOR NERVE, SKELETAL MUSCLE AND CARDIAC TISSUE REPAIR: A KEY TOOL IN REGENERATIVE MEDICINE

Presented by: Michela Morano Tutor: Prof. Stefano Geuna

PhD Coordinator: Prof. Giuseppe Saglio

Scientific disciplinary sector: BIO 16

2012-2016

CONTENT

ABSTRACT 5

OUTLINE 11 ABBREVIATIONS 15

CHAPTER 1 PERIPHERAL NERVE INJURY AND REPAIR 17 Introduction and Scientific Background Aim of the Research Scientific Publications Discussion and Future Directions

CHAPTER 2 SKELETAL MUSCLE DENERVATION 159 Introduction and Scientific Background Aim of the Research Scientific Publication Discussion and Future Directions

CHAPTER 3 CARDIAC MUSCLE: ISCHEMIA/REPERFUSION INJURY 205 Introduction and Scientific Background Aim of the Research Scientific Publication Discussion and Future Directions

GENERAL CONCLUSIONS 247

ACKNOWLEDGMENTS 255

ABSTRACT

INTRODUCTION: The regenerative medicine is a continuously evolving interdisciplinary field aimed to enhance the regenerative capability of the body itself and to guide the regeneration process taking advantage of endogenous repair mechanisms to restore a tissue or organ morphology and function. Two types of approaches can be distinguished in regenerative medicine: cell based- and cell free- therapies. Particular attractive are growth factors-based therapies, which developed, thank to bioengineering and technical science influences, into a more integrated therapeutic approaches based on biomaterial, implant, engineered factors and controlled drug release. Here a multiple analysis is presented: i) the first part examines the in vitro and in vivo investigation of a set of engineered growth factors combined with biomaterials for promoting peripheral nerve regeneration; ii) the second part focus on the study of Neuregulin 1 (NRG1) growth factor and its receptors in two models of muscle injury (denervated skeletal muscle and ischemia/reperfusion cardiac muscle), aimed to better understand the properties of this factors for future in vivo manipulation of this system in muscle recovery after denervation and amelioration of the existing therapies in cardioprotection field. Traumatic injury in peripheral nerves, as for the central nervous system, results in high morbidity with great changes in patient's life and, obviously, elevated socioeconomic cost. Beside the elevated regenerative abilities of peripheral nervous system, the nerve regeneration often fails, with aberrant sprouting and the development of a neuroma at the proximal nerve stump. When regeneration process takes long time, the neurotrophic factors supply decays over the time, failing to sustain neuron regeneration. In most of nerve injury cases is necessary a surgeon intervention, and the standard approach used is the nerve autograft, which entails the suture of a nerve achieved from the patient himself to bridge the gap of another more important nerve. However autograft often gives unsatisfactory functional recovery, therefore increasing efforts have been made to seek new surgical alternative approaches. Among alternative strategies, the tubulization technique, which consist in the insertion of biological or artificial tubular structure between the nerve stumps, obtained good clinical results. Hollow tube are effective specifically for short nerve defect (≤ 3-4 cm). Researcher are now focus on conduits with more complex design in order to increase nerve conduit performance. Some of the solutions adopted are the use of extracellular matrix structural components that increase cell adhesion and invasion, internal framework, conductive polymers, matrix releasing growth factors and supportive cells. The production of neurotrophic growth factors and the expression of their receptors changes considerably after nerve injury in different nerve cell types and is now accepted that neurotrophic factors play an important role in peripheral nerve injury, influencing and controlling several aspect of nerve regeneration. Thus growth factors controlled and prolonged release inside artificial nerve conduit is actually a goal in peripheral nerve tissue

5 engineering. Incorporation of exogenous growth factors in nerve conduit can be done directly (in solution) in the tube or using an hydrogel, as collagen or agarose, which acts as a scaffold releasing the drug in the lumen. Among factors studied for nerve regeneration there is the neuregulin1 (NRG1) a widely express growth factor existing in numerous isoforms as a result of alternative splicing; it can be a soluble or transmembrane protein, that mediates various cellular process through ErbB tyrosine kinase receptors. Nrg1 isoforms have been shown upregulated after injury. They drive the dedifferentiation of Schwann cells and their migration in the site of injury to create the Bands of Büngner, a tubular structure that direct growing axons to their original targets. Moreover a transmembrane NRG1 expressed by the axons guides the deposit of myelin layers by Schwann cells, regulating remyelination process. Therapies for nerve regeneration should monitor also nerve target behaviour, avoiding muscle atrophy and promoting the correct reinnervation of muscle fibres to obtain total functional recovery. After nerve injury the target muscle undergoes to a molecular and morphological changes that result in muscle atrophy, and when denervation persists permanent changes occur, reducing the possibility to recover the complete functionality after reinnervation. The denervation activates the ubiquitin-proteosome machinery and the autophagy-lysosome machinery that are responsible for protein breakdown. At the same time satellite cells, the resident stem cells, became activated, proliferate and then differentiate and fuse in myofibres. Meanwhile the role of NRG1 and ErbB receptors is well defined in nerve, little is known about NRG1/ErbB system in muscle after nerve injury. It is clear that NRG1 has a role in muscle development and controls spindle maintenance, glucose uptake and neuromuscular junction formation. Moreover ErbB2/ErbB3 expression in satellite cells are able to induce pro-survival signalling in activated cells. How NRG1/ErbB system is influenced by nerve acute injury, and if the system could be a good therapeutic target to maintain the muscle receptive for nerve reinnervation remain to be investigated. In another muscular tissue, the cardiac tissue, is becoming more and more clear that NRG1/ErbB system is a potential target for heart failure therapy. NRG1/ErbB system is essential for a correct cardiac development, furthermore, it is now clear that also in adult heart this signalling plays a critical role in the normal function as well as in ischemia or other pathological conditions. In adult heart, cardiac microvascular endothelial cells (EC) express soluble NRG1 isoforms, which stimulate cell survival and growth, glucose uptake, protein synthesis and “hypertrophic” gene expression in cardiomyocytes, expressing ErbB receptors. The deletion of NRG1 from EC increases the infarct area and the number of TUNEL positive cells after ischemia and reperfusion (I/R) injury. For its pro-survival effect Nrg1 has been proposed as a potential drug for heart failure treatment. Several preclinical studies in rat or mouse models of heart failure and several clinical trials demonstrated that intravenous administration of recombinant soluble NRG1 improved

6 cardiac contractility and relaxation, left ventricular remodelling, decreased apoptosis and attenuated mitochondrial dysfunction. However, the molecular bases of this beneficial effect remain unclear, as well as how the downregulation of NRG1/ErbB system detected in cardiac chronic disease is a cause or an effect of the pathological status.

MATERIALS AND METHODS: The first part of the study is related to nerve injury. We investigated in vitro and in vivo a set of engineered growth factors with the final goal of in vivo long-time release inside an artificial nerve guide. NGF, FGF and GDNF, together with the extracellular domain of NRG1 beta were covalently conjugated to iron-oxide nanoparticles (10nm±2 diameter). We tested in vitro the retention of factor bioactivity after nanoparticles conjugation, analyzing neurite outgrowth in adult or neonatal dorsal root ganglion (DRG) cultures. Ascertained the bioactivity, we recreated in vitro a possible environment present in artificial scaffolds, culturing neonatal DRGs inside a layer of NVR (a biocompatible hydrogel composed mainly by laminin and hyaluronic acid), and we evaluated neurite outgrowth. Furthermore we tested the stability of the conjugated factors comparing the effects of non-conjugated or conjugated factors left some 2/4 weeks at 37°C in cell medium and then used to stimulate neurite outgrowth or cell migration. In collaboration with our partner in Israel an in vivo pilot study was performed in rat using GDNF factor. A comparison of nerve regeneration was done after sciatic nerve injury (15 mm gap) and nerve repair among the following groups: 1. autograft (gold standard/control); 2. hollow chitosan tube (17 mm length); 3. chitosan tube filled with NVR gel; 4. chitosan tube filled with NVR plus GDNF; 5. chitosan tube filled with NVR gel plus conjugated GDNF. Functional analysis were performed after one, three and five months from surgery. After 5 months animals were sacrificed and morphological parameter (fibre diameter, axon diameter, number of myelinated fibres, myelin thickness) were analyzed on regenerated nerves inside the tube.

In parallel we planned an in vivo time-course study of NRG1/ErbB expression (mRNA and protein) in muscle after denervation and reinnervation process, in order to obtain a detailed overview of the system as a starting point for therapeutic approach. Three different nerve injury were performed in rat median nerve: one slight lesion with fast functional recovery (crush lesion), one wider injury with surgical repair (end-to-end repair, E-E) and a denervation without reinnervation. We monitor the expression of NRG1 (alpha and beta isoforms), ErbB2, ErbB3 and ErbB4 together with atrophy marker (like Foxo3, Atrogin, Murf1) in superficial digital flexor muscle at different time points from surgery (from 1 day to some months), with a Real-Time PCR analysis and western blot protein analysis. Using the in vitro model of C2c12 myotubes we performed similar expression analysis in presence of dexamethasone, a corticosteroid that induce muscle atrophy and is extensively used for in vitro studies. In addiction we investigated the effects of NRG1 beta or alpha administration in atrophic condition given by dexamethasone treatment, analyzing the myotube diameter (index for atrophic condition) and the expression of atrophic marker.

In the last part of the study we investigated how NRG1/ErbB system in heart is affected in the early phase of I/R injury or after I/R injury followed by the protective postconditioning procedure (PostC, brief cycles of ischemia/reperfusion carried out after a sustained ischemia). We used the Langendorff’s heart ex-vivo model to mimic an I/R injury. Male rat hearts were randomly divided in three groups: (1) control group (Sham), hearts were subjected to 150 min perfusion only; (2) I/R group, hearts underwent 30 min of global ischemia (aorta occlusion) and then a period of 120 min full reperfusion; (3) PostC group, after 30 min ischemia, hearts underwent a PostC protocol (5 cycles of 10 sec reperfusion and 10 sec of global ischemia) then a period of 120 min full reperfusion. We analysed ischemic marker (LDH, infarct area) to confirm the benefit of PostC treatment; then the mRNA level of Nrg1 and ErbB receptors and Nrdp1, a ubiquitin ligase targeting ErbB3 was evaluated. We next investigated the protective role of ErbB3 in H9c2 cells in oxidative stress condition to assess how this receptors can contribute to the protective effect give by Nrg1.

RESULTS: Results can be summarized as follow: -nerve perspective: neurotrophic factors (NTFs) after conjugation with iron oxide- nanoparticles preserve their bioactivity and are able to induce neurite outgrowth as the free-NTFs, also when they are administered in a matrix, in our case NVR gel. The bioactivity is comparable to the non conjugated factors. Moreover the conjugation with iron oxide nanoparticles stabilizes the factor: conjugated GDNF maintains its ability to induce neurite outgrowth in adult rat DRGs after 2 weeks of incubation at 37°C; conjugated Nrg1 is able to stimulate cell migration after 6 weeks, meanwhile the non conjugated factors lose their effects. The in vivo pilot study using the combination of chitosan tube enriched with NVR gel plus GDNF, for sciatic nerve repair, reveals that the enrichment with NVR implement the tube performance, improving the functional and morphological outcome. Comparing the morphological analysis of NVR gel plus conjugated or not GDNF with NVR gel alone we can see an increase in number of myelinated fibres, moreover there are no statistically relevant differences between the tube with conjugated GDNF and the tube with free GDNF. Anyway no one of the chitosan tube groups achieve the results obtained with autograft repair. We can conclude that iron oxide nanoparticles are potentially a good candidate for nerve device enrichment. However the best way to administer conjugated factors in the tube needs further investigation, as well as the effects of long time exposition to these factors. -skeletal muscle perspective: in order to monitoring the NRG1/ErbB system in muscle during denervation and reinnervation process we performed a time course expression analysis in three different rat models with increased gravity of nerve injury. The analysis shows a large alteration of the system in all the three models. In particular ErbB2 receptor mRNA is up-regulated during the denervation, but return to basal levels after reinnervation in crush and E-E groups. Similar profile was observed for ErbB4 cyt1 isoform. NRG1 was found upregulated after 1 day both alpha and beta isoform in crush and E-E groups,

8 meanwhile in denervated group Nrg1 is down-regulated from 1days to the end of our timeline (8 weeks). Our in vitro analysis with C2c12 reveals that also after dexamethasone treatment ErbB2 is upregulated (mRNA and protein). Dexamethasone induces an atrophic condition detectable by the upregulation of Atrogin and Murf1 and the reduction of C2c12 myotube diameter, curiously the administration of NRG1 (alpha or beta isoform) rescue these effects. Moreover Nrg1 treatment downregulates Foxo3 protein, a transcription factor responsible for Atrogin and Murf1 expression. The levels of P-Foxo3 protein, corresponding to the inactive form of the protein, were also analyzed and results show that Nrg1 alpha but not beta increases this protein compared to the dexamethasone-treated cells and control cells. These data suggest that Nrg1 alpha could have a role in muscle response to atrophy. -cardiac muscle perspective: taking advantage of the ex vivo Langendorff’s heart model, we investigated the changes that occur in Nrg1 and ErbB expression in heart tissue following a global ischemia/reperfusion challenge (I/R group), or following the postconditioning procedure (PostC group). The expression analysis shows that the four ErbB receptor genes are transcripted in adult rat heart. After I/R injury, the mRNA levels of ErbB1 and ErbB2 did not differ significantly from the Sham, regardless of PostC procedure. Similar results were obtained for the four ErbB4 isoforms. Conversely, ErbB3 mRNA displayed a 2.8 fold increase after I/R, compared to the Sham. For PostC samples the upregulation was not statistically relevant (P=0.081 with respect to Sham). Nrg1 transcription was also investigated, but both alpha and beta isoform expression was unvaried. We found that ErbB3 mRNA is upregulated after I/R, while for ErbB3 protein only a tendency to downregulation in I/R group, but not in PostC samples, can be observed. This tendency could be the effect of a different post-transcriptional regulation or balance between protein synthesis and degradation in the protected and non protected heart. We analyzed the expression of a E3 ligase known as “neuregulin receptor degradation protein-1” (Nrdp1), which targets specifically ErbB3. Nrdp1 mRNA results upregulated after I/R injury, in line with literature data about Nrdp1 protein, but not after postconditioning procedure. Nrdp1 degrades ErbB3 and we can postulate that this mechanism is perturbed after PostC treatment. Directly or indirectly mechanical PostC can influence ErbB3 protein levels, suggesting that ErbB3 signalling pathway might be part of the protective signalling given by postconditioning treatment.

CONCLUSIONS: Nanoparticles changed the field of drug release and they are now an attractive tool for regenerative medicine. They have the advantage that can be easily incorporated in biomaterials or devices and can carry the molecule of interest. Iron oxide nanoparticles are considered to be nontoxic and are already used for various biomedical applications, such as diagnostics, cell labelling, magnetic resonance imaging. In nerve regeneration field np-NTF are promising, specially for their long term stability useful for in vivo prolonged release. However many aspects need to be addressed, such as long-term

9 toxicity, best way of combination with nerve device and the possibility to use nanoparticles as marker for high-resolution imaging of nerve regeneration. The NRG1/ErbB system is involved in several cellular response, included tissue repair. Its role is strictly correlated to tissue type and cellular population. In peripheral nerve Nrg1 pathways are a potential target to induce Schwann cells migration in the early phase and myelination in a later phase of regenerative process. In cardiac tissue NRG1 treatments are actually study in clinical trials for heart failure and myocardial infarction treatments. Our study evidences a pivotal role for ErbB3 in the context of I/R and redox biology, highlighting the importance to study this receptors not only for the optimization of NRG1 treatment for heart diseases, but also for the development of new cancer therapies with no side effects on heart. Moreover our data concerning denervated muscle show a regulation of the NRG1/ErbB system also in skeletal muscle. Future experiments are necessary to address the in vivo efficacy of NRG1 in skeletal muscle. The open question remain the exact role exerted by each single NRG1 isoform.

OUTLINE by Michela Morano

The regenerative medicine is a continuously evolving interdisciplinary field aimed to enhance the regenerative capability of the body itself and to guide the regeneration process taking advantage of endogenous repair mechanisms to restore a tissue or organ morphology and function. Generally, two main approaches can be distinguished in regenerative medicine: cell-based and cell-free therapies. Among the cell-free approaches the delivery of growth factors results very promising with a wide range of applications. In the last decades, the promiscuous dialogue among biomaterial science, bio-engineering, clinic and cellular biology has given rise to more integrated therapeutic approaches based on biomaterial, implant, engineered factors and controlled drug release.

During my PhD course, I had the opportunity to work in the field of regenerative medicine, investigating the potentiality of novel biomaterial for nerve tissue reconstruction and new strategies to amend and control the intrinsic nerve regenerative process thank to growth factor delivery. Moreover, part of my work was dedicated to the analysis of a specific growth factor - Neuregulin 1 - and its receptors – ErbBs - in injured skeletal and cardiac tissues, to investigate its therapeutic potential.

Since different regenerative processes have been taken into account, this thesis is divided in three main chapters, each of them dedicated to a specific tissue and injury model. Each chapter is composed by a short introduction to get into the topic, followed by the "Scientific Background" section, which contains an overview of the current knowledge of the described biological process and the limits of existing clinical approaches, providing to the reader the necessary information to understand the work. The "Aim of the Research" attempts to summarize, in short sentences, the goal of the research, whereas the section "Scientific Publications" represents the core of the chapter, containing the published or in preparation articles produced during the PhD course. In the last paragraph, entitled "Discussion and Future Direction", the obtained results are summarized and critically assessed in relationship with literature data, integrated with overall considerations regarding current knowledge and future directions in the specific field presented.

Chapter 1 deals with the peripheral nerve regeneration, which represents the main topic of my research. In the background section of this chapter, I describe the traumatic injuries of peripheral nerves and the intrinsic mechanisms driving nerve regeneration. Despite several microsurgery techniques have been developed in the past decades, the clinical outcome after nerve injury is still suboptimal and functional recovery is often partial. Thus, in the nerve regenerative medicine several approaches have been explored in order to improve nerve regeneration and functional outcome, such as artificial scaffolds, biomaterials, gene therapy, growth factor delivery and cell therapy. I present an overview of these proposed approaches, ranging from vein graft to artificial nerve device based on tubulization technique. A special section is dedicated to the growth factors involved in the regenerative process and their delivery inside nerve devices. In Chapter 1, I describe the study of the extracellular fragment of ErbB4 receptors, whose biological activity was first investigated

13 in vitro taking advantage of a glial cell model and later in vivo exploring the gene therapy approach for factor release by muscle tissue used as filler into vein graft. Moreover, the study of three well known growth factors (NGF, GDNF and FGF-2) conjugated with iron- oxide nanoparticles is also reported in this chapter. These engineered factors were characterized in vitro to investigate their application in nerve regeneration field, and further employed in a chitosan-based nerve device, using an hydrogel delivery system, for the reconstruction of long nerve defect in the rat animal model.

Traumatic nerve injuries affect the connection between nerve and target muscle, determining a loss of motor and sensory function. The prolonged time required for nerve to regrowth up to the target often results in morphological changes in skeletal muscle tissue that can afterwards compromise the reinnervation and the functional recovery. In Chapter 2, I describe the nerve injury from the target muscle prospective, illustrating the molecular and morphological changes occurred and the proposed approaches to maintain the muscle tissue in optimal trophic state before reinnervation. In this part of the thesis, I explain an unpublished study regarding the expression of the growth factor Neuregulin 1 and its receptors in muscle after denervation and reinnervation processes. The therapeutic potential of Neuregulin1 is here discussed, with final remarks on muscle reinnervation and future directions to obtain the total functional recovery.

Neuregulin 1 is a pleiotropic growth factor expressed in several tissues. Its numerous isoforms can signal in autocrine, justacrine or paracrine manner, playing a role during the development of various organs such as peripheral nervous system and cardiac system. It is now clear that in the adult life these proteins are involved in cellular response and repair process after tissue damage. Since years, in the laboratory of Professor Geuna, in collaboration with the Prof. Perroteau and Prof. Gambarotta's team, the Neuregulin 1/ErbB system is studied in peripheral nerves during the regenerative process to understand its complex modulation and the role of different Neuregulin 1 isoforms. Recently, we have started a collaboration with the group of Prof. Pagliaro of our Department to enlarge the Neuregulin 1 investigation to heart field, where the therapeutic potential of this protein is becoming clear. In Chapter 3 I describe the role of Neuregulin 1 in heart tissue and the ongoing clinical trials with this growth factor, highlighting the necessity to improve our knowledge about this complex cellular signalling in order to ameliorate the current therapeutic approach. This chapter shows the results of our analysis of the Neuregulin 1/ErbB system expression in heart tissue after ischemia/reperfusion injury and the in vitro study of the ErbB3 receptor role in oxidative stress.

The last section, the "Conclusion", collects some final considerations about growth factor delivery-based medicine approach, its clinical applicability and possible future developments.

ABBREVIATIONS

AAV Adeno-Associated Virus AChR Acetylcholine Receptor AMI Acute Myocardial Infarction AKT Protein kinase B BACE β-site of Amyloid precursor protein Cleaving Enzyme BMSC Bone Marrow Stem Cells CNS Central Nervous System DRG Dorsal Root Ganglia E-E End-to-End repair ECM Extracellular Matrix Ecto-ErbB4 Extracellular domain of ErbB4 ERK Extracellular Signal-Regulated Kinase EC Endothelial Cells FDA Food and drugs administration FES Functional Electrical Stimulation FGF-2 Fibroblast Growth Factor 2 GDNF Glia-derived Neurotrophic Factor HA Hyaluronic Acid HF Heart Failure IONP Iron Oxide Nanoparticles I/R Ischemia/reperfusion LVEF Left Ventricular Ejection Fraction MAG Myelin associated glycoprotein MIV Muscle-in-vein mPTP mitochondrial permeability transition pore MSC Mesenchymal Stem Cells NGF Nerve Growth Factor NMJ Neuromuscular Junction np-NTF Neurotrophic Factors conjugated with iron oxide nanoparticles Nrdp1 neuregulin receptor degradation protein-1 NRG1 Neuregulin 1 NTF Neurotrophic factors NVR Neurovascular Research PNR Peripheral Nerve Regeneration PNS Peripheral Nervous System PostC Postconditioning RNS Reactive nitrogen species ROS Reactive oxygen species SC Schwann cells TNFα Tumor necrosis factor alpha

CHAPTER 1 PERIPHERAL NERVE INJURY AND REPAIR

CONTENT:

1.1 INTRODUCTION AND SCIENTIFIC BACKGROUND ...... 19 1.1.1 NERVE ANATOMY AT A GLANCE ...... 19 1.1.2 TYPE OF NERVE INJURY ...... 21 1.1. 3 MOLECULAR AND MORPHOLOGICAL CHANGES IN NERVE DEGENERATION AND REGENERATION ...... 23 1.1.4 REPAIR STRATEGIES ...... 25 1.1.4.1 Nerve conduit ...... 28 1.1.4.2 Implementation of nerve conduit: internal modification...... 32 1.1.5 GROWTH FACTORS INVOLVED IN PERIPHERAL NERVE REGENERATION ...... 36 Neurotrophic factors (neurotrophins) ...... 37 Other growth factors...... 39 1.2 AIM OF THE RESEARCH ...... 43 1.3 SCIENTIFIC PUBLICATIONS ...... 45 Characterization of glial cell models and in vitro manipulation of the neuregulin1/ErbB system ...... 45 Local delivery of the Neuregulin1 receptor ecto-domain (ecto-ErbB4) has a positive effect on regenerated nerve fiber maturation...... 69 Nanotechnology versus stem cell engineering: in vitro comparison of neurite inductive potentials ...... 87 Effect of Local Delivery of GDNF Conjugated Iron Oxide Nanoparticles on Nerve Regeneration along Long Chitosan Nerve Guide ...... 113 1.4 DISCUSSION AND FUTURE DIRECTIONS ...... 127

1.5 REFERENCES ...... 139

CHAPTER 1: Introduction and Scientific Background

1.1 INTRODUCTION AND SCIENTIFIC BACKGROUND 1 The peripheral nervous system (PNS) is composed by a network of differentiated cells that expand across the body, reaching all tissues and organs, to provide sensory or motor signalling from and versus the central nervous system. The great extension and exposition of nerves explains the elevate incidence of damages interesting nerve tissue. In Europe more than 300000 cases of nerve injury are recorded in one year, in U.S.A. nerve trauma affect 360000 persons for year. Most of nerve injuries are due to motor vehicle and sport accident, metal laceration, compression or secondary effects of bone fractures, focal contusion, tumour resection, electrical injury or drug injection injury. Traumatic injury in peripheral nerves, as for the central nervous system, results in high morbidity with great changes in patient's life and, obviously, elevated socioeconomic cost. Thus, the optimal repair of peripheral nerve injury and the functional recovery, besides being an important and defensible social goal of human community, represents a great challenge in repair medicine.

1.1.1 NERVE ANATOMY AT A GLANCE

Peripheral nerves are organs that originate from the central nervous system and, thanks to multiple branches, reach all body district transmitting sensory or motor stimuli. Nerves origin from the spinal cord and are generally divided in two categories: the cranial nerves (12 pairs) and the spinal nerves (33 pairs). Except for the two first cranial nerves, the olfactory and the optic nerve, the morphology observed in all nerves is similar (Figure 1.1).

The nerve fibre represents the smallest functional unit of nerves, and it is composed by a single axon and its glial component associated. Different classification methods are used for nerve fibres, based on various parameters as the function, the conduction velocity, the diameter and so on. Depending on glial cell/axon relationship, nerve fibres can be anatomically divided in unmyelinated or myelinated fibres. In PNS the glial component is gives by Schwann cells (SC). In unmyelinated fibres a single SC form one sheath around multiple axons that results grouped by the SC. In the second case, the myelinated fibres, a series of SC embrace a single axon and the plasmalemma of each single cell wraps several times the axon forming a plurilamellar myelin sheath. The axon regions occupied by single SC are defined internodes and the space between two internodes represents the Ranvier's node, where the axon is uncovered by SC. The length of Ranvier's nodes can varied from 150 µm to 1500 µm corresponding to the axon diameter. This unmyelinated portion permit the saltatory conduction of the impulses along the fibre. Normally large diameter axons are myelinated, while small axons are unmyelinated.

Moreover, according to the type and the direction of the transported signal, nerve fibres can be classified as follow: efferent or motor fibres, afferent or sensitive fibres and special sensory fibres. The efferent fibres transport motor signal from central nervous system

CHAPTER 1: Introduction and Scientific Background

(CNS) to PNS; they can be somatic fibres - controlling voluntary movements of skeletal muscle- or visceral fibres - controlling involuntary movements of cardiac tissue, smooth muscle and glands. The afferent fibres carry on sensory impulses from the periphery to CNS. Among these type of fibres the somatic fibres transport signals derived from receptors in skin, muscle and joints; while visceral sensitive fibres connect receptors located into viscera to CNS.

Figure 1.1 Peripheral nerve anatomy1

Meanwhile, the information from sense organs arrive to the CNS thank to special sensory fibres. Depending on fibre type composition, peripheral nerve can be divided in three main categories: sensory, motor and mixed nerves. Another way to identify fibres is based on cell body size and axons length. According to this methods fibres are divided in: A fibres, myelinated fibres with diameter of 3-22 µm; B fibres, myelinated fibres with diameter of 1.5-3 µm; C fibres, unmyelinated fibres with small diameter (0.3-1.5 µm).

The other main component of peripheral nerve is given by the connective tissue, which sustains structurally the nerve and gives resistance to compression or stretch force. Three connective layers are visible in nerves: the endoneurium, the perineurium and the epineurium1. Each axon and their surrounding Schwann cells is surrounded by the endoneurium. This protective layer is composed by loose connective tissue of collagen I fibres that run parallel to the axon, and a small number of fibroblasts and macrophages.

CHAPTER 1: Introduction and Scientific Background

Groups of endonevrium-enveloped fibres are surrounded by a multilayer dense sheath, the perineurium, forming the nerve fascicles. The perineurium is given by alternating layers of 1 collagen/extracellular matrix and flattened polygonal cells (perineurial cells) and its extremely resistance to pressure. While glucose and small electrolytes can pass, most macromolecules cannot diffuse through the perineurium, suggesting a role in regulating endoneurial milieu similar to the blood-nerve barrier. The outmost connective layer is the epineurium: it surrounds the nerve, defining its boundaries, and supports the inner nerve structure holding the fascicles. It is composed by dense collagen fibres, fibroblast and adipocytes. Large blood vessel are visible in this layer, where they run parallel to fascicles and they form numerous capillaries that, through the other connective layers, supply the oxygen and nutrients to each cellular component.

The spinal nerves extend from a nerve root generated from the lateral-ventral furrow of the vertebral canal and occupy the intervertebral foramina. The anterior root is formed by axon from motorneurons located in the anterior horn of spinal cord. While, a swelling of each posterior hood, called dorsal root ganglia (DRG), hold the soma of sensory neuron (pseudo-unipolar ganglion neuron) which project one bifurcated axon extending one branch to the peripheral target and the other branch to the spinal cord. DRG neurons are divided into distinct subpopulation based on cell body morphology, gene expression pattern and sensory modalities2,3. According to morphology two main group are distinguished: large neurons, light with granular cytoplasm (A neurons) and small, dark neurons with dense cytoplasm (B neurons)4. Different sensory neurons subtype required specific neurotrophic factors, for example distinct subpopulation of DRG neurons express TrkA, TrkB or TrkC receptors, responding to specific neurotrophin5,6.

1.1.2 TYPE OF NERVE INJURY

According to Seddon's classification, dated 1943, three type of nerve lesions are described: i) neuroprassia, characterized by local damage of myelin and axon, without interruption of endoneurium sheath. Although the nerve results anatomically intact (no interruption of axon continuity), it has lost its function and is not able to transmit nerve impulses. This type of lesion occurs to heat, cold, irradiation or electrical exposition. Surgery is unnecessary and the recovery occurs within hours, days, or up to a few months. ii) axonotmesis, a damage or destroy of axons, which lost their continuity, without any alteration of connective tissue; axon damage usually is given by nerve crushing, pinching or prolonged pressure. All the axon portions located after the site of injury undergo Wallerian degeneration (discussed later), while the proximal stump of nerve can re-grow inside the intact endoneurial tubes. The nerve functionality is lost and the recovery depend to the degree of the damage and to the distance to the nerve target. Since connective tissue is intact, surgery is unnecessary.

CHAPTER 1: Introduction and Scientific Background iii) neurotmesis, a severe nerve injury affecting both the neuronal and the connective components. The distal nerve stump undergo Wallerian degeneration, while retrograde degeneration occurs at the proximal stump. Due to the extended endoneurial tube disruption, spontaneous regeneration is slow and functional recovery does not easily occur; however successful regeneration might be achieved with surgical intervention.

In a most recent classification (Sunderland, 1951) neurotmesis injury were additional divided in three levels with increasing severity (Figure 1.2). This classification identifies five degrees of nerve lesions7. The first and the second grade correspond to neuroprassia and axonotmesis, respectively. Lesions of grade 3 correspond to a neurotmesis with damage at axon and endoneurium level, with intact epineurium and perineurium. In grade 4 lesions nerve trunk continuity is maintained but connective tissue is largely damage and only epineurium results intact. The grade 5 refers to the complete transaction of nerve trunk. Beside Sunderland grades can be describe only histologically, this classification has limited clinical utility. Thus the most nerve injuries are mixed, Mackinnon and Dellon included in Sunderland's classification another type of injury which is a combinatiof of all grade (grade 6)8.

Figure 1.2 Schematic representation of the five degrees of nerve injury described by Sunderland classification 7.

CHAPTER 1: Introduction and Scientific Background

1.1. 3 MOLECULAR AND MORPHOLOGICAL CHANGES IN NERVE DEGENERATION AND REGENERATION 1 After peripheral nerve injury, several morphological and molecular changes occur immediately and can be prolonged in time, at multiple levels: the nerve at the site of injury (proximal and distal stump), but also at cell body level (dorsal and ventral root) and target organ level (Figure 1.3). Here we describe the events that occur after axonotmesis or neurotmesis. Immediately the synaptic transmission is lost and the target organ does not receive nerve signals.

Nerve portion placed distally to the injury site undergoes degeneration known as Wallerian degeneration (Figure 1.3A). Characterized for the first time by A. V. Waller in 1950 9, this degeneration is a sequence of events that results in the axon and myelin fragmentation and, finally, in nerve debris elimination. The increase in free intracellular Ca++ into axoplasm led to the activation of Ca-sensible proteases, named calpains, which start the degradation of neurofilament protein, microtubulin and other cytoskeletal components, determining the axon fragmentation in 48 h 10,11.

SC loose axon-contact and changes their phenotype, they isolate lipid droplets made my myelin and start its degradation (Figure 1.3B). The dedifferentiation of SC is given by a changes of gene expression, that occurs within 48 hours after nerve injury and implicates a downregulation of myelination genes (P0, MBP) or gene related to node and internode organization (connexin 32, E-cadherin) and the concomitant upregulation of regeneration- associated genes (growth factors receptors, growth factor and adhesion molecules)7,12. After few days from injury activated SC proliferates and later on they line up to form a tubular structure, aligned with endoneurial tubes, on the basal lamina known as Band of Bṻngner (Figure 1.3C). This structure physically supports re-growing axons and directs the regeneration through adhesion molecules and neurotrophic factors (NGF, BDNF and NT4) provided also by SC13,14. Besides, SC of the distal nerve stump are also the main source of cytokines that attract macrophages. Indeed few days after nerve injury the resident endoneurium macrophages become activated and move in the site of injury. Moreover, the dysfunction of the blood-nerve barrier make possible a consistent infiltration of haematogenous macrophages in the inner compartment of nerve, with a subsequence massive myelin removal15. Myelin clearance is one of the most important precondition for axon re-growth, whereas when it does not occur properly the myelin components, like myelin associated glycoprotein (MAG), work as an inhibitors and the environment do not support the regeneration. In addition to the receptor-dependent phagocytosis of axon and myelin debris, macrophages release also some mitogenic factors active on fibroblast and SC15,16. A large body of evidence shows that also other resident and infiltrated immune cells play a role in degeneration process: few hours after the damage mast cells release histamine and chemokines recruiting neutrophils and monocytes/macrophages from blood vessels; neutrophils invasion occurs within the first week and is, together with

CHAPTER 1: Introduction and Scientific Background lymphocytes action, responsible of neuropathic pain. The Wallerian degeneration is so associated with a strong inflammatory response17.

Morphological changes occurs also at the soma level, where information arrives thank to the retrograde signalling from proximal nerve stump18. In cells with axon damage, the nucleus became peripherally located, Nissl bodies disappear, cell volume increases and dendrites are retracted, the so called "chromatolytic changes"19 (Figure 1.3B). Besides, to support the new roles (survival and growth), neuron modifies its mRNA transcription profile. Indeed, the shifting from a signalling state to a growing state require the upregulation of proteins related to cell growth, like cytoskeletal components and growth factor receptors20. Motor and sensory neurons are differently affected by nerve injury: motor neurons show good survival after axotomy, meanwhile sensory neurons have been reported to be more sensitive and the percentage of cell death after nerve injury is higher respect to motor neurons. Nevertheless literature results about neuron death after injury are conflicting and some authors report no neuron death after peripheral nerve damage. DRG neurons often are characterized by a lowered firing thresholds or continuous spontaneous firing following injury, which may be related to phenomenon of sensibilization and neuropathic pain 3.

Proximal nerve stump is interested by an initial retrograde degradation process, consequently the axon retracts back to first node of Ranvier, reduces its diameter, then the neuronal regeneration process begins. Proximal stump maintains the contact with soma, from whom it receives materials through the fast and slow axoplasmic transport. Axon sprouting begin in different directions on basal lamina provided by SC, guided by neurotrophic factors and adhesion molecules. The growth is about 1-3 mm for day, but it can vary depending on injury site location21. Distal nerve stump and denervated organs influence nerve regeneration by the release of chemotactic molecules.

Simultaneously, in the target organ several molecular changes occurs. The timing of reinnervation affects negatively the entity of morphological alterations and the functional activity of the organ; in fact, prolonged denervation usually results in muscle loss of function or incomplete recovery. The modification interesting the target organ will be discussed in details in the next chapter.

When regeneration occurs successfully, the axons restore the contact with the target organ. The SC-axon interaction induces the myelin formation around regenerated nerves that initially show shorter internodes. Soma modification reverse as well as target muscle atrophy (Figure 1.3D).

CHAPTER 1: Introduction and Scientific Background

Figure 1.3 Schematic representation of degeneration and regeneration events 7

Besides, nerve reconnection with the target sometimes is related with incomplete or absent reinnervation, with no functional recovery. Regenerative process often fails, with aberrant sprouting and the development of a neuroma at the proximal nerve stump (Figure 1.3E). When regeneration process takes long time, or in chronic axotomy, the neurotrophic factors supply decays over the time, since SC or target skeletal muscle cannot sustain a long term release of factors 22. Indeed untreated severe nerve injury entails further morphological and cellular changes, as the reductions of SC numbers and the corresponding increases of fibroblast pool, with the formation of great scar tissue and fibrosis in nerve tissue. Moreover the degree of functional recover after nerve injury depends on multiple factors, for example the site of injury and the type of nerve, further ageing is related to less favourable prognosis after nerve injury 23, moreover some diseases as diabetic are associated with decreased nerve regenerative capacity and high incident of neuropathies 24.

1.1.4 REPAIR STRATEGIES

In order to restore nerve morphology and, especially, nerve functionality, in most of the neurotmesis cases is necessary a surgeon intervention (Figure 1.4). When nerve transection occurs without losing nerve portion, or when the gap is short (less than 5 mm in human),

CHAPTER 1: Introduction and Scientific Background tension-free primary neurorrhaphy is the repair method used. The surgeon reconnect the two free nerve stumps (end-to-end repair), suturing together the two stumps through the epineurial tissue (standard epineural repair technique) or placing sutures into the interfascicular epineurium connecting group of fascicles (grouped fascicular repair technique) or using the more complex epineneural sleeve repair technique1. For larger gap nerve damage, the standard approach used in clinic is the nerve autograft, which entails the suture of a nerve achieved from the patient himself (usually a sensory nerve, as the sural nerve) to bridge the gap of another more important nerve (as motor nerve)25. Besides it remains the gold standard technique for peripheral nerve repair, the autologous nerve graft is associated with morbidity, including possible neuroma formation at the donor site, and low regenerative successful rate25. Furthermore the calibre of donor nerve or its fibres composition often do not match perfectly for grafting, influencing negatively the functional outcome.

During the last three decade, increasing efforts have been made to potentiate nerve autograft success, on one hand, and to seek new surgical alternative approaches, on the other hand.

The first option regard the investigation of strategies to increase intrinsic regenerative capacity and limit the extrinsic factors negative affecting the nerve regeneration. Among tested strategies, the electrical stimulation of the injured nerve gets researcher and clinical attention. It has been demonstrated that brief electrical stimulations induce axon outgrowth across injury sites and enhance functional recovery, also in cases of delayed surgical repair of peripheral nerves, in animals model and human26. Moreover the electrical stimulation upregulates neurotrophic factors expression and growth-associated genes27,28. Another approach is the continuous low-power laser therapy. Phototherapy has been shown to improve axon regeneration and remyelination, beside the mechanism of this beneficial effect remain to be determined29. Also physiotherapy and voluntary exercise have been proposed to ameliorate autograft outcome. These non-invasive therapies improves neurite growth, inducing an high number and longer regenerated fibres, however the mechanisms responsible of these effects remain largely obscure 7. Another promising approach is the gene therapy, the introduction of exogenous genetic materials (DNA/RNA) in living cells in injured peripheral nerve. Recent gene therapy research has focused on the delivery of neurotrophic factors to improve autograft or maintain active SC in distal nerve stump and prevent target muscle atrophy30,31. Gene delivery can be performed by direct in vivo injection of viral vector, with desired molecule coding sequence, in nerve or by transducing transplantable cells in vitro. Particularly adeno-associated viral (AAV)-vectors are well-tolerated and safe, and are already used in clinical trial for neural disease31. Various in vivo studies were performed to investigated the gene therapy application in nerve repair, targeting SC, fibroblast and DRG neurons31,32. Future experiments will elucidate how gene therapy is a valid clinical approach for peripheral nerve repair.

CHAPTER 1: Introduction and Scientific Background

Regarding alternatives technique to autograft, a valid option is the nerve allograft, in which the donor nerve derives from human cadaver. Nerve allograft avoids the donor site 1 morbidity and is free from size limitation; however a concomitant immunosuppressant treatment is necessary, increasing the cost and the complexity of this technique. Currently, the allograft is the most used clinical approach for treating long gap nerve injury (up to 5 cm)33. For reducing nerve immunogenicity various protocols have been tested to get an acellular nerve graft. Enzymatic digestion, detergent processing or irradiation process are used to remove the cellular component and maintain the internal structure and topography of the nerve, able alone to give physical support to cell invasion and axon elongation 33. Curiously the type of protocol used can influence nerve regeneration outcome 34. Avance® Nerve Graft (AxoGen, Alachua, FL, USA), a human prepared allograft, received the Food and Drug Administration (FDA).

Another branches of peripheral nerve research focus the attention on the use of non nervous tube (tubulization technique). This technique involve the insertion of an tubular conduit, sutured to the two nerve stumps, to bring the gap. The rational for the use of a nerve conduit is that a closed system permits the accumulation of neurotrophic factors, facilitates a directional axon elongation and limits fibroblast infiltration, besides giving a physical support for regenerating nerve. Further the regeneration process is favoured when the surgical trauma and scar tissue formation are limited. The nerve guide characteristics, their use in clinics and their future implementation are described in the following chapters.

Figure 1.4 Scheme of surgical approaches for nerve injury

CHAPTER 1: Introduction and Scientific Background

1.1.4.1 Nerve conduit

The nerve conduit requires specific characteristic as non toxic material (avoiding great inflammation and rejection), good permeability (ensuring diffusion of nutrient and oxygen), flexibility and good malleability for the surgery (avoiding nerve compression). During the tube preparation various parameters have to be taken in account, as well as wall thickness and tube size that influence the biomechanical properties of the conduit; or the biodegradability and the reabsorbing time of the material used 35. The success of the conduit mostly depend on the rate of Schwann cells invasion and the deposition of extra cellular matrix; further the limited scar tissue formation and low fibroblast infiltration positively influence the regenerative process. Indeed, the internal surface has to be principally suitable for SC migration and growth factors diffusion. Latest, the conduit production process need to be relative ease, sterilize and low cost. Various fabrication techniques are used, like injection molding, electrospinning, photolithography and extrusion.

Different materials has been proposed for nerve conduits, both biological and synthetic ones. A brief overview of the state of the art is illustrated below.

AUTOGENOUS BIOLOGICAL CONDUITS

Autogenous biological graft took the advantage of the fact that the tissue used is harvested from the patients themselves, it is a biological tissue, immunologically compatible and is characterized by a structure that orients the regeneration of the nerve. Particular attention was given at vein conduits, further improved, to avoid vein collapse, with a filler of skeletal muscle fibres (muscle-in-vein conduits). These conduits achieved similar results to autograft in digital nerve repair 36. Besides the preparation of the conduit is strongly surgeon dependent and the application is still related to short gap nerve injury (less than 3 cm gap) 37. Also soft tissue as muscle alone or tendon have been proposed for nerve gap, but only muscle has been used in clinic with the concept that muscle fibre basal lamina and extracellular matrix gives a support for cell migration and axon elongation 38. However, some limits of autogenous biological graft are the requirement of a donor site and the efficacy only for repairing short nerve gaps 37.

NONAUTOLOGENOUS BIOLOGICAL CONDUITS

This category of conduits includes nerve guide made by biological material derived from human tissues (allogenic) or other species tissues (xenogenic). Natural molecules, as peptides or polysaccharides, are used to prepare a nerve conduit, with the advantaged of a good cell adhesion, migration and proliferation, due to their biocompatibility, without cytotoxic effects. Further, natural polymers can be usually degraded by enzymatic process, making the tube absorbable and so avoiding a second surgery to remove it. A special attention is given by the purification and the sterilization process, while common

CHAPTER 1: Introduction and Scientific Background sterilization method might alter the molecular structure of the materials used for the conduit. 1 Among natural polymers proposed for nerve reconstruction there are proteins, polysaccharides and polyesters. A short list of these polymers is illustrated below.

- Collagen and others ECM components

Collagen is the main component of extracellular matrix (ECM) and is abundant in nerve tissue. As the other components of ECM, collagen promotes cell migration and proliferation and is known as one of the factors driving axon elongation39. Collagen is a fibrous protein formed by three peptide chains; it forms aggregates that vary on the base of collagen molecule arrangement. The most common collagen aggregate structures are: collagen type I, II, III and IV. Thanks to easily manipulation, non toxicity during degradation and induction of a minimal foreign body response, collagen is one of the most popular materials in bioengineering and has been used for various biomedical application as medical implants such as artificial skin, wound dressings and cosmetic surgery40. In literature data demonstrated the effectiveness of nerve guide prepared with purified bovine or porcine collagen type I or type III applied to peripheral nerve repair in rats and non human primates, also for 5-mm gap41. Furthermore, the FDA approves for clinical use five collagen type I -based commercial conduits: NeuraGen® (Integra Lifesciences Cor., USA); NeuraWrap® (Integra Lifesciences Cor., USA); NeuroFlex® (Collagen Matrix Inc., USA); NeuroMatrix® (Collagen Matrix Inc., USA); NeuroMed® (Collagen Matrix Inc., USA) 34. Among these conduits, NeuraGen® obtain encouraging clinical data for sensory and motor recovery 42.

Other ECM molecule-derived materials, as fibronectin or laminin, are also used for the preparation of nerve conduits, principally as lumen fillers. Fibronectin is found in interstitial matrices and has been shown to influence cell migration and axonal growth. During embryogenesis laminin is the first ECM proteins expressed and it guides neurites elongation. Nerve conduits made by laminin or fibronectin were used to bridge 8-10 mm gap in rat sciatic nerve, with promising results 43,44.

- Gelatin

Gelatin is a natural biodegradable polymer derived by the thermal denaturation of collagen. Animal derived gelatin has been widely investigated for medical application, specially for its peculiarity as biocompatible materials, its plasticity and adhesiveness. Gelatin is less expensive then collagen and, as collagen, gelatin-based conduit has been also tested for nerve repair. The solution in water and the easy collapse of gelatin conduits make necessary some molecular modification of gelatin structure, using cross-linking agents (as glycidoxypropyltrimethoxysilane and genipin), resulting in the alteration of its physical and mechanical properties with a control of the degradation rate45. A biodegradable genipin-

CHAPTER 1: Introduction and Scientific Background cross-linked gelatin based conduit was used in sciatic nerve defect (10-mm gap) in rat with poor success, since after 8 weeks most of regenerated axons were not myelinated46, the same authors showed that an increases of conduit porosity can implement its performance, with good regeneration in rat animal, giving real prospective for future studies in longer defects47. Moreover gelatin-based neural scaffolds can be easily integrated with additional bioactive molecules, which can be gradually released during the scaffold degradation48. Various scaffold has been prepared used gelatin combined with other materials49–51.

- Silk fibroin

Silk fibroin is a protein, purified from silkworm Bombyx mori, that presents special elastic properties, great strength and toughness. Fibroin compatibility with nerve cells has been widely demonstrated and silk fibroin- based conduits achieve good results in sustaining and inducing axon regeneration and tissue repair, but only for small defects 7,52,53. Another source for fibroin is given by spiders. Nephila spider silk, composed mainly by two proteins with glycine and alanine repeated unit, is biocompatible and supports SC adhesion and proliferation in vitro54. Anyway in vivo data of spider silk conduit are missing.

- Chitosan and other natural polysaccharides

Among the new families of bio macromolecules, chitosan receives great attention by researchers. Differently to the materials observed before, chitosan is a linear polysaccharides derived by chitin, made up of copolymer of D-glucosamine and N-acetyl- D-glucosamine. Chitin can be found in arthropods, cuticles of insects, shellfish (crabs and shrimps) and fungi cell wall. Its commercial form is usually purified from marine crustaceans, largely available from waste of food processing, with a low cost of production. Chitosan has an antitumor activity, analgesic, hemostatic and antimicrobial effect, and it shows antioxidant properties; thus it has been proposed for different medical application (wound-healing agent, scaffold for bone, cartilage and skin, additive for lung surfactants)55,56. Numerous in vitro data demonstrated that chitosan is a favourable support for nerve cells, promoting cell survival and neurite outgrowth55,57,58. Moreover, chitosan- based nerve conduit has been widely investigated for spinal cord and peripheral nerve repair in several animal models 59. An advantage of chitosan is that its degradation can be regulated by the degree of deacetylation: high percentage of deacetylation give rise to a non degradable material, meanwhile partially deacetylation correspond to a degradable chitosan more adapt for nerve conduit57,60. Additionally the transparency of the material make easier the manipulation of the chitosan conduit by the surgeon. Clinical trials with artificial chitosan/PGA nerve graft, used for 30/35-mm median nerve defects, obtain promising results: a successful motor and sensory functional recovery was observed in patients during long follow-up (up to 3 years)61,62. The chitosan based Reaxon® Nerve Guide (Medovent, Germany), CE- approved, is on the market and was implanted in few cases of median nerve injury in Germany. At the moment a clinical study for repairing

CHAPTER 1: Introduction and Scientific Background short-gap digital nerve lesion is ongoing in Germany. Recently the chitosan based Reaxon® Plus (Medovent, Germany) obtain also the FDA approvals. Since these very promising 1 results, the efficacy of chitosan-based conduit for reconstruction of major peripheral nerves defects are expected to be explored.

Other linear polysaccharides proposed for nerve repair are alginate and agarose. Both can be derived from algae and needed a great purification process to avoid immunogenicity of non purify preparation. Despite their biodegradability, thermo stability and non-toxicity, alginate and agarose show low mechanical resistance, and for this reason they were later proposed as internal filler of the nerve guidance (discussed later).

- Polyesters

Biological polyesters are derived from microorganism. A class of polyesters are polyhydroxyalkonates (PHAs), which are biocompatible and biodegradable and can be found in blue-green algae and soil bacteria, where they represent an intracellular energy and carbon storage63. The PHAs principally used for nerve conduit are P3HB and poly(3- hydroxybutyrate-co-3-hydroxyvalerate) (PHBV)64. These materials are tested in animal models in combination with other components, giving a more complex conduit (see "Implementation of nerve conduit: internal modification" ).

NON BIOLOGICAL CONDUITS

Another class of materials used in nerve engineering is the ones of synthetic materials which have the advantage to possess specific chemical and physical properties that help the fabrication and manipulation of the tube itself. Despite this, they are often hydrophobic and do not show bioactive molecules to promote cell-substrate interaction, showing incompatibility with cell adhesion and proliferation. Moreover the induction of immune and inflammatory response has to be taken in account and tested before studying the synthetic polymers. Although cost of production usually are higher respect to biological materials, various polymers has been tested for nerve repair and some of them obtain the FDA approvals for clinical use34,64. Synthetic polymers are usually divided in nonabsorbable and absorbable materials, based on their biodegradability properties.

Among non-absorbable synthetic materials, silicone was one of the first proposed, for its elastic properties and its thermal and chemical stability. Silicon conduits were tested to repair both short and long nerve defects with good results, even if in some clinic cases it induces tricky tissue reactions65 . Others type of non-degradable conduits were produced using plastic polymers, as polyethylene, elastomer, acrylic polymer and expanded polytetrafluoroethylene (ePTFE) (Gore-Tex1)38. Non-absorbable synthetic conduits have the main disadvantage of not been degraded by host body and remaining in situ until their removal, causing often extended scar tissue and chronic inflammatory response.

CHAPTER 1: Introduction and Scientific Background

To overcome the necessity of a second surgery for tube removal, absorbable synthetic conduits have been developed. The degradation kinetics is variably among the materials used, anyway a rapid degradation should be avoid because fail in supporting nerve regeneration, but too slow degradation may give unwanted effects as inflammatory reaction and nerve compression. Some of absorbable materials suitable for nerve reconstruction application are aliphatic polyesters and copolyesters, as poly-glycolic acid (PGA), poly-L-lactic acid (PLLA), poly-L-lactide-coglycolide (PLGA), poly- (caprolactone) (PCL) and poly-lactic acid-ε-caprolactone38,64. The PGA-based NeuroTube® obtained the FDA approvals and clinical studies on digital nerve repair with short or moderate defect (up to 4-cm) showed positive results, with motor recovery, similar to autograft or allograft66–68.

1.1.4.2 Implementation of nerve conduit: internal modification

All existing nerve conduits approved for clinical use are hollow tubes. This basic design demonstrated its efficacy specifically for short nerve defect (≤ 3-4 cm), where functional outcome is similar to auto- and allograft technique. Indeed, the inefficiency for long nerve-gaps repair limits the application of nerve conduit and leaves still open the challenge of a new surgical approaches for satisfactory outcome after severe nerve injury. The increasingly close cooperation among surgeons, biologist and engineers resulted in further developments of manufactured nerve conduits. Conduits with more complex design are now prepared and tested in animal models. Tissue engineering technique transferred to nerve tubulisation basically attempts to modify the inner space of the conduit in order to mimic the native structure of a nerve, or use the tube as a scaffold for prolonged drug release. The nerve guide engineering led to endless possibilities, limited only by technical and material hindrances of working on nanoscale. Some of the solutions adopted are ECM structural components that increase cell adhesion and invasion, internal framework, conductive polymers, growth factors and supportive cells (Figure 1.5)69. It is now clear that combined approach might be the best solution.

Here we briefly examine the different approaches to potentiate nerve conduit with internal modification.

Structural component

The ECM components have been described fundamental for SC migration and axon elongation, able to modify cell behavior. Therefore, researchers investigated several surface modification techniques, to work up biomimetic materials which mimic the properties of native ECM70. Presently, three different strategies are used to obtain biomimetic materials: (i) inclusion of bioactive molecules, later released; (ii) modification of material surface with ECM macromolecules or specific binding motifs; and (iii) nanoscale structural motif of the materials. Laminin, collagen and fibrinogen were used for

CHAPTER 1: Introduction and Scientific Background biomaterial coating to facilitate cell attachment and migration 71–73. Moreover from ECM various binding motif were isolated and immobilized on biomaterials in a sequence of 1 short linear peptide. It is now clear that also the pattern of molecule disposition on biomaterial influence the cell response and the outcome of nerve regeneration70.

Another structural improvement for synthetic conduit is enhancing material porosity, with the presence of hole on nerve conduit wall 74–76. The size and the morphology of hole need to be controlled, since macroporous materials (> 80% porosity) facilitate macrophages infiltration and let to increased fibrotic tissue respect to intermediated porosity ( 70% porosity)77. Moreover large pores accelerate the hydrolysis of the absorbable conduits with easy obstruction of conduit lumen. Thanks to the improvements in nanofabrication field, we are now able to control, at microscale and nanoscale level, the pores size in nerve conduit. These lead to the production of asymmetric walls conduits with micropores inside and macropores in the outer layer, which permit to differently control the inflow and the outflow of the tube, promoting SC proliferation inside and blocking fibroblast proliferation outside 77,78.

Figure 1.5 Modification of hollow tube for nerve regeneration69

Internal framework

The lack of internal structure in hollow conduit might be one of the causes of poor success in long gap repair. An appropriate internal filler or framework inside the conduit is an efficient strategy to improve the outcome of nerve regeneration.

CHAPTER 1: Introduction and Scientific Background

A simply way to potentiate synthetic conduit is fill the tube with fibrin, laminin, collagen and other ECM components supply as a hydrogel. These hydrogels favourite cell invasion and axon elongation inside the conduit, and allow neurotrophic factors diffusion. Collagen based hydrogel and other hydrogels has been used in vivo for nerve repair79–83. Since the filler can slow down nerve regeneration physically obstructing the growth, it is necessary a fine control of distribution and density of the hydrogel itself. Currently, researchers are focused on mechanical qualities of the hydrogel, like material stiffness which seems able to modulate tension exerted by neurites and growth cones influencing axon elongation and altering gene expression84,85.

Multiple type of structures, as films, filaments, nano and micro fibres, micro-channels, have also been tested to control cell attachment, migration and proliferation inside nerve conduit. Really promising strategy is the insertion of fibres, which act in the tubular lumen as physical support and guidance for SC migration. Collagen, gelatin, chitosan, polycaprolactone (PCL) and other materials have been used to prepare fibres86–88, preferentially with the electrospinning technique. Fibre size, topography and orientation affects cells behaviours88,89. In vitro has been demonstrated that, on aligned fibres, neurites growth parallel to fibres and SC show an elongated morphology follow fibre long axis88,90,91. Moreover, in vivo studies show that aligned nanofibres reach to superior recovery in rat nerve injury, respect to random oriented fibres92. These data stress the importance of aligned structure inside the conduit.

Conductive polymers

It has been demonstrated that electric stimulation promotes neurite outgrowth and glial cells proliferation in vitro 93,94 and in vivo 95,96. The incorporation of electrical conducting material in nerve conduit results difficult, but, recently, some advancements have been done using organic conducting polymers (OCPs). One of the first conductive polymer proposed for tissue engineering applications is polypyrrole, which is biocompatible and induces axon elongation in vitro and in vivo97–99. Besides, as other OCPs, polypyrrole is not biodegradable, so various studies explore its incorporation in biodegradable polymers. Recently a polypyrrole/chitosan-based nerve conduits has been used to repair 15-mm gap in rats in combination with electrical stimulation, and great results were achieved100. Since the electro-conductivity of conductive polymers declines after long time actually researches are testing more stable electro-conductive materials.

Growth factors and other exogenous agent

The important role exerted by released growth factors in all phases of nerve regeneration has been already mentioned, here we shortly examine their application in tissue engineering, illustrated in detail in the next paragraph. Nerve regeneration failure over long distance is due largely by decay of growth factors releasing by SC and target organs22.

CHAPTER 1: Introduction and Scientific Background

Since growth factors enhance nerve regeneration through several mechanisms their controlled and prolonged release inside artificial nerve conduit is actually a goal in 1 peripheral nerve tissue engineering. Incorporation of exogenous growth factors in nerve conduit can be done directly (in solution) in the tube or using an hydrogel, as collagen or agarose, which act as a scaffold releasing the drug in the lumen101,102. This delivery system has been shown to improve the efficacy of nerve conduits103. Also absorption to microspheres104–106 or the immobilization on nanoparticles107 made by various materials have been investigated, with the advantages that they can be added to the structure of the conduit, as wall or internal structure.

Also hormones and other small molecules were suggested for pharmacological treatment of nerve injury. Steroids, like progesterone, can modulate SC behaviour108; thyroid hormone has been shown to improve myelination in rat sciatic nerve injury109, neurotransmitters influence neuronal-glia interaction110.

Theoretically gradients of factors can be generated over the tube length to sustain nerve regeneration avoiding trapping of axons. However, our still little knowledge about release kinetics and biological activity of incorporated factors/drug, as well as their side effects, limits the clinical application of these delivery systems. Further experiments in vitro and in vivo are needed, together with the improvement of technical ability to work at nanoscale level.

Supportive cells

The exploration of cells-based therapies receives great attention by researches and cell transplantation in nerve conduit is actually consider an engaging alternative to growth factors release. Transplanted cells are a source of growth factors and might help to create permissive environment for axon regeneration, also, producing ECM proteins111,112.

Given their key role in regenerative process, SC have been proposed as transplantable cells in animal models of nerve injury. The addiction of exogenous SC improves nerve regeneration in vein conduits or acellular nerve grafts in rat113,114, but cells survival for long time in host body required a concomitant immunosuppression therapy. Two clinical cases with autologous SC transplant to repair large sciatic nerve defect (> 5 cm-gap) using autograft are reported and the long follow-up showed a regain of motor function, partial sensory recovery and complete resolution of neuropathic pain115. Anyway, their limited source (from autologous or exogenous nerve), the morbidity associated and the long time (at least two weeks) required for in vitro expansion of SC before their transplantation limit SC use. Thus, regenerative medicine moved toward to other cell type, as stem cells or induced pluripotent stem cells (iPSC)111,116. The iPSC can be prepared by human fibroblast easily harvested, but require an in vitro reprogramming and have a tendency to form teratomas117,118.

CHAPTER 1: Introduction and Scientific Background

Among stem cells proposed for nerve repair there are different sources and type of cells:

 embryonic stem cells (ESC): ESC can differentiate in SC-like cells and in rat sciatic nerve injury they improve the number of regenerated axons119. However there are some disadvantages which restrict ESC clinic application, as their limited availability, the variability of differentiation rate and the ethical concern with the use of ESC. In addition, the research on human embryonic stem cells is strongly subject to individual state's laws ranging from full governmental support to strong restriction.

 adipose-derived stromal cells (ADSC): ADSC can be easily achieved using liposuction technique from the patient itself, without serious morbidity or immunogenic reactions. In addition to elevated ability to differentiate in SC-like cells, ADSC secrete various neurotrophic and angiogenic factors. Besides, pre-clinical studies indicated variable results and in vivo the differentiation rate was reduced.

 neural crest stem cells: neural crest cells are obtained from skin or hair follicle. Skin- derived autologous neural crest cells were used in collagen conduit for repair long nerve defect (8/10 cm-gap) in human and, despite the nerve regeneration was observed across the long gap, motor recovery was poor120.

 bone marrow stromal stem cells (BMSC): BMSC are harvested from bones, they can differentiate in SC-like cells and secrete several growth factors like BDNF and GDNF. BMSC have been widely tested in vivo to implement nerve conduit performance, obtaining successful results, as reduced muscle atrophy and increased number of regenerated axons121–124. Promising results were obtained also in bigger animal models for long nerve-gap repair (> 40 cm)125. Besides only a small rate of implanted cells differentiate in SC-like cells, most of the benefits derived from BMSCs do not appear to depend exclusively on cell differentiation and might be due to cell secretome126,127. Currently, the collected data about BMSCs-enriched nerve conduits showed that the achieved outcomes are comparable, and not improved, to the outcomes of autograft.

1.1.5 GROWTH FACTORS INVOLVED IN PERIPHERAL NERVE REGENERATION

Growth factors are small diffusible proteins and polypeptides produced by cells as signalling molecules regulating cell survival, migration, proliferation and cell differentiation, though endocrine, paracrine or autocrine mechanisms. Normally the signal transduction is initiated by growth factor binding to its receptor on the surface of target cells. The specificity and the type of signal is controlled both by receptor expression and spatial/temporal release of growth factor. The localization of growth factors and their receptors has been extensively investigated in CNS or PNS and numerous efforts have been made to understand growth factors role in neuronal tissue development and maintenance128. As mentioned before, the production of neurotrophic growth factors and

CHAPTER 1: Introduction and Scientific Background the expression of their receptors changes considerably after nerve injury in different nerve cell types3,129 and is now accepted that neurotrophic factors play an important role in 1 peripheral nerve injury, influencing and controlling several aspect of nerve regeneration. This view is supported by several in vitro and in vivo studies about exogenous neurotrophic factors which promote neurite outgrowth, axon elongation, SC migration and axon remyelination. It is worth mentioning that sensory and motor neurons express different receptors and they require different neurotrophic support for survival or growth130.

Various growth factors have been proposed to treat neuronal injury or neuronal disease in CNS and PNS and neurotrophins is the group of factors receiving most attention131. In mammals neurotrophic family is composed by four members (NGF, BDNF, NT3 and NT4/5) all synthesized as precursor proteins and cleaved by convertases into small peptide of 13-15 kDa. The mature proteins are realised as stable non-covalent homodimers and bind to two structurally unrelated receptors the tropomyosin receptor kinase (TrkA, -B, and -C) and the neurotrophin receptor p75NTR 132. Moreover other growth factors have been investigated for nerve repair in combination with biomaterial that control their release in the site of injury. The main studied growth factors are listed below, with a short description of their physiological role in nerve tissue and their use in peripheral nerve regeneration medicine.

NEUROTROPHIC FACTORS (NEUROTROPHINS)

NERVE GROWTH FACTOR (NGF)

NGF is the first neurotrophin discovered in 1950 by Rita Levi Montalcini and Viktor Hamburger, binding the TrkA receptors. This small protein promotes survival and differentiation of sensory and sympathetic neurons, neurite outgrowth and SC migration129. Normally, NGF mRNA and protein are not detected in adult rat nerve, but its expression is induced after injury, in non-neuronal cells at the distal stump of transected nerve with a biphasic curve133,134. Curiously NGF is not expressed and not upregulated after injury in motoneurons135,136. NGF release in silicone nerve conduit increased the number of myelinated axons and the thickness of reform myelin sheaths137. The NGF release by microspheres in chitin conduit shows similar results: improvement of myelinated fibres maturation and functional recovery in the long term138. Other controlled NGF release methods has been successfully tested in vivo using various biomaterial, obtaining good nerve regeneration and functional recovery51,105,139,140.

BRAIN-DERIVED NEUROTROPHIC FACTOR (BDNF)

BDNF is the most abundant neurotrophin in CNS. Its action is mediated by TrkB, even if BDNF can bind with low affinity also p75 receptor. BDNF expression was found in DRG neurons, in various subpopulation of neurons, as TrkA positive cells and some large TrkC- positive neurons. In intact motorneurons BDNF is expressed at low levels, but it is rapidly upregulated early after injury, as well as in distal nerve stump135,141,142. In vivo several

CHAPTER 1: Introduction and Scientific Background experiments proved that BDNF is a potent survival and trophic factor for motor neurons, suggesting its main role in mediating motoneuronal response to injury143–145. Although, as demonstrated by in vitro experiments, BDNF induces neurite outgrowth of both motor and sensory neurons146,147. The endogenous BDNF plays a role in controlling axon myelination, as underlined by negative effects given by anti-BDNF antibodies application, which, in addition to reduce regenerated axon length and number of sensory neurons, leads to a lower number of myelinated axons148. Controversial results have been obtained with exogenous BDNF administration: in rat spinal root avulsion model the high level of BDNF immediately after injury promotes long-term survival of adult motorneurons and is able to induce motor axonal growth149, however continuous administrations of low doses of BDNF did not affect the number of regenerated motorneurons and functional recovery150. The divergent effects observed might be due to delivery methods (through hydrogel or as viral vector) and factor concentration. Several strategies were investigated for BDNF release in nerve conduits: a study in rat sciatic nerve injury showed that PHEMA-MMA porous tubes filled with collagen matrix with BDNF improved nerve regeneration respect to empty tube or matrix alone 151, furthermore other studies with collagen conduit loaded with recombinant BDNF linked with laminin hydrogel for laryngeal nerve repair 152, or silicone tube filled with collagene matrix and microspheres releasing BDNF in rat sciatic nerve injury 153 reported similar results.

NEUROTROPHIN-3 (NT-3)

NT-3 neurotrophin was discovered in 1990 and was found in brain, peripheral nerve as well as in kidney, lung and other organs154. The main source of NT-3 in peripheral nerve are satellite glial cells and SC, while TrkC, NT-3 main receptor, is expressed by sympathetic and sensory neurons, as well as in SC130. In DRG or spinal cord neurons the expression of TrkC and NT3 is downregulated after severe injury129. At distal nerve NT-3 mRNA levels decrease shortly after nerve injury and then return to normal level after 2 weeks135,155,156. Besides being important for the survival of several groups of neurons157– 159, NT-3 is described as inductor of SC survival and migration160 and is considered a negative modulator of myelination161. Several studies explored the application of NT-3 in peripheral nerve repair. Direct application of NT-3 in nerve injury site has been shown to prevent the reduction of sensory and motor neuron conduction velocity130. The release of NT-3 by fibronectin matrix in 10 mm nerve gap increases the number of regenerated axons162, moreover NT-3 increases the size and the number of neuromuscolar junctions for fast fibres163,164. Porous poly-2-ydroxyethyl methacrylate-methyl methacrylate (PHEMA- MMA) containing collagen matrix with NT-3 was used to repair 10 mm rat sciatic nerve gap, demonstrating that the enriched conduit let to better nerve regeneration respect to hollow or matrix without factor-filled tube151. A PLGA nerve conduit combined with exogenous NT-3 has been shown to enhance axonal regrowth and improve functional outcome after spinal cord injury165. Moreover a sustained release of NT-3 was achieved

CHAPTER 1: Introduction and Scientific Background immobilizing the factor on silk fibroin scaffold enriched with neural stem cells, the nerve guide was then implanted in rats with transected SCI showing increased axonal 1 regeneration, improved functional outcomes and higher neuronal stem cell survival and differentiation166.

NEUROTROPHIN-4/5 (NT-4/5)

The latest discovered neurotrophins are NT-4 and NT-5, which bind preferentially TrkB and whose expression and role in injured nerve is less studied and still unclear. It has been suggested that the expression of NT-4/5 decreases in motorneurons after nerve injury and their mRNA decline in distal nerve 6-12 h after rat sciatic nerve injury to then increase and became upregulated after 2 weeks135. Despite our little knowledge, NT-4/5 were used to stimulate regeneration of peripheral nerves. Direct application of fibrin glue mixed with NT-4 on nerve injury site improves nerve regeneration, giving higher number of regenerated axons, larger axonal diameter and myelin thickness167. Another study reported that exogenous NT-4 exerts its effect principally on motoneurons innervating type 1 and 2a muscle fibres168. Further NT-4/5 are implicated in the positive effects given by treadmill training after peripheral nerve injury in mice169.

OTHER GROWTH FACTORS

FIBROBLAST GROWTH FACTOR 2 (FGF2)

FGF-2 is a member of FGF family, which counts other 22 members, and exerts its action mainly through high-affinity tyrosine transmembrane receptors (FGF receptors, FGFR). In CNS FGF-2 is expressed by both glial and neuronal cells. In peripheral nerve low levels of FGF-2 and FGFR-1,2 and 3 are founded in intact nerves, but mRNA and protein expression were shown to increase as early as 5 h after injury in proximal and distal nerve stumps, as well as in ganglia, and remain elevated for weeks170. SC, endothelial vascular cells and macrophages were indicated as the main source of these FGF-2 and FGFR, together with sensory neurons170. It has been postulated that FGF-2 might induce SC proliferation, blocking differentiation program, and promotes neovascularisation during nerve repair171,172. Various studies investigated the FGF-2 efficacy in promoting peripheral nerve regeneration. A silicone tube filled with SC overexpressing the 21-23 kDa form of FGF-2 was shown to support sensory recovery in a long gap nerve injury, on the contrary the FGF-218kDa isoform release inhibited axon myelination173. Moreover another groups demonstrated the efficacy of FGF-2 in long gap repair in a study of collagen conduit to bridge 35 mm nerve gap in minipig model, showing elevated number of regenerated axons and increased myelination174. A poly-ε-caprolactone conduit, combined with local delivery of a plasmid vector for FGF-2, was used to bridge 5 mm-gap in mouse sciatic nerve, and results showed increased number of myelinated fibres and S-100 positive cells, better recovery for sensory and motor function respect to conduit alone175.

CHAPTER 1: Introduction and Scientific Background

GLIAL CELL LINE-DERIVED NEUROTROPHIC FACTOR (GDNF)

GDNF, respect to neutrophins, has been recently discovered. It was isolated from supernatant of rat glial cell line B49 culture and later was described as a potent survival factor for motorneurons and midbrain dopaminergic cells176–179. GDNF is not expressed in intact or injured motorneurons180,181, and the main source of this factor in nerve is given by SC. GDNF receptors (c-RET and GFR-α) are found in motorneurons and sensory neurons, and in axotomized motorneurons they are upregulated181,182. Schwann cells express only GFR-α receptor181. Nerve injury induces in distal nerve trunk an increase of GDNF expression together with receptors mRNA upregulation22,181. Exogenous application of GDNF has been shown to improve motor axon regeneration183,184, anyway a single local dose produced a transient effect and, moreover, elevated levels of GDNF generated trapping of regenerating axons and might induce the formation of nerve coils185,186. Several studies investigated the release of GDNF inside nerve conduits and different strategies has been used to control factor release, as adenoviral187,188 or lentiviral-mediated gene transfection189 of motorneurons and Schwann cells; polymer-encapsulated cells producing GDNF190; microsphere releasing GDNF191,192. Generally the regeneration of both motor and sensory neurons was shown to be enhanced by factor release in conduit, with higher number of regenerated axons and improvement of functional recovery. Moreover GDNF has been shown to regulate myelination in vitro193 and in vivo experiments demonstrated that exogenous GDNF increases the myelination of regenerating axons194 and promotes SC proliferation195.

NEUREGULIN 1 (NRG1)

Neuregulins are a family of soluble and transmembrane proteins widely express in tissues, encoded by four genes. The most studies is NRG1 which exists in several isoforms (more than 30 in human) thank to alternative slicing. Most of these isoforms are produced as transmembrane proteins, after enzymatic cleavage the active soluble form is released (only NRG1-type III remain transmembrane after the cleavage) and through autocrine or pararine-manner it binds ErbBs receptors (ErbB3 and ErbB4). In nerves, NRG1 is expressed both by neuron and SC, and in adulthood it controls some aspects of myelination and regeneration process196. In particular soluble NRG1 induces SC dedifferentiation, proliferation, migration and re-myelination196–198. Meanwhile transmembrane NRG1 expressed by axons is required for axon remyelination and for controlling myelin thickness199,200. Due to numerous isoforms, the NRG1 modulation after nerve injury proves to be complex and related to regeneration phases and type of injury201. Generally, the expression of soluble NRG1 is induced after injury in distal nerve stump and remain elevated for at least 30 days, strictly correlated with SC197,202; ErbB2 and ErbB3 receptors are strongly downregulated early after nerve injury202 but result upregulated one week after nerve injury201,203,204; transmembrane NRG1 expression initially decreases, while is upregulated in a tardive phase201,205. Since NRG1 controls axon remyelination and SC

CHAPTER 1: Introduction and Scientific Background behaviours, it has been proposed as therapeutic approach for peripheral and spinal nerve injury repair206. Several delivery systems were tested: PHB conduit filled with alginate 1 matrix containing recombinant NRG1 (type II) was used to bridge 4 cm nerve gap in rabbit peroneal nerve and results showed an increase of SC number, improved axonal regeneration and muscle reinnervation compared to matrix alone or empty tube207; injection at distal and proximal nerve stump of recombinant adenoviruses expressing NRG1 (EGF-like domain of NRG1-type I) has been shown to promote nerve regeneration, as indicated by longer axon lengths and thicker calibers leading to improved sensory and 208 motor functions ; the bridge of 10 mm- gap rat sciatic nerve injury was performed with silicone tube filled with collagen gel containing adipose-derived regenerative cells releasing NRG1 and VEGF, resulting in promotion of SC proliferation and migration209; further transfected SC overexpressing NRG1 were implanted in rat after hemisection spinal cord injury and reported results showed ameliorate recovery, related to increased glial cells proliferation and protection of neuronal apoptosis210. Moreover a gelatin hydrogel releasing NRG1 implanted in rat facial nerve resulted in faster muscular function recovery and higher density of large-diameter axon respect no factor therapy or single NRG1 injection103.

VASCULAR ENDOTHELIAL GROWTH FACTOR (VEGF)

VEGF, mainly associated to angiogenesis, has been recently investigated for its role in nerve regeneration, given the close relationship between blood vessels and nerve fibres211. VEGF family includes 5 members and VEGF A (here indicated as VEGF) is the most studied. VEGF is a homodimeric glycoprotein that exists, thank to posttranslational modifications, in various isoforms characterized by different biochemical and biological properties (solubility, affinity for heparin, bioavailability)212. VEGF exerts its effect via binding to specific tyrosine kinase VEGFR1 and VEGFR2. It is well established that VEGF has a trophic effect on different type of neurons, included DRG neurons, promoting cell survival and neurite outgrowth; further VEGF induces proliferation in microglia and astrocyte cells and stimulates SC survival and migration213. Moreover VEGF is implicated in neuropatic pain and neuroprotection214,215. Induction of VEGF expression has been described after traumatic spinal cord injuries and in DRG neurons after nerve injury216–218. Several studies reported that VEGF releasing nerve grafts or VEGF releasing matrix fillers219, transplantation of transfected cells220 or delivery of VEGF with plasmid injections221 are good strategies to enhanced nerve regeneration. An advantage of VEGF use is its action on both vascular and nervous components in nerve regeneration process.

CHAPTER 1: Aim of The Research

1.2 AIM OF THE RESEARCH 1 Whereas tubulization technique has been proven effective for short nerve gap repair, to now the best clinical approach for long nerve defect repair remains the autograft. The failure of nerve conduit to support regeneration over long distances seems related to the lack of internal framework and the difficult diffusion of growth factors in the inner part of the conduit, thus suggesting that the implementation of tube structure might result in increased ability to sustain the nerve regenerative process.

The aim of the work here presented is to investigate, in vitro and in vivo, various strategies to implement nerve device performance, principally based on growth factor release for trigger and potentiate intrinsic regenerative processes. All growth factors taken into account are differently implicated in the peripheral nerve regeneration process and their release naturally occurs during the reinnervation process.

Specific questions addressed by this part of my thesis:

 Does the recombinant soluble extracellular domain of ErbB4 trap NRG1 and influence glial cell behaviour?  Does the recombinant soluble extracellular domain of ErbB4 release in muscle-in- vein conduit affect peripheral nerve regeneration?  Could neurite induction be increased by incorporation of iron-oxide nanoparticle conjugated growth factors into NVR hydrogel?  Does the combination of iron-oxide nanoparticles conjugated GDNF and NVR hydrogel be effective in enhancing chitosan tube promotion of nerve regeneration after long-gap nerve injury?

CHAPTER 1: Scientific Publications

1.3 SCIENTIFIC PUBLICATIONS 1

Characterization of glial cell models and in vitro manipulation of the neuregulin1/ErbB system

Davide Pascal,1* Alessia Giovannelli1, Sara Gnavi1,2, Stefan Hoyng3,4, Fred de Winter3,4, Michela Morano1, Federica Fregnan1, Paola Dell’Albani5, Damiano Zaccheo6, Isabelle Perroteau1,7, Rosalia Pellitteri5 and Giovanna Gambarotta1,7

1 Department of Clinical and Biological Sciences, Nerve Regeneration Group, University of Torino, Italy 2 Neuroscience Institute Cavalieri Ottolenghi (NICO), Torino, Italy 3 Department of Neuroregeneration, Netherlands Institute for Neuroscience, Royal Netherlands Academy of Arts and Sciences (KNAW), Amsterdam, the Netherlands 4 Department of Neurosurgery, Leiden University Medical Center, Leiden, the Netherlands 5 Institute of Neurological Sciences, National Research Council (CNR), Section of Catania, Italy 6 Department of Experimental Medicine, Section of Anatomy, University of Genova, Italy 7 Neuroscience Institute of Torino (NIT), Interdepartmental Centre of Advanced Studies in Neuroscience, University of Torino, Italy

published in BioMed Research International DOI: 10.1155/2014/310215

ABSTRACT

The neuregulin1/ErbB system plays an important role in Schwann cell behavior both in normal and pathological conditions. Upon investigation of the expression of the neuregulin1/ErbB system in vitro, we explored the possibility to manipulate the system in order to increase the migration of Schwann cells, that play a fundamental role in the peripheral nerve regeneration. Comparison of primary cells and stable cell lines shows that both primary olfactory bulb ensheathing cells and a corresponding cell line express ErbB1- ErbB2 and neuregulin1, and that both primary Schwann cells and a corresponding cell line express ErbB2-ErbB3, while only primary Schwann cells express neuregulin1. To interfere with the neuregulin1/ErbB system, the soluble extracellular domain of the neuregulin1 receptor ErbB4 (ecto-ErbB4) was expressed in vitro in the neuregulin1 expressing cell line, and an unexpected increase in cell motility was observed. In vitro experiments suggest that the back signaling mediated by the transmembrane neuregulin1 plays a role in the migratory activity induced by ecto-ErbB4. These results indicate that ecto-ErbB4 could be used in vivo as a tool to manipulate the neuregulin1/ErbB system.

CHAPTER 1: Scientific Publications

INTRODUCTION

The neuregulin1/ErbB system plays an important role in Schwann cell (SC) behavior both in normal and pathological conditions [1] and the possibility to manipulate it gives new perspectives to improve post-traumatic nerve regeneration [2-4].

The ErbB receptor family consists of four tyrosine kinase receptors: Epidermal Growth Factor Receptor (EGFR, also called ErbB1 or HER1), ErbB2, ErbB3 and ErbB4 [5]. ErbB receptors bind several structurally related growth factors. Among them, neuregulin1 (NRG1) is the most characterized and studied in the peripheral nervous system (PNS) for its role in axon-glial signaling and SC activity. The NRG1 gene codes for more than 20 different isoforms [1, 6-8] which differ because of alternatively spliced exons. Actually, soluble and transmembrane isoforms were described, that differ in the presence of N- terminal domains and their signaling mode: soluble isoforms (types I and II) are mainly released by glial cells and signal in a paracrine/autocrine manner, transmembrane isoforms (type III) are mainly expressed by axons and signal in a juxtacrine manner [8]. NRG1 are further classified as alpha and beta isoforms, according to the characteristics of their EGF- like domain.

It has been shown that transmembrane ligand-receptor interactions may lead to a process of back-signaling, which is mediated by the action of a -secretase enzyme which causes the release of a cytoplasmic fragment able to translocate into the nucleus [9, 10].

In order to better study the role of the NRG1/ErbB system in the peripheral nerve, four different in vitro models were analyzed: primary rat SC harvested from sciatic nerve, a SC line (RT4-D6P2T) [11], primary glial cells of the olfactory nerve, known as Olfactory Ensheathing cells (OEC) and a Neonatal Olfactory Bulb Ensheathing cell line (NOBEC) [12]. In vitro experiments were carried out to address these questions:

Are NRG1 and ErbB receptor expressed in these four cellular models? What are the in vitro effects of manipulating the NRG1/ErbB system by expression of the soluble extracellular domain of the NRG1 receptor ErbB4 (ecto-ErbB4) in glial cells?

MATERIALS AND METHODS

In vitro assays

Cultures of NOBEC, RT4-D6P2T and COS7. Neonatal Olfactory Bulb Ensheathing Cells (NOBEC) line, derived from primary cells dissociated from neonatal rat olfactory bulb and immortalized by retroviral transduction of SV40 large T antigen [12], was kindly provided by Dr. Jacobberger (Comprehensive Cancer Center, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 4106-4944, USA). Rat RT4-D6P2T [11] and COS7 were provided by the American Type Culture Collection (ATCC). Cell lines were grown as monolayer at 37°C in a humidified atmosphere of 5% CO2/air, in Dulbecco’s modified

CHAPTER 1: Scientific Publications

Eagle’s medium (DMEM, Invitrogen, UK) supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, 1 mM sodium pyruvate, β mM l-glutamine, and 10% heat-inactivated 1 foetal bovine serum (FBS; Invitrogen). Recombinant NRG11 has been produced in the laboratory as a His-tag fusion protein in E. coli [13].

Cultures of Primary Schwann cells (SC). To harvest Schwann cells (SC), rat sciatic nerves were exposed, removed and kept in DMEM plus glutamax (Invitrogen, UK) containing 100 U/ml penicillin, 100 μg/ml streptomycin. Nerves were then dissected in trunks, de- sheathed and finally chopped in 1 mm segments. The segments were then plated in a Petri dish in SC growth medium (DMEM plus glutamax containing 100 U/ml penicillin, 100 μg/ml streptomycin, 14 μM forskolin and 100 ng/ml NRG11, R&D Systems, UK). Cells were incubated for 2 weeks at 37 °C with fresh medium added approximately every 72 h. After these 2 weeks medium was aspirated and 0.125% (w/v) collagenase type IV and 117 U/mg dispase were added to the Petri dish. After 24 hours (h) incubation, cell suspension was filtered through a 70-mm cell strainer (Falcon; BD Biosciences Discovery Labware, Bedford, MA), centrifuged at 100 x g for 5 min to obtain the cell pellet. Finally, the cell pellet was re-suspended in SC growth medium, seeded into a Petri dish pre-coated with poly-D-lysine (Sigma, St Louis, MO, USA) and incubated in the same conditions. The following day, the medium was changed and cells were left to proliferate. When confluent, SC were purified by an antibody complement method to eradicate the remaining fibroblasts [14-16].

Cultures of primary Olfactory Ensheathing Cells (OEC). OEC were isolated from 2-day-old rat pups (P2) olfactory bulbs using an already described method [17]. Ten neonatal rats were used to produce each batch of OEC. Initial steps involved peeling away the olfactory nerve layer from the rest of the bulb and digesting the tissue in MEM-H containing 0.03% collagenase and 0.25% trypsin for 15 min at 37 °C. This step was repeated twice with a fresh solution. Trypsinization was stopped by adding 10% foetal calf serum (FCS)- DMEM. The digested tissue was mechanically dissociated by trituration and filtrated through a 80-μm nylon mesh followed by centrifugation at 500 × g for 10 min. Cells were resuspended and plated in flasks, fed with fresh 10% FCS-DMEM, supplemented with 2mM L-glutamine, 50U/ml penicillin and 50μg/ml streptomycin. β4h after initial plating, 10μM antimitotic agent cytosine arabinoside was added to reduce the number of dividing fibroblasts. OEC cultures were further processed by passing cells from one flask to another. This step reduces contaminating cells because they adhere more readily to plastic than OEC. In the last passage OEC were plated on 25cm2 flasks and cultured in 10% FCS- DMEM supplemented with bovine pituitary extract. OEC purity was verified using immunocytochemistry with p75 and S-100. The percentage of S-100/p75 positive cells in the cultures was ~ 85-90 % (data not shown). Cells were incubated at 37°C in 10% FCS- DMEM and the medium was changed twice a week.

CHAPTER 1: Scientific Publications

GDNF stimulation assay. Purified primary OEC and NOBEC were plated on 14-mm diameter poly-L-lysine (PLL, 10 g/ml, Sigma) coated glass coverslips at a final density of 3 x 103 cells/coverslip and grown both in 10% FCS-DMEM and in serum-free DMEM. Cells were cultured with the addition of Glial Derived Neurotrophic Factor (GDNF, 1 ng/ml, Chemicon) for eight days, changing the medium twice. Control cultures received medium with no addition of trophic factor (CTR). Then, cells were processed by immunocytochemical procedures and total RNA extraction.

Immunocytochemistry. OEC and NOBEC were fixed in 0.1 M phosphate buffer pH 7.4 (PBS) containing 4% paraformaldehyde (PAF) for 30 minutes (min). After washing in PBS cells were treated with PBS containing 5% normal goat serum (NGS), 0.1% Triton X-100 at room temperature (RT) for 15 min. OEC and NOBEC were incubated overnight at 4°C with the following primary antibodies: anti-S-100 (mouse, working dilution/w.d. 1:100; Sigma Aldrich), anti-nestin (rabbit, w.d. 1:100; Immunological Science), anti-vimentin (mouse, w.d. 1:50; Dako). After washing, cells were incubated for 45 min at RT with the correspondent anti-mouse and anti-rabbit fluorescent secondary antibodies to visualize primary antibodies. The immunostained coverslips were analyzed on a Zeiss fluorescence microscope and images were captured with an Axiovision Imaging System. No staining of cells was observed in control incubations in which the primary antibodies were omitted.

Cell Transfection. For transient transfection of plasmidic DNA, NOBEC, RT4-D6P2T and COS7 cells were transfected with Lipofectamine β000, according to manufacturer’s instructions. Efficient expression of the recombinant protein was assessed by western blot analysis.

Cell migration assay. Transwell assays were performed 48 h after DNA transfection as previously described [18]. To inhibit the -secretase, cells were pre-treated with 100 M DAPT (-secretase inhibitor compound IX, Calbiochem) for 3 days. Since DAPT was diluted with DMSO, the control was carried out by treating cells with the same volume of this solvent. For each Transwell four images were analyzed and the amount of migrated cells was evaluated as the total area of migration (in pixel2) calculated with the Image J software and expressed as percentage of the total number of migrated cells for each single experiment. Cells were discriminated by the pores of the Transwell membrane by applying a threshold of 300 pixel2. For each experimental condition a technical triplicate was made and each experiment was repeated at least 3 times.

RNA isolation and cDNA preparation. Total RNA was extracted with TRIzol (Invitrogen) according to the manufacturer's instructions, adding 5μg glycogen as a carrier to facilitate RNA precipitation. 1 μg total RNA was subjected to a reverse transcriptase (RT) reaction in β5 μl reaction volume containing: 1X RT-Buffer (Fermentas); 0.1μg/μl bovine serum albumin (BSA, Sigma); 0.05% Triton X-100; 1mM dNTPs; 7.5μM random exanucleotide primers (Fermentas); 1U/μl RIBOlock (Fermentas) and β00U RevertAid™ M-MuLV

CHAPTER 1: Scientific Publications reverse transcriptase (Fermentas). The reaction was performed for 10min at 25°C, 90min at 42°C, 10min at 90°C. Control reactions “RT-“ (without the enzyme RT) and H2O, without 1 RNA, were carried out.

Quantitative real-time PCR (qRT-PCR) analysis. Quantitative real-time PCR analysis was performed using Sybr Green chemistry; data were analyzed by ΔΔCt relative quantification method normalizing to the housekeeping gene Tata-box Binding Protein (TBP) (forward primer: 5’-TAAGGCTGGAAGGCCTTGTG-γ’; reverse primer: 5’- TCCAGGAAATAATTCTGGCTCATAG-γ’). Real-time PCR reactions were performed using the 7300 real-time PCR system (Life Technology). Each sample was run in triplicate on 96-well optical PCR plates (Life Technology). In each well a PCR reaction was carried out on 5 μl cDNA corresponding to β0 ng starting RNA (1 μg RNA retrotranscribed in β5 μl, diluted 1:10 in water), Sybr Green PCR Master Mix (Life Technology) and γ00nM primers (Life Technology) in a reaction volume of β5 μl. Specific primers designed to amplify ErbB1, ErbB2, ErbB3, ErbB4, NRG1, p75NGFR, GFAP, S100 are listed in Table 1. After an initial denaturation step for 10 min at 95°C, denaturation in the subsequent 40 cycles was performed for 15 seconds (sec) at 95°C followed by primer annealing and elongation at 60°C for 1 min. Relative amount of mRNA that had been retrotranscripted into cDNA was calculated by comparative (ΔΔCt) method. In the first step of the method, the difference between Ct values of target and housekeeping gene was calculated (ΔCt), whereas in the second step the difference between the ΔCt values of the samples and the calibrator was determined (ΔΔCt). For each gene, the cells with the highest level of expression were chosen as calibrator. The normalized relative quantity (NRQ) was determined using the formula: NRQ=2-(ΔΔCt). Results were expressed as mean + SEM.

CHAPTER 1: Scientific Publications

HybondTM C Extra membrane as previously described [19]. Primary antibodies used are: ErbB1 (#sc-03), ErbB2 (#sc-284), ErbB3 (#sc-285), ErbB4 (#sc-283), NRG1 (#sc-347, #sc-348), w.d. 1:1000, from Santa Cruz; p-AKT (#4051), AKT (#9272), p-ERK (#9106), ERK (#9102) w.d. 1:1000, from Cell Signaling; flag (#F7425) w.d. 1:10000, from Sigma; p75-NGFR (#ab52987), w.d. 1:2000, from Abcam; GAPDH (#4300), w.d. 1:20000, from Ambion; secondary antibodies used are: horseradish peroxidase-linked anti-rabbit (#NA934) and anti-mouse (#NA931) w.d. 1:40000, from GE Health.

Nuclear and cytoplasmic protein extraction. Cell plates were washed 3 times with PBS on ice. On each plate, 300µl of lysis buffer (20mM Tris-HCl pH8.0, 20mM NaCl, 0.5% NP40, plus an anti-protease cocktail, Roche) were added for 10min at 4°C. After lysis, cells and nuclei were collected and centrifuged for 30min at 3000 rpm at 4°C, in order to separate the nuclear and cytoplasmic material; the supernatant, which contains the cytoplasmic extract, was separated from the pellet, syringed with a G26 needle and then centrifuged for 5min at 12000rpm at 4°C. Supernatant was collected, aliquoted and frozen at -80°C. The pellet obtained after the first centrifuge, containing nuclear material, was washed 3 times with lysis buffer and centrifuged, in order to remove cytoplasmic protein traces. Pellet was resuspended in γ0μl buffer C (β0mM Hepes pH8.0; 1.4βmM NaCl, 1.5mM MgCl2, 1.2mM EDTA); then, γ0μl Buffer C containing 50% glycerol was added, and nuclear proteins were incubated on ice for 30min. The treatment with this hyperosmotic buffer causes cell nucleus collapse and nuclear protein spillage. After incubation, extracts were centrifuged for 90sec at 12000rpm at 4°C. The supernatant containing nuclear proteins was collected, centrifuged again, diluted 1:3 in 20mM Hepes pH 8.0 and stored at -20°C.

Construct preparation

Cloning strategies for NRG1-ICD-NLS and NRG1-ICD-NLS constructs. Because the C terminal cytoplasmic domain is common to all NRG1 isoforms, the cDNA coding for NRG1-typeI-1a (kindly provided by K. Lay, accession number NM_01γ956) was used as template to clone the cytoplasmic domain of NRG1. To allow expression of the protein, an artificial start codon (ATG, shown in bold in the primer sequence) was added on the forward primer, inserted in a Kozac sequence. To obtain an Intra-Cellular Domain (ICD) containing the Nuclear Localization Signal (NLS) the following primers were used: forward: 5’- TAGCCTGCAGCATGGGCAAAACCAAGAAACAGCG-γ’; reverse: 5’- ATCGATATCTACAGCAATAGGGTCTTGGTTAG. To obtain an Intra-Cellular Domain (ICD) missing the Nuclear Localization Signal (ΔNLS) -the following primers were used: forward: 5’- TAGCCTGCAGCATGGAGCTTCATGATCGGCTCC -γ’; reverse: 5’- ATCGATATCTACAGCAATAGGGTCTTGGTTAG. Restriction enzymes sequences (underlined) were added to the primers to facilitate the subcloning (PstI in the forward primer, EcoRV in the reverse primer). Amplification reactions were carried out using 0,8ng template (NRG1-tipoIII-1a) and the AmpliTaq Gold enzyme, following manufacturer’s instructions, in the Thermal Cycler GeneAmp PCR System β400 (Perkin 50

CHAPTER 1: Scientific Publications

Elmer). The amplification was performed according to the following protocol: 5min at 94°C; then 40 cycles: 30sec at 94°C, 30sec at 60°C, 90sec at 72°C; 20min at 72°C. 1 Amplification products were cloned into the pGEM-T vector, using chemically competent JM-109 cells (Promega), following manufacturer’s instructions. Two clones, corresponding to the constructs ICD-NLS and ICD-ΔNLS, were completely sequenced (BMR Genomics Laboratories). Inserts were recovered by EcoRV and NcoI digestions, sticky ends were blunted and inserted into pIRESpuro2 (Clontech) previously cut with NotI, blunted and dephosphorylated.

Cloning strategies for ecto-ErbB4-FLAG construct and subcloning into lentiviral vector. To obtain a construct to express the extracellular domain of ErbB4 (ecto-ErbB4) fused with a FLAG epitope, the ErbB4 extracellular domain - recovered from the pIRES-puro2- ErbB4-JMa-cyt2 construct [18] by EcoRV and BbsI digestion and blunting - was subcloned into the EcoRV site of the pCMV-Tag4a vector (Stratagene). To obtain a lentiviral vector to express ecto-ErbB4-FLAG, the insert with the FLAG and the following STOP codon was recovered from pCMV-Tag4c vector by EcoRV and KpnI digestion and blunting and subcloned into the multiple cloning site (mcs) of the lentiviral vector pRRL- CMV-mcs-WPRE, ﬂanked by the constitutively active CMV promoter and the woodchuck hepatitis virus posttranscriptional regulatory element (WPRE).

Lentivirus production and RT4-D6P2T infection. LV stocks were generated as previously described [20-22]. Briefly, two 15cm diameter Petri dishes containing 1.25 x 107 HEKβ9γT seeded in Iscoves modified Dulbecco’s medium (IMDM) containing 10% FCS, 100U/ml penicillin, 100μg/ml streptomycin (PS) and Glutamax (Invitrogen) were prepared. Using branched polyethylenimine (Sigma, St Louis, MO) a triple transfection with the LV transfer, packaging (pCMVdeltaR8.74) and envelope (pMD.G.2) plasmid was performed (ratio γ:β:1, total DNA 90 μg/plate). After 14h, the medium was replaced by IMDM containing 2% FCS, 1% PS and Glutamax. After 24h, the medium was harvested, filtered through a 0.ββμm filter and concentrated by ultracentrifugation at β0.000rpm for β.5h in a SW32Ti rotor (Beckman Coulter B, The Netherlands). Viral pellets were resuspended in PBS pH 7.4 aliquoted and stored at -80°C until further use. Serial dilutions (10-2,-3,-4 and 10-2 to -7 for LV-GFP) of all viral stocks were used to infect 2 x 105 HEK293T in IMDM 2% FCS, 1% P/S and Glutamax seeded in poly-L-lysine (PLL) coated 24-well culture plates. After 48 h the number of transducing units per ml (TU/ml) for the LV-GFP stock was manually quantified by counting transduction events in the LV-GFP transduced cells using a fluorescence microscope (10-5,-6,-7) and genomic DNA (gDNA) of all samples was extracted and measured for viral integrating events by quantitative PCR (10-2,-3,-4). Briefly, cells were harvested and gDNA was extracted (DNeasy Blood & Tissue Kit, Qiagen, Venlo, the Netherlands). Viral mediated transgene integration was measured using primers directed against the lentiviral WPRE on an ABI 7900HT detection system (Applied Biosystems) using the SYBR green PCR kit (Applied Biosystems). WPRE primers sequences were: 5’-TTCCCGTATGGCTTTCATTT-γ’ and 5’- 51

CHAPTER 1: Scientific Publications

GAGACAGCAACCAGGATTTA-γ’. All expression values were normalized to that of the reference gene GAPDH. GAPDH primers were as follows: 5’- TGCACCACCAACTGCTTAGC-γ’ and 5’-CGCATGGACTGTGGTCATGA-γ’. The ratio between the TU/ml and gDNA WPRE content of the LV-GFP stock was used to calculate relative TU/ml titers for all stocks on the basis of their gDNA WPRE content.

RT4-D6P2T cells seeded at 1 x 103 cells/ well in a 96-well plate were infected using LV- sGFP or a combination of LV-sGFP and LV-Ecto-ErbB4 FLAG. Cells were infected three times at a MOI of 100. Confluent cells were harvested, expanded and frozen.

Western Blot and immunocytochemistry were performed on confluent cells in order to evaluate virus infection efficiency and sGFP and ecto-ErbB4-FLAG expression.

Confluent cells were fixed by incubation with 4% PFA and permeabilized with 0.2% Triton X-100 diluted in PBS by 45 min incubation at RT. Blocking solution containing 5% FCS diluted in 0.2% Triton X-100 PBS was applied for 1h at RT. Monoclonal anti-GFP antibody (1:500 in blocking solution, Abcam) and rabbit anti-FLAG primary antibody (1:500 in blocking solution,Sigma) were incubated over night (o/n) at 4°C. Following 3 washes of 15min each goat-anti-rabbit IgG (H+L) Cy3 and goat-anti-mouse IgG (H+L) Alexa 488 secondary antibody (Invitrogen, diluted 1:200 in PBS) were incubated 2h at RT. Following 3 washes of 15min nuclei were stained using Hoechst (Sigma) diluted 1:1000 in PBS. Fluorescent images were acquired using an inverted optical microscope (Axiovert 200, Zeiss). Western blot analysis was performed as described above.

Ecto-ErbB4-FLAG protein purification

Co-Immunoprecipitation assay. Ecto-ErbB4-FLAG protein for co-immunoprecipitation assay was obtained from conditioned medium of COS7 transiently expressing the protein. After three days of culture, medium was collected, centrifuged and 1 ml aliquots were incubated with either 100 ng or β00 ng recombinant NRG11-Hys protein and immunoprecipitated as previously described [18] using an anti FLAG polyclonal antibody (SIGMA #F7425). Immunoprecipitated proteins were analyzed by western blot, using an antibody directed to the hystidine tag.

Ecto-ErbB4-FLAG protein production and purification. Ecto-ErbB4-FLAG protein for in vitro experiments was purified from conditioned medium of RT4-D6P2T stably expressing the protein. After four days of culture in the presence of serum free medium, medium was collected, centrifuged and filtered (0,ββμm filter). Target protein was then purified using ANTI-FLAG M2 affinity gel (SIGMA #A2220), and eluted by competitive elution using 100μg/ml FLAG peptide. Eluted solution was collected in 10 fractions and a western blot was performed to confirm the presence of target protein. Only positive fractions were frozen in liquid nitrogen, adding 15% glycerol to prevent protein damage, and stored at - 80°C.

CHAPTER 1: Scientific Publications

Statistical analysis

Quantitative data are presented as mean + SEM. All data were statistically analyzed using 1 t-test or one-way analysis of variance and post-hoc analysis by the Bonferroni test (SPSS software).

RESULTS

Comparative characterization of primary glial cell cultures and glial cell lines

Cellular differentiation of OEC and NOBEC upon GDNF treatment under different growth conditions. To characterize these cell models in vitro, a stimulation assay was carried out with Glial cell-derived neurotrophic factor (GDNF), a well-recognized growth factor for glial cells [23], to both primary OEC and NOBEC cell line cultures in order to evaluate its effect on cell survival and expression of glial differentiation markers. Concentration and exposure time of the cultures to the GDNF were previously established [24]. In the presence of serum, the majority of OEC and NOBEC exhibited both star and spindle shapes which are their in vitro typical morphological features (Figure 1).

Results show a different intensity level in the expression of some glial markers both in OEC and in NOBEC, grown in different conditions and treated with GDNF: S-100 immunoreactivity was higher in OEC and NOBEC grown in the presence of serum in comparison with cells grown in serum free medium (SFM); when GDNF was added to cultures in the presence of serum, an increased number of positive S-100 both in OEC and in NOBEC was observed.

Figure 1. OEC and NOBEC respond similarly to GDNF stimulation. Figure shows representative fields of OEC and NOBEC immuno-stained with anti-S-100, anti-nestin and anti-vimentin antibodies. Cells were grown for eight days after plating with and without serum and GDNF. Fields were chosen to clearly show both the morphological aspect and the specific marker expression. Scale bars 50 µm.

CHAPTER 1: Scientific Publications

Nestin immunoreactivity showed a more intense expression in NOBEC compared with OEC in all culture conditions and the addition of GDNF determined an up-regulation of nestin both in serum and in SFM in NOBEC and OEC. Vimentin was more expressed in OEC grown in the presence of serum and GDNF addition increased the expression of vimentin in OEC grown both in serum and in SFM. NOBEC showed a major expression of vimentin compared with OEC, especially when they were grown with GDNF both in serum and in SFM (Figure 1). This different expression suggests that NOBEC cell line is more resistant to stress than OEC primary cultures and that NOBEC might represent a good glial cell model for in vitro assays.

Expression analysis of NRG1/ErbB system and glial markers. The mRNA expression level of the NRG1/ErbB system (Figure 2) and of glial markers (Figure 3) was examined by quantitative real-time PCR (qRT-PCR) in primary cultures of Schwann cells (SC) and olfactory bulb ensheathing cells (OEC) and was compared with the corresponding stable cell lines RT4-D6P2T (derived from a Schwannoma) and NOBEC (Neonatal Olfactory Bulb Ensheathing Cells).

For each gene, the normalized relative quantity (NRQ) was determined using as calibrator (NRQ=1) the sample with the highest level of expression; therefore, the relative (and not absolute) gene expression shown in the graphs cannot be compared among different genes. Results show that ErbB1 is expressed by OEC and NOBEC, while SC and RT4-D6P2T do not express this receptor; ErbB2 is expressed by all cell types; ErbB3 is expressed by SC and RT4-D6P2T cells, barely detectable in OEC and NOBEC; ErbB4 mRNA is barely detectable only in SC and OEC.

To investigate the presence of mRNA coding for NRG1, different primer pairs were used that allow the amplification of different NRG1 isoforms [25]. Results show that RT4- D6P2T cells do not express any NRG1 isoform, whereas OEC and NOBEC express different NRG1 isoforms (type I/II, α and , type III); SC express only NRG1 type I/II isoforms (Figure 2). The glial gene GFAP is expressed by SC, OEC, RT4-D6P2T; S100 and p75 are expressed by all cell types, mostly by SC and RT4-D6P2T (Figure 3).

Immunoblot analysis was performed to confirm protein expression of some of the genes analyzed by qRT-PCR (Figure 4). A faint band corresponding to ErbB1 protein is visible in OEC and NOBEC samples; ErbB2 is expressed, at different levels, in all samples; ErbB3 is strongly expressed by SC and RT4-D6P2T, lowly expressed by NOBEC; ErbB4 is not detectable and for this reason a positive control was added in the Western Blot.

Different NRG1 isoforms are detectable in OEC, NOBEC and SC. RT4-D6P2T cell line does not express NRG1.

CHAPTER 1: Scientific Publications

Figure 2. Primary cultures and cell lines derived from OEC and SC express different levels of NRG1 isoforms and ErbB receptors. Graphs show normalized relative quantification (NRQ) of the different NRG1 isoforms and ErbB receptors obtained by qRT-PCR. For each gene, the cells with the highest level of expression were chosen as calibrator (NRQ=1). Data are presented as mean + SEM.

Ecto-ErbB4 stimulates NOBEC migration. It has been shown that SC migrate following stimulation with soluble NRG1 and that the removal of soluble NRG1 with recombinant soluble receptors, such as the soluble extracellular fragment of ErbB3, strongly reduces SC migration [26].

CHAPTER 1: Scientific Publications

Figure 3. Primary cultures and cell lines derived from OEC and SC express different levels of glial genes. Graphs show normalized relative quantification (NRQ) of GFAP, S100 and p75 obtained by qRT- PCR. For each gene, the cells with the highest level of expression were chosen as calibrator (NRQ=1). Data are presented as mean + SEM.

NOBEC are characterized by high migratory activity [27], which could be due to an autocrine stimulation mediated by endogenous soluble NRG1 on endogenous ErbB3-ErbB2 heterodimer.

To verify this hypothesis, a construct to express the soluble extracellular domain of ErbB4 (ecto-ErbB4) tagged with a FLAG was produced (see Materials and Methods). By western blot analysis the expression of ecto-ErbB4-FLAG was assayed both in the cell extract and in the supernatant of transiently transfected COS7 cells, to verify that the soluble extracellular domain of ErbB4 was expressed and released in the extracellular environment (Figure 5A).

A co-immunoprecipitation assay was performed to verify the ability of ecto-ErbB4-FLAG to interact with soluble NRG1: different amounts of soluble recombinant NRG11 tagged with 6 histidine [13] were incubated with conditioned medium containing recombinant ecto-ErbB4 tagged with FLAG. A co-immunoprecipitation against FLAG (ecto-ErbB4) was performed and co-immunoprecipitated proteins were analyzed by western blot using an antibody against histidine (NRG11). Data confirm that ecto-ErbB4 and NRG11 interact (Figure 5B).

A three-dimensional migration assay (transwell assay) was performed using NOBEC transiently transfected with the expression vector for ecto-ErbB4-FLAG (Figure 6A). Contrary to what expected, transwell assays showed that the expression of the soluble extracellular portion of the receptor ErbB4 increased cell migration. The same results were

CHAPTER 1: Scientific Publications obtained when NOBEC were transiently transfected with an expression vector for ecto- ErbB3-FLAG (Figure 6B). 1

Figure 4. Western Blot analysis confirms that primary cultures and cell lines derived from OEC and SC express different levels of ErbB receptors and glial proteins. Western Blot analysis of proteins extracted from OEC, NOBEC, RT4-D6P2T, SC were analyzed with antibodies directed to ErbB1, ErbB2, ErbB3, ErbB4, NRG1, p75 and GAPDH. Different NRG1 isoforms are expressed by OEC, NOBEC, SC. An asterisk indicates a positive control for ErbB4 expression (a cell line stably expressing ErbB4 [18]).

ecto-ErbB4 - NRG1 interaction. The pro-migratory activity elicited by ecto-ErbB4 could be explained by two different models:

1 - ecto-ErbB4, sequestering endogenous soluble NRG1, stimulates migration. This hypothesis would be confirmed if soluble NRG1 would have an inhibitory effect on NOBEC migration. To test this hypothesis a transwell assay was performed in the presence of 50 ng/ml soluble recombinant NRG11. Transwell assay analysis demonstrates that NRG11 does not inhibit migration and, indeed, it slightly stimulates migration (Figure 6C). Therefore, the first model does not explain the ecto-ErbB4 mediated migration.

2 - ecto-ErbB4, interacting with transmembrane NRG1, stimulates migration. Actually, it is known that transmembrane NRG1 type III isoforms are able to mediate reverse signaling [9, 10]: upon interaction with ErbB receptor, transmembrane NRG1 undergoes a proteolytic cleavage mediated by a -secretase, which releases a cytoplasmic fragment which is able to translocate into the nucleus, mediating reverse signaling.

CHAPTER 1: Scientific Publications

Figure 5. Recombinant ecto-ErbB4-FLAG is expressed and released by cells and is able to interact with soluble NRG1. Panel A- The correct expression of ecto-ErbB4-FLAG was assayed both in the cell extract and in the conditioned medium (surnatant) of transiently transfected COS7 cells (ecto-ErbB4). Mock samples are COS7 cells transfected with the empty vector. The asterisk indicates an unspecific band. Panel B- Different amounts of soluble recombinant NRG11 tagged with 6 histidine were incubated with conditioned medium of COS7 cells transiently expressing recombinant ecto-ErbB4-FLAG. A FLAG co- immunoprecipitation was performed, followed by western blot against histidine to recognize NRG11. The asterisk indicates the band corresponding to the primary antibody used for co-immunoprecipitation.

To verify the expression of NRG1 type III isoforms, an RT-PCR was carried out using specific primers that amplify these isoforms on RNA extracted by NOBEC wild type and NOBEC expressing ecto-ErbB4. RT-PCR confirmed that these NRG1 isoforms are expressed by NOBEC (data not shown), both wild type and expressing ecto-ErbB4.

To validate the reverse-signaling hypothesis, a transwell migration assay was performed by treating NOBEC expressing ecto-ErbB4 and wild type (WT) NOBEC with DAPT, a specific inhibitor of -secretase, the enzyme which mediates NRG1 proteolytic cleavage, releasing a cytoplasmic fragment.

In both cell lines, treatment with DAPT inhibits the migration in a consistent way. However, the migration is inhibited more strongly in NOBEC expressing ecto-ErbB4, supporting the hypothesis that this soluble fragment, binding to NRG1-typeIII, leads to the production of a cytoplasmic fragment responsible of the stimulus for migration. The migration is inhibited by 75% in NOBEC expressing ecto-ErbB4 (Figure 6E), while only by 25% in wild type NOBEC (Figure 6D). Actually, there are numerous signaling pathways that require the presence of -secretase for the propagation of the signal; therefore, treatment with DAPT could interfere also with the signaling produced by other signal transduction pathways.

NRG1 cytoplasmic fragment stimulates NOBEC migration. To investigate the role of the cytoplasmic fragment of NRG1 produced after interaction with ErbB receptors, two different constructs were prepared to express the NRG1 intracellular domain (ICD): 58

CHAPTER 1: Scientific Publications

NRG1-ICD-NLS, containing the nuclear localization sequence (NLS), and NRG1-ICD- NLS, missing NLS (Figure 7A). NLS is a sequence consisting of eight amino acids 1 (KTKKQRKK) following the transmembrane portion of the protein, present in all NRG1 isoforms containing an ICD [9].

Specific primers were designed with the insertion of an artificial ATG to allow the expression of the protein in transfected cells; in one primer the nuclear localization sequence was included, in the other one it was omitted. To verify that the nuclear localization sequence (NLS) inserted into the construct NRG1-ICD-NLS works properly, a western blot analysis of nuclear and cytoplasmic proteins was performed (Figure 7B).

Western blot data show that cells transfected with the construct NRG1-ICD-NLS express the protein both in the cytoplasm and in the nucleus, while cells transfected with the construct NRG1-ICD-NLS, lacking the nuclear localization sequence, express the protein predominantly in the cytoplasm. A barely detectable signal can be observed also in the nucleus; this could be due to the fact that this fragment interacts with proteins, such as LIMK1 [28], able to enter the nucleus carrying associated proteins.

Transwell migration assays were performed with wild type NOBEC transiently transfected with NRG1-ICD-NLS and NRG1-ICD-NLS (Figure 7C).

Migration assay analysis shows that transient expression of the NRG1 cytoplasmic domain able to translocate into the nucleus, confers a migratory activity that is significantly higher compared to control cells. Unexpectedly, the expression of the NRG1 cytoplasmic fragment lacking NLS, confers to the cells a migratory activity that is significantly higher than control, and of the cells expressing the isoform with NLS. These data suggest that NRG1-ICD plays a role in the stimulation of the migratory activity, when it is in the cytoplasm.

NRG1 cytoplasmic fragment does not stimulate RT4-D6P2T migration. To understand if the cytoplasmic fragment of NRG1 type III is able to confer migratory activity to cells that do not express endogenously any isoform of NRG1, the cell line RT4-D6P2T was transiently transfected with vectors to express NRG1-ICD-NLS and NRG1-ICD-NLS and migratory activity was analyzed by transwell assays (Figure 7D).

Data show that transient expression of NRG1-ICD does not confer migration activity, suggesting that these cells lack pivotal factors necessary to mediate NRG1 migratory signal transduction.

Signal transduction pathways activated by ecto-ErbB4 on the NOBEC cell line. It has been shown that in neurons - expressing transmembrane NRG1-typeIII - stimulation with soluble ErbB4 fragment produces an increase in AKT phosphorylation, while ERK phosphorylation remains unchanged [29].

CHAPTER 1: Scientific Publications

Figure 6. NOBEC transiently expressing ecto-ErbB4 and ecto-ErbB3 migrate more than control cells and their migration is inhibited by DAPT treatment. Migration activity of NOBEC transiently transfected with empty vector (AAV) was compared with migration activity of NOBEC expressing ecto-ErbB4 (ErbB4, Panel a) or ecto-ErbB3 (ErbB3, Panel b). Soluble recombinant NRG11 stimulates NOBEC wild type (WT) migration (Panel c). NOBEC wild type (Panel d) or NOBEC expressing ecto-ErbB4 (Panel e) were pre- treated for three days with 100 M DAPT (-secretase inhibitor) or DMSO (mock control). Cell migration was assessed by transwell assays. Values represent the average of three biological replicates performed as technical triplicates. Values of each replicate are expressed in percentage with respect to the total number of cells that migrated in that experiment (**, p ≤ 0.01; ***, p ≤ 0.001). 60

CHAPTER 1: Scientific Publications

To further investigate the reverse signaling mediated by ecto-ErbB4 in NOBEC cells, a time course of stimulation was performed. 1 .

Figura 7. The expression of NRG1 intracellular domain stimulates cell migration. (a) To express the NRG1 intracellular domain (ICD), two constructs were prepared, one containing the nuclear localization sequence (NLS), one lacking it (ΔNLS). ICD is shown in blue, NLS in yellow and the inserted ATG in red. (b) Validation of nuclear and cytoplasmic localization of NRG1-ICD fragments. Nuclear and cytoplasmic proteins were extracted from mock (CTR) and NRG1 transfected COS7 cells and subjected to SDS-PAGE and western blot analysis. Membranes were incubated with anti-NRG1 (sc-348) antibody. Asterisk indicates an unspecific band. (c) NOBEC transiently expressing the NRG1 intracellular domain (ICD), containing (NLS) or lacking (ΔNLS) the nuclear localization sequence were assayed for migration activity; data show that the cytoplasmic protein confers a migratory activity higher than the migratory activity conferred by the nuclear protein. (d) RT4-D6P2T cells transiently expressing the NRG1 intracellular domain (ICD), containing (NLS) or lacking (ΔNLS) the nuclear localization sequence were assayed for migration activity. No statistical difference between cells transfected with the empty vector and cells transfected with the two constructs was observed. Values represent the average of three biological replicates performed as technical triplicates. Values of each replicate are expressed in percentage with respect to the total number of cells that migrated in that experiment (*, p≤0.01; **, p≤0.01; ***, p≤ 0.001).

CHAPTER 1: Scientific Publications

Figure 8. Following lentivirus (LV) infection, RT4-D6P2T successfully express GFP and ecto-ErbB4- FLAG. Confluent cells infected with LV-sGFP only (Panels A-D) or with LV-sGFP and LV-ecto-ErbB4- FLAG (Panels E-H) were stained with anti-GFP (green) and anti-FLAG (red) antibody. Nuclei were stained with Hoechst (blue). Scale bar 100 μm. Panel I- western blot analysis of RT4-D6P2T infected with LV- sGFP only, or with LV-sGFP and LV-ecto-ErbB4-FLAG. Panel J- recombinant ecto-ErbB4-FLAG peptide was purified from RT4-D6P2T conditioned medium using ANTI-FLAG affinity gel and eluted using FLAG peptide. Eluted fractions were analyzed by western blot to identify the positive fractions to be frozen and used for the following experiments.

To obtain pure and highly concentrated recombinant protein to stimulate cells, RT4-D6P2T were transduced with a lentivirus expressing ecto-ErbB4-FLAG (Figure 8). Conditioned medium was collected from RT4-D6P2T-ecto-ErbB4-FLAG cells to purify ecto-ErbB4- FLAG through a chromatographic column (Figure 8J). NOBEC cells were starved 24 h in serum-free medium, then stimulated with the purified recombinant protein for 5, 10, 15, 30 and 60 min. By western blot analysis AKT and ERK 1/2 phosphorylation was analyzed (Figure 9). Data analysis shows that in this glial cell model, contrary to neuron cells, AKT phosphorylation does not change, while ERK phosphorylation changes. In particular, there is a strong increase in ERK2 phosphorylation, and a low increase in ERK1 phosphorylation (Figure 9C).

DISCUSSION

The study of peripheral nerve repair and regeneration is an emerging issue in biomedicine since, although peripheral nerves retain a significant capacity for spontaneous regeneration

CHAPTER 1: Scientific Publications after lesion, regeneration is almost never complete [30] and functional recovery is often unsatisfactory [31]. 1 Innovative strategies to improve peripheral nerve regeneration are expected to arise from the manipulation of the several trophic factor loops playing a role in this process. Among them the neuregulin1/ErbB system is attracting much attention because of its action on SC [3, 4]. The aim of this study was to further explore the role of the neuregulin1/ErbB system and the potential of its manipulation.

Neuregulin1 isoforms and ErbB receptors are differentially expressed in SC and OEC (both primary cultures and cell lines). The NRG1/ErbB system is involved in the development of the central and peripheral nervous system, in which different members of the ErbB receptor family and different NRG1 isoforms control processes opposed to each other, such as proliferation and cell death [32]. NRG1/ErbB signaling plays a fundamental role in SC precursor growth and in interactions between SC (expressing ErbB2-ErbB3 and soluble type I/II NRG1) and axons (mainly expressing transmembrane NRG1-type III). NRG1-type III plays an instructive role on myelination and SC development, determining the ensheathment fate of axons: its reduced expression causes hypo-myelination [33]. Moreover, NRG1-type III absence [34] or the lack of co-receptors ErbB3 [35] or ErbB2 [36, 37] gives rise to mice without or with severely reduced amount of SC precursors.

Figure 9. Ecto-ErbB4 stimulates ERK phosphorylation in NOBEC. Western blot analysis of NOBEC cells stimulated with recombinant soluble ecto- ErbB4-FLAG for 0, 5, 10, 15, 30, 60 min. Western Blot were analyzed with antibodies anti p-AKT and AKT (Panel A) and anti p-ERK 1/2 and ERK (Panel B). Panel C- Bands were analyzed by quantifying the intensity of the pixels per mm2 (ImageJ). The values of the bands corresponding to phosphorylated proteins were normalized to the intensity of the bands corresponding to the total proteins.

. 63

CHAPTER 1: Scientific Publications

In this paper, primary cultures of SC and OEC and immortalized cultures of SC (RT4- D6P2T) and OEC (NOBEC), commonly used in in vitro analysis, were compared and characterized in terms of ErbB receptors, NRG1 isoforms, P75, S100 and GFAP expression.

Data presented here have shown that RT4-D6P2T cells are very similar to primary SC, since they express mRNAs for ErbB2, ErbB3, and the glial genes GFAP, P75, S100, and do not express mRNA for ErbB1 and NRG1 type III. RT4-D6P2T do not, however, show detectable expression of NRG1, while SC express NRG1 type I/II. OEC and NOBEC cell lines express all NRG1 isoforms; ErbB1 and ErbB2 are expressed in both OEC and NOBEC, while ErbB3 is slightly expressed. ErbB4 mRNA is barely expressed in OEC and SC, and its protein is not detectable.

Ecto-ErbB4 in vitro expression increases NOBEC migration. Following the characterization of the four cell populations as well as the immunohistochemical assay to compare OEC and NOBEC, we decided to proceed with the study of NRG1/ErbB system manipulation using the NOBEC line, which expresses the transcripts for different NRG1 isoforms.

Data previously obtained in our laboratory showed that NOBEC have a high migratory activity [27]. We speculated whether the high migration was due to an autocrine loop induced by self-produced NRG1; in fact, it is known that glial cells migrate following stimulation with NRG1 [26]. To negatively interfere with NOBEC migration, we planed to subtract soluble NRG1 using the recombinant soluble extracellular fragment of ErbB3 or ErbB4 (ecto-ErbB). Contrary to what expected, results showed that ecto-ErbB3 and ecto- ErbB4 expression significantly increases cellular migration.

We hypothesized that this migration increase was due to the interaction between the soluble portion of ErbB4 receptor and the NRG1 type III expressed by the NOBEC cell line, through a process of back-signaling that this transmembrane isoform of NRG1 is able to generate. Usually, this NRG1 isoform is expressed by axons in which the back-signaling is mediated by the action of a -secretase that, following ligand-receptor interaction, causes the release of a NRG1 cytoplasmic fragment able to translocate into the nucleus [9], where it can regulate transcription of genes involved in neuronal development [38].

Actually, we demonstrated that NRG1 intracellular domain (NRG1-ICD) stimulates migration. Particularly, we saw that the fragment localized in the cytoplasm, more than the fragment localized in the nucleus, plays an important role in the stimulation of the migratory activity. This action could be due to the interaction between this intracellular fragment and cytoplasmic binding partners, such as the LIM kinase1/LIMK1 [28], that is able to shuttle between the cytoplasm and the nucleus, by regulating gene transcription through interaction with the actin cytoskeleton [39]. Interaction between NRG1-ICD and

CHAPTER 1: Scientific Publications

LIMK1 has been found at the level of synapses, where LIMK1 is involved in the regulation of the reorganization of actin filaments in the neuritic protuberances [28, 40, 41]. 1 Transfection of NRG1-ICD in cells that do not express any isoform of NRG1, like RT4- D6P2T cells line, does not affect migration, thus suggesting that these cells lack elements necessary for the signal transduction mediated by NRG1.

Moreover, we found that stimulation of NOBEC with the soluble portion of ErbB4 increases ERK phosphorylation without affecting AKT phosphorylation, contrary to what happens in neuron cells, in which there is an increase of AKT phosphorylation [29]; these results suggest that the back-signaling mediated by NRG1 type III activates signal transduction pathways which differ according to the cell type.

CONCLUSION

This study shows that stable cell lines and the corresponding primary cultures have many characteristics in common, thus suggesting that cell lines are a good model for in vitro studies. On the other hand, these results show that glial olfactory ensheathing cells and Schwann cells differ for the expression of some proteins, thus suggesting that the choice of the cell model for in vitro studies should be done carefully, after investigating the expression of the genes of interest.

Finally, these results suggest that recombinant ecto-ErbB4 can be used not only to sequester soluble NRG1, but also to activate transmembrane NRG1 through a back signaling pathway that can stimulate cell migration. Thus, ecto-ErbB4, a protein fragment endogenously released by cells expressing the cleavable isoform of the NRG1 receptor ErbB4 [42], turns out to be a potential tool to manipulate in vivo the neuregulin1/ErbB system. Nevertheless, further studies are required to design a strategy for a finely tuned ecto-ErbB4 delivery, to investigate the possibility to promote post-traumatic peripheral nerve regeneration.

ACKNOWLEDGEMENTS. The research leading to this paper has received funding from the European Community’s Seventh Framework Programme (FP7-HEALTH-2011) under grant agreement no. 278612 (BIOHYBRID), from MIUR and from Compagnia di San Paolo (MOVAG).

REFERENCES

1. Britsch, S., The neuregulin-I/ErbB signaling system in development and disease. Adv Anat Embryol Cell Biol, 2007. 190: p. 1-65. 2. Fricker, F.R. and D.L. Bennett, The role of neuregulin-1 in the response to nerve injury. Future Neurol, 2011. 6(6): p. 809-822.

CHAPTER 1: Scientific Publications

3. Gambarotta, G., et al., Neuregulin 1 role in Schwann cell regulation and potential applications to promote peripheral nerve regeneration. Int Rev Neurobiol, 2013. 108: p. 223-56. 4. Taveggia, C., M.L. Feltri, and L. Wrabetz, Signals to promote myelin formation and repair. Nat Rev Neurol, 2010. 6(5): p. 276-87. 5. Yarden, Y. and M.X. Sliwkowski, Untangling the ErbB signalling network. Nat. Rev. Mol. Cell Biol., 2001. 2(2): p. 127-37. 6. Talmage, D.A., Mechanisms of neuregulin action. Novartis Found Symp, 2008. 289: p. 74-84; discussion 84-93. 7. Falls, D.L., Neuregulins: functions, forms, and signaling strategies. Exp Cell Res, 2003. 284(1): p. 14-30. 8. Mei, L. and W.C. Xiong, Neuregulin 1 in neural development, synaptic plasticity and schizophrenia. Nat. Rev. Neurosci., 2008. 9(6): p. 437-52. 9. Bao, J., et al., Back signaling by the Nrg-1 intracellular domain. J Cell Biol, 2003. 161(6): p. 1133-41. 10. Bao, J., et al., Activity-dependent transcription regulation of PSD-95 by neuregulin-1 and Eos. Nat Neurosci, 2004. 7(11): p. 1250-8. 11. Hai, M., et al., eds. Comparative analysis of Schwann cell lines as model systems for myelin gene transcription studies. J Neurosci Res. . Vol. 69. 2002. 497-508. 12. Goodman, M.N., J. Silver, and J.W. Jacobberger, Establishment and neurite outgrowth properties of neonatal and adult rat olfactory bulb glial cell lines. Brain Res, 1993. 619(1-2): p. 199-213. 13. Mautino, B., et al., Bioactive recombinant neuregulin-1, -2, and -3 expressed in Escherichia coli. Protein Expr Purif, 2004. 35(1): p. 25-31. 14. Mosahebi, A., et al., Retroviral labeling of Schwann cells: in vitro characterization and in vivo transplantation to improve peripheral nerve regeneration. Glia, 2001. 34(1): p. 8-17. 15. Tohill, M.P., et al., Green fluorescent protein is a stable morphological marker for schwann cell transplants in bioengineered nerve conduits. Tissue Eng, 2004. 10(9- 10): p. 1359-67. 16. Caddick, J., et al., Phenotypic and functional characteristics of mesenchymal stem cells differentiated along a Schwann cell lineage. Glia, 2006. 54(8): p. 840-9. 17. Chung, R.S., et al., Olfactory ensheathing cells promote neurite sprouting of injured axons in vitro by direct cellular contact and secretion of soluble factors. Cell Mol Life Sci, 2004. 61(10): p. 1238-45. 18. Gambarotta, G., et al., ErbB4 expression in neural progenitor cells (ST14A) is necessary to mediate neuregulin-1beta1-induced migration. J Biol Chem, 2004. 279(47): p. 48808-16. 19. Fregnan, F., et al., Eps8 involvement in neuregulin1-ErbB4 mediated migration in the neuronal progenitor cell line ST14A. Exp Cell Res, 2011. 317(6): p. 757-69. 20. Henderson, C.E., et al., GDNF: a potent survival factor for motoneurons present in peripheral nerve and muscle. Science, 1994. 266(5187): p. 1062-4. 21. Blits, B., et al., Rescue and sprouting of motoneurons following ventral root avulsion and reimplantation combined with intraspinal adeno-associated viral

CHAPTER 1: Scientific Publications

vector-mediated expression of glial cell line-derived neurotrophic factor or brain- derived neurotrophic factor. Exp Neurol, 2004. 189(2): p. 303-16. 1 22. Eggers, R., et al., Neuroregenerative effects of lentiviral vector-mediated GDNF expression in reimplanted ventral roots. Mol Cell Neurosci, 2008. 39(1): p. 105-17. 23. Liu, Y., et al., Combined effect of olfactory ensheathing cell (OEC) transplantation and glial cell line-derived neurotrophic factor (GDNF) intravitreal injection on optic nerve injury in rats. Mol Vis, 2010. 16: p. 2903-10. 24. Pellitteri, R., et al., Olfactory ensheathing cells represent an optimal substrate for hippocampal neurons: an in vitro study. Int J Dev Neurosci, 2009. 27(5): p. 453-8. 25. Ronchi, G., et al., ErbB2 receptor over-expression improves post-traumatic peripheral nerve regeneration in adult mice. PLoS One, 2013. 8(2): p. e56282. 26. Yamauchi, J., et al., ErbB2 directly activates the exchange factor Dock7 to promote Schwann cell migration. J Cell Biol, 2008. 181(2): p. 351-65. 27. Audisio, C., et al., Morphological and biomolecular characterization of the neonatal olfactory bulb ensheathing cell line. J. Neurosci. Methods., 2009. 185(1): p. 89-98. 28. Wang, J.Y., et al., Transmembrane neuregulins interact with LIM kinase 1, a cytoplasmic protein kinase implicated in development of visuospatial cognition. J Biol Chem, 1998. 273(32): p. 20525-34. 29. Canetta, S.E., et al., Type III Nrg1 back signaling enhances functional TRPV1 along sensory axons contributing to basal and inflammatory thermal pain sensation. PLoS One, 2011. 6(9): p. e25108. 30. Muratori, L., et al., Can regenerated nerve fibers return to normal size? A long- term post-traumatic study of the rat median nerve crush injury model. Microsurgery, 2012. 32(5): p. 383-7. 31. Tos, P., et al., Future perspectives in nerve repair and regeneration. Int Rev Neurobiol, 2013. 109: p. 165-92. 32. Breuleux, M., Role of heregulin in human cancer. Cell Mol Life Sci, 2007. 64(18): p. 2358-77. 33. Taveggia, C., et al., Neuregulin-1 type III determines the ensheathment fate of axons. Neuron, 2005. 47(5): p. 681-94. 34. Wolpowitz, D., et al., Cysteine-rich domain isoforms of the neuregulin-1 gene are required for maintenance of peripheral synapses. Neuron, 2000. 25(1): p. 79-91. 35. Riethmacher, D., et al., Severe neuropathies in mice with targeted mutations in the ErbB3 receptor. Nature, 1997. 389(6652): p. 725-30. 36. Morris, J.K., et al., Rescue of the cardiac defect in ErbB2 mutant mice reveals essential roles of ErbB2 in peripheral nervous system development. Neuron, 1999. 23(2): p. 273-83. 37. Woldeyesus, M.T., et al., Peripheral nervous system defects in erbB2 mutants following genetic rescue of heart development. Genes. Dev., 1999. 13(19): p. 2538- 48. 38. Barakat, A., et al., Decreased Neuregulin 1 C-terminal fragment in Brodmann's area 6 of patients with schizophrenia. Schizophr Res, 2010. 124(1-3): p. 200-7.

CHAPTER 1: Scientific Publications

39. Geneste, O., J.W. Copeland, and R. Treisman, LIM kinase and Diaphanous cooperate to regulate serum response factor and actin dynamics. J Cell Biol, 2002. 157(5): p. 831-8. 40. Wang, J.Y., et al., LIM kinase 1 accumulates in presynaptic terminals during synapse maturation. J Comp Neurol, 2000. 416(3): p. 319-34. 41. Aizawa, H., et al., Phosphorylation of cofilin by LIM-kinase is necessary for semaphorin 3A-induced growth cone collapse. Nat Neurosci, 2001. 4(4): p. 367-73. 42. Veikkolainen, V., et al., Function of ERBB4 is determined by alternative splicing. Cell Cycle, 2011. 10(16): p. 2647-57.

CHAPTER 1: Scientific Publications

1 Local delivery of the Neuregulin1 receptor ecto-domain (ecto-ErbB4) has a positive effect on regenerated nerve fiber maturation.

Giovanna Gambarotta1,2, Davide Pascal1, Giulia Ronchi1,3, Michela Morano1, Sara Buskbjerg Jager4, Silvia Moimas5,6, Lorena Zentilin5, Mauro Giacca5,6, Isabelle Perroteau1,2, Pierluigi Tos7, Stefano Geuna1,3 and Stefania Raimondo1,3

1 Department of Clinical and Biological Sciences, Nerve Regeneration Group, University of Torino, Italy 2 Neuroscience Institute of Torino (NIT), Interdepartmental Centre of Advanced Studies in Neuroscience, University of Torino, Italy 3 Neuroscience Institute Cavalieri Ottolenghi (NICO), Torino, Italy 4 The Lundbeck Foundation Research Center MIND and Danish Research Institute of Translational Neuroscience DANDRITE, Aarhus University, Denmark 5 Molecular Medicine Laboratory, International Centre for Genetic Engineering and Biotechnology (ICGEB), Trieste, Italy 6 Department of Medical Sciences, Faculty of Medicine, University of Trieste, Trieste, Italy 7 Reconstructive Microsurgery, Orthopedics Department, Trauma Center Hospital (CTO), University of Torino, Italy published in Gene Teraphy DOI: 10.1038/gt.2015.46

ABSTRACT

The Neuregulin/ErbB system plays an important role in the peripheral nervous system both in normal and pathological conditions. We previously demonstrated that expression of soluble ecto-ErbB4, the released extracellular fragment of the ErbB4 receptor, stimulated glial cell migration in vitro. In this study we examined the possibility to manipulate this system in vivo in order to improve injured peripheral nerve regeneration. Transected rat median nerves of adult female Wistar rats were repaired with a 10-mm-long graft made by muscle-in-vein combined nerve guide previously transduced with either the adeno- associated viral (AAV) vector AAV2-LacZ or AAV2-ecto-ErbB4. Autologous nerve grafts were used as control. Both stereological and functional analysis were performed to assess nerve regeneration. Data show that delivery of soluble ecto-ErbB4 by gene transfer in the muscle-in-vein combined nerve guide has a positive effect on fiber maturation, suggesting that it could represent a potential tool to improve peripheral nerve regeneration.

CHAPTER 1: Scientific Publications

INTRODUCTION

The peripheral nervous system (PNS) is a complex network that enable the cross-talk between peripheral organs and the central nervous system (CNS). Peripheral nerves are present in almost all tissues of the human body to ensure motor sensory and autonomic functions 1. Peripheral nerve injury can lead to a reduction, or even a loss, of motor and sensory function if not properly treated. After nerve injury, axon continuity is interrupted, and the loss of functional connection between motor neurons and skeletal muscle fibers can cause progressive muscle atrophy 2. When nerve substance loss occurs, a bridging strategy is necessary to restore the axonal continuity and the functional neuromuscular connection 3. Different experimental and surgical approaches exist; one of them is the muscle-in-vein combined technique that repairs a nerve gap with an engineered combined conduit, in which a vein segment is enriched with fresh skeletal muscle fibers 4, 5. Experimental studies in animal models demonstrate that muscle-in-vein conduits mimic the nervous environment since Schwann cells rapidly colonize the graft and maintain their proliferative capacity 6, 7. The efficacy of this technique is confirmed by its clinical application in primary crush injuries 8 and digital nerve repair 9-11.

In spite of the spontaneous regeneration potential of nerve fibers and the new surgical approaches used to treat peripheral nerve lesions, clinical results are still unsatisfactory and a complete recovery of nerve function almost never occurs 12, 13.

For this reason, and due to the advances in the study of molecular partners that regulate the nerve regeneration process, surgical procedures have been combined with new biomedical strategies such as the production of new devices and the induction of specific factors to enhance axonal regeneration and to reduce the atrophy of denervated muscle 3. In the field of tissue engineering, gene therapy has interesting perspectives; indeed, lentivirus (LV) and adeno-associated virus (AAV) can efficiently deliver therapeutic genes to muscle, nerve and Schwann cells, thus stimulating regeneration 14-16 and preventing muscular atrophy 17.

Among the different trophic factors that regulate the regeneration process, the Neuregulin1 (NRG1)/ErbB system plays an important role in Schwann cell behavior, both in myelination and in remyelination processes 18-20; for these reasons, the manipulation of this system opens new perspectives to improve post-traumatic nerve regeneration 21-23. Ecto- ErbB4, a protein fragment endogenously released by cells expressing the cleavable isoform of the NRG1 receptor ErbB4 24, can stimulate glial cell migration in vitro by activating a back signaling pathway after binding the transmembrane isoform of NRG1 25. These result suggest that ecto-ErbB4 could be a potential tool to manipulate the NRG1/ErbB system in vivo.

CHAPTER 1: Scientific Publications

In this article, we carried out in vivo experiments to investigate the effects of ecto-ErbB4 fragment expression inside muscle-in-vein combined nerve conduits used to repair a rat 1 median nerve defect.

MATERIALS AND METHODS

Cell culture and protein analysis. Neonatal Olfactory Bulb Ensheathing Cells (NOBEC) line, derived from primary cells dissociated from neonatal rat olfactory bulb and immortalized by retroviral transduction of SV40 large T antigen 42, was kindly provided by Dr. Jacobberger (Comprehensive Cancer Center, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 4106-4944, USA). Cell were grown and transfected with AAV-ecto-ErbB4-FLAG construct as previously described 25. NSC-34 motor neuron-like cell line, kindly provided by Sebastian Rademacher, MSc, and Prof. Dr. Peter Claus (Institute of Neuroanatomy, Hannover Medical School, Germany), is derived by fusion of embryonic mouse spinal cord cells with mouse neuroblastoma. Cells were maintained in a monolayer at 37°C in humidified atmosphere of 5% CO2/air in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen, Waltham, Massachusetts, USA) supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, 1 mM sodium pyruvate, β mM l-glutamine, and 10% heat-inactivated foetal bovine serum (FBS; Invitrogen). For cell differentiation, proliferation medium was changed to differentiation medium containing 1:1 DMEM/F-12 Ham (Sigma, Saint Louis, MO, USA), 1% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin and 1 μM all-trans retinoic acid (Sigma). Medium was changed every two days and experiments were performed on the fifth day according to the protocol described by Maier and colleagues 43. Recombinant ecto-ErbB4 was purified according to the protocol previously described 25, and used to stimulate differentiated NSC-34 cells.

Total protein extraction and Western Blot analysis. Total proteins were extracted by solubilizing cells in boiling Laemmli buffer (2.5% SDS, 0.125M Tris–HCl pH6.8), followed by 3 min at 100°C. Protein concentration was determined by the BCA method, and equal amounts of proteins (denaturated at 100°C in 240 mM 2-mercaptoethanol and 18% glycerol) were loaded onto each lane, separated by SDS-PAGE, transferred to a HybondTM C Extra membrane as previously described 44. Primary antibodies used are: FLAG (#F7425) w.d. 1:10000, from Sigma; p-AKT (#4051), AKT (#9272), p-ERK (#9106), ERK (#9102) w.d. 1:1000, from Cell Signaling (Danvers, MA, USA); secondary antibodies used are: horseradish peroxidase-linked anti-rabbit (#NA934) and anti-mouse (#NA931) w.d. 1:40000, from GE Health (Little Chalfont, Buckinghamshire, UK).

Cloning strategies for ecto-ErbB4-FLAG construct and subcloning into viral vectors. To obtain a construct to express the extracellular domain of ErbB4 (ecto-ErbB4) fused with a FLAG epitope, the ErbB4 extracellular domain - recovered from the pIRES-puro2-ErbB4- JMa-cyt2 construct 45 by EcoRV and BbsI digestion and blunting - was subcloned into the EcoRV site of the pCMV-Tag4a vector (Stratagene Santa Clara, CA, USA). To obtain an

CHAPTER 1: Scientific Publications adenoassociated virus (AAV) to express ecto-ErbB4-FLAG, the insert with the FLAG and the following STOP codon was recovered from pCMV-Tag4c vector by EcoRV and KpnI digestion and blunting and subcloned into the SalI site (previously blunted) of the adeno- associated virus vector pAAV-MCS (Stratagene).

Cloning strategies for ecto-ErbB4 constructs and subcloning into viral vectors. Ecto- ErbB4 sequence was achieved from pCMV-Tag4a-ecto-ErbB4-FLAG vector 25 by PCR reaction, using a forward primer containing an EcoRI site (underlined in the primer sequence) and a reverse primer excluding the FLAG sequence and containing an additional STOP codon (showed in bold in the primer sequence) and a HindIII site (underlined in the primer sequence). Both restriction enzyme sites were already present in the pCMV-Tag4a vector; forward primer: 5’-CCCGGGCTGCAGGAATTCG-γ’; reverse primer 5’-TCGATAAGCTTGATGAGTCAGTCTTCCATTTTCTC-γ’.

To obtain an AAV to express ecto-ErbB4, the amplified sequence was digested with EcoRI and HindIII and subcloned into the corresponding restriction sites of the AAV vector pAAV-MCS (Stratagene) and completely sequenced (BMR Genomics Laboratories, Padova, Italy).

Recombinant AAV vectors. The recombinant AAV vectors used in this study were produced by the AAV Vector Unit (AVU) at ICGEB Trieste (http://www.icgeb.org/avu- core-facility.html), according to the previously described protocol 17. The vectors used in this study express the cDNAs (ecto-ErbB4-FLAG, ecto-ErbB4 and -galactosidase) under the control of the constitutive cytomegalovirus immediate early promoter. All viral stocks used in this study had a titer ≥ 1 × 1012 viral genome particles per ml. The proper expression of transgenes was tested by real-time PCR quantification of the transgene messenger RNAs in transduced tissues.

Surgery. For this study, adult female Wistar rats (5 for each experimental group) weighing approximately 200g were used. All animals were maintained in an acclimatized atmosphere under hygienic conditions, in a temperature- and humidity-controlled room with 12–12 h light/dark cycles, and were fed with standard chow and water ad libitum. Adequate measures were taken to minimize pain and discomfort taking into account human endpoints for animal suffering and distress. All surgeries were performed with the approval of the local Institution Animal Care and Ethics Committee and in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC). Animals were operated under deep anesthesia using Tiletamine and Zolazepam (Zoletil) i.m. (3 mg/kg) with the aid of an operating microscope (Zeiss OPMI 7, Milano, Italy). They were divided into 3 experimental groups: (i) muscle-in-vein (MIV) infected with AAV-Lac-Z (MIV-LacZ), (ii) MIV infected with AAV-ecto-ErbB4 (MIV-ecto-ErbB4) and (iii)

CHAPTER 1: Scientific Publications autograft. A preliminary assay was carried out using AAV-ecto-ErbB4-FLAG, to assess the production of the protein by the infected muscle fibers. 1 For the animals belonging to the groups with AAV injection the protocol previously described was used 17. Briefly, the right pectoralis major muscle, used two weeks later for the conduit, was exposed and injected with either AAV2-LacZ or AAV2-ecto-ErbB4. Three injections of β0 μl (corresponding to 1 x 1011 viral genome particles per rat) were performed. The area of injection in the pectoralis major muscle was marked by two suture stitches in order to allow its identification at the main surgery. Two weeks after the AAV injection, the median nerve of the left forelimb was approached from the axillary region to the elbow with a longitudinal skin approach and was carefully exposed and cut. Transected median nerve was immediately repaired with a 10-mm-long graft made by MIV-combined nerve guide. This conduit was constituted by the epigastric vein filled with pectoral muscle fibers previously transduced with either AAV2-LacZ or AAV2-ecto-ErbB4 7. The graft was sutured using three or four stitches of 9–0 monofilament nylon for each stump. In the autograft group, 1 cm segment of the left median nerve was resected, rotated 180° and then sutured at the proximal and distal nerve stumps.

In order to prevent interferences during the grasping test, the contralateral median nerve (right median nerve) of all the animals was transected at the middle third of the brachium and its proximal stump was sutured in the pectoralis major muscle to avoid spontaneous reinnervation 46. Finally, the skin was sutured and the animals were allowed to recover.

Functional analysis. All rats were subjected to a grasping test for functional evaluation, as previously described 46. The device for the grasping test (BS-GRIP Grip Meter - 2Biological Instruments, Varese, Italy) is constituted by a precision balance connected to a grid the rat can grip. The test is carried out by holding the rat by its tail and putting it close enough to the device to grip it. The rat is allowed to pull on the bar until it loses the grip. The balance records the maximum weight that the animal manages to hold before losing the grip. Animals were tested every three weeks until the sacrifice (week 12); each animal was tested three times and the average value was recorded.

Histology and stereology. A 15-mm-long segment of the repaired median nerve (10mm of graft and 5mm of distal nerve) was removed 12 weeks after surgery. Nerve specimens were fixed by immediate immersion in 2.5% glutaraldehyde in 0.1M PBS pH7.4 for up to 6h at 4°C. Samples were then post-fixed in 2% osmium tetroxide for 2h and carefully dehydrated in passages in ethanol from 30% to 100%. Samples were then cleared in propylene oxide and embedded in Glauerts’ embedding mixture of resins consisting of equal parts of Araldite M and Araldite Harter, HY 964 (Merck, Darmstad, Germany), containing 0.5% of the plasticizer dibutyl phthalate and 1–2% of the accelerator 964, DY 064 (Merck). For high-resolution light microscopy, 2.5µm thick series of semi-thin transverse sections were cut starting from the distal stump of each median nerve specimen,

CHAPTER 1: Scientific Publications using an Ultracut UCT ultramicrotome (Leica Microsystems, Wetzlar, Germany). Sections were stained by toluidine blue (1%) and used for qualitative and quantitative morphological analysis, performed using a DM4000B microscope equipped with a DFC320 digital camera and an IM50 image manager system (Leica Microsystems). On one randomly selected toluidine blue stained semi-thin section, the total cross-sectional area of the whole nerve was measured and 12–16 sampling fields were selected using a systematic random sampling protocol 47-49. In order to avoid the ‘edge effect’, a two dimensional dissector procedure which is based on sampling the “tops” of the fibers was adopted48, 50. Mean fiber density was then calculated by dividing the total number of nerve fibers within the sampling field by its area (N/mm2). Total fiber number (N) was estimated by multiplying the mean fiber density by the total cross-sectional area of the whole nerve cross section. Moreover, both fiber and axon area were measured and the diameter of fiber (D) and axon (d) were calculated. These data were used to calculate myelin thickness [(D−d)/β] and the g-ratio (d/D). Analysis of the sampling scheme was performed by calculating the coefficient of error (CE) that is a measure of the precision of the quantitative estimates.

As regards quantitative estimate on fiber number, the following formula for computing the CE(n) was used (according to Schmitz, 1998 51):

1 CE n)(  Q' where Q' is number of counted fibers in all dissectors.

For size estimate, the coefficient of error of the mean size was estimated (according to Geuna et al., 2001 52) as:

SEM CE z)(  Mean where SEM = standard deviation of the mean.

During the pilot study phase, the sampling scheme was established in order to keep the CE below 0.10, a value that allows to obtain a sufficient estimate precision for neuromorphological studies (Pakkenberg and Gundersen, 1997 53).

Statistical analysis. Quantitative data are presented as mean + SEM or SD. All data were statistically analyzed using t-test or one-way analysis of variance and post-hoc analysis by the Bonferroni’s test (SPSS software). Regression lines were analyzed by Student’s t test using the Prism Software Package (GraphPad).

CHAPTER 1: Scientific Publications

RESULTS

In motor neuron-like cells AKT is phosphorylated following recombinant ecto-ErbB4 1 fragment stimulation. To assess the ability of motor neurons to activate a signal transduction pathway following recombinant ecto-ErbB4 fragment stimulation, AKT and ERK 1/2 protein phosphorylation levels were evaluated in a motor neuron-like cell line (NSC-34) (Figure 1, panels A and B). This cell line expresses only the NRG1 transmembrane isoform, which is known to be able to activate a back signaling pathway after ecto-ErbB4 fragment interaction26. As shown by western blot, ecto-ErbB4 enhances AKT protein phosphorylation twenty minutes after stimulus, while ERK1/2 phosphorylation is not affected.

Ecto-ErbB4 expression does not affect functional recovery after nerve reconstruction. Serotype 2 adeno-associated viral vector (AAV2), characterized by muscle tissue tropism, was used to express ecto-ErbB4 in the right pectoralis major muscle, used two weeks later for the muscle-in-vein (MIV) conduit. To confirm local protein delivery, a preliminary assay was carried out using an AAV vector expressing ecto-ErbB4-FLAG. Western blot analysis showed the production of the recombinant protein by transplanted muscle fibers (Figure 1, panel C). To conﬁrm AAV-ecto-ErbB4 muscle transduction, transgene mRNA expression was assessed by qRT-PCR in a muscle portion withdrawn from the injection area. The expression of the transgene was detectable in all samples (data not shown).

All animals underwent functional recovery assessment by grasping test every three weeks (Figure 2). Rats belonging to the autograft group showed a significant better functional recovery than the other two groups (MIV-LacZ, the mock control, and MIV-ecto-ErbB4), at six and nine weeks post-surgery. The difference in functional recovery from the autograft group is more significant with MIV-LacZ group at six (p ≤ 0.001) and nine (p ≤ 0.01) weeks, than with the MIV-ecto-ErbB4 group (p ≤ 0.05). At the end of the observation time (week-12), the differences among the groups are not statistically significant.

Morphological and morphometrical analyses show that ecto-ErbB4 has a positive effect on nerve maturation.. To determine if structural parameters were influenced by ecto- ErbB4 fragment delivery, we compared semi-thin sections of the regenerated nerves stained with toluidine blue (Figure 3). Morphological and morphometrical analyses were performed 12 weeks after the surgery for all experimental groups. Myelinated axons have a normal post-operatory morphological appearance in the autograft group (Figure 3A), which is the gold standard technique used when substance loss occur and here is used as a positive control of nerve regeneration3. In MIV-LacZ (Figure 3B) and MIV-ecto-ErbB4 (Figure 3C) groups, fibers are more fasciculated and a higher number of small fibers, surrounded by more connective tissue than the autograft group, is observed.

CHAPTER 1: Scientific Publications

Figure 1. Ecto-ErbB4 stimulates AKT phosphorylation in NSC-34 motor neuron-like cells, and is expressed by the infected muscle. Western blot analysis of NSC-34 cells stimulated with purified recombinant ecto-ErbB4 for 10 and 20 min. Western blot was analyzed with anti-p-AKT and AKT (panel A) and anti-p-ERK and ERK (panel B) antibodies. Ecto-ErbB4 expression was assessed by western blot analyzed with antibody anti-FLAG (panel C): NOBEC cells (1), NOBEC cells transfected with AAV-ecto-ErbB4-FLAG (2), muscle infected with AAV-LacZ (3), and muscle infected with AAV-ecto- ErbB4-FLAG (4). Asterisk indicates an unspecific band.

Quantification was carried out in the distal part of the regenerated median nerve. Results are summarized in figure 4. The number of myelinated fibers is not significantly different among the three experimental groups (Figure 4A). The graph for fiber diameter (Figure 4B) shows statistically significant (p ≤ 0.05) higher values for the autograft group if compared with the MIV-LacZ experimental group, but not with the MIV-ecto-ErbB4 group.

The data on axon diameter do not show statistically significant differences among the three experimental groups (Figure 4C).

Concerning the parameters related to myelination, a significantly (p ≤ 0.01) thinner myelin sheath in MIV-LacZ group only was observed, in comparison to the autograft group. No significant differences were detectable between the MIV-ecto-ErbB4 group and the other two experimental groups (Figure 4D). We then plotted g-ratio with axon diameter, a method which allows to determine fiber maturation. Indeed, the slope of the linear regression is an indirect parameter of nerve fiber maturation: the higher the slope, the higher the ratio between the g-ratio and the axon diameter and the lower the fiber maturation. Data show that in the MIV-LacZ group, the linear regression has a higher slope that is significantly (p ≤ 0.0001) different when compared to the autograft (Figure 5A) and the MIV-ecto-ErbB4 (Figure 5B) experimental group. On the contrary, there is no difference between the autograft and the MIV-ecto-ErbB4 group (Figure 5C).

CHAPTER 1: Scientific Publications

Figure 2. Functional analysis shows that animals belonging to the autograft group have better functional recovery than the other groups (MIV-LacZ and MIV-ecto-ErbB4). Functional recovery (grasping test) was assessed every three weeks until week-12. Data are presented as mean + SEM; data were statistically analyzed using one-way analysis of variance (* = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001). The autograft group was used as a positive control of nerve regeneration.

DISCUSSION

After peripheral nerve injury, restoring connections between neurons and muscle is a key factor to obtain good clinical outcome, avoiding deficits in muscle mass and force27. Various molecular partners are involved in the different steps of the nerve regeneration process, and there is evidence of the important role played by the Neuregulin1(NRG1)/ErbB system in myelination and re-myelination processes22, and in regulation of several Schwann cell cellular processes such as survival, development, maturation, differentiation and migration28, 29. We recently demonstrated that the released extracellular fragment of the ErbB4 receptor (ecto-ErbB4) improves glial cell migration, through the activation of a back signaling mediated by the transmembrane isoform of NRG1 25. Because of these encouraging data, we manipulated the NRG1/ErbB system in vivo testing the effect of ecto-ErbB4 on nerve regeneration, through gene transfer inside muscle-in-vein (MIV) combined nerve conduits, a good model for AAV-mediated local molecule delivery. In a previous study we demonstrated the presence of the viral DNA and the expression of the transgene in the muscular tissue transplanted inside the vein 17. Here, we show that also the protein is expressed in the transplanted infected muscle fibers. In this model the protein delivery is only transient, because most skeletal muscle fibers degenerate within one month after surgery30. Results of this in vivo study were twofold. On one side ecto-ErbB4 releasing does not affect functional recovery (Figure 2). On the other side a positive effect on fiber maturation (g-ratio/axon diameter) can be observed when MIV- ecto-ErbB4 group is compared to MIV-LacZ group (Figure 5).

CHAPTER 1: Scientific Publications

Figure 3. Morphological analysis of regenerated nerves shows thicker myelin in ecto-ErbB4 expressing samples. Representative toluidine blue- stained semithin sections of regenerated nerves observed in light microscopy at 12 weeks after surgery: Autograft (A); MIV-LacZ (B); MIV-ecto- ErbB4 (C).

The discrepancy between functional and morphological results after nerve lesion is a common and recurrent observation31, 32. The absence of positive effects observed in the functional test can be caused by two variables. Measurements could be affected by the high intra/inter biological variability of the animals belonging to the experimental groups, and by the high functional outcome given by the MIV surgical technique, that does not allow to appreciate small differences between the experimental groups.

CHAPTER 1: Scientific Publications

Figure 4. Stereological analysis shows that fiber diameter and myelin thickness of ecto-ErbB4 expressing samples are not statistically different from autograft samples. The number of myelinated fibers (A), the fiber diameter (B), the myelin thickness (C) and the axon diameter (D) were evaluated 12 weeks after the surgery. All data are presented as average + SD; data were statistically analyzed using one-way analysis of variance and post-hoc analysis by the Bonferroni test (* = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001).

On the other side, we hypothesize that positive effects on myelination, observed in morphological assays, could be due to the ability of the ecto-ErbB4 fragment to interfere with several signaling pathways mediated by both axons and Schwann cells, at different steps of the regeneration process. Indeed, ecto-ErbB4 can act on both transmembrane and soluble NRG1 isoforms.

The first target of ecto-ErbB4 is the transmembrane NRG1 expressed on axons surface. It has been shown that ligand interaction with transmembrane NRG1 expressed by axons mediates a back signaling through the production of an intracellular fragment that migrates from the cytoplasm to the nucleus33, 34. In the first weeks after a nerve injury, the myelin sheet does not coat the regrowing axon surface and the ecto-ErbB4 fragment can interact with axonal transmembrane NRG1. In this early phase axon back-signaling stimulation could enhance neuron survival. Indeed, we show that ecto-ErbB4 stimulation can activate 79

CHAPTER 1: Scientific Publications

AKT phosphorylation in a motor neuron-like cell line expressing transmembrane NRG1. This activated signaling pathway can stimulate neuron survival by the inactivation of several regulators of the apoptotic process26, 34.

In the following regenerative steps, axonal transmembrane NRG1 is important to determine axon ensheathment, stimulating Schwann cell to myelinate through ErbB2/ErbB3 heterodimer interaction20. The lack of positive functional effects could be due not only to the variables discussed above, but also to the continuous interaction between the released fragment and the axonal NRG1. In this case, ecto-ErbB4 could act as a dominant negative, reducing the interaction between ErbB2-ErbB3 expressed by Schwann cell and transmembrane NRG1 expressed by axon, limiting the improvement of the myelination process. Inducible gene transfer might timely regulate fragment release, overcoming this problem35, 36.

The second target of ecto-ErbB4 is the soluble NRG1 released by the grafted muscle 37 and by Schwann cells38. NRG1 isoforms can be α or β, and it has been shown that α isoforms are less active than β39. Thanks to the affinity for the soluble NRG1 isoforms 25, ecto- ErbB4 fragment could sequester the α isoform released by the grafted muscle 37. If this isoform competes with the NRG1 β isoform released by Schwann cells for binding ErbB2/ErbB3 receptors, the titration of this isoform by the ecto-ErbB4 fragment could elicit a positive effect.

Nevertheless, the action of ecto-ErbB4 as dominant negative, binding and sequestering the soluble NRG1 isoforms, could reduce the ligand concentration resulting in a double-side effect. Indeed, it has been demonstrated that variations of soluble NRG1 concentration affect the myelination process: low ligand concentrations have a myelination-promoting effect by PI3-kinase pathway activation, high concentration levels inhibit the process by MAP-kinase pathway activation40. By reducing the soluble NRG1 concentration, ecto- ErbB4 could elicit a positive effect on myelination.

The regenerative process is a very complex and stepped process. The stimulation of a precise signaling pathway in the correct regenerative step can improve nerve regeneration; on the contrary, an incorrect activation at the wrong regenerative step can affect negatively nerve regeneration41.

We can conclude that the manipulation of the NRG1/ErbB system by gene therapy could be a promising tool to improve peripheral nerve regeneration. However, to obtain stronger positive effects, a more finely tuned release of the ecto-ErbB4 fragment might be required35, 36.

AKNOWLEDGMENTS. This work was supported by grants from the European Community's Seventh Framework Programme (FP7-HEALTH-2011) under grant agreement n°278612 (BIOHYBRID) and from Compagnia di San Paolo (MOVAG).

CHAPTER 1: Scientific Publications

Figure 5. Scatter plots show that ecto-ErbB4 expressing samples are not statistically different from the autograft samples. The graphs display g-ratio of individual fiber in relation to axon diameter (obtained from more than 250 myelinated axons per group). All the experimental groups were compared: autograft with MIV-LacZ (A); MIV-ecto-ErbB4 with MIV-LacZ (B); autograft with MIV-ecto-ErbB4 (C). (* = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001).

CHAPTER 1: Scientific Publications

CONFLICT OF INTEREST. The authors declare that no conflict of interest exist.

REFERENCES

1. Geuna S, Raimondo S, Ronchi G, Di Scipio F, Tos P, Czaja K et al. Chapter 3: Histology of the peripheral nerve and changes occurring during nerve regeneration. Int. Rev. Neurobiol. 2009; 87: 27-46. 2. Tamaki H, Tomori K, Yotani K, Ogita F, Sugawara K, Kirimto H et al. Electrical stimulation of denervated rat skeletal muscle retards trabecular bone loss in early stages of disuse musculoskeletal atrophy. J Musculoskelet Neuronal Interact 2014; 14(2): 220-8. 3. Deumens R, Bozkurt A, Meek MF, Marcus MA, Joosten EA, Weis J et al. Repairing injured peripheral nerves: Bridging the gap. Prog Neurobiol 2010; 92(3): 245-76. 4. Brunelli GA, Battiston B, Vigasio A, Brunelli G, Marocolo D. Bridging nerve defects with combined skeletal muscle and vein conduits. Microsurgery 1993; 14(4): 247-51. 5. Battiston B, Tos P, Cushway TR, Geuna S. Nerve repair by means of vein filled with muscle grafts I. Clinical results. Microsurgery 2000; 20(1): 32-6. 6. Fornaro M, Tos P, Geuna S, Giacobini-Robecchi MG, Battiston B. Confocal imaging of Schwann-cell migration along muscle-vein combined grafts used to bridge nerve defects in the rat. Microsurgery 2001; 21(4): 153-5. 7. Raimondo S, Nicolino S, Tos P, Battiston B, Giacobini-Robecchi MG, Perroteau I et al. Schwann cell behavior after nerve repair by means of tissue-engineered muscle-vein combined guides. J. Comp. Neurol. 2005; 489(2): 249-59. 8. Tos P, Battiston B, Ciclamini D, Geuna S, Artiaco S. Primary repair of crush nerve injuries by means of biological tubulization with muscle-vein-combined grafts. Microsurgery 2012; 32(5): 358-63. 9. Battiston B, Geuna S, Ferrero M, Tos P. Nerve repair by means of tubulization: literature review and personal clinical experience comparing biological and synthetic conduits for sensory nerve repair. Microsurgery 2005; 25(4): 258-67. 10. Marcoccio I, Vigasio A. Muscle-in-vein nerve guide for secondary reconstruction in digital nerve lesions. J Hand Surg Am 2010; 35(9): 1418-26. 11. Manoli T, Schulz L, Stahl S, Jaminet P, Schaller HE. Evaluation of sensory recovery after reconstruction of digital nerves of the hand using muscle-in-vein conduits in comparison to nerve suture or nerve autografting. Microsurgery 2014; 34(8): 608-15. 12. Gordon T, Sulaiman O, Boyd JG. Experimental strategies to promote functional recovery after peripheral nerve injuries. J Peripher Nerv Syst 2003; 8(4): 236-50. 13. Shieh SJ, Lee JW, Chiu HY. Long-term functional results of primary reconstruction of severe forearm injuries. J Plast Reconstr Aesthet Surg 2007; 60(4): 339-48. 14. Allodi I, Mecollari V, Gonzalez-Perez F, Eggers R, Hoyng S, Verhaagen J et al. Schwann cells transduced with a lentiviral vector encoding Fgf-2 promote motor neuron regeneration following sciatic nerve injury. Glia 2014; 62 (10): 1736-46.

CHAPTER 1: Scientific Publications

15. Arsic N, Zacchigna S, Zentilin L, Ramirez-Correa G, Pattarini L, Salvi A et al. Vascular endothelial growth factor stimulates skeletal muscle regeneration in vivo. 1 Mol Ther 2004; 10(5): 844-54. 16. Homs J, Ariza L, Pages G, Udina E, Navarro X, Chillon M et al. Schwann cell targeting via intrasciatic injection of AAV8 as gene therapy strategy for peripheral nerve regeneration. Gene Ther 2011; 18(6): 622-30. 17. Moimas S, Novati F, Ronchi G, Zacchigna S, Fregnan F, Zentilin L et al. Effect of vascular endothelial growth factor gene therapy on post-traumatic peripheral nerve regeneration and denervation-related muscle atrophy. Gene Ther 2013; 20(10): 1014-21. 18. Fricker FR, Bennett DL. The role of neuregulin-1 in the response to nerve injury. Future Neurol 2011; 6(6): 809-822. 19. Fricker FR, Lago N, Balarajah S, Tsantoulas C, Tanna S, Zhu N et al. Axonally derived neuregulin-1 is required for remyelination and regeneration after nerve injury in adulthood. J Neurosci 2011; 31(9): 3225-33. 20. Taveggia C, Zanazzi G, Petrylak A, Yano H, Rosenbluth J, Einheber S et al. Neuregulin-1 type III determines the ensheathment fate of axons. Neuron 2005; 47(5): 681-94. 21. Taveggia C, Feltri ML, Wrabetz L. Signals to promote myelin formation and repair. Nat Rev Neurol 2010; 6(5): 276-87. 22. Ronchi G, Gambarotta G, Di Scipio F, Salamone P, Sprio AE, Cavallo F et al. ErbB2 receptor over-expression improves post-traumatic peripheral nerve regeneration in adult mice. PLoS One 2013; 8(2): e56282. 23. Gambarotta G, Fregnan F, Gnavi S, Perroteau I. Neuregulin 1 role in Schwann cell regulation and potential applications to promote peripheral nerve regeneration. Int Rev Neurobiol 2013; 108: 223-56. 24. Veikkolainen V, Vaparanta K, Halkilahti K, Iljin K, Sundvall M, Elenius K. Function of ERBB4 is determined by alternative splicing. Cell Cycle 2011; 10(16): 2647-57. 25. Pascal D, Giovannelli A, Gnavi S, Hoyng SA, de Winter F, Morano M et al. Characterization of Glial Cell Models and In Vitro Manipulation of the Neuregulin1/ErbB System. Biomed Res Int 2014; 2014: 310215. 26. Canetta SE, Luca E, Pertot E, Role LW, Talmage DA. Type III Nrg1 back signaling enhances functional TRPV1 along sensory axons contributing to basal and inflammatory thermal pain sensation. PLoS One 2011; 6(9): e25108. 27. Aydin MA, Mackinnon SE, Gu XM, Kobayashi J, Kuzon WM, Jr. Force deficits in skeletal muscle after delayed reinnervation. Plast Reconstr Surg 2004; 113(6): 1712-8. 28. Maki Y. Specificity in peripheral nerve regeneration: a discussion of the issues and the research. J Orthop Sci 2002; 7(5): 594-600. 29. Garratt AN, Britsch S, Birchmeier C. Neuregulin, a factor with many functions in the life of a schwann cell. Bioessays 2000; 22(11): 987-96.

CHAPTER 1: Scientific Publications

30. Tos P, Battiston B, Nicolino S, Raimondo S, Fornaro M, Lee JM et al. Comparison of fresh and predegenerated muscle-vein-combined guides for the repair of rat median nerve. Microsurgery 2007; 27(1): 48-55. 31. Munro CA, Szalai JP, Mackinnon SE, Midha R. Lack of association between outcome measures of nerve regeneration. Muscle Nerve 1998; 21(8): 1095-7. 32. Wolthers M, Moldovan M, Binderup T, Schmalbruch H, Krarup C. Comparative electrophysiological, functional, and histological studies of nerve lesions in rats. Microsurgery 2005; 25(6): 508-19. 33. Bao J, Lin H, Ouyang Y, Lei D, Osman A, Kim TW et al. Activity-dependent transcription regulation of PSD-95 by neuregulin-1 and Eos. Nat Neurosci 2004; 7(11): 1250-8. 34. Bao J, Wolpowitz D, Role LW, Talmage DA. Back signaling by the Nrg-1 intracellular domain. J Cell Biol 2003; 161(6): 1133-41. 35. Hoyng SA, Gnavi S, de Winter F, Eggers R, Ozawa T, Zaldumbide A et al. Developing a potentially immunologically inert tetracycline-regulatable viral vector for gene therapy in the peripheral nerve. Gene Ther 2014; 21(6): 549-57. 36. Shakhbazau A, Mohanty C, Shcharbin D, Bryszewska M, Caminade AM, Majoral JP et al. Doxycycline-regulated GDNF expression promotes axonal regeneration and functional recovery in transected peripheral nerve. J Control Release 2013; 172(3): 841-51. 37. Nicolino S, Raimondo S, Tos P, Battiston B, Fornaro M, Geuna S et al. Expression of alpha2a-2b neuregulin-1 is associated with early peripheral nerve repair along muscle-enriched tubes. Neuroreport 2003; 14(11): 1541-5. 38. Stassart RM, Fledrich R, Velanac V, Brinkmann BG, Schwab MH, Meijer D et al. A role for Schwann cell-derived neuregulin-1 in remyelination. Nat Neurosci 2013; 16(1): 48-54. 39. Eckert JM, Byer SJ, Clodfelder-Miller BJ, Carroll SL. Neuregulin-1 beta and neuregulin-1 alpha differentially affect the migration and invasion of malignant peripheral nerve sheath tumor cells. Glia 2009; 57(14): 1501-20. 40. Syed N, Kim HA. Soluble Neuregulin and Schwann Cell Myelination: a Therapeutic Potential for Improving Remyelination of Adult Axons. Mol Cell Pharmacol 2010; 2(4): 161-167. 41. Hoyng SA, De Winter F, Gnavi S, de Boer R, Boon LI, Korvers LM et al. A comparative morphological, electrophysiological and functional analysis of axon regeneration through peripheral nerve autografts genetically modified to overexpress BDNF, CNTF, GDNF, NGF, NT3 or VEGF. Exp Neurol 2014; 261C: 578-593. 42. Goodman MN, Silver J, Jacobberger JW. Establishment and neurite outgrowth properties of neonatal and adult rat olfactory bulb glial cell lines. Brain Res 1993; 619(1-2): 199-213. 43. Maier O, Bohm J, Dahm M, Bruck S, Beyer C, Johann S. Differentiated NSC-34 motoneuron-like cells as experimental model for cholinergic neurodegeneration. Neurochem Int 2013; 62(8): 1029-38.

CHAPTER 1: Scientific Publications

44. Fregnan F, Petrov V, Garzotto D, De Marchis S, Offenhauser N, Grosso E et al. Eps8 involvement in neuregulin1-ErbB4 mediated migration in the neuronal 1 progenitor cell line ST14A. Exp Cell Res 2011; 317(6): 757-69. 45. Gambarotta G, Garzotto D, Destro E, Mautino B, Giampietro C, Cutrupi S et al. ErbB4 expression in neural progenitor cells (ST14A) is necessary to mediate neuregulin-1beta1-induced migration. J Biol Chem 2004; 279(47): 48808-16. 46. Papalia I, Tos P, Stagno d'Alcontres F, Battiston B, Geuna S. On the use of the grasping test in the rat median nerve model: a re-appraisal of its efficacy for quantitative assessment of motor function recovery. J Neurosci Methods 2003; 127(1): 43-7. 47. Larsen JO. Stereology of nerve cross sections. J. Neurosci. Methods 1998; 85(1): 107-18. 48. Geuna S, Tos P, Battiston B, Guglielmone R. Verification of the two-dimensional disector, a method for the unbiased estimation of density and number of myelinated nerve fibers in peripheral nerves. Ann. Anat. 2000; 182(1): 23-34. 49. Piskin A, Kaplan S, Aktas A, Ayyildiz M, Raimondo S, Alic T et al. Platelet gel does not improve peripheral nerve regeneration: an electrophysiological, stereological, and electron microscopic study. Microsurgery 2009; 29(2): 144-53. 50. Schmitz C. Variation of fractionator estimates and its prediction. Anat. Embryol. (Berl) 1998; 198(5): 371-97. 51. Schmitz C. Variation of fractionator estimates and its prediction. Anat Embryol (Berl) 1998; 198(5): 371-97. 52. Geuna S, Tos P, Guglielmone R, Battiston B, Giacobini-Robecchi MG. Methodological issues in size estimation of myelinated nerve fibers in peripheral nerves. Anat Embryol (Berl) 2001; 204(1): 1-10. 53. Pakkenberg B, Gundersen HJ. Neocortical neuron number in humans: effect of sex and age. J Comp Neurol 1997; 384(2): 312-20.

CHAPTER 1: Scientific Publications

Nanotechnology versus stem cell engineering: in vitro 1 comparison of neurite inductive potentials

Michela Morano1*, Sandra Wrobel2,3*, Federica Fregnan1, Ofra Ziv-Polat4, Abraham Shahar4 Andreas Ratzka2, Claudia Grothe2,3, Stefano Geuna1 and Kirsten haastert-Talini2

1 Department of clinical and Biological sciences, Università Degli studi di Torino, Orbassano, Italy 2 Institute of Neuroanatomy, Hannover Medical School, Hannover, Germany 3 Center for systems Neuroscience (ZsN), Hannover, Germany 4 NVR research ltd, Ness-Ziona, Israel

*These authors contributed equally to this work and share first authorship

Published in International Journal of Nanomedicine DOI: https://doi.org/10.2147/IJN.S71951

ABSTRACT

Purpose: Innovative nerve conduits for peripheral nerve reconstruction are needed in order to specifically support peripheral nerve regeneration (PNR) whenever nerve autotransplantation is not an option. Specific support of PNR could be achieved by neurotrophic factor delivery within the nerve conduits via nanotechnology or stem cell engineering and transplantation. Methods: Here, we comparatively investigated the bioactivity of selected neurotrophic factors conjugated to iron oxide nanoparticles (np- NTF) and of bone marrow-derived stem cells genetically engineered to overexpress those neurotrophic factors (NTF-BMSC). The neurite outgrowth inductive activity was monitored in culture systems of adult and neonatal rat sensory dorsal root ganglion neurons as well as in the cell line from rat pheochromocytoma (PC-12) cell sympathetic culture model system. Results: We demonstrate that np-NTF reliably support numeric neurite outgrowth in all utilized culture models. In some aspects, especially with regard to their long-term bioactivity, np-NTF are even superior to free NTF. Engineered NTF-BMSC proved to be less effective in induction of sensory neurite outgrowth but demonstrated an increased bioactivity in the PC-12 cell culture system. In contrast, primary non transfected BMSC were as effective as np-NTF in sensory neurite induction and demonstrated an impairment of neuronal differentiation in the PC-12 cell system. Conclusion: Our results evidence that nanotechnology as used in our setup is superior over stem cell engineering when it comes to in vitro models for PNR. Furthermore, np-NTF can easily be suspended in regenerative hydrogel matrix and could be delivered that way to nerve conduits for future in vivo studies and medical application. Keywords: iron oxide nanoparticles, conjugated neurotrophic factors, bone marrow-derived mesenchymal stem cells, genetic cell engineering, neurite outgrowth.

CHAPTER 1: Scientific Publications

INTRODUCTION

Tissue engineering of peripheral nerves is an active field in research and development. Tissue-engineered nerves could become a valuable alternative to autologous nerve grafts, which are used as the gold standard for peripheral nerve reconstruction surgeries.1 This type of surgery is indicated whenever complete transection injuries of a peripheral nerve cannot be repaired by tension-free end-to-end coaptation of the severed nerve ends2. Tissue engineering of peripheral nerves in order to bridge nerve defects and provide an optimized regenerative milieu is a complex effort that can only be achieved in a multidisciplinary setting.3 One important task is the delivery of regeneration-promoting molecules such as neurotrophic factors (NTF) into the nerve defect or, more specifically, their application together with the artificial nerve graft. The main drawback in NTF application is the short half-life time, causing minimal efficacy of single NTF application at the time of reconstructive surgery or systemic application. Different ways to ensure extended availability of added NTF at the site of nerve reconstruction have been attempted in recent years, including gene therapy via transplanted Schwann cells4 or nanotechnology approaches.5 Ex vivo gene therapy can be used to genetically induce the overexpression of selected NTF in cells that are later transplanted as part of tissue-engineered nerve grafts.6 The usefulness of this approach has been proven already for transplanted Schwann cells overexpressing different isoforms of fibroblast growth factor-2 (FGF-2) in the rat sciatic nerve model.7,8 Schwann cells are crucially involved in successful peripheral nerve regeneration (PNR), but they are not easy to harvest and propagate for cell transplantation strategies. Therefore, mesenchymal stem cells (MSCs) have been addressed as an easy to access and potentially unlimited cell source for tissue-engineered nerve grafts.9 In this study, we performed stem cell engineering by nonviral genetic modification of bone marrow-derived mesenchymal stem cells (BMSC), resulting in overexpression of selected NTF. Another option to ensure availability of NTF within an artifical nerve graft is the conjugation of the proteins to nanoparticles and their delivery within a hydrogel matrix for axonal regeneration. This strategy was also evaluated in the current work as an alternative strategy to cell-based delivery of NTF. Three well-defined NTF have been analyzed in the presented study. Nerve growth factor (NGF) is known as the neurotrophin with the strongest effect on sensory neurite outgrowth with regard to both axonal elongation and sprouting.10 The second neurotrophin analyzed was gliaderived neurotrophic factor (GDNF), which exerts the most prominent effects on regenerating motor neurons.11 The growth factor FGF-2, in particular its low molecular weight isoform FGF-218kDa, has demonstrated significant support on axonal elongation along with reduced sprouting.12 For the study presented, all three NTF named were conjugated to iron oxide nanoparticles while overexpression of exclusively FGF-218kDa or GDNF was induced in BMSC. Sensory neurite outgrowth assays as well as the cell line from rat pheochromocytoma cells (PC-12) neuronal differentiation assay were utilized to

CHAPTER 1: Scientific Publications quantitatively analyze the regeneration-promoting potential of NTF-overexpressing BMSC or nanoparticle-conjugated NTF in comparison with free recombinant NTF proteins. 1 MATERIALS AND METHODS

Synthesis of np-NTF. Dextran-coated iron oxide magnetic nanoparticles of, on average, 10 nm dry diameters were prepared as described previously.13 The proteins NGF, GDNF, and FGF-218kDa (PeproTech Asia, Rehovot, Israel) were covalently conjugated to the dextran coating layer through the divinyl sulfone binding reagent, as described previously.5,14,15 Utilizing the appropriate enzyme-linked immunosorbent assay kits ( PeproTech Asia, Rehovot, Israel, and Boster Immunoleader, Pleasanton, CA, USA), the concentration of each protein conjugated to the nanoparticles was determined. Stock aqueous nanoparticle suspensions, with a final concentration of 10 ng/µL of each NTF (np- NTF), were used for the in vitro evaluations.

In vitro techniques. The harvest of material for tissue and cell culture was approved by the local animal protection committees in Torino, Piemonte, Italy, and Hannover, Lower- Saxony, Germany. Figure 1 summarizes details about the different culture systems used in this study.

Adult dorsal root ganglion explant cultures. The spinal column of adult Hannover Wistar rats (approximately 250 g body weight, Harlan, Verona, Italy) was removed after euthanasia. To reach the spinal cord, the vertebral bodies were cut off and removed. Using fine scissors, the vertebral canal was opened through a double cut on both sides of the vertebral bodies. On both sides of the spinal cord, dorsal root ganglia (DRG) were reached along the posterior roots and extracted from the intervertebral foramina. Harvested DRG were maintained in Leibovitz’s medium (Invitrogen, Monza, Italy) for 1 hour under sterile conditions. The connective tissue capsules were then mechanically reduced and the explants were half cut. The DRG explants were cultured on matrigel-coated (BD Biosciences, Bedford, MA, USA) glass coverslips in serum-free medium (SFM16): 50% (v/v) F12 nutrient mixture, 50% (v/v) Basal Medium Eagle, 2×10-3 M glutamine, 0.5% (v/v) 100× penicillin– streptomycin–neomycin antibiotic mixture (all Invitrogen), 10 mg/mL bovine albumin serum, 1×10-4 M putrescine, 0.1 mg/mL transferrin, 3×10-8 selenium, 0.005 mg/mL insulin, 3.8×10-5 M vitamin C, 7.5 mg/mL glucose (all Sigma, Milan, Italy) in culture-grade water (Figure 1A). To evaluate the stability of NTF bioactivity in vitro, free recombinant GDNF (Peprotech, Hamburg, Germany) and np- GDNF were selected, and adult DRG explant cultures were either incubated with fresh or preincubated pure SFM or supplemented with either GDNF type. Preincubated SFM formulations were prepared by incubating the medium for 1 week or 2 weeks at 37°C and

5% CO2. Afterward, this medium was used for DRG explant cultures.

Preparation of NVR-gel for neurite outgrowth assay. NVR-Gel17 was used as a matrix for seeding drops of dissociated neonatal rat DRG (Figure 1B). NVR-Gel is composed of high 89

CHAPTER 1: Scientific Publications molecular weight hyaluronic acid (3×106 Da, BTG Polymers, Kiryat Malachi, Israel) and laminin (Sigma, Rehovot, Israel). For cell cultivation, hyaluronic acid of 1% was further diluted 1:1 either with N2 cell culture medium (see Figure 1B) or PC12 differentation medium (see Figure 1C) to a final concentration of 0.5%. Laminin was then added (10 ng/mL) to complete the NVR-Gel composition. For bioactivity assays, NVR-Gel was either left pure or enriched with 50 ng/mL of single free recombinant NTF (NGF, Invitrogen, Darmstadt, Germany; FGF-2 and GDNF, Peprotech, Hamburg Germany) or single np-NTF or a pool of either the free NTF or np-NTF (50 ng/mL each). In the alternative setting, nontransfected BMSC or BMSC overexpressing selected NTF (see next section) were seeded into the NVR-Gel (Figure 1B).

Harvest and culture of rat BMSC. Isolation and culturing of rat BMSC from the femur bone marrow of adult Hannover Wistar rats were performed as described.18 Therefore, BMSC were cultured in noncoated cell culture flasks in Minimum Essential Medium Alpha (Gibco, Darmstadt, Germany) supplemented with 15% fetal calf serum (FCS) (PAA Laboratories GmbH, Cölbe, Germany) and 1% penicillin/streptomycin (pen/strep, Gibco). Propagated cell cultures were processed to genetic modification or left nontransfected and maintained in Minimum Essential Medium Alpha supplemented with 10% FCS and 1% pen/strep (BMSC culture medium). The nonviral plasmids for genetic modification of BMSC were derived from the pCAGGS-empty vector.19 The plasmids encoding for enhanced green fluorescent protein (EGFP, pCAGGS-EGFP-Flag) and FGF-218kDa (pCAGGS-FGF-218kDa-Flag, NCBI GenBank accession NM_019305.2, 533-994 bp) have been previously described.20,21 The pCAGGS-GDNF-Flag plasmid was constructed by polymerase chain reaction-based cloning of the rat GDNF coding sequence (NCBI GenBank accession NM_019139.1, 50-682 bp) using primers introducing MfeI and XbaI cloning sites and removing the stop codon. The in-frame cloned C-terminal 3× Flag tag of the pCAGGS-Flag vector backbone enabled convenient detection of all three proteins (EGFP-Flag, FGF-218kDa-Flag, GDNF-Flag) with the same antibody (anti-Flag M2, F1804 Sigma-Aldrich, Seelze, Germany). Internal reference numbers for the plasmids are given in parentheses for pCAGGS-empty (R399, used for control transfections), pCAGGS- EGFP-FLAG (R412), pCAGGS-FGF-218kDa-FLAG (R417), and pCAGGS-GDNF-FLAG (R415). For the purpose of cell transfection, plasmid deoxyribonucleic acid (DNA) has been purified using the QIAfilter Plasmid Maxi Kit (Qiagen, Hilden, Germany). For the nonviral transfection, 2.5×106 viable (trypan blue negative) BMSC were suspended in 90 µL basic transfection solution (Human MSC Kit, LONZA, Cologne, Germany) and mixed with 5 µg of the selected plasmid DNA in AMAXA specific cuvettes using the program A-33 (AMAXA II device, LONZA). The reaction was stopped by adding 900 µL RPMI 1640 medium (Gibco) supplemented with 10% FCS, and the cell suspension was then transferred into new 15 mL falcon tubes. After centrifugation (5 minutes, 235× g), cell pellets were prepared for seeding into 0.5% NVR-Gel (based on either N2 medium or on PC-12 differentiation medium, see Figure 1B and 1C). The transfection efficiency, as

CHAPTER 1: Scientific Publications exemplarily determined for EGFP-Flag-BMSC in immunofluorescence microscopy, reached about 60% (Figure S1 in “Supplementary materials”). We assume the same 1 transfection rate also for the other transfections performed.

Drop cultures of dissociated neonatal Drg (sensory neurite outgrowth assay). Neonatal DRGs were harvested from P1–P3 Hannover Wistar or Sprague Dawley rat pups (breeding pairs obtained from Janvier, Le Gernest-Saint Isle, France). Briefly, after decapitation, the skin on the back was removed and the vertebral canal was longitudinally opened from the neck to the tail. The spinal cord was taken off and the DRGs were collected from intervertebral foramina at both sides of the vertebral canal. DRGs were kept for up to 1 hour in Hank’s balanced salt solution (w/o Mg and Ca, PAA Laboratories GmbH) prior to enzymatic dissociation at γ7°C for β0 minutes in Hank’s balanced salt solution added with 0.125% trypsin (Gibco) and 0.05% DNase (Roche, Mannheim, Germany). Then, an additional 0.0075% collagenase type IV (0.1%, PAA Laboratories) was added and the dissociation continued for another 20 minutes. DRGs were then mechanically dissociated using a fire-polished glass pipette before dissociation was stopped by adding 2.5 mL N2 cell culture medium (Dulbecco’s Modified Eagle’s Medium [DMEM] F-12, 1% pen/strep, 0.25% bovine serum albumin, 2 mM L-glutamine, 1 mM sodium pyruvate, 1× N2 supplement; all PAA Laboratories) supplemented with 3% fetal bovine serum. Cells were then counted and placed as drop culture into the middle of 0.5% NVR-Gel (containing N2 medium) (Figure 1B).

Neuronal differentiation assay with Pc-12 cells. During propagation of the PC-12 cell line from rat phaeochromocytoma cells of sympathoadrenal origin, a specific proliferation medium was used: DMEM high glucose + 10% horse serum + 5% FCS + 4 µM L- glutamine + 1 µM sodium pyruvate + 100 U/mL pen/strep.22 To analyze the neurite inductive potential of np-NTF, np-NGF or np-FGF-218kDa was mixed into 0.5% NVR-Gel containing PC-12 cell differentation medium: DMEM high glucose + 1% horse serum + 1% FCS + 4 µM L-glutamine + 1 µM sodium pyruvate + 100 U/mL pen/strep. To analyze the neurite inductive potential of nontransfected (nT-BMSC) and FGF-218kDa overexpressing BMSC (FGF-218kDa-BMSC), the respective cells were seeded into 0.5% NVR-Gel containing PC-12 cell differentiation medium (Figure 1C). In order to enable specific monitoring of neuronal differentiation of PC-12 cells, the latter were nonvirally transfected to express EGFP. Therefore, pCAGGS-EGFP-FLAG (R412) was introduced to the cells using the AMAXA nucleofection technique similar to the procedures described (Cell Line Nucleofection Kit, program U-29, LONZA). For the neurite outgrowth assay, EGFP-PC12 cells were mixed with np-NTF or nT-BMSC or FGF-218kDa-BMSC. After 5 days of cultivation, the lengths of neuritic processes of EGFP-PC-12 cells were analyzed with fluorescence microscopy (Olympus IX 70 microscope, Olympus, Hamburg, Germany).

CHAPTER 1: Scientific Publications

Figure 1 Schematic illustration of culture models used in this study. Notes: (A) To evaluate the bioactivity (neurite outgrowth inductive potential) of np-NTF, half-cut adult DRG explants were maintained for 3 days in SFM at 37°c with 5% cO2 and sister cultures were stimulated with three different quantities (10 ng/ml, 50 ng/ml, or 100 ng/ml) of each single np-NTF (np-NGF, np-FgF-218kDa, or np-GDNF) or with a pool of the three np-NTF (final concentration of each np-NTF 30 ng/mL or 50 ng/mL). (B, upper row) Neonatal DRGs were dissociated and seeded as a drop (2,500 neurons in 5 µl medium) into a layer of 800 or 1,000 µl of NVR gel (pure or supplemented with either single free recombinant NTF or np-NTF or a

CHAPTER 1: Scientific Publications

pool of either the free NTF or np-NTF). (B, lower row) In addition, DRG drop cultures were seeded into NVR-gel containing non transfected BMSC or BMSC overexpressing selected NTF at a density of 40×104 cells/1,000 µl of 0.5% NVR-gel. cultures were maintained for 1 day. (C) a density of 30×104 EGFP-PC-12 1 cells was mixed with either np-NTF or 40×104 nT-BMSC or FGF-218kDa-BMSC in a volume of 1 ml 0.5% NVR-gel (containing PC-12 cell differentiation medium) and transferred into a 24-multiwell culture well. cultures were maintained for 5 days. Abbreviations: np, nanoparticle; NTF, neurotrophic factors; DRG, dorsal root ganglion; SFM, serum-free medium; NGF, nerve growth factor; FGF, fibroblast growth factor; GDNF, glia-derived neurotrophic factor; BMSC, bone marrow-derived mesenchymal stromal cells; EGFP, enhanced green fluorescent protein; PC-12 cells, cell line from rat pheochromocytoma cells; PSN, penicillin–streptomycin–neomycin; BSA, bovine serum albumin; DMEM, Dulbecco’s Modified Eagle’s Medium; MEM, Minimum essential Medium; FCS, fetal calf serum; pen/strep, penicillin/streptomycin.

Sodium dodecyl sulfate gel electrophoresis and Western blot analyses. Overexpression of selected NTF by nonvirally transfected BMSC was controlled using Western blot analyses of cell lysates. Therefore, nT-BMSC, FGF-218kDa-BMSC, and GDNF-BMSC, as well as empty vector-transfected BMSC (empty-BMSC), were cultured with BMSC culture medium as listed in Figure 1C (passage 3, 4, and 5). After 24 hours, proteins were extracted from sister culture cells and samples prepared for Western blotting. Additionally, cell culture supernatants (300 µL) were collected 24 hours, 72 hours, and 120 hours after seeding of nT-BMSC or after genetic modification (empty-BMSC, FGF-218kDa-BMSC, or GDNF-BMSC) and concentrated using the speedvac system Con 1000 (LTF Labortechnik, Wasserburg, Germany). A calculated protein concentration of 50 µg was prepared in Laemmli buffer following sodium dodecyl sulfate polyacrylamide gel electrophoresis (15% gels). The separated proteins were electrophoretically transferred to a nitrocellulose membrane Amersham Bioscience, Freiburg, Germany), as described20.

Immunodetection of FGF-218kDa or GDNF was performed in the Enhanced Chemiluminescence Imager System system (ECL, Intas, Science Imaging, Göttingen, Germany) using monoclonal anti-FGF-2 antibody (1:1,000; MerckMillipore, Darmstadt, Germany) or polyclonal anti-GDNF antibody (1:100; Santa Cruz, Heidelberg, Germany) followed by incubation with the respective antimouse immunoglobulin G (IgG) or antirabbit IgG horseradish peroxidase-coupled secondary antibodies (1:4,000; both Amersham Bioscience Europe, Freiburg, Germany). Protein signals were detected in a horseradish peroxidase chemiluminescent reaction (Immobilon, Merck-Millipore, Schwalbach, Germany).

Immunocytochemistry. Transfection efficiency for nonviral genetic modification of BMSC was determined using EGFP-Flag BMSC. Cells were fixed for 30 minutes in 4% paraformaldehyde (PFA, Sigma-Aldrich). Prior to immunostaining, fixation was stopped by washing with phosphate-buffered saline. Adult DRG explant cultures were also fixed for 15 minutes in 4% PFA, while neonatal DRG drop cultures and PC-12 cell cultures in 0.5% NVR-Gel were fixed for 30 minutes in 6% PFA. Primary antibodies used were anti- III-tubulin (monoclonal mouse, 1:400, Sigma-Aldrich), anti-S100 antibody (polyclonal

CHAPTER 1: Scientific Publications goat, 1:1,000, Dako, Glostrup, Denmark), and anti-Flag antibody (1:1,000, Sigma- Aldrich). Secondary antibodies used were antimouse IgG Alexa-Fluor-488 conjugated (for adult DRG explants and BMSC assays, 1:500–1:2,000, Molecular Probes, Eugene, OR, USA) or antimouse IgG Alexa-Fluor-555 conjugated (for neonatal DRG drop cultures and EGFP-Flag-BMSC, 1:500, Invitrogen) and rabbit antigoat IgG Alexa-Fluor-488 (1:500; Invitrogen). All primary antibodies were incubated overnight at 4°C and secondary antibodies for 1 hour at room temperature. In addition, to identify BMSC, the actin filaments were stained for 30 minutes with phalloidin-tritc fluorescence stain (1:500; Sigma-Aldrich). All samples were stained with the nuclear marker 4′,6-diamidino-2- phenylindole (1:1,000, 10 minutes, room temperature; Sigma-Aldrich).

Quantification of sensory neurite outgrowth. Adult DRG explants were scanned horizontally and images were merged in order to obtain three-dimensional pictures. The samples were observed with an LSM 510 confocal laser microscopy system (Zeiss, Jena, Thuringia, Germany), which incorporates two lasers (argon and HeNe) and is equipped with an inverted Axiovert 100 M microscope. All images captured were digitally adjusted for optimal resolution. During photomicrography, entire neonatal DRG drop cultures were taken in six or more overlapping image sections using the 4× objective of an inverted fluorescence microscope (Olympus IX 70, Olympus, Hamburg, Germany). The image sections were then automatically combined using Adobe Photoshop software to obtain a new aligned picture of the entire neuritic network. Images of both adult DRG explants and neonatal DRG drop cultures were processed using image analysis software (ImageJ, National Institutes of Health, Bethesda, MD, USA) to evaluate the number of extending neurites in certain distances from the center of the cultures. For each sample, the perimeter of the DRG explant or the DRG drop was manually drawn (Polygon selection tool) and automatically enlarged (Edit–Selection–Enlarge tool) to create concentric curves at a fixed distance (50–200 µm for adult DRG and 500–1,100 µm for neonatal DRG) between each other. The number of intersections of neuritic processes with the different curves was then counted and represented as mean values. Additionally, the maximum neurite lengths originating from neonatal DRG drop cultures were analyzed for the ten longest neurites.

Statistical analysis. Three and four biological replicates were performed for experiments with adult DRG explants and neonatal DRG drop cultures, respectively. Each culture condition was analyzed for five adult DRG explants or two to three neonatal DRG drop cultures in each biological replicate. PC-12 cell assays were carried out in three biological replicates with two sister cultures per condition in each replicate. Western blot analyses were performed in three biological replicates. Results are reported as a mean ± standard deviation. Intergroup comparisons were made using one-way or two-ways analysis of variance followed by Bonferroni posttest. A probability value lower than 0.05 was considered as statistically significant.

CHAPTER 1: Scientific Publications

RESULTS

Np-NTF demonstrate a prolonged bioactivity in vitro in comparison with free NTF. To 1 evaluate whether the bioactivity of NTF is improved by conjugation of the proteins to nanoparticles in comparison with free recombinant proteins, we cultured adult rat DRG explants over 3 days in the presence of fresh and preincubated medium formulations prepared with 50 ng/mL of free NTF or NTF conjugated to iron oxide nanoparticles (np- NTF) and determined the numbers of extending neurites at certain distances from the DRG explant. GDNF was selected as reference NTF because it has demonstrated high bioactivity on adult DRG explant cultures, although with levels below NGF treatment (see Figure S2A). This made GDNF a candidate NTF to also detect minimal changes in bioactivity over time. As is depicted in Figure 2, incubation of adult DRG explant cultures for 3 days with fresh free GDNF or np-GDNF induced significantly increased numeric neurite outgrowth (Figure 2A) and extension of neuritic processes (Figure 2B) in comparison with control conditions with fresh medium without GDNF supplementation. Representative photomicrographs are given in Figure 2C. Fresh free GDNF additionally demonstrated an increased bioactivity over that of fresh np-GDNF. While the neurite outgrowth was not significantly affected after 1-week preincubation of medium formulations (data not shown), prolonged bioactivity of np-GDNF compared with free GDNF was clearly detectable after a 2-week preincubation period. After this period, free GDNF no longer increased numeric neurite outgrowth when compared with control conditions, and we detected a significant reduction of bioactivity by 65% of the bioactivity of fresh free GDNF. In contrast, np-GDNF revealed its bioactivity also after a 2-week preincubation period, as demonstrated by the maintained neurite outgrowth inductive activity, which was then still significantly increased over control conditions (Figure 2A and B). Additionally, we demonstrated that at a certain dosage, all np-NTF tested (np-NGF, np- FGF-218kDa, and np-GDNF) significantly increased adult sensory neurite outgrowth over the values seen in control cultures (cultured in SFM medium alone). All np-NTF further demonstrated a dose dependency of their neurite inductive effects. The respective data are presented as supplementary material, including Figure S2. np-NTF exert the same bioactivity as free recombinant proteins when incorporated into a regenerative hydrogel matrix. After we could demonstrate that np-NTF exert a neurite inductive effect on adult DRG explants, we further compared effects of free NTF and np- NTF on neurite outgrowth from neonatal DRG preparations placed into the nerve regenerative matrix NVR-Gel. NVR-Gel is composed of high molecular weight hyaluronic acid and laminin and has been described as regenerative matrix for nervous system repair.17,23 The next experiment was performed in order to determine whether the suspension of np-NTF in the regenerative NVR-Gel matrix still allows them to exert inductive effects on neurite outgrowth, as seen for their suspension in medium. Therefore, the in vitro neurite outgrowth assay was analyzed after placing dissociated neonatal DRG 95

CHAPTER 1: Scientific Publications

Figure 2. Conjugation of NTF to iron oxide nanoparticles maintains their bioactivity over 2 weeks in vitro. Notes: (A) Bar graph depicting the number of neuritic extensions from adult DRG explants (at distance 0 µm) treated with free GDNF or np-GDNF. Medium formulations were used either freshly prepared or after a period of in vitro preincubation of 2 weeks. Fresh free GDNF has the strongest effect on numeric neurite outgrowth. however, during in vitro preincubation, free gDNF loses all its bioactivity, while np-gDNF maintains its increased neurite inductivity in comparison with control conditions. Bars represent mean values ± standard deviations (*P≤0.05, **P≤0.01). (B) line graph representing the number of intersections of the neuritic processes with circles drawn in distinct distances from the center of the DRG (x-axis) and their distance from the center of the DRG explants (y-axis). (C) representative photomicrographs of adult Drg explant cultures in SFM supplemented with 50 ng/ml of free GDNF or np-GDNF freshly prepared or after a period of in vitro preincubation of 2 weeks. Neurites marked with anti-III-tubulin in green, scale bars: 200 µm. Abbreviations: NTF, neurotrophic factors; DRG, dorsal root ganglion; GDNF, glia-derived neurotrophic factor; np, nanoparticle; SFM, serum-free medium; CTR, control.

CHAPTER 1: Scientific Publications drop cultures into 0.5% NVR-Gel supplemented with np-NTF. To further elucidate whether conjugation of NTF to iron oxide nanoparticles has an impact on the bioactivity of 1 the NTF proteins, we compared in this experiment the effects of np-NTF with those of free recombinant NTF proteins. As is depicted in Figure 3, the supplementation of NVR-Gel with either free NTF or np-NTF resulted in significant effects on neurite outgrowth induction from neonatal DRG drop cultures (Figure 3A, C, and E). Both free NGF and np- NGF induced significantly higher numeric neurite outgrowth in distances up 700 µm from the center of the DRG drop culture than the other NTF (Figure 3A and B).However, analysis of the ten longest neurites extending from each DRG drop culture did not reveal any significant difference among all treatments (Figure 3G). No statistically significant difference was detectable between the neurite outgrowth inductive effects of np-NTF compared with free NTF, suggesting that when incorporated into a regenerative hydrogel matrix, np-NTF are able to exert the same bioactivity as free NTF. Interestingly, for FGF- 218kDa, an even slightly increased neurite inductive potential over short distances was visible for np-FGF-218kDa in comparison with free FGF-218kDa (Figure 3C and D). The opposite was seen when comparing np-GDNF with free GDNF (Figure 3E and F). Because we could demonstrate that np-NGF exerts the most prominent neurite inductive effect in both adult DRG explant cultures and neonatal DRG drop cultures, we were interested in whether this effect could be further increased by a possible synergistic effect of the other np-NTF. Therefore, all three np-NTF (np-NGF + np-FGF-218kDa + np-GDNF) were mixed to an np-NTF pool. As is shown in the supplementary material, including Figures S3 and S4, we observed that neither np-NTF pool concentration was able to significantly induce neurite outgrowth over the value achieved with the treatment with np-NGF alone.

BMSC overexpressing NTF induce less sensory neurite outgrowth than np-NTF. BMSC were nonvirally transfected to induce overexpression of NTF. Successful genetic modification was monitored in Western blot analysis from cell lysates and cell culture supernatants (Figure 4). Cell lysates of engineered BMSC reveal strong expression of the respective FGF-218kDa-Flag or pro-GDNF-Flag (Figure 4A). Cell culture supernatants collected from cultures of GDNF-BMSC contain increased levels of pro-GDNF with increasing culture time as well as mature GDNF protein after 72–120 hours of culture ( Figure 4B). On the contrary, nT-BMSC culture supernatants contain low levels of endogenous pro- GDNF, while culture supernatants of the empty-BMSC demonstrate a decreased expression of pro-GDNF (Figure 4B). With regard to FGF-2, neither endogenous FGF-218kDa nor FGF-218kDa-Flag was detectable in cell culture supernatants of nT-BMSC or genetically engineered BMSC (data not shown). Genetically engineered BMSC were suspended in NVR-Gel and analyzed with respect to their neurite inductive effect on neonatal DRG drop cultures.

CHAPTER 1: Scientific Publications

Figure 3. After incorporation into NVr-gel, np-NTF demonstrate the same bioactivity as free NTF. Notes: (Left column) representative photomicrographs of neonatal Drg drop cultures in NVr-gel supplemented for 24 hours with 50 ng/ml of np-NGF (A), np-FGF-218kDa (C), or np-GDNF (E). Neurites marked with anti-III- tubulin in red, schwann cells marked with anti-s100 in green, and nuclear staining (DaPI) in blue. scale bars: 500 µm. (Right column) Bar graphs depicting the number of neurite intersections (y-axis) with circles drawn at distinct distances (x-axis) from the neonatal Drg drop culture. The neurite inductive effect of each free NTF was compared with its np-NTF counterpart and the nontreated cTr: (B) NGF, (D) FGF-218kDa, (F) GDNF (*P≤0.05, **P≤0.01, ***P≤0.001). (G) Bar graph illustrating the mean length ± standard deviation of the ten longest neurites extending from each DRG drop culture. No significant difference was detectable among all treatments. Abbreviations: np, nanoparticle; NTF, neurotrophic factors; DRG, dorsal root ganglion; NGF, nerve growth factor; FGF, fibroblast growth factor; GDNF, glia-derived neurotrophic factor; DAPI, 4′,6- diamidino-2-phenylindole; CTR, control

CHAPTER 1: Scientific Publications

Figure 4 Western blot results for protein detection after nonviral transfection of BMSC. Notes: (A) analysis of cell lysates of nT-BMSC or genetically engineered empty-BMSC, FGF-218kDa-BMSC, or GDNF- BMSC (β4 hours, α-FGF-β, and α-GDNF antibody). (B) analysis of cell culture supernatants from nT- BMSC or genetically engineered empty-BMSC, or GDNF-BMSC harvested after 24 hours, 72 hours, and 1β0 hours (α-GDNF antibody). Abbreviations: BMSC, bone marrow-derived mesenchymal stromal cells; np, nanoparticle; FGF, fibroblast growth factor; GDNF, glia-derived neurotrophic factor; nT, non transfected.

As is depicted in Figure 5, all types of genetically engineered BMSC, but especially nT- BMSC, demonstrated nonsignificant increase in numeric neurite outgrowth from DRG drop cultures over the value achieved in pure NVR-Gel (control [CTR]), and no significant difference was detected between the different cell types analyzed (Figure 5A). The maximum neurite lengths were almost not affected by the presence of nT-BMSC (Figure 5C) in comparison with control levels (CTR), while engineered BMSC (FGF-218kDa-BMSC and GDNF-BMSC) demonstrated nonsignificant impairment of distance outgrowth, which was most prominent in the presence of GDNF-BMSC (Figure 5B). When comparing numeric neurite outgrowth from DRG drop cultures in the presence of engineered BMSC with the incubation with np-FGF-218kDa or np-GDNF, respectively, no significant differences were detectable ( Figure 6A). However, it is noteworthy that while nT-BMSC had a very similar effect on maximum neurite lengths as np-NTF, co-culture with engineered BMSC, especially GDNF-BMSC, significantly impaired distance outgrowth of neurites (Figure 6B).

CHAPTER 1: Scientific Publications

Figure 5. Engineered neurotrophic factor-overexpressing BMSC have less neurite inductive bioactivity in Drg cultures than nontransfected cTr cells. Notes: Neurite outgrowth was quantitatively analyzed with regard to (A) the number of neurite intersections (y-axis) with circles drawn at distinct distances (x-axis) from the neonatal DRG drop culture co-cultured with nT-BMSC, FGF-218kDa-BMSC, or GDNF-BMSC and (B) mean length ± standard deviation of the ten longest neurites extending from each Drg drop culture. (C) Sample photomicrograph of fluorescence-labeled DRG drop culture (-III-tubulin, green) co-cultured with nT-BMSC (F-actinphalloidin, red). Abbreviations: BMSC, bone marrow-derived mesenchymal stromal cells; DRG, dorsal root ganglion; CTR, control; nT, nontransfected; FGF, fibroblast growth factor; GDNF, glia-derived neurotrophic factor.

Neurite formation by PC-12 cells is supported to a higher extent in the presence of FGF- 218kDa-BMSC than in the presence of np-FGF-218kDa. In order to directly compare the neurite inductive potential of np-NTF and engineered NTF-BMSC again in another cell system, we utilized the PC-12 cell neuronal differentiation assay as a third in vitro model. PC-12 cells are highly sensitive for treatment with NGF or FGF-218kDa. Therefore, free NGF, np-NGF, free FGF-218kDa, np-FGF-218kDa, or FGF-218kDa-BMSC were incorporated in NVR-Gel together with PC-12 cells. As is shown in Figure 7, 50 ng/mL of np-NGF demonstrated a slightly stronger neurite outgrowth inductive potential than 50 ng/mL free NGF (positive control condition) or free FGF-218kDa and np-FGF-218kDa (Figure 7A). Interestingly, the presence of nT-BMSC significantly impaired neurite outgrowth from PC- 100

CHAPTER 1: Scientific Publications

12 cells when compared with all other conditions, while FGF-218kDa-BMSC (Figure 7B) demonstrated a significantly increased effect on neuronal differentiation of PC-12 cells in 1 comparison with free FGF-218kDa or np-FGF-218kDa.

Figure 6. Engineered neurotrophic factor-overexpressing BMSC have less neurite inductive bioactivity than np-NTF. Notes: (A) Bar graphs depicting the number of neurite intersections from the neonatal Drg drop culture under cTr conditions or treatment with 50 ng/ml np-NTF or co-cultured with genetically engineered BMSC. (B) Bar graph illustrating the mean length ± standard deviation of the ten longest neurites extending from each neonatal Drg drop culture. Abbreviations: BMSC, bone marrow- derived mesenchymal stromal cells; np, nanoparticle; NTF, neurotrophic factors; Drg, dorsal root ganglion; cTr, control; nT, nontransfected; FGF, fibroblast growth factor; GDNF, glia-derived neurotrophic factor.

DISCUSSION

In this study, we compared in vitro two different possible approaches to deliver NTF within artificial grafts for peripheral nerve reconstruction. The first was the nanotechnology approach of conjugation of NTF with iron oxide nanoparticles, which was supposed to increase the stability of the conjugated NTF, but also to ensure local and slow release of NTF in the lumen of nerve conduits5. The second approach we studied was stem cell engineering. Here, we induced NTF overexpression in BMSC with nonviral genetic modification. The genetically engineered BMSC could potentially be transplanted within tissue-engineered nerve conduits, as has been successfully done before with genetically engineered Schwann cells4.

101

CHAPTER 1: Scientific Publications

In order to compare the feasibility of optimized NTF delivery increasing the neurite outgrowth in vitro, we utilized adult DRG explant organotypic cultures as well as neonatal DRG drop cultures. The DRG neurite outgrowth model is widely used to assess the neurite outgrowth in a specific environment.24,25 Using adult DRG explant cultures, we demonstrate that the bioactivity of np-NTF is comparable with that of free NTF when incorporated into a regenerative hydrogel matrix. The latter is a good prerequisite for future in vivo application. From previous studies, we already have proof that NTF conjugated to iron oxide nanoparticles demonstrate prolonged stability over free NTF.14 Here, we clearly demonstrate that not only the stability of the proteins but also, most importantly, their bioactivity is stabilized by conjugation of NTF to nanoparticles. Our study comparing the neurite inductive potential of fresh with preincubated medium formulations evidenced that only minimal reduction of bioactivity occurs for np-GDNF during a 2-week preincubation period, while free GDNF lost its neurite inductive activity in the same time span. We could further demonstrate that np-NTF are able to stimulate neuronal outgrowth in a dose-dependent manner and that np-NTF have similar bioactivity as free NTF when suspended in a regenerative hydrogel matrix. The latter system mimics a way to apply np- NTF with nerve conduits for peripheral nerve reconstruction surgery. As hydrogel matrix, we used NVR-Gel, which has been described before as a suitable matrix for neuronal cultures17 and which seemed to be appropriate as luminal filler for artificial nerve conduits.

Figure 7. Neuronal differentiation of Pc-12 cells is seen to the highest extent in co-cultures with FgF-218kDa-BMSC. Notes: (A) Neurite formation by EGFP-PC-12 cells was quantitatively analyzed in the presence of either 50 ng/ml of free NTF (NGF or FGF-218kDa) or np-NTF (NGF or FGF-218kDa) or in co-culture of FgF-218kDa-BMSC. While no significant differences were detectable between treatment with free NTF or np-NTF, nT-BMSC demonstrated a significantly reduced neurite inductive bioactivity in comparison with all other conditions (**P≤0.01, ***P≤0.001). In contrast, maximal neuronal differentiation by PC-12 cells was detected in co-culture with FGF-218kDa-BMSC, with significant difference to free FGF- 218kDa and np-FGF-218kDa (*P≤0.05). (B) sample photomicrograph of EGFP-Pc12 cells co-cultured with FgF-218kDa-BMSC. Abbreviations: PC-12 cells, cell line from rat pheochromocytoma cells; FGF, fibroblast growth factor; BMSC, bone marrow-derived mesenchymal stromal cells; EGFP, enhanced green fluorescent protein; NTF, neurotrophic factors; NGF, nerve growth factor; np, nanoparticle; nT, non transfected 102

CHAPTER 1: Scientific Publications

Using neonatal DRG drop cultures we demonstrate that np-NTF are effective in inducing the numeric neurite outgrowth activity also when suspended in NVR-Gel. In NVR-Gel 1 enriched with np-NTF, the elongation of neurites, however, seems not to be stimulated. We suppose that the presence of the gel slowed down neurite elongation, but further analysis using long-term cultures is necessary to elucidate how neurite distance outgrowth is influenced by NVR-Gel. With regard to potential in vivo application of np-NTF, another finding from our study has to be considered. The dose dependency of np-NTF-induced sensory neurite outgrowth varied among the NTF investigated. While 50 ng/mL of np-NGF or np-GDNF induced maximal neurite outgrowth among the concentrations chosen, the lower concentration of only 10 ng/mL np-FGF-218kDa showed a trend to be more efficient than higher concentrations (data presented as supplementary material). All np-NTF investigated, however, demonstrated an induction of neurite outgrowth activity at the concentration of 50 ng/mL, which can be concluded as a useful concentration for future in vivo application. From the acquired in vitro data, it cannot be finally decided which np-NTF would give maximum support of axonal regeneration in vivo. DRG neuron populations are heterogeneous with regard to the quality of sensory transmission, and different growth factors act selectively on different subpopulations of DRG neurons.26–29 Therefore, an emerging concept in the field of artificial nerve graft development is to use a combination of NTF with the aim to obtain a widespread effect on all sensory, motor, and sympathetic neuronal subpopulations.30–32 It is, however, still under debate which NTF combination will result in the most successful support of PNR while simultaneously avoiding negative influences in the process.33 Therefore, we also investigated the effect of an np-NTF pool on the neurite outgrowth activity in adult DRG explant and neonatal DRG drop cultures. In our hands, an np-NTF pool of np-NGF, np-FGF-218kDa, and np-GDNF did not demonstrate higher efficiency in promoting neurite outgrowth than 50 ng/mL np-NGF alone. This indicates that there is no synergistic effect among these three selected NTF. However, considering the limits of an in vitro approach, it is possible that in our culture systems, 50 ng/mL np-NGF treatment already induced the maximum neurite outgrowth with no possibility to demonstrate a synergy for the additionally added np-FGF-218kDa or np- GDNF.

Evaluation of the second innovative approach of NTF delivery to nerve conduits and stem cell engineering showed that primary nT-BMSC have a higher potency to induce sensory neurite outgrowth than engineered NTF-BMSC. The basis for this may be the secretome of BMSC that provides a variety of growth factors to the co-cultured neurons.34 We assume that the secretome has, to some extent, been negatively influenced by the genetic modification during BMSC engineering. Further analysis is needed to elucidate this in more detail prior to potential in vivo application, but, although only seen in one out of three independent analyses, the Western blot analysis shown in Figure 4B already indicates

103

CHAPTER 1: Scientific Publications that the GDNF expression might be altered between nT-BMSC and genetically engineered empty-BMSC.

When comparing the two innovative approaches of nanotechnology and stem cell engineering with regard to sensory neurite outgrowth, however, np-NTF have demonstrated a slightly increased bioactivity over NTF-BMSC. Additionally, using a different neuronal differentiation model,22,35 the PC-12 cell neurite outgrowth induction assay, NTF-BMSC demonstrated an increased bioactivity over np-NTF. We therefore assume that the secretome together with the overexpressed FGF-218kDa was, in this setting, specifically supportive for PC-12 cell sympathetic neurite outgrowth.

One last point has to be considered when thinking of future in vivo application of nanotechnological NTF delivery to peripheral nerve guidance conduits. The conduits will be sutured between the transected nerve ends and, especially the distal nerve end, will undergo Wallerian degeneration. As this process will provide Schwann cell ingrowth into the nerve conduits, we were also interested in whether Schwann cell migration is modified in our in vitro models. The data are not shown because we have no hints that Schwann cell migration is modified by adding np-NTF or BMSC to the system. Whether synergistic effects would develop for np-NTF delivered within the conduits and those additionally liberated during Wallerian degeneration of the distal nerve end, is only accessible in vivo and must be the subject of future studies.

CONCLUSION

Both approaches demonstrated their potential to induce neurite outgrowth in different neuronal models. They may be chosen specifically in the context of peripheral nerve tissue engineering. However, the variability of the BMSC secretome upon genetic engineering of the cells needs to be studied in more detail prior to their in vivo application. Our comparative in vitro study provides evidence that np-NTF have longterm bioactivity and reliably support neurite outgrowth in vitro also after periods when free NTF have already lost their bioactivity. With regard to their therapeutical use, iron oxide nanoparticles are considered to be nontoxic and are already in use for various biomedical applications, such as diagnostics, cell labeling, magnetic resonance imaging, and X-ray contrast agents, and for hyperthermia.5 Therefore, npNTF are good candidates for the in vivo evaluation of their PNR-promoting potential in future studies.

ACKNOWLEDGMENTS. This project has received funding from the European Union’s Seventh Programme for Research, Technological Development and Demonstration under Grant Agreement No 278612. For excellent technical assistance, we thank Maike Wesemann from the Institute of Neuroanatomy at Hannover Medical School. For support in bone marrow-derived mesenchymal stem cell cultures, we thank Antonio J Salgado from the University of Minho, Braga, Portugal.

104

CHAPTER 1: Scientific Publications

DISCLOSURE. The authors report no conflicts of interest in this work

REFERENCES 1

1. Tos P, Ronchi G, Geuna S, Battiston B. Future perspectives in nerve repair and regeneration. Int Rev Neurobiol. 2013;109:165–192. 2. Deumens R, Bozkurt A, Meek MF, et al. Repairing injured peripheral nerves: bridging the gap. Prog Neurobiol. 2010;92(3):245–276. 3. Geuna S, Gnavi S, Perroteau I, Tos P, Battiston B. Tissue engineering and peripheral nerve reconstruction: an overview. Int Rev Neurobiol. 2013;108:35–57. 4. Haastert K, Grothe C. Gene therapy in peripheral nerve reconstruction approaches. Curr Gene Ther. 2007;7(3):221–228. 5. Ziv-Polat O, Skaat H, Shahar A, Margel S. Novel magnetic fibrin hydrogel scaffolds containing thrombin and growth factors conjugated iron oxide nanoparticles for tissue engineering. Int J Nanomedicine. 2012; 7:1259–1274. 6. Madduri S, Gander B. Schwann cell delivery of neurotrophic factors for peripheral nerve regeneration. J Peripher Nerv Syst. 2010;15(2): 93–103. 7. Haastert K, Lipokatic E, Fischer M, Timmer M, Grothe C. Differentially promoted peripheral nerve regeneration by grafted Schwann cells over-expressing different FGF-2 isoforms. Neurobiol Dis. 2006; 21(1):138–153. 8. Haastert K, Ying Z, Grothe C, Gomez-Pinilla F. The effects of FGF-2 gene therapy combined with voluntary exercise on axonal regeneration across peripheral nerve gaps. Neurosci Lett. 2008;443(3):179–183. 9. Oliveira JT, Mostacada K, de Lima S, Martinez AM. Bone marrow mesenchymal stem cell transplantation for improving nerve regeneration. Int Rev Neurobiol. 2013;108:59–77. 10. Johnson EO, Charchanti A, Soucacos PN. Nerve repair: experimental and clinical evaluation of neurotrophic factors in peripheral nerve regeneration. Injury. 2008;39 Suppl 3:S37–S42. 11. Catrina S, Gander B, Madduri S. Nerve conduit scaffolds for discrete delivery of two neurotrophic factors. Eur J Pharm Biopharm. 2013;85(1):139–142 12. Klimaschewski L, Hausott B, Angelov DN. The pros and cons of growth factors and cytokines in peripheral axon regeneration. Int Rev Neurobiol. 2013;108:137–171. 13. Molday RS, MacKenzie D. Immunospecific ferromagnetic iron-dextran reagents for the labeling and magnetic separation of cells. J Immunol Methods. 1982;52(3):353–367. 14. Ziv-Polat O, Shahar A, Levy I, et al. The role of neurotrophic factors conjugated to iron oxide nanoparticles in peripheral nerve regeneration: in vitro studies. Biomed Res Int. 2014;2014:1–10. 15. Shahar A, Ziv-Polat O, Margel S, Freier T. Developing a customized tissue-engineered implant for peripheral nerve reconstruction: role of nanoparticles. In: Bhoop BS, ed. Medical Nanobiotechnology. Vol 5. Houston, TX: Studium Press LLC; 2014.

105

CHAPTER 1: Scientific Publications

16. Fueshko S, Wray S. LHRH cells migrate on peripherin fibers in embryonic olfactory explant cultures: an in vitro model for neurophilic neuronal migration. Dev Biol. 1994;166(1):331–348. 17. Shahar A, Nevo Z, Rochkind S, inventors. Cross-linked hyaluronic acid-laminin gels and use thereof in cell culture and medical implants. United States patent US 00243732001. November 13, 2001. 18. Wrobel S, Serra SC, Ribeiro-Samy S, et al. In vitro evaluation of cellseeded chitosan films for peripheral nerve tissue engineering. Tissue Eng Part A. 2014;20(17–18):2339– 2349. 19. Niwa H, Yamamura K, Miyazaki J. Efficient selection for highexpression transfectants with a novel eukaryotic vector. Gene. 1991; 108(2):193–199. 20. Ratzka A, Kalve I, Ozer M, et al. The colayer method as an efficient way to genetically modify mesencephalic progenitor cells transplanted into 6-OHDA rat model of Parkinson’s disease. Cell Transplant. 2012;21(4):749–762. 21. Ratzka A, Baron O, Grothe C. FGF-2 deficiency does not influence FGF ligand and receptor expression during development of the nigrostriatal system. PLoS One. 2011;6(8):e23564. 22. van Bergeijk J, Haastert K, Grothe C, Claus P. Valproic acid promotes neurite outgrowth in PC12 cells independent from regulation of the survival of motoneuron protein. Chem Biol Drug Des. 2006;67(3): 244–247. 23. Rochkind S, Shahar A, Fliss D, et al. Development of a tissue-engineered composite implant for treating traumatic paraplegia in rats. Eur Spine J. 2006;15(2):234–245 24. Melli G, Hoke A. Dorsal root ganglia sensory neuronal cultures: a tool for drug discovery for peripheral neuropathies. Expert Opin Drug Discov. 2009;4(10):1035–1045. 25. Bilsland J, Rigby M, Young L, Harper S. A rapid method for semiquantitative analysis of neurite outgrowth from chick DRG explants using image analysis. J Neurosci Methods. 1999;92(1–2):75–85. 26. Benedetti M, Levi A, Chao MV. Differential expression of nerve growth factor receptors leads to altered binding affinity and neurotrophin responsiveness. Proc Natl Acad Sci U S A. 1993;90(16):7859–7863. 27. Bennett DL, Michael GJ, Ramachandran N, et al. A distinct subgroup of small DRG cells express GDNF receptor components and GDNF is protective for these neurons after nerve injury. J Neurosci. 1998; 18(8):3059–3072. 28. Wright DE, Snider WD. Neurotrophin receptor mRNA expression defines distinct populations of neurons in rat dorsal root ganglia. J Comp Neurol. 1995;351(3):329–338. 29. Xiao J, Wong AW, Willingham MM, et al. BDNF exerts contrasting effects on peripheral myelination of NGF-dependent and BDNFdependent DRG neurons. J Neurosci. 2009;29(13):4016–4022.

106

CHAPTER 1: Scientific Publications

30. Sharma HS. Neurotrophic factors in combination: a possible new therapeutic strategy to influence pathophysiology of spinal cord injury and repair mechanisms. Curr Pharm 1 Des. 2007;13(18):1841–1874. 31. Madduri S, di Summa P, Papaloizos M, Kalbermatten D, Gander B. Effect of controlled co-delivery of synergistic neurotrophic factors on early nerve regeneration in rats. Biomaterials. 2010;31(32):8402–8409. 32. Blesch A, Fischer I, Tuszynski MH. Gene therapy, neurotrophic factors and spinal cord regeneration. Handb Clin Neurol. 2012;109:563–574. 33. Madduri S, Gander B. Growth factor delivery systems and repair strategies for damaged peripheral nerves. J Control Release. 2012; 161(2):274–282. 34. Ribeiro CA, Salgado AJ, Fraga JS, Silva NA, Reis RL, Sousa N. The secretome of bone marrow mesenchymal stem cells-conditioned media varies with time and drives a distinct effect on mature neurons and glial cells (primary cultures). J Tissue Eng Regen Med. 2011;5(8):668–672. 35. van Bergeijk J, Rydel-Konecke K, Grothe C, Claus P. The spinal muscular atrophy gene product regulates neurite outgrowth: importance of the C terminus. Faseb J. 2007;21(7):1492–1502.

107

CHAPTER 1: Scientific Publications

SUPPLEMENTARY MATERIALS

Neurotrophic factors conjugated to iron oxide nanoparticles induce a dose-dependent neurite outgrowth from adult dorsal root ganglion explants. Adult rat dorsal root ganglion (DRG) explants were cultured over 3 days in the presence of increasing amounts of neurotrophic factors conjugated to iron oxide nanoparticles (np-NTF), and the numbers of extending neurites measured at certain distances from the DRG explant. As is depicted in Figure S2, at a certain dosage, all np-NTF tested significantly increased adult sensory neurite outgrowth over the values seen in control cultures (cultured in serum-free medium alone). All np-NTF demonstrated a dose dependency of their neurite inductive effects. The strongest effect on numeric neurite outgrowth (Figure S2A) was seen for the treatment with 50 ng/mL of either np-NGF or np-glia-derived neurotrophic factor (GDNF). Lower or higher medium concentrations of both np-NGF and np-GDNF showed reduced effects with significantly lower neurite numbers for the reduced concentration of 10 ng/mL of np-NGF and the elevated concentration of 100 ng/ mL of np-GDNF, respectively. Only minimal differences in the numeric neurite outgrowth were detectable for the treatment with different concentrations of np-fibroblast growth factor (FGF)-218kDa. While a significant increase of neurite numbers over control levels was already detectable for the lowest concentration of 10 ng/mL np-FGF-218kDa, the elevated concentrations no longer induced significant effects on neurite outgrowth. As is depicted in Figure S2B, the strongest effect on neurite elongation was detected under treatment with np-NGF by which, again, 50 ng/mL induced the maximum neurite lengths in the highest number of neurites. But treatment with 50 ng/mL np-FGF-218kDa also induced outgrowth of longer neurites, up to a distance of 450 µm from the centre of the DRG, than all other treatments except the one with np-NGF. Under control conditions, outgrowing neurites reached a distance of only 300 µm from the DRG centre.

Figure S1 Non transfected and EGFP-Flag-BMSC in culture. Notes:(A) nT-BMSC stained for F-actin filaments (red), demonstrating their regular fibroblast-like shape. (B) anti-Flag (red) stained BMSC after nonviral transfection with EGFP-Flag. Nuclear staining with 4′,6-diamidino-2-phenylindole (blue). Abbreviations:EGFP, enhanced green fluorescent protein; BMSC, bone marrow-derived mesenchymal stromal cells; nT, non transfected

108

CHAPTER 1: Scientific Publications

Combination of three different neurotrophic factors conjugated to np-NTF fails to further increase the neurite inductive effect of np-NGF alone. Because a concentration of 1 50 ng/mL np-NGF had demonstrated the most prominent neurite inductive effect, we were interested in whether this effect could be further increased by a possible synergistic effect of the other np-NTF. Therefore, all three np-NTF (np-NGF + np-FGF218kDa + np-GDNF) were mixed to an np-NTF pool of two different concentrations: 30 ng/mL (each np-NTF) and 50 ng/mL (each np-NTF). As is demonstrated in Figure S3, we observed that neither np-NTF pool concentration was able to significantly induce neurite outgrowth over the value achieved with the treatment with 50 ng/mL np-NGF alone. The numeric neurite outgrowth from adult DRG explants was significantly increased over control levels in both np-NTF pool treatments, which were not significantly different from each other (Figure S3A). Significantly longer neurites than under control conditions were induced by both np- NTF pool treatments as well (Figure S3B). However, treatment with 50 ng/mL np-NGF alone demonstrated a tendency to induce an even more increased number of long neurites (up to > 500 µm from the center of the DRGs). Additionally to testing the neurite inductive effects of np-NTF, we again analyzed the bioactivity of an np-NTF pool (50 ng/mL of each np-NGF, np-FGF 218kDa, and np-GDNF) suspended in NVR-Gel. Similar to the results after incubation of adult DRG explant cultures with an np-NTF pool, we observed no synergistic effect. As is shown in Figure S4, the neurite inductive potential of the np-NTF pool was similar to that of 50 ng/mL np-NGF alone.

109

CHAPTER 1: Scientific Publications

Figure S2 Neurotrophic factors conjugated to iron oxide nanoparticles stimulate neurite outgrowth from adult DRG explants in a dose-dependent manner. Notes:(A) Bar graph representing the number of neuritic extensions from adult DRG explants (at distance 0 µm) treated with different amounts of np-NTFs. Statistical analysis revealed the strongest effect on neurite outgrowth when explants were treated with 50 ng/mlof either np-NgF or np-GDNF. Treatment with 10 ng/ml np-FgF-218kDa already evoked a significant increase of neurite outgrowth when compared with untreated controls. Bars represent mean values ±standard deviations (**P≤0.01, ***P≤0.001). (B) line graph representing the number of intersections of the neuritic processes with circles drawn in distinct distances from the centre of the DRG (x-axis) and their distance from the centre of the DRG explants (y-axis). Abbreviations: DRG, dorsal root ganglion; np, nanoparticle; NTF, neurotrophic factors; NGF, nerve growth factor; GDNF, glia-derived neurotrophic factor; FGF, fibroblast growth factor; CTR, control

110

CHAPTER 1: Scientific Publications

Figure S3 combined administration of the three np-NTFs fails to increase the number of neurites extending from adult DRG explants. Notes: (A) Bar graph representing the number of neuritic extensions of adult DRG explants (at distance 0 µm) treated with np-NGF (50 ng/ml) or with an np-NTF pool (np-NGF +np- FgF-218kDa+np-gDNF) in the concentration of 30 ng/ml (each np-NTF) or 50 ng/ml (each np-NTF), respectively. Bars represent mean values ±standard deviations (*P≤0.05, ***P≤0.001). (B) line graph representing the number of intersections of the neuritic processes with circles drawn in distinct distances from the center of the DRG (x-axis) and their distance from the center of the DRG explants (y-axis). No significant increase over the neurite outgrowth inductive effect seen with treatment with 50 ng/ml np-NgGF alone could be achieved by treatment with any np-NTF pool. With regard to neurite lengths, the strongest effect has rather been detected for the single treatment with np-NGF. Abbreviations:np, nanoparticle; NTF, neurotrophic factors; DRG, dorsal root ganglion; NGF, nerve growth factor; FGF, fibroblast growth factor; GDNF, glia-derived neurotrophic factor; CTR, control

111

CHAPTER 1: Scientific Publications

Figure S4 No synergistic effect is induced by supplementing NVR-gel with a pool of np-NTFs. Notes:(A) Bar graphs depicting the number of neurite intersections (y-axis) with circles drawn at distinct distances (x-axis) from the neonatal DRG drop culture under control conditions or treatment with 50 ng/mL np- NGF or the 50 ng/mL np-NTF pool. No statistically significant difference between the neurite inductive potential of np-NGF and np-NTF pool was detectable (*P≤0.05, **P≤0.01, ***P≤0.001). (B) Bar graph illustrating the mean length ±standard deviation of the ten longest neurites extending from each of the DRG drop cultures. No significant difference was detectable between both treatments. Abbreviations:np, nanoparticle; NTFs neurotrophic factors; DRG, dorsal root ganglion; NGF, nerve growth factor; CTR, control

112

CHAPTER 1: Scientific Publications

1 Effect of Local Delivery of GDNF Conjugated Iron Oxide Nanoparticles on Nerve Regeneration along Long Chitosan Nerve Guide

Federica Fregnan1, Ofra Ziv-Polat2, Michela Morano1, Mira M. Mandelbaum-Livnat3, Moshe Nissan3, Tolmasov Michael3, Akiva Koren3, Tali Biran3, Yifat Bitan3, Evgeniy Reider3, Mara Almog3, Nicoletta Viano1, Shimon Rochkind3, Stefano Geuna1 and Abraham Shahar2

1 Department of Clinical and Biological Sciences, Cavalieri Ottolenghi Neuroscience Institute, University of Turin, Turin, Italy 2 NVR Research Ltd., Ness Ziona, Israel 3 Division of Peripheral Nerve Reconstruction, Department of Neurosurgery, Research Center for Nerve Reconstruction, Tel Aviv‐ Sourasky Medical Center, Tel Aviv University, Tel Aviv, Israel

‐ Published in INTECH "Peripheral Nerve Regeneration - From Surgery to New Therapeutic Approaches Including Biomaterials and Cell-Based Therapies Development", Chapter 8

ABSTRACT Local delivery of neurotrophic factors is a pillar of neural repair strategies in the peripheral nervous system. The main disadvantage of the free growth factors is their short half‐life of few minutes. In previous studies, it was demonstrated that conjugation of various neurotrophic factors to iron oxide nanoparticles (IONP) led to stabilization of the growth factors and to the extension of their biological activity compared to the free factors. In vitro studies performed on organotypic dorsal root ganglion (DRG) cultures seeded in NVR gel (composed mainly of hyaluronic acid and laminin) revealed that the glial cell–derived neurotrophic factor (GDNF) conjugated to IONP‐enhanced early nerve fiber sprouting and accelerated the onset and progression of myelin significantly earlier than the free GDNF and other free and conjugated factors. The present article summarizes results of in vivo study, aimed to test the effect of free versus conjugated GDNF on regeneration of the rat sciatic nerve after a severe segment loss. We confirmed that nerve device enriched with a matrix with GDNF gives more successful results in term of regeneration and functional recovery in respect to the hollow tube; moreover, there are no detectable differences between free versus conjugated GDNF.

113

CHAPTER 1: Scientific Publications

INTRODUCTION The incidence of nerve injury is quite high in the world, related also to the exposition of nerve tissue in the body. Despite the good regeneration ability retained by peripheral nerve system in adulthood, nerve injuries are associated with high morbidity and deep alteration of patient’s life [1]. Severe transection injuries, in which nerve continuity is lost, require a surgical approach, specially where the distance between the proximal and the distal nerve stumps is extended. In order to avoid negative side effect of the autologous nerve graft, the current gold standard technique, and to improve the functional recovery and reduce neuropathic pain, various biosynthetic nerve grafts have been developed to bridge the two nerve stumps [2–4]. Some of the artificial nerve guides have been approved for clinical use, and indeed, their use in reconstructive surgery is restricted because they still show limitations, principally for longdistance repair [5]. One category of these nerve conduits is defined absorbable conduit, according to their characteristic to be degraded in the host body; among these, the chitosan tube has been widely used in pre‐clinical studies obtaining promising results [4, 6, 7]. Chitosan is a polysaccharide derived from chitin, it is biocompatible, and it can promote glial cells survival and neurite outgrowth [8–10]. It has been shown that hollow chitosan tube can be effective as autologous nerve graft for bringing 10‐mm gap in rat sciatic nerve model [7, 11, 1β]. In order to increase the performance of chitosan tube for long defect repair; here, we filled the tube with a hydrogel, called NVR gel, composed mainly by hyaluronic acid and laminin. Previous works demonstrated that NVR gel is a promising hydrogel that allow neurite outgrowth and cell survival in vitro [13]. Furthermore, we used NVR gel as a carrier for neurotrophic factors. In particular, we focus our attention on glial cell line–derived neurotrophic factor (GDNF), which is mainly involved in motor neuron regrowth and remyelination [14–16]. After nerve injury, endogenous growth factors are released by neuronal and glial cells from the distal nerve stump, and they can stimulate and guide axon regeneration; yet, this support is ineffective in regeneration over long distance and extended in time due to the short life of neurotrophic factors and the decline of their production [17, 18]. The implementation of the nerve conduit with neurotrophic factors has the aim to maintain elevated the concentration of these growth factors and to extend their action for the prolonged time needed for axons to reach the target. For this purpose, we use the NVR gel as a scaffold for the release of GDNF conjugated to iron oxide nanoparticles (GDNF‐IONP). We have previously demonstrated that conjugation of various neurotrophic factors to iron oxide nanoparticles (IONP) led to stabilization of the growth factors and to prolonged biological activity respect to the non‐conjugated factors [19]. IONP are considered biocompatible and biodegradable, they are actually used for various biomedical application [20–22], and thus, they were suggested for in vivo use for peripheral nerve regeneration [23]. Moreover, in vitro experiments showed that the combination of NVR gel and IONP represents a permissive environment to neurite outgrowth [19], and GDNF‐IONP has been shown to accelerate the onset and progression of myelin in

114

CHAPTER 1: Scientific Publications organotypic dorsal root ganglion (DRG) cultures seeded in NVR gel, significantly earlier than the free GDNF and other free and conjugated factors [24]. 1 In the present study, we explored the efficacy of chitosan tube enriched with NVR gel and GDNF‐IONP to the repair of critical length, 15 mm, sciatic nerve defect in rat model. We evaluated the functional and morphological outcome of nerve regeneration at 5 months after nerve injury, analyzing the effects given by the presence of NVR gel alone, NVR gel with GDNF‐IONP or free GDNF inside the chitosan tube respect to the empty nerve device.

MATERIALS AND METHODS

Synthesis of GDNF‐IONP. Dextran‐coated iron oxide magnetic nanoparticles of 10 nm diameter were prepared as described previously [3]. The dextran coating was used to covalently bind the protein GDNF using divinyl sulfone binding reagent [1].

NVR gel preparation. NVR gel is a matrix supporting cell growth and survival composed by high‐molecularweight hyaluronic acid (γ × 106 Da, BTG Polymers, Kiryat Malachi, Israel) and laminin (Sigma, Rehovot, Israel) [4]. For the in vivo application as a tube filler, the NVR gel was diluted in nutrient medium corresponding to Dulbecco’s modified eagle medium‐nutrient mixture F‐1β (DMEM‐F1β), supplemented with 10% fetal calf serum (FCS), 2 nM glutamine, 6 g/L d‐glucose, β5 μg/mL gentamycine and 50 ng/mL IGF‐I. The final concentration of the solution is 0.5% to render more suitable gel manipulation. NVR gel was filled into the chitosan tubes during the surgical implantation using a syringe.

Preparation of chitosan conduit. Chitosan tubes were manufactured by Medovent GmbH (Mainz, Germany) under ISO 13485 conditions from chitin tubes made following three main procedures: the extrusion process, distinctive washing, and hydrolysis steps to reach the required medium degree of acetylation (DAII). Tubes were finally cut into the length of 17 mm and treated with ethylene oxide for sterilization.

Animals. All animal experiments were approved by the Institutional Animal Care and Usage Committee (IACUC) and adhered strictly to the Animal Care guidelines. Female Wistar rats were brought to the vivarium β weeks prior to the surgery and housed two per cage with a 12 h light/dark cycle, with free access to food and water.

Experiment design and surgical technique. Fifty female Wistar rats, weighing 200–β50 g each, were studied using an experimental model for producing a complete peripheral nerve injury with massive nerve defect that has recently been described [7]. During the 2 weeks before surgery and for the entire study, the rats were given Amitriptyline in their drinking water, in order to decrease autotomy—self‐eating of toes—in the operated limb. Before surgery, a general anesthesia was induced with intraperitoneal injection of xylazine (15 mg/kg) and ketamine (50 mg/kg). During the surgical procedures, a high magnification microscope was used. The left sciatic nerve was uncovered and disconnected from biceps 115

CHAPTER 1: Scientific Publications femoris and semimembranous muscles, beginning from the area of branches to the glutei and hamstring muscles and distally to the trifurcation into peroneal, tibial, and sural nerves. At the third femur level, the sciatic nerve was fully transected using microsurgical scissors, and a 15‐mm nerve segment was removed. Afterward, the rats were divided into five experimental groups according to the type of implant: empty tubes (n = 10); tubes filled with NVR gel (n = 10); tubes filled with NVR gel enriched with free GDNF (n = 10); tubes filled with NVR gel enriched with GDNF‐IONP (n = 10); and autologous nerve grafts (n = 10). For nerve reconstruction in treatment groups, a 17‐mm chitosan empty tube was located between the proximal and the distal sides of the transected sciatic nerve. Both proximal and distal sides of the sciatic nerve were positioned 1 mm into the tube ends, providing a 15‐mm nerve gap between the proximal and distal ends. Then, the tube ends were sutured to the epineurium at the proximal and distal nerve stumps using a 9‐0 nonabsorbable suture. In control group (autologous nerve graft reconstruction), after exposition the left sciatic nerve, a 15‐mm nerve segment was severely cut, using micro scissors, at the femur level, below the superior gluteal nerve and above the dissection of the sciatic nerve into the tibial and peroneal nerves. Immediately thereafter, inverse end‐to‐end coaptation of the nerve segment was performed using β–γ nonabsorbable 10‐0 sutures. Coaptation of nerve fascicles was performed to preserve all the fascicles within the epineural sac. Then, the muscular, subcutaneous, and skin layers were closed.

End‐point assessments. Assessments before and after surgery (γ0, 90, and 150 days postoperatively) were carried out in a blinded manner without disclosure of rat’s affiliation to the evaluating team. The assessments consisted of functional motor assessment of the sciatic nerve utilizing sciatic function index (SFI), somatosensory evoked potentials (SSEP), ultrasound evaluation, and morphological analysis.

Electrophysiological evaluation. SSEPs were recorded in both the operated and intact limbs using a Dantec™ KEYPONT® PORTABLE. Conductivity of the sciatic nerve was measured by stimulating the sciatic nerve at the level of the tarsal joint with simultaneous recording from the skull over the somatosensory cortex in anesthetized rats. Two subcutaneous needle electrodes were placed under the skin of the skull, when the active electrode was placed above the somatosensory cortex along the midline, and the reference electrode was placed between the eyes. The ground electrode was placed subcutaneously on the dorsal neck. Stimulation of the sciatic nerve was conducted by a set of two polarized electrodes located on the lateral aspect of the tarsal joint. The sciatic nerve was stimulated by γ00 pulses of 0.β ms in duration with a rate of γ Hz. The intensity of the stimulus was set to 2–5 mA, causing a slight twitching of the limb under observation. A response to the stimulus was considered positive if an evoked potential appeared in at least two consecutive tests.

Ultrasound evaluation (US).During the observational period, imaging studies employing ultrasonography were carried out, in anesthetized rats, for real‐time evaluation of nerve 116

CHAPTER 1: Scientific Publications regeneration inside the chitosan tubes. The lateral aspect of the right leg was shaved to improve transducer contact. The sonographic scanning technique included longitudinal and 1 transverse sections. US examinations were performed using conventional US units equipped with color Doppler capabilities using 7–15 MHz linear transducer, yielding an axial resolution of 0.2–0.4 mm. The identification of the chitosan conduit on the US image was based on the recognition of a hyperechoic structure of a tubular shape in the longitudinal axis and a circular shape on the transverse section.

Sample resin embedding. All the nerve samples were harvested 5 months after the surgical implantation. The complete tube was collected taking care to preserve part of the proximal and the distal nerve and was fixed for 4–6 h at 4°C in 0.1‐M phosphate buffer (pH 7.4) with 2.5% glutaraldehyde. After the fixation, all the tubes were opened using a scalpel and removed in order to free and clean the regenerated nerve inside the tube and prepared it to be processed for resin inclusion. The post‐fixation of nerves regenerated inside tubes was done using β% osmium tetroxide for β h, and then, the samples were dehydrated in ethanol from γ0 to 100%. Two washings of 7 m using propylene oxide were performed, and then, samples were embedded in a mixture of propylene oxide and Glauerts’ mixture of resins (50% Araldite HY964, 50% Araldite M and 0.5% dibutylphthalate) mixed in equal parts and left 1 h at room temperature. A second embedding with Glauerts’ mixture of resins alone for an overnight was performed. The Glauerts’ mixture of resins was changed, and samples were left γ7°C for 1 h. Two following samples embedding were done using resin with 2% of accelerator 964. Samples were left for 3 days at 60°C. For stereological analysis, resin‐embedded nerve was cut from the distal stump using Ultracut UCT ultramicrotome (Leica Microsystems, Wetzlar, Germany) to obtain transverse semi‐thin sections (β.5 μm of thickness) for optical analysis.

Morphometrical analysis. Semi‐thin transverse nerve sections were stained with 1% toluidine blue and analyzed in high‐resolution light microscopy. The qualitative and quantitative morphological analysis was performed with DM4000B microscope and DFC320 digital camera, using IM50 image manager system (Leica Microsystems). One randomly selected semi‐thin section was used to measure the total cross‐section area of the nerve. Using a systematic protocol described in [25], 12–16 fields were selected, and the following parameters were measured: mean fiber density (number of fiber/field area), total number of myelinated fibers (mean fiber density × area of the nerve section) fiber area, axon area, fiber diameter (D) and fiber axon (d), myelin thickness [(D − d)/β], and g‐ratio (d/D).

Statistical analysis. Functional and electrophysiological analysis and calculations were done using MatLab software (Ver. 2008b, The MathWorks, Inc.). Nonparametric statistics were used in this study. Hence, all figures are presented with median ± mad. Significance levels were calculated using a Mann‐Whitney U test and a Wilcoxon signed rank test. SSEP responses were analyzed as categorical parameters using χβ test. 117

CHAPTER 1: Scientific Publications

For the stereological analysis, more than five sciatic nerves were analyzed for each single groups (autograft, n = 5; tube with NVR gel, n = 8; tube with NVR gel enriched with GDNFIONP, n = 8; tube with NVR gel enriched with free GDNF, n = 7). The ANOVA one way followed by Bonferroni post hoc test was performed using SPSS Statistic Program. Results are reported as mean +SD.

RESULTS

All the five defined animal groups (autograft, DAII empty tube, DAII tube with NVR gel, DAII tube with NVR gel and GDNF‐IONP, DAII tube with NVR gel with GDNF) were followed‐up until 5 months after nerve injury. During the follow‐up period, the regeneration and the position of the implanted tubes were evaluated using ultrasound observation that enabled to observe in vivo the condition of the tube without scarifying the rats. All implants were found to be complete and in correct position (Figure 1). Five months after nerve guide implantation, animals were sacrificed, and all the chitosan tubes were removed for nerve regeneration evaluation. For morphological and morphometrical analysis, the autograft group was considered as the control group, since in the comparison with healthy nerve already revealed the best regenerative aspect [26, 27]. A nerve regeneration was observed in all of the investigated conditions: a higher percentage of regenerated nerves was detected for tubes containing NVR gel with GDNF‐IONP (100%), while a lower percentage of regenerated nerves (56%) was observed for tubes containing NVR gel with GDNF; the lowest percentage of regenerated nerves was found for empty tubes (11%) or tubes filled with NVR gel alone (30%).

Figure 1. Ultrasound evaluation of implanted chitosan tube. (A) 20‐day post‐tube implantation—no regeneration is visible inside the tube and (B) 90 days post‐tube implantation—nerve filaments are present inside the tube reconnecting nerve stumps (white arrows).

Functional analysis

In order to assess the functional recovery, we used the electrophysiological evaluation. We recorded SSEP at different time points (0, γ0, 90 and 150 days after injury) calculating SSEP peak to peak (P2P) amplitude (NR type) in which the operated limb was compared to the intact limb (Figure 2). Regarding the operated limb, all groups showed a decrease at

118

CHAPTER 1: Scientific Publications the first follow‐up (γ0 days), followed by an increase, except for the tube filled with NVR gel and NVR gel enriched with GDNF‐IONP, which recovered only after 90 days. 1 Interestingly, for what concerns the intact limb, the animal group treated with tube filled with NVR gel enriched with GDNF‐IONP also exhibited a decrease in amplitude following the operation and then recovered, while both the autologous nerve graft reconstruction and the tube filled with NVR gel enriched with free GDNF groups demonstrated an increase in P2P.

It was interesting to observe the comparison of the P2P among the different treatments in both the operated and intact limbs (Figure γ). At γ0 days after injury, PβP amplitude is significantly higher in the group of animals in which truncated nerves were repaired with tubes filled with NVR gel and GDNF‐IONP respect to those repaired with autologous nerve graft (p < 0.05).

The same treatment seemed to provoke significant decrease in P2P amplitude at the intact limb but only during the first time point investigated. The difference was observed in relation to the empty tube group (p < 0.05), and the tube filled with NVR gel enriched with free GDNF (p < 0.05).

Figure 2. Comparison of somatosensory evoked potentials (SSEP) peak‐to‐peak (P2P) amplitude among various treatments. Data were gathered from each limb separately (operated and intact) at four follow‐up periods: 0 (pre‐operatively), 30, 90, and 150 days. Values are presented as median ± mad. Statistical analysis: Wilcoxon signed rank test. *p < 0.05.

119

CHAPTER 1: Scientific Publications

Figure 3. Comparison among the various treatments. Data were gathered from each limb separately (operated and intact) at four follow‐up periods: 0 (pre‐operatively), 30, 90, and 150 days. Values are presented as median ± mad statistical analysis: Mann‐Whitney rank‐sum test. *p < 0.05

In order to normalize the results, we decided to subtract P2P measurement in the operated limb with the P2P measurement in the intact limb (Figure 4) in each of the four follow‐up time points (0, γ0, 90, and 150 days). The “0” value indicates similar amplitude in both intact and operated limbs. Since a dramatic decrease in amplitude is an indicator of neurological dysfunction, we assume that a large shift down from “0” is a marker for this pathology. As expected, at time point 0 (pre‐operation), all groups exhibited an amplitude difference very close to “0.” During the follow‐up, most treatments exhibited a decrease at the amplitude. The most robust and sustained decrease was found at the empty tube group. Surprisingly, the rat group that was treated with tube filled with NVR gel enriched with GDNF‐IONP had no decrease.

Morphological and morphometrical analysis

To determine how structural aspect of regenerated nerves is influenced by the choice of the dispositive used to repair the long gap, we performed morphological and morphometrical analysis on nerve regenerated inside the chitosan tube 5 months after the surgery. Beside only two animals showed regeneration in empty tube group, this group was excluded from morphometrical analysis. The morphometric quantification was carried out in the distal part of the regenerated nerve inside the tube (Figure 5A). Semi‐thin section of regenerate nerves was used for the analysis and compared to the distal part of regenerated nerves of autograft group, our positive control (Figure 5B–E).

120

CHAPTER 1: Scientific Publications

It is possible to notice the similarity among experimental groups in which nerve cross sections are organized in fascicles with small fibers. 1 The number of myelinated fibers is significantly higher in autograft group (p ≤ 0.001) that remains the gold standard of regeneration (Figure 6); nevertheless, it is important to notice that in the experimental group repaired with the tube enriched with NVR gel and GDNFIONP, the number of myelinated fibers is statistically significant higher (p ≤ 0.05) respect to the group repaired with the conduit functionalized with NVR gel alone (Figure 6).

Figure 4. P2P measurement normalized to the intact limb. The graph shows the data obtained through the subtraction of P2P measurement in the operated limb with P2P measurement in the intact limb. The “0” value indicates similarity between the two limbs. Values are presented as median ± mad.

The myelin thickness parameter, recorded for the group repaired using a tube with NVR gel alone, shows a statistically relevant reduction compared to autograft group (p ≤ 0.05). As regard the other parameters considered, it is interesting to note that there are no differences among the experimental groups.

DISCUSSION

A wide range of conduits with different internal design have been tested for reconstruction of peripheral nerves in animal models [28]. The efforts concern the attempt to replace the current gold standard, the autologous nerve graft, above all in the case of very serious injury and to avoid negative side‐effects [β, 4]. Chitosan conduits, among the absorbable forms, have been shown to be able of supporting the peripheral nerve regeneration with values of recovery fully comparable to the surgical gold standard technique, with regard to sciatic nerve lesions of 10 mm in the rat animal model [7, 11]. Despite this, in case of 121

CHAPTER 1: Scientific Publications nerve injuries with a considerable loss of substance, it is necessary to functionalize the conduit, to provide a filler, in order to prevent the collapse of the walls, and to provide a releasing system of neurotrophic molecules able to accelerate the regenerative processes [26]. In this study, we used the NVR gel, made of hyaluronic acid and laminin [13], as internal filler and for the release of the neurotrophic factor GDNF, which has been further enhanced by stabilizing its duration through a covalent conjugation with IONP [19].

Figure 5. Morphological analysis of regenerated nerves inside chitosan tube. (A) Representative photos of sciatic rat regenerated nerves harvested inside chitosan tube 5 months after the implantation. (B–E) The figure shows representative semi‐thin transversal nerve sections of regenerated sciatic nerves used for morphometrical analysis. For autograft group (B), the distal nerve is represented; for tube with NVR gel alone (C), tube with NVR gel enriched with GDNF‐IONP (D) and tube with NVR gel enriched with free GDNF (E); the pictures refer to the nerve found inside the chitosan tube. For all the groups, a good regeneration is visible.

We have previously demonstrated that the administration of this factor induces an early myelination in organotypic cultures of neonatal DRG [24]. The whole device has been proposed for the repair of a rat sciatic nerve lesion with severe loss of substance (15 mm). In this study, the nerve regeneration and the correct position of the device have been constantly monitored. Ultrasound observations and functional analysis have allowed us to carry out regular follow‐up (γ0, 90, and 150 days after surgery).

The ultrasound analysis carried out on the implants revealed that none of these have undergone a shift from the correct position by the moment of the surgery. The two

122

CHAPTER 1: Scientific Publications representative figures show the implanted tube at β0 and 90 days after surgery; at the experimental time 90 days, nerve regeneration appears through nerve fascicles. 1 Electrophysiological analysis on the five experimental groups demonstrates a nerve conductivity recovery (PβP amplitude) by the end of the follow‐up period (150 days), when the tube enriched either with NVR gel or NVR gel and GDNF‐IONP treatments displayed a complete recovery, as time 0, whether other treatments showed only a partial recovery.

Figure 6. Morphometrical analysis of regenerated nerve inside the chitosan tube. The analysis was performed on semithin transverse sections of the regenerated nerves 5 months after implantation. For autograft group (used as control), the distal part of the nerve was analyzed; for all the other groups, the values in the graphs are referred to the distal regenerated nerve found inside the chitosan tube. The graphs show the total number of number of myelinated fibers, the axon diameter, the fiber diameter, the myelin thickness, and the g‐ratio. In the filled tubes, there is less myelinated fibers respect to the autograft, but all the other parameters are comparable with the autograft. The two groups with GDNF factors, free or conjugated to iron‐oxide nanoparticles, are similar. All data are presented as mean + SD. Statistical analysis: One‐way ANOVA with post hoc Bonferroni Test. *p < 0.05; **p < 0.01; ***p < 0.001.

Normalizing the SSEP results of the operated limb (left) to the intact one (right), all groups showed a decrease after surgery followed by an increase, as expected, except for the group treated using a tube enriched with NVR gel and GDNF‐IONP that did not showed any decrease during the follow‐up period after surgery. These electrophysiological findings suggest that the treatment, in which a tube is enriched with NVR gel, is comparable to the autologous nerve graft treatment. 123

CHAPTER 1: Scientific Publications

The morphological aspect of regenerated nerves when, after 5 months, rats were sacrificed and tubes explanted revealed for all the cases analyzed a good nerve regeneration.

The morphometric quantification was referred on the distal part of the regenerated nerve inside the tube. Semi‐thin sections of nerves were used for the analysis and compared to the ones from the distal part of regenerated nerves of autograft group, our positive control. All the nerves revealed the same cross section structure: fascicles with small fibers, a typical nerve regeneration framework.

It was interesting to observe similar values for the measured parameters, such as axon diameter, myelin thickness, fiber diameter, and g‐ratio, demonstrating an equal good regeneration, because compared with the gold standard, the autograft. The only parameter in which the superiority of our positive control was found was the total number of the fibers, although the group represented by the tube enriched with NVR gel and GDNF‐IONP had a statistically significant greater value compared to the tubes enriched with only NVR gel.

CONCLUSIONS

In this work, we investigated the efficacy, in the repair of a nerve injury with a considerable loss of substance, of a conduit functionalized (i) with a factor that has been demonstrated to accelerate nerve regeneration, (ii) with a hydrogel, which has proven to be a good substrate for neurite outgrowth, (iii) and with the system of the IONP, able to stabilize the signaling of the factors. The analysis carried out has not shown a great improvement in nerve regeneration in animals treated with this functionalized device compared to the other devices investigated. IONPs are potentially a good candidate for nerve device enrichment; however, the best way to administer neurotrophic factors IONP in the tube needs further investigation, as well as the effects of their long time exposition.

ACKNOWLEDGEMENTS. This study was financially supported by the European Community’s Seventh Framework Program (FP7‐HEALTH‐β011), grant agreement n° 278612 (BIOHYBRID). Altakitin SA (Lisbon, Portugal) supplied medical grade chitosan for nerve guides, which were produced by Medovent GmbH (Mainz, Germany).

REFERENCES

[1] Ciaramitaro P, et al. Traumatic peripheral nerve injuries: Epidemiological findings, neuropathic pain and quality of life in 158 patients. Journal of the Peripheral Nervous System. 2010;15(2):120–127 [2] Battiston B, et al. Nerve repair by means of tubulization: Literature review and personal clinical experience comparing biological and synthetic conduits for sensory nerve repair. Microsurgery. 2005;25(4):258–267 [3] Gu XS, et al. Construction of tissue engineered nerve grafts and their application in peripheral nerve regeneration. Progress in Neurobiology. 2011;93(2):204–230

124

CHAPTER 1: Scientific Publications

[4] Konofaos P, Halen JPV. Nerve repair by means of tubulization: Past, present, future. Journal of Reconstructive Microsurgery. 2013;29(3):149–163 1 [5] Isaacs J, Browne T. Overcoming short gaps in peripheral nerve repair: Conduits and human acellular nerve allograft. Hand (N Y). 2014;9(2):131–137 [6] Gnavi S, et al. The use of chitosan‐based scaffolds to enhance regeneration in the nervous system. Tissue Engineering of the Peripheral Nerve: Biomaterials and Physical Therapy. 2013;109:1–62 [7] Shapira Y, et al. Comparison of results between chitosan hollow tube and autologous nerve graft in reconstruction of peripheral nerve defect: An experimental study. Microsurgery. 2016;36(8):664–671 [8] Simoes MJ, et al. In vitro and in vivo chitosan membranes testing for peripheral nerve reconstruction. Acta Medica Portuguesa. 2011;24(1):43–52 [9] Yuan Y, et al. The interaction of Schwann cells with chitosan membranes and fibers in vitro. Biomaterials. 2004;25(18):4273–4278 [10] Freier T, et al. Controlling cell adhesion and degradation of chitosan films by N‐acetylation. Biomaterials. β005;β6(β9):587β–5878 [11] Haastert‐Talini K, et al. Chitosan tubes of varying degrees of acetylation for bridging peripheral nerve defects. Biomaterials. 2013;34(38):9886–9904 [12] Stenberg L, et al. Nerve regeneration in chitosan conduits and in autologous nerve grafts in healthy and in type2 diabetic Goto‐Kakizaki rats. European Journal of Neuroscience. 2016;43(3):463–473 [1γ] Shahar A, Nevo Z, Rochkind S. Cross‐linked hyaluronic acid‐laminin gels and use thereof in cell culture and medical implants. U.S. patent. Editor. 2001: US [14] Boyd JG, Gordon T. Neurotrophic factors and their receptors in axonal regeneration and functional recovery after peripheral nerve injury. Molecular Neurobiology. 2003;27(3):277–323 [15] Hoke A, et al. Glial cell line‐derived neurotrophic factor alters axon Schwann cell units and promotes myelination in unmyelinated nerve fibers. Journal of Neuroscience. 2003;23(2):561–567 [16] Santos D, et al. Dose‐dependent differential effect of neurotrophic factors on in vitro and in vivo regeneration of motor and sensory neurons. Neural Plasticity. 2016;2016:4969523 [17] Gordon T. The role of neurotrophic factors in nerve regeneration. Neurosurgical Focus. 2009;26(2):E3 [18] Hoke A, et al. A decline in glial cell‐line‐derived neurotrophic factor expression is associated with impaired regeneration after long‐term Schwann cell denervation. Experimental Neurology. 2002;173(1):77–85 [19] Morano M, et al. Nanotechnology versus stem cell engineering: In vitro comparison of neurite inductive potentials. International Journal of Nanomedicine. 2014;9:5289–5306 [20] Chertok B, et al. Iron oxide nanoparticles as a drug delivery vehicle for MRI monitored magnetic targeting of brain tumors. Biomaterials. 2008;29(4):487–496 [21] Corot C, et al. Recent advances in iron oxide nanocrystal technology for medical imaging. Advanced Drug Delivery Reviews. 2006;58(14):1471–1504

125

CHAPTER 1: Scientific Publications

[ββ] Maier‐Hauff K, et al. Intracranial thermotherapy using magnetic nanoparticles combined with external beam radiotherapy: Results of a feasibility study on patients with glioblastoma multiforme. Journal of Neuro‐Oncology. β007;81(1):5γ–60. Epub 2006 Jun 14 [βγ] Ziv‐Polat O, Margel S, Shahar A. Application of iron oxide anoparticles in neuronal tissue engineering. Neural Regeneration Research. 2015;10(2):189–191. DOI: 10.4103/ 167γ‐5γ74.15βγ64 [β4] Ziv‐Polat O, et al. The role of neurotrophic factors conjugated to iron oxide nanoparticles in peripheral nerve regeneration: In vitro studies. BioMed Research International. 2014;2014:267808 [25] Geuna, S., et al., Appreciating the difference between design‐based and model‐based sampling strategies in quantitative morphology of the nervous system. J Comp Neurol. 2000;427(3):333–9 [β6] Meyer C, et al. Chitosan‐film enhanced chitosan nerve guides for long‐distance regeneration of peripheral nerves. Biomaterials. 2016;76:33–51 [β7] Meyer C, et al. Peripheral nerve regeneration through hydrogel‐enriched chitosan conduits containing engineered schwann cells for drug delivery. Cell Transplantation. 2016;25(1):159–182 [28] Tos P, et al. Future perspectives in nerve repair and regeneration. International Review of Neurobiology. 2013;109:165–192

126

CHAPTER 1: Discussion and Future Directions

1.4 DISCUSSION AND FUTURE DIRECTIONS 1 During my PhD course, I studied different approaches to improve nerve regeneration inside a nerve conduit. My work ranges from in vitro analysis to investigate the mechanism of action of molecules operating in the regeneration process and their manipulation, to the combination of these molecules with biomaterials, like nanoparticles or hydrogel, and the exploration of their in vivo applications. In the following three paragraphs I will discuss the results of my studies, highlighting the relationship with literature data and I will make some overall considerations regarding our current knowledge and future directions in the field of nerve regeneration.

NRG1/ErbB SYSTEM MANIPULATION FOR PERIPHERAL NERVE REPAIR

Innovative studied approaches to sustain peripheral nerve regeneration are based on the signalling manipulation of several growth factors, known for their pivotal role in the regenerative process. The in vivo interference with a signalling pathway may have two implications: on one hand it is useful to define the role of a molecule in a precise biological process step, on the other hand the manipulation permits to potentiate an endogenous signalling pathway to force and guide a biological process.

We focused our attention on the NRG1/ErbB system. NRG1 signalling has been shown to control survival and migration of SC, furthermore it is involved in axon remyelination regulating the crosstalk between SC and neuron196–198,200. Professor Geuna's group previously demonstrated that muscle fibres, used as filler in muscle-in-vein (MIV) technique, express high levels of soluble NRG1222. Authors hypothesized that soluble NRG1 release by muscle might exert, together with other factors, a positive effect on nerve regeneration inside the vein conduit, potentiating the endogenous NRG1 signalling after nerve injury. In order to demonstrate the validity of this assumption we decided to interfere with muscle released NRG1 through the recombinant ecto-ErbB4 protein. Ecto-ErbB4 is an extracellular fragment of the NRG1 receptor ErbB4, which is endogenously released by cells expressing the cleavable isoform (ErbB4 JMa) of this receptor and whose role is still unclear223. Since this fragment contains the binding site for NRG1, we hypothesized that ecto-ErbB4 might trap NRG1 and might be a useful tool to manipulate NRG1/ErbB system. We successfully set up a cellular system for ecto-ErbB4 production and release, and we demonstrated that in vitro ecto-ErbB4 interacts with NRG1.

To study the effect of recombinant ecto-ErbB4 in vitro we decided to use glial cells. We investigated the expression of the components of our target NRG1/ErbB system, together with common glial markers, in SC and OEC primary cells compared with the corresponding immortalized cell lines RT4-D6P2T and NOBEC, respectively. Our analysis revealed some similarities among cell lines and primary cultures, but evidenced differences regarding NRG1 and ErbB expressions. These data stress the importance to choose the optimal in vitro cell model for each experimental design. Since NOBEC cells express

127

CHAPTER 1: Discussion and Future Directions several NRG1 isoforms and possess an elevated migration ability, we selected this cell line to test the action of ecto-ErbB4. NOBEC cells show elevated motility after NRG1 treatment. The transient expression of ecto-ErbB4 in NOBEC increases cell migration, contrary to any expectations. We hypothesized that the higher migration rate might be due to the interaction between ecto-ErbB4 and NRG1 type III (the transmembrane isoform) expressed by NOBEC. In fact, it has been demonstrated that in neurons ecto-ErbB4 interaction with transmembrane NRG1 activates a back-signalling mediated by the intracellular domain of this NRG1 (NRG1-ICD)224. NRG1-ICD releasing, given by - secretase cleavage, and its subsequent translocation into nucleus, determines the transcription of genes involved in neuron survival224,225. In NOBEC cells we demonstrated that NRG1-ICD induces migration and its release block, obtained by treatment with - secretase enzyme inhibitor, determined a reduction in cell migration rate, especially for NOBEC expressing ecto-ErbB4. Moreover, we showed that the higher cell motility is not dependent upon NRG1-ICD nuclear fraction, but upon cytoplasmic NRG1-ICD. Since an interaction between NRG-ICD and LIM kinase (LIMK) has been shown at synapse level, where LIMK regulates actin reorganization226–228, we suggest that similar mechanisms might be involved in NOBEC migration. Despite the release of extracellular domain of ErbB receptors is known since long time, our knowledge about the function of these fragments is lacking. As mentioned before, Canetta and collaborators showed how ecto-ErbB4 regulates the NRG1 type III back signalling in neurons224. Other researchers studied the ability of these soluble fragments to inhibit ErbB2 or ErbB1. For example, soluble ErbB3 was shown to inhibit ErbB2/ErbB3 dimers in human breast cancer cell line MCF7229. We demonstrated that ecto-ErbB4 is able to interact with soluble NRG1 and with transmembrane NRG1 inducing NOBEC migration.

Our in vivo experiment showed that the expression of ecto-ErbB4 by muscle fibres, used for preparing the MIV conduit, has positive effect on fibre maturation, as indicated by increased fibres diameter and myelin thickness respect to control MIV. However, this morphology does not correspond to a functional improvement. Overall data showed that ecto-ErbB4 does not negatively affect the regenerative process. This unexpected effect can be explained by the interaction of ecto-ErbB4 with transmembrane NRG1 expressed by neurons, enhancing cell survival in the early phase of the regenerative process. Anyway, since a low level of soluble NRG1 has been associated with the promotion of the remyelination process, whereas opposite effects were detected at high dose230, the positive effects of ecto-ErbB4 might also be given by the sequestering of soluble NRG1 and the consequent reduction of its concentration up to remyelination permissive-range. Further investigation about ErbB extracellular fragment function will clarify their potential application in nerve repair.

The peripheral nerve regeneration is a complex and stepped process, given by the activation and modulation of several signalling pathways. The recent attention on the

128

CHAPTER 1: Discussion and Future Directions

NRG1/ErbB system as potential therapeutic target in the field of peripheral nerve disease and injury is highlighted by the elevated number of comprehensive reviews196,198,206,231. 1 The involvement of this system in nerve repair is now definitely clear. Several authors observed a strongly upregulation of NRG1 after nerve injury197,201,203 and it has been suggested that nerve damage triggers the release of soluble NRG1 by SC, promoting SC survival and dedifferentiation196,197. Moreover, the transmembrane NRG1-type III expressed by neurons has been identified as a positive regulator of the re-myelination process197,199,232. Since this pivotal role of NRG1 in nerve repair, it has been postulated that treatment with recombinant soluble NRG1 might improve peripheral nerve regeneration and the functional recovery. Several strategies have been used for NRG1 delivery, such as direct injection, release by biomaterials, transplantation of cells overexpressing NRG1 or injection of AAV for NRG1 expression (reviewed by Gambarotta and colleagues206). Most of the authors reported positive effects of the NRG1 treatment, indicated by increased axon number, number of myelinated fibres, SC migration and improvement in functional recovery206.

Indeed, the NRG1/ErbB system is complex and we have to consider the presence of several isoforms, with specific or overlapping roles, the activity of enzymes regulating NRG1 release and NRG1 back-signalling, and finally the intracellular back-signalling itself (see Chapter 2, paragraph 2.1.4). All these aspects need to be further investigated in order to potentiate NRG1 treatment for nerve traumatic injury or nerve disease. Various NRG1 isoforms have been used for in vivo delivery, however the precise role of each isoform remains to be elucidated. Recently, my colleagues demonstrated that several isoforms are differently regulated in nerve tissue after injury201. In particular, soluble NRG1 alpha and beta were found upregulated at 1 day after injury, suggesting that also the less studied NRG1 alpha isoform might have a role in regeneration. Curiously, it was believed that the difference between alpha and beta isoforms was only quantitative (with beta isoform 10-fold more active than alpha isoform), but recently new insights in NRG1 alpha actions suggested a qualitative difference: indeed, alpha and beta isoforms differentially influence the migration of malignant peripheral nerve sheath tumour cells233. It has been shown that NRG1 alpha regulates the expression of genes involved in acetylation, phosphorylation and alternative splicing in lymphoblastoid cells234. As suggested by our studies on ecto-ErbB4, also the NRG1 back-signalling can be considered for a therapeutic approach. Indeed, also NRG1 isoforms type a and type b (which differ for their intracellular fragment ICD) are regulated after nerve injury, suggesting that different back-signalling might be activated in nerve cells. Moreover, also the transmembrane isoform NRG1 type III might be the target of therapeutic approach to improve nerve repair, for its implication in the myelination process. Recombinant soluble NRG1 type III has been produced and used for in vivo delivery with good results206. Anyway, a suitable strategy may be the modulation of secretases responsible for NRG1 type III cleavage. It is known that NRG1-type III is cleaved by α-secretases (eg “tumour necrosis factor-α-

129

CHAPTER 1: Discussion and Future Directions converting enzyme”/TACE or “a disintegrin and metalloprotease” /ADAM) or by - secretases (”-site of amyloid precursorprotein-cleaving enzyme”/ BACE-1). The cleavage by -secretase results in stimulation of myelination or re-myelination235–237; whereas the α- secretase action has an inhibitory effect238. A modulation of this enzyme through inhibitor treatment might be effective to induce myelination in later phase of nerve regeneration. As observed for other growth factors, the inappropriate activation of the NRG1/ErbB system may be dangerous and may determine demyelinating neuropathies or peripheral tumours, as suggested by in vitro and in vivo experiments239–241. Thus, the future research in the field of peripheral nerve regeneration will be focused on the spatial and temporal control of growth factor release in nerve conduit, in order to better sustain the regenerative process and avoid side effects given by the activation of specific signalling pathways in unwanted step of the process or their over-activation.

As demonstrated by our study, muscle fibres might be a suitable carrier for factor delivery in the vein conduit. Muscle-in-vein conduit represents a valid alternative to autograft technique, as demonstrated by the optimal regeneration obtained for digital nerve repair, comparable to autograft36. The basal lamina of muscle fibres provides a good substrate for SC migration and organization in bands of Büngner. Despite successful results achieved and the economic convenience of MIV(all the components are easily available at nerve injury site242), the preparation of the conduit is strongly surgeon dependent and the application is still limited to short gap nerve injury (less than 3 cm gap)37. Gene therapy is emerging as a suitable approach to improve autograft or nerve conduit by the delivery of specific growth factors or molecules promoting SC survival and migration, axon elongation and re-myelination or preventing muscle atrophy30,31. Thus, MIV might be potentiated by gene delivery, given by direct in vivo injection of viral vector. In this regard, AAV-vectors are particularly attractive: they are well-tolerated and safe, permit long-term gene expression and are actually used in clinical trials for neural diseases31. Moreover, most of AAV serotype vectors showed a good tropism for animal and human skeletal muscle tissue243,244. However, it has to be taken into account that muscle tissue inside vein graft undergoes degradation and around one month after nerve repair most of muscle fibres are degenerated in the graft colonized by nerve cells245. Thus, the protein delivery by muscle fibres is transient, with a short time window, and it must be considered in the design of factor release.

IRON OXIDE NANOPARTICLES AND NVR GEL COMBINATION FOR THE DEVELOPMENT OF COMPOSITE NERVE DEVICE

The decay of neurotrophic factors (NTF) during prolonged nerve injury has been described as a crucial point in nerve regeneration failure over long distances22. Growth factors enhance nerve regeneration through several mechanisms and it has been postulated that exogenous administration of NTF could sustain the regeneration process for prolonged time246. It has been established that nerve devices successfully support nerve regeneration

130

CHAPTER 1: Discussion and Future Directions in short nerve gaps, whereas the reconstruction of nerve for critical size defects is not completely achieved with hollow conduits and further supports are needed. The low 1 concentration or absence of NTF in the inner part of nerve conduits has been indicated as one of the limiting factors of hollow tube performance247,248. Thus, the controlled release of NTF inside artificial nerve device has been proposed as a strategy to obtain potentiated conduit but, despite several pre clinical researches, it remains a challenge in regenerative nerve medicine.

As the main limitation of NTF in vivo release is given by their short live-span, which reduces their temporary window of action, we focused our attention on IONP-NTF. It was reported that thrombin conjugation with IONPs increases its half-life maintaining the biological activity of the compound249. We attempted to stabilize, through IONP conjugation, three well studied NTFs: NGF, GDNF and FGF-218kDa. First, we investigated the bioactivity of the three IONP-NTF produced, testing their effects on neuronal growth. We demonstrated that IONP-NTF stimulate neuronal outgrowth in adult DRG explants in a dose-dependent way. According with our data, recently M. Marcus and colleagues showed that NGF conjugated to iron oxide nanoparticles is able to promote PC12 differentiation into motor-neuron-like cells250. These data indicated that IONP-NTF are a suitable tool for neuronal growth.

Several strategies might be used for NTF release in nerve conduits. An easy way is the injection in the lumen of the conduit of an hydrogel which releases the selected growth factor. Hydrogels are a semi-gel biomaterial with high water content, which possesses versatile properties, like mechanical stability, permeability for small particles, compositional and structural similarity to ECM. Moreover, hydrogels can be easily mixed with NTF or cells for their controlled and prolonged release, ideal for medical application102. Among natural hydrogels we studied NVR gel: a transparent and biodegradable hydrogel, composed mainly by hyaluronic acid and laminin, both components of ECM. Previous studies reported that NVR gel supports survival, growth and differentiation of several cells, included neuronal cells251. We investigated whether the NTF release by NVR gel in vitro might be effective in inducing neurite growth. The hydrogel composition has the advantages to allow 3D culture, more resembling in vivo environment inside nerve device.

Our data show that NVR gel mixed with IONP-NTF supports neonatal DRG and PC12 cell survival. We observed neurite outgrowth in all studied conditions, demonstrating that growth factors are still accessible for neurons when mixed in 0.5 % NVR gel and the neurite growth is not physically prevented by hydrogel. At the same time, our collaborators demonstrated that organotypic DRG cultures, seeded in 0.3-0.5% NVR gel mixed with IONP-GDNF or free GDNF, show neurite sprouting and axon myelination252. Accordingly with other authors, our data have shown that hydrogels resembling ECM are a good candidate for factor release101. Hyaluronic acid (HA) and laminin, which mainly

131

CHAPTER 1: Discussion and Future Directions constituted NVR gel, are ECM components. HA modulates cell adhesion and migration253 and it is the prominent component of the bands of Büngner during Wallerian degeneration254. Several studies showed that HA and laminin have positive effects on peripheral nerve regeneration255–257.

In our experiments conjugated and non-conjugated factors showed comparable bioactivity, whereas other authors reported increased bioactivity of the factors after conjugation. For example, it has been published that IONP-NGF has higher ability to induce PC12 differentiation and faster TrkA receptor phosphorylation respect to free NGF250. Also Skaat and colleagues showed enhanced migration, growth, and differentiation of the NOM (Nasal Olfactory Mucosa) cells treated with FGF-2 conjugated with IONPs respect to un- conjugated FGF-2251. These differences might be related to the specific coating and the conjugation methods used, characteristics that need to be considered during IONP preparations. Moreover, also the factor release strategy might influence its activity. It can be possible that an increased action of IONP-NTF is not detectable for limitation of our in vitro model, which has already reached the highest effect with no further improvement of neurite outgrowth at the used factor concentration. This fact explains also the absence of synergistic effect when factors were administrated as a mix. Since neuron populations are varying and response to different growth factors, it is becoming clear that a multi-factor approach might be the most effective in promoting mixed nerve regeneration. Some authors demonstrated that cocktail of NTFs (GDNF and NGF, or GDNF, NGF and CNTF, or NGF and NT3) results in synergistic effect, inducing higher neurite growth and SC migration respect to single factor administration in vivo258–260. However, other authors report no synergistic effects for NGF and BDNF or NGF, GDNF and CNTF combination260,261, indicating that factor combination and single factor concentration need to be carefully investigated to obtain the maximum effect in nerve regeneration and that not all NTFs may work synergistically.

Noteworthy, we demonstrated the successful stability of conjugated GDNF and NRG1 by an indirect way, assessing their biological activity after long time IONP-NTF storage in cell medium at 37°C. At the same time, Ziv-Polat and colleagues analysed directly, through ELISA technique, the concentration of the three studied IONP-NTFs in cell medium confirming higher factor stability respect to free-NTFs, even if differences among factors were detected252. The increased stability makes IONP-NTF particularly adapt for a prolonged release inside the nerve device. Similar results have been obtained by Marcus and colleagues for IONP-NGF, whose concentration remains stable for at least 7 days of incubation at 37 °C250. This extended half life of conjugated factors or peptide has been observed also for other nanoparticle types, like silver262 or polymersome nanoparticles263, and can be explained by a less accessibility of growth factors to proteases.

Hydrogels are suitable for cell encapsulation102. Release of NTF in nerve device can be achieved also by cell transplantation. In our work we compared the two strategies of

132

CHAPTER 1: Discussion and Future Directions nanotechnology and cell engineering. Specifically, genetically modified bone marrow stem cells (BMSC) expressing GDNF or FGF-218KDa were studied. IONP-NTF did not 1 demonstrate significantly higher potency to induce sensory neurite outgrowth respect to NTF-BMSC. However, cell FGF-218KDa delivery was more effective to induce neurite elongation in PC12 cell culture respect to IONP- FGF-218KDa. NVR gel resulted suitable for cell encapsulation, coherently with what observed by Meyer and colleagues: authors demonstrated that engineered SC survive, migrate and assume bipolar morphology when seeded in NVR gel-filled chitosan tube in vitro264. Nevertheless, our in vitro studies did not provide evidence of the superiority of one factor release strategy respect to the other. Assunção-Silva and colleagues performed similar analysis: authors observed that tested hydrogel (modified gellan gum hydrogel) ability to support neurite growth was further potentiated by the addition of cells or IONP-GDNF; however, even if nanoparticles action seems stronger, there are no statistically relevant differences between the two release strategies265. Cell mediated factor release, through engineered cells or taking advantage of stem cell secretome, is a widely investigated and promising strategy; however, safeness, efficacy and clinical translation of such approach need to be further investigated.

IONP application in nerve regeneration field has numerous advantages. For example, IONP distribution can be spatially controlled using an external magnetic field266. This magnetic properties of IONP might be used to finely control IONP-NTF distribution or remotely guide them in a nerve device. Moreover, it has been proposed that IONP can induce and direct peripheral nerve regeneration267. Riccio et al. demonstrated that differentiated PC12 cells exposed to IONP-NGF and a magnetic field have oriented neurite growth, with axons aligned with the direction of magnetic force268. Moreover, Poggetti and colleagues reported that preliminary data about rat median nerve repair seem indicate that IONP are able to guide the migration of SC, which internalized nanoparticles, from proximal to distal nerve267. These effects are probably due to the mechanical force exerted by IONP. The importance of mechanical factors in nerve regeneration became clear only recently, with the emerging evidences that growth cones generate forces and the increasingly relevance of substrate stiffness in nerve device performance269–271.

However, the toxicity of IONPs need to be considered. We did not check specifically neurotoxicity, but the absence of differences between conjugated factors and free counterpart in neurite induction suggests no wide side effects of IONP at the used concentrations. Low doses of IONP were shown to be safe in several neuronal cell models and primary cells, with no effects on cell survival or growth in vitro268, even if Schwann cells resulted more sensitive to IONP cytotoxicity272. Accordingly with our data, IONP- NTF were found to be internalized by cells, with no damage to nucleus or cytoplasmatic organelles268. In the cytoplasm, IONP-NTF have been found in lysosomes, indicating the cell ability to destroy the nanoparticles, avoiding potentially dangerous accumulations268,272. IONP are generally classified as biocompatible; however, several

133

CHAPTER 1: Discussion and Future Directions types of IONP can be produced with different surface modifications (as coating and functional groups) that influence IONP cytotoxicity273,274. The risk related to IONP-NTF prolonged exposition need to be further investigated, since the collected information are very limited and conflicting, as well as the criteria to evaluate and define toxicity. Nevertheless, the nanoparticle concentration required for achieving an effect in nerve regeneration might be relatively low, due to high stability of IONP-NTF and their high bioactivity, thus avoiding possible side effects. To this regard, a recent in vivo study aimed to localize and assess the safety of IONP release after rat median nerve injury reported no side effects in animals: the direct injection of high doses of polyethylenimine (PEI)-coated iron oxide nanoparticles (33 µg) in a nerve device did not interfere with the regeneration process and did not trigger any inflammatory response107. Accordingly, our in vivo data did not evidence regeneration impairment in IONP-GDNF treated animals respect to un- conjugated GDNF release.

FACTOR RELEASE AND INTERNAL FILLER: NEW FRONTIERS IN NERVE DEVICE DESIGN

Since commercially available hollow nerve devices failed to match regeneration ability of nerve autograft in long nerve defects, a new generation of conduits has been developed. The concept of passive scaffold is now unseated by an active tube that, as well as providing a protected space for the regeneration process, guides and accelerates the axon elongation and tissue reconstruction. Intraluminal surface topography and biological cue release are some of the strategies investigated in pre-clinical studies. Among materials used to prepare nerve devices, chitosan possesses suitable characteristics for regenerative medicine application, such as biocompatibility, low toxicity, antimicrobial effect and low cost of production, and literature data provide significant evidences that chitosan tubes support nerve regeneration59,275–277. The degree of chitosan acetylation regulates its degradation and it has been shown that medium acetylation corresponds to a degradable chitosan suitable for nerve conduit57,60. Moreover, a chitosan-based tube (Reaxon® Nerve Guide, Medovent, Germany) recently obtained the CE-approval and the FDA-approval for clinical use. Thus, chitosan- based conduits are promising for future clinical applications in the repair of long nerve gaps.

We investigated the efficacy of chitosan tube enriched with NVR gel releasing IONP- GDNF to repair a long nerve defect (15 mm) in rat model. Overall results suggest that the NVR gel filler, with or without GDNF, can ameliorate the nerve regeneration compared to hollow tube. Nonetheless, best nerve regeneration was achieved with the autograft repair. In fact, the morphological analysis of the regenerated nerve inside the conduit showed that fibre diameter is comparable to autograft, while the total number of myelinated fibres is lower, especially with NVR gel alone as a filler. The release of factors improves the NVR gel action, as indicated by higher fibre number , myelin thickness and increased functional recovery. Coherently with our data, other authors showed that matrix alone achieved poor nerve regeneration, whereas growth factor enrichment gives better results respect to empty tube or matrix alone207,278,279. A positive effect of exogenous GDNF was also reported by

134

CHAPTER 1: Discussion and Future Directions other authors183,184,191,192 and, according with our results, GDNF has been shown to increase myelination in vitro193 and in vivo194. 1

Given our promising in vitro data about long term-stability of IONP-GDNF, a difference between the effects of conjugated and non-conjugated factor release in chitosan conduit was expected. Anyway, no differences were detected. We cannot exclude that a longer time-window of nerve regeneration observation might reveal a superior effect of IONP- GDNF respect to the free counterpart. Indeed, the hydrogel embedding might exert a protective action on growth factors, avoiding factor rapid loss due to exchange of fluid between external tissue and lumen, and thus increasing factor stability279. In fact, it has been demonstrated that the free-recombinant FGF-2 suspension in hydrogel permits its long term releasing and improves the regeneration of facial nerve after decompression, whereas a single injection of FGF-2 alone was not effective, probably due to factor rapid loss280. The hydrogel factor-protection effect could also explain the similar neurite induction observed for IONP-NTF and free-NTF in vitro. Despite hydrogels represent an easy and suitable way for factor delivery in nerve conduit, other strategies could be investigated to completely take advantage of IONP-NTF potentialities.

We demonstrated that NVR gel in vitro supports the neurite outgrowth in DRG culture, moreover our in vivo analysis showed that nerve regeneration occurs in nerve device enriched with this hydrogel. On the contrary, Meyer and colleagues reported that a chitosan tube enriched with SC encapsulated in NVR gel, used for the reconstruction of a 15-mm nerve defect in rat model, does not support the nerve regeneration. Authors suggested that NVR gel may act as a physical barrier for the regenerating axons. Curiously, the authors reported an increase of axon regeneration inside composite tube containing engineered SC releasing FGF-2, indicating that NVR gel supports cell survival and growth factor release, according with their and our in vitro experiments. The differences observed in the in vivo experiments might be due to the variability in NVR gel production and not standardized method for its injection in the tube during the surgery, which might give rise to different hydrogel density and gel swelling in the tube. Noteworthy, beside various studies demonstrated that hydrogel used as nerve conduit filler can improve nerve regeneration79,81–83,103,281,282, several authors reported negative results associated with hydrogel and matrix enrichment, which slow down the nerve regeneration207,279,283,284. In particular, the injection of a chitosan/glycerol-ß-phosphate disodium salt hydrogel into chitosan conduit resulted in nerve regeneration impediment, data not in agreement with previously in vitro evidences285. Despite hydrogels have been extensively studied for their suitable characteristics and give the unique opportunity to perform 3D in vitro cultures, it is now clear that for in vivo experiments more attention has to be pointed on hydrogel density, quantity of injected material and hydrogel stiffness.

The presence of hydrogel in the lumen gives the advantage of having a substrate on which cells can migrate and axon can grows; moreover, as we demonstrated, hydrogel can be

135

CHAPTER 1: Discussion and Future Directions used as scaffold for growth factor release. However, hydrogel-enriched conduits have not achieved results equal or better than autograft and do not offer a suitable clinical alternative to autograft approach. A fine study of mechanical qualities of the hydrogels may improve their ability to sustain nerve regeneration, avoiding unwanted effects as physical obstacle to axon growth. It is now clear that material stiffness is able to modulate tension exerted by neurites and growth cones, influencing axon elongation and altering gene expression84,85. Moreover, another crucial point is the ability to direct the axon elongation: current injectable hydrogels do not possess an organized template for directional growth control. It has been demonstrated that alignment of hydrogel components can be obtained by the addition of magnetical iron oxide particles or microgels containing IONP and the successive application of magnetic force286–288. It has been demonstrated that this component alignment can direct cell migration and neurite outgrowh288. Moreover, an agarose hydrogel containing a laminin gradient has been shown to direct the neurite outgrowth in DRG cultures289. Alternative to hydrogel are the sponges. Unlike hydrogels, which are a network of polymer chains, sponges are porous scaffolds and they can solve the physical obstruction given by hydrogel. Gelatin and collagene-based sponges were used for growth factor delivery in nerve device and they demonstrated a good nerve regeneration potentiality290,291. Moreover, other delivery systems have been investigated and need to be considered as alternative to hydrogels. Particularly attractive is the incorporation of growth factors in the conduit wall. Since usually a single polymeric material cannot fulfil two functions (structural and release function), this strategy entails the use of multicompartment design: the external part of the conduit gives the supportive structure and the inner layer represents the delivery system. Several solutions were adopted, as reviewed by Ramburrun et al.248. However, this system eliminated the internal support for cell migration given by hydrogel and may be unsuitable for long nerve defect reconstruction. To overcame this support lacking, multichannel guidance can be prepared, in which the lumen is divided in several smaller channels292,293. Further studies will be necessary to demonstrate the validity of these approaches and their clinical transposition.

Regarding growth factor delivery several questions remain open. One concerning is related to the ideal choice of the factor to release in nerve device. As already mentioned, several growth factors play a role in the regenerative process and all of them show a distinct expression pattern after injury3,129. The choice of the factor may be related to the type of repaired nerve, motor or mixed nerve, using the factors that better stimulate survival and growth of injured neurons. Despite single factor release can trigger nerve regeneration, better results might be achieved with a mix of factors acting both on sensory and motor neurons, as previously referenced. However, the correct mix of factors need to be further explored.

136

CHAPTER 1: Discussion and Future Directions

Anyway, which is the optimal quantity of exogenous factor to be released in the nerve device to obtain the maximum effect avoiding side effects? This question remains still 1 open. It is clear that uncontrolled initial burst release (large and immediate release) of factor has to be avoided, since it induces strong sprouting affecting the following axon elongation248; however, suboptimal doses of growth factors are not effective. Barras and colleagues demonstrated that rod embedding GDNF incorporated in nerve device wall produces a strong release of factors in the first few days that impair nerve regeneration. This unwanted effects was avoided by the pre-incubation of nerve conduit before implantation, which determines a constant and continuous factor release and an improvement of therapeutic performance294. In a study performed by Dodla and colleagues, a nerve guide was enriched with four sequential agarose hydrogels forming a gradient of NGF and laminin-1. Each hydrogel contains an higher amount of NGF respect to the previous layer, with low concentration at proximal stump and high concentration at the distal stump. Authors demonstrated that this conduit design supports nerve regeneration in long nerve gap (20 mm rat sciatic nerve); however, it fails to match nerve transfer outcome295. Thus, factor concentration is strictly correlated to the concept of spatial and temporal control of factor release. The distinctive expression patter of each growth factor after nerve injury suggests that a well controlled timing of growth factor release is fundamental to obtain optimal nerve regeneration and functional recovery. As previously described, a gradient of factors inside nerve conduit can be a solution; however, further attempts to engineer nerve conduit with controlled factor delivery need to be performed, specially to manage the contemporaneous release of various factors with different kinetics. An interesting approach is the affinity-based delivery system, which is based on factor immobilization on nerve conduit wall or internal matrix, avoiding its random diffusion. Great attention were pointed out on heparin-based affinity system, where heparin fragment is crosslinked to biomaterial and can trap growth factors that possess heparin-binding domain296–298. Working on heparin concentration it is possible to change growth factor concentration in the conduit247. These methods are associated with minor burst release and provide sustained factor release248. Multiple layer conduits are promising for controlled release of several growth factors: each layer of the tube wall can be used to encapsulate a different factor and the progressive layer degradation changes the factor availability inside the conduit248. Moreover, the engineered factors or microspheres containing factors can be useful to spatial control of single factor release in nerve conduit. The distribution of IONP- NTFs, as mentioned before, can be controlled by magnetic field266,268. Microspheres are spherical particles that can be incorporated in matrix and carry on different factors or factor mixes with variable concentration248. Until now, the spatial and temporal control of factor delivery remains a goal in nerve regenerative medicine; however, future technical progress will probably help to achieve an optimal release control.

Finally, the growing literature on mathematical models attempting to describe nerve regeneration processes and factor diffusion in nerve conduits will probably help to build an

137

CHAPTER 1: Discussion and Future Directions optimal delivery system299,300. In fact, mathematical models can accelerate the process of novel conduit design, providing new rationale for experimental decisions.

138

CHAPTER 1: References

1.5 REFERENCES 1 1. Siemionow, M. & Brzezicki, G. in International review of neurobiology 87, 141– 172 (2009). 2. Price, J. An immunohistochemical and quantitative examination of dorsal root ganglion neuronal subpopulations. J. Neurosci. 5, 2051–9 (1985). 3. Richner, M. et al. Peripheral Nerve Injury Modulates Neurotrophin Signaling in the Peripheral and Central Nervous System. Molecular Neurobiology (2014). doi:10.1007/s12035-014-8706-9 4. Lawson, S. N. The postnatal development of large light and small dark neurons in mouse dorsal root ganglia: a statistical analysis of cell numbers and size. J. Neurocytol. 8, 275–94 (1979). 5. Kashiba, H., Noguchi, K., Ueda, Y. & Senba, E. Coexpression of trk family members and low-affinity neurotrophin receptors in rat dorsal root ganglion neurons. Brain Res. Mol. Brain Res. 30, 158–64 (1995). 6. Wright, D. E. & Snider, W. D. Neurotrophin receptor mRNA expression defines distinct populations of neurons in rat dorsal root ganglia. J. Comp. Neurol. 351, 329–38 (1995). 7. Deumens, R. et al. Repairing injured peripheral nerves: Bridging the gap. Prog. Neurobiol. 92, 245–276 (2010). 8. Mackinnon, S. E. & Dellon, L. A. Nerve repair and nerve graft. Surg. Peripher. Nerve (1988). 9. Waller, A. 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. Philos. Trans. R. Soc. London 140, 423–429 (1850). 10. Coleman, M. Axon degeneration mechanisms: commonality amid diversity. Nat. Rev. Neurosci. 6, 889–898 (2005). 11. Ma, M. et al. Calpains mediate axonal cytoskeleton disintegration during Wallerian degeneration. Neurobiol. Dis. 56, 34–46 (2013). 12. Hall, S. The response to injury in the peripheral nervous system. J. Bone Joint Surg. Br. 87, 1309–19 (2005). 13. Madl, C. & Heilshorn, S. Matrix interactions modulate neurotrophin-mediated neurite outgrowth and pathfinding. Neural Regen. Res. 10, 514 (2015). 14. Höpker, V. H., Shewan, D., Tessier-Lavigne, M., Poo, M. & Holt, C. Growth-cone attraction to netrin-1 is converted to repulsion by laminin-1. Nature 401, 69–73 (1999). 15. Chen, P., Piao, X. & Bonaldo, P. Role of macrophages in Wallerian degeneration and axonal regeneration after peripheral nerve injury. Acta Neuropathol. 130, 605– 618 (2015). 16. Klein, D. & Martini, R. Myelin and macrophages in the PNS: An intimate relationship in trauma and disease. Brain Res. 1641, 130–138 (2016). 17. Gaudet, A. D., Popovich, P. G. & Ramer, M. S. Wallerian degeneration: gaining perspective on inflammatory events after peripheral nerve injury. J.

139

CHAPTER 1: References

Neuroinflammation 8, 110 (2011). 18. Rishal, I. & Fainzilber, M. Axon–soma communication in neuronal injury. Nat. Rev. Neurosci. 15, 32–42 (2013). 19. Hanz, S. & Fainzilber, M. Retrograde signaling in injured nerve--the axon reaction revisited. J. Neurochem. 99, 13–9 (2006). 20. Faroni, A., Mobasseri, S. A., Kingham, P. J. & Reid, A. J. Peripheral nerve regeneration: Experimental strategies and future perspectives. Adv. Drug Deliv. Rev. 82, 160–167 (2015). 21. Trojaborg, W. Rate of recovery in motor and sensory fibres of the radial nerve: clinical and electrophysiological aspects. J. Neurol. Neurosurg. Psychiatry 33, 625– 38 (1970). 22. Höke, A., Gordon, T., Zochodne, D. & Sulaiman, O. A decline in glial cell-line- derived neurotrophic factor expression is associated with impaired regeneration after long-term Schwann cell denervation. Exp. Neurol. 173, 77–85 (2002). 23. Kovačič, U., Sketelj, J. & Bajrović, F. F. in International review of neurobiology 87, 465–482 (2009). 24. Yasuda, H. et al. Diabetic neuropathy and nerve regeneration. Prog. Neurobiol. 69, 229–85 (2003). 25. Lee, S. K. & Wolfe, S. W. Peripheral nerve injury and repair. J. Am. Acad. Orthop. Surg. 8, 243–52 26. Gordon, T. Electrical Stimulation to Enhance Axon Regeneration After Peripheral Nerve Injuries in Animal Models and Humans. Neurotherapeutics 13, 295–310 (2016). 27. Al-Majed, A. A., Brushart, T. M. & Gordon, T. Electrical stimulation accelerates and increases expression of BDNF and trkB mRNA in regenerating rat femoral motoneurons. Eur. J. Neurosci. 12, 4381–4390 (2000). 28. Al-Majed, A. A., Siu, L. T. & Gordon, T. Electrical stimulation accelerates and enhances expression of regeneration-associated genes in regenerating rat femoral motoneurons. Cell. Mol. Neurobiol. 24, 379–402 (2004). 29. de Oliveira, R. F. erreira et al. Benefits of laser phototherapy on nerve repair. Lasers in medical science 30, 1395–1406 (2015). 30. Haastert, K. & Grothe, C. Gene therapy in peripheral nerve reconstruction approaches. Curr. Gene Ther. 7, 221–8 (2007). 31. de Winter, F. et al. Gene therapy approaches to enhance regeneration of the injured peripheral nerve. Eur. J. Pharmacol. 719, 145–52 (2013). 32. Busuttil, F., Rahim, A. A. & Phillips, J. B. Combining Gene and Stem Cell Therapy for Peripheral Nerve Tissue Engineering. Stem Cells Dev. 26, 231–238 (2017). 33. Safa, B. & Buncke, G. Autograft Substitutes: Conduits and Processed Nerve Allografts. Hand Clin. 32, 127–140 (2016). 34. Jones, S., Eisenberg, H. M. & Jia, X. Advances and future applications of augmented peripheral nerve regeneration. Int. J. Mol. Sci. 17, 1–17 (2016). 35. Gu, X., Ding, F., Yang, Y. & Liu, J. Construction of tissue engineered nerve grafts and their application in peripheral nerve regeneration. Prog. Neurobiol. 93, 204–230 (2011).

140

CHAPTER 1: References

36. Manoli, T., Schulz, L., Stahl, S., Jaminet, P. & Schaller, H.-E. Evaluation of sensory recovery after reconstruction of digital nerves of the hand using muscle-in-vein 1 conduits in comparison to nerve suture or nerve autografting. Microsurgery 34, 608–15 (2014). 37. Geuna, S. et al. Update on nerve repair by biological tubulization. J. Brachial Plex. Peripher. Nerve Inj. 9, 3 (2014). 38. Konofaos, P. & Ver Halen, J. Nerve Repair by Means of Tubulization: Past, Present, Future. J. Reconstr. Microsurg. 29, 149–164 (2013). 39. Myers, J. P., Santiago-Medina, M. & Gomez, T. M. Regulation of axonal outgrowth and pathfinding by integrin-ecm interactions. Dev. Neurobiol. 71, 901–923 (2011). 40. Pawelec, K. M., Best, S. M. & Cameron, R. E. Collagen: a network for regenerative medicine. J. Mater. Chem. B. Mater. Biol. Med. 4, 6484–6496 (2016). 41. Keilhoff, G., Stang, F., Wolf, G. & Fansa, H. Bio-compatibility of type I/III collagen matrix for peripheral nerve reconstruction. Biomaterials 24, 2779–87 (2003). 42. Farole, A. & Jamal, B. T. A Bioabsorbable Collagen Nerve Cuff (NeuraGen) for Repair of Lingual and Inferior Alveolar Nerve Injuries: A Case Series. J. Oral Maxillofac. Surg. 66, 2058–2062 (2008). 43. Kauppila, T., Jyväsjärvi, E., Huopaniemi, T., Hujanen, E. & Liesi, P. A Laminin Graft Replaces Neurorrhaphy in the Restorative Surgery of the Rat Sciatic Nerve. Exp. Neurol. 123, 181–191 (1993). 44. Whitworth, I. H., Brown, R. A., Doré, C., Green, C. J. & Terenghi, G. Orientated mats of fibronectin as a conduit material for use in peripheral nerve repair. J. Hand Surg. Br. 20, 429–36 (1995). 45. Tonda-Turo, C. et al. Comparative analysis of gelatin scaffolds crosslinked by genipin and silane coupling agent. Int. J. Biol. Macromol. 49, 700–706 (2011). 46. Chen, Y.-S. et al. An in vivo evaluation of a biodegradable genipin-cross-linked gelatin peripheral nerve guide conduit material. Biomaterials 26, 3911–8 (2005). 47. Chang, J.-Y. et al. Highly permeable genipin-cross-linked gelatin conduits enhance peripheral nerve regeneration. Artif. Organs 33, 1075–85 (2009). 48. Chen, P.-R., Chen, M.-H., Lin, F.-H. & Su, W.-Y. Release characteristics and bioactivity of gelatin-tricalcium phosphate membranes covalently immobilized with nerve growth factors. Biomaterials 26, 6579–87 (2005). 49. Vatankhah, E. et al. Artificial neural network for modeling the elastic modulus of electrospun polycaprolactone/gelatin scaffolds. Acta Biomater. 10, 709–21 (2014). 50. Binan, L. et al. Differentiation of neuronal stem cells into motor neurons using electrospun poly-L-lactic acid/gelatin scaffold. Biomaterials 35, 664–74 (2014). 51. Koudehi, M. F. et al. Preparation and evaluation of novel nano-bioglass/gelatin conduit for peripheral nerve regeneration. J. Mater. Sci. Mater. Med. 25, 363–73 (2014). 52. Madduri, S., Papaloïzos, M. & Gander, B. Trophically and topographically functionalized silk fibroin nerve conduits for guided peripheral nerve regeneration. Biomaterials 31, 2323–2334 (2010). 53. Sivak, W. N. et al. Delivery of chondroitinase ABC and glial cell line-derived

141

CHAPTER 1: References

neurotrophic factor from silk fibroin conduits enhances peripheral nerve regeneration. J. Tissue Eng. Regen. Med. 11, 733–742 (2017). 54. Allmeling, C., Jokuszies, A., Reimers, K., Kall, S. & Vogt, P. M. Use of spider silk fibres as an innovative material in a biocompatible artificial nerve conduit. J. Cell. Mol. Med. 10, 770–7 55. Kumar, M. N. V. R., Muzzarelli, R. A. A., Muzzarelli, C., Sashiwa, H. & Domb, A. J. Chitosan chemistry and pharmaceutical perspectives. Chem. Rev. 104, 6017–84 (2004). 56. Elieh-Ali-Komi, D. & Hamblin, M. R. Chitin and Chitosan: Production and Application of Versatile Biomedical Nanomaterials. Int. J. Adv. Res. 4, 411–427 (2016). 57. Freier, T., Koh, H. S., Kazazian, K. & Shoichet, M. S. Controlling cell adhesion and degradation of chitosan films by N-acetylation. Biomaterials 26, 5872–8 (2005). 58. Simões, M. J. et al. In vitro and in vivo chitosan membranes testing for peripheral nerve reconstruction. Acta Med. Port. 24, 43–52 59. Gnavi, S. et al. The use of chitosan-based scaffolds to enhance regeneration in the nervous system. Int. Rev. Neurobiol. 109, 1–62 (2013). 60. Tomihata, K. & Ikada, Y. In vitro and in vivo degradation of films of chitin and its deacetylated derivatives. Biomaterials 18, 567–75 (1997). 61. Fan, W. et al. Repairing a 35-mm-long median nerve defect with a chitosan/PGA artificial nerve graft in the human: a case study. Microsurgery 28, 238–42 (2008). 62. Gu, J. et al. Surgical repair of a 30 mm long human median nerve defect in the distal forearm by implantation of a chitosan-PGA nerve guidance conduit. J. Tissue Eng. Regen. Med. 6, 163–8 (2012). 63. Ali, I. & Jamil, N. Polyhydroxyalkanoates: Current applications in the medical field. Front. Biol. (Beijing). 11, 19–27 (2016). 64. Arslantunali, D., Dursun, T., Yucel, D., Hasirci, N. & Hasirci, V. Peripheral nerve conduits: Technology update. Med. Devices Evid. Res. 7, 405–424 (2014). 65. Danielsson, P. A., Adolfsson, L. & Dahlin, L. B. Different effect on axonal outgrowth of application of non-absorbable or absorbable tubes around a nerve repair. Scand. J. Plast. Reconstr. Surg. hand Surg. 35, 347–53 (2001). 66. Weber, R. A., Breidenbach, W. C., Brown, R. E., Jabaley, M. E. & Mass, D. P. A randomized prospective study of polyglycolic acid conduits for digital nerve reconstruction in humans. Plast. Reconstr. Surg. 106, 1036-45–8 (2000). 67. Navissano, M., Malan, F., Carnino, R. & Battiston, B. Neurotube® for facial nerve repair. Microsurgery 25, 268–271 (2005). 68. Rosson, G. D., Williams, E. H. & Dellon, A. L. Motor nerve regeneration across a conduit. Microsurgery 29, 107–114 (2009). 69. de Ruiter, G. C. W., Malessy, M. J. A., Yaszemski, M. J., Windebank, A. J. & Spinner, R. J. Designing ideal conduits for peripheral nerve repair. Neurosurg. Focus 26, E5 (2009). 70. Rahmany, M. B. & Van Dyke, M. Biomimetic approaches to modulate cellular adhesion in biomaterials: A review. Acta Biomater. 9, 5431–5437 (2013). 71. Lewandowska, K., Pergament, E., Sukenik, C. N. & Culp, L. A. Cell-type-specific

142

CHAPTER 1: References

adhesion mechanisms mediated by fibronectin adsorbed to chemically derivatized substrata. J. Biomed. Mater. Res. 26, 1343–63 (1992). 1 72. Nakayama, K. et al. Enhancement of peripheral nerve regeneration using bioabsorbable polymer tubes packed with fibrin gel. Artif. Organs 31, 500–8 (2007). 73. Toba, T. et al. Evaluation of peripheral nerve regeneration across an 80-mm gap using a polyglycolic acid (PGA)--collagen nerve conduit filled with laminin-soaked collagen sponge in dogs. Int. J. Artif. Organs 25, 230–7 (2002). 74. Zhang, B. G. X. et al. [REVIEWD] Recent advances in nerve tissue engineering. Int. J. Artif. Organs 37, 277–291 (2014). 75. Rodríguez, F. J., Gómez, N., Perego, G. & Navarro, X. Highly permeable polylactide-caprolactone nerve guides enhance peripheral nerve regeneration through long gaps. Biomaterials 20, 1489–1500 (1999). 76. Meek, M. F. & Den Dunnen, W. F. A. Porosity of the wall of a Neurolac nerve conduit hampers nerve regeneration. Microsurgery 29, 473–478 (2009). 77. Chang, C. J., Hsu, S. H., Yen, H. J., Chang, H. & Hsu, S. K. Effects of unidirectional permeability in asymmetric poly(DL-lactic acid-co-glycolic acid) conduits on peripheral nerve regeneration: An in vitro and in vivo study. J. Biomed. Mater. Res. - Part B Appl. Biomater. 83, 206–215 (2007). 78. Chang, C. J. & Hsu, S. H. The effect of high outflow permeability in asymmetric poly(DL-lactic acid-co-glycolic acid) conduits for peripheral nerve regeneration. Biomaterials 27, 1035–1042 (2006). 79. de Ruiter, G. C. W., Spinner, R. J., Yaszemski, M. J., Windebank, A. J. & Malessy, M. J. A. Nerve Tubes for Peripheral Nerve Repair. Neurosurg. Clin. N. Am. 20, 91– 105 (2009). 80. Verdú, E. et al. Alignment of collagen and laminin-containing gels improve nerve regeneration within silicone tubes. Restor. Neurol. Neurosci. 20, 169–79 (2002). 81. Apel, P. J. et al. Peripheral Nerve Regeneration Using a Keratin-Based Scaffold: Long-Term Functional and Histological Outcomes in a Mouse Model. J. Hand Surg. Am. 33, 1541–1547 (2008). 82. Kalbermatten, D. F. et al. Fibrin matrix for suspension of regenerative cells in an artificial nerve conduit. J. Plast. Reconstr. Aesthetic Surg. 61, 669–675 (2008). 83. Katayama, Y. et al. Coil-reinforced hydrogel tubes promote nerve regeneration equivalent to that of nerve autografts. Biomaterials 27, 505–518 (2006). 84. Man, A. J., Davis, H. E., Itoh, A., Leach, J. K. & Bannerman, P. Neurite outgrowth in fibrin gels is regulated by substrate stiffness. Tissue Eng Part A 17, 2931–2942 (2011). 85. Koch, D., Rosoff, W. J., Jiang, J., Geller, H. M. & Urbach, J. S. Strength in the periphery: Growth cone biomechanics and substrate rigidity response in peripheral and central nervous system neurons. Biophys. J. 102, 452–460 (2012). 86. Tonda-Turo, C. et al. Crosslinked gelatin nanofibres: preparation, characterisation and in vitro studies using glial-like cells. Mater. Sci. Eng. C. Mater. Biol. Appl. 33, 2723–35 (2013). 87. Zhang, K., Zheng, H., Liang, S. & Gao, C. Aligned PLLA nanofibrous scaffolds coated with graphene oxide for promoting neural cell growth. Acta Biomater. 37,

143

CHAPTER 1: References

131–142 (2016). 88. Tian, L., Prabhakaran, M. P. & Ramakrishna, S. Strategies for regeneration of components of nervous system: scaffolds, cells and biomolecules. Regen. Biomater. 2, 31–45 (2015). 89. Yu, Y., Meng, D., Man, L. & Wang, X. The Interactions Between Aligned Poly(L- Lactic Acid) Nanofibers and SH-SY5Y Cells In Vitro. J. Nanosci. Nanotechnol. 16, 6407–13 (2016). 90. Gnavi, S. et al. The Effect of Electrospun Gelatin Fibers Alignment on Schwann Cell and Axon Behavior and Organization in the Perspective of Artificial Nerve Design. Int. J. Mol. Sci. 16, 12925–12942 (2015). 91. Gnavi, S. et al. In vitro evaluation of gelatin and chitosan electrospun fibers as artificial guide in peripheral nerve repair: a comparative study. J. Tissue Eng. Regen. Med. (2016). doi:10.1002/term.2351 92. Neal, R. A. et al. Alignment and composition of laminin-polycaprolactone nanofiber blends enhance peripheral nerve regeneration. J. Biomed. Mater. Res. A 100, 406–23 (2012). 93. Huang, J. et al. Electrical regulation of Schwann cells using conductive polypyrrole/chitosan polymers. J. Biomed. Mater. Res. A 93(1), 164–74 (2010). 94. Kimura, K., Yanagida, Y., Haruyama, T., Kobatake, E. & Aizawa, M. Electrically induced neurite outgrowth of PC12 cells on the electrode surface. Med. Biol. Eng. Comput. 36, 493–8 (1998). 95. Gordon, T., Udina, E., Verge, V. M. K. & de Chaves, E. I. P. Brief electrical stimulation accelerates axon regeneration in the peripheral nervous system and promotes sensory axon regeneration in the central nervous system. Motor Control 13, 412–41 (2009). 96. Udina, E. et al. Electrical stimulation of intact peripheral sensory axons in rats promotes outgrowth of their central projections. Exp. Neurol. 210, 238–47 (2008). 97. Wang, X. et al. Evaluation of biocompatibility of polypyrrolein vitro andin vivo. J. Biomed. Mater. Res. 68A, 411–422 (2004). 98. Thompson, B. C. et al. Conducting polymers, dual neurotrophins and pulsed electrical stimulation — Dramatic effects on neurite outgrowth. J. Control. Release 141, 161–167 (2010). 99. Xu, H. et al. Conductive PPY/PDLLA conduit for peripheral nerve regeneration. Biomaterials 35, 225–235 (2014). 100. Huang, J. et al. Electrical stimulation to conductive scaffold promotes axonal regeneration and remyelination in a rat model of large nerve defect. PLoS One 7, e39526 (2012). 101. Katz, J. S. & Burdick, J. A. Hydrogel mediated delivery of trophic factors for neural repair. Wiley Interdisciplinary Reviews: Nanomedicine and Nanobiotechnology 1, 128–139 (2009). 102. Carballo-Molina, O. A. & Velasco, I. Hydrogels as scaffolds and delivery systems to enhance axonal regeneration after injuries. Front. Cell. Neurosci. 9, 13 (2015). 103. Yasui, G. et al. Neuregulin-1 released by biodegradable gelatin hydrogels can accelerate facial nerve regeneration and functional recovery of traumatic facial

144

CHAPTER 1: References

nerve palsy. J. Plast. Reconstr. Aesthet. Surg. 69, 328–34 (2016). 104. Zeng, W. et al. Incorporation of chitosan microspheres into collagen-chitosan 1 scaffolds for the controlled release of nerve growth factor. PLoS One 9, (2014). 105. Santos, D., Giudetti, G., Micera, S., Navarro, X. & del Valle, J. Focal release of neurotrophic factors by biodegradable microspheres enhance motor and sensory axonal regeneration in vitro and in vivo. 1636, 93–106 (2016). 106. Zhao, Q. et al. Polylactic-co-glycolic acid microspheres containing three neurotrophic factors promote sciatic nerve repair after injury. Neural Regen. Res. 10, 1491–7 (2015). 107. Giannaccini, M. et al. Magnetic Nanoparticles for Efficient Delivery of Growth Factors: Stimulation of Peripheral Nerve Regeneration. Adv. Healthc. Mater. 6, 1601429 (2017). 108. Faroni, A. & Magnaghi, V. The Neurosteroid Allopregnanolone Modulates Specific Functions in Central and Peripheral Glial Cells. Front. Endocrinol. (Lausanne). 2, 103 (2011). 109. Devesa, P. et al. Growth hormone treatment enhances the functional recovery of sciatic nerves after transection and repair. Muscle Nerve 45, 385–392 (2012). 110. Magnaghi, V., Procacci, P. & Tata, A. M. in International review of neurobiology 87, 295–315 (2009). 111. Jiang, L., Jones, S. & Jia, X. Stem Cell Transplantation for Peripheral Nerve Regeneration: Current Options and Opportunities. Int. J. Mol. Sci. 18, 94 (2017). 112. Fairbairn, N. G., Meppelink, A. M., Ng-Glazier, J., Randolph, M. A. & Winograd, J. M. Augmenting peripheral nerve regeneration using stem cells: A review of current opinion. World J. Stem Cells 7, 11–26 (2015). 113. Hoben, G. et al. Comparison of Acellular Nerve Allograft Modification with Schwann Cells or VEGF. HAND 10, 396–402 (2015). 114. Strauch, B. et al. Autologous Schwann cells drive regeneration through a 6-cm autogenous venous nerve conduit. J. Reconstr. Microsurg. 17, 589-95–7 (2001). 115. Gersey, Z. C. et al. First human experience with autologous Schwann cells to supplement sciatic nerve repair: report of 2 cases with long-term follow-up. Neurosurg. Focus 42, E2 (2017). 116. Ikeda, M. et al. Acceleration of peripheral nerve regeneration using nerve conduits in combination with induced pluripotent stem cell technology and a basic fibroblast growth factor drug delivery system. J. Biomed. Mater. Res. Part A 102, 1370–1378 (2014). 117. Nakagawa, M. et al. Generation of induced pluripotent stem cells without Myc from mouse and human fibroblasts. Nat. Biotechnol. 26, 101–106 (2007). 118. Okita, K., Ichisaka, T. & Yamanaka, S. Generation of germline-competent induced pluripotent stem cells. Nature 448, 313–317 (2007). 119. Cui, L. et al. Transplantation of Embryonic Stem Cells Improves Nerve Repair and Functional Recovery After Severe Sciatic Nerve Axotomy in Rats. Stem Cells 26, 1356–1365 (2008). 120. Grimoldi, N. et al. Stem cell salvage of injured peripheral nerve. Cell Transplant. 24, 213–22 (2015).

145

CHAPTER 1: References

121. Chen, C.-J. et al. Transplantation of bone marrow stromal cells for peripheral nerve repair. Exp. Neurol. 204, 443–453 (2007). 122. Mohammadi, R. et al. The use of undifferentiated bone marrow stromal cells for sciatic nerve regeneration in rats. Int. J. Oral Maxillofac. Surg. 41, 650–656 (2012). 123. Hundepool, C. A., Nijhuis, T. H. J., Mohseny, B., Selles, R. W. & Hovius, S. E. R. The effect of stem cells in bridging peripheral nerve defects: a meta-analysis. J. Neurosurg. 121, 195–209 (2014). 124. Haghighat, A., Mohammadi, R. & Amini, K. Transplantation of undifferentiated bone-marrow stromal cells into a vein graft accelerates sciatic nerve regeneration in streptozotocin induced diabetic rats. Curr. Neurovasc. Res. 11, 230–41 (2014). 125. Xue, C. et al. Joint use of a chitosan/PLGA scaffold and MSCs to bridge an extra large gap in dog sciatic nerve. Neurorehabil. Neural Repair 26, 96–106 (2012). 126. Oliveira, J. T., Mostacada, K., de Lima, S. & Martinez, A. M. B. in International review of neurobiology 108, 59–77 (2013). 127. Kumar, A., Kumar, V., Rattan, V., Jha, V. & Bhattacharyya, S. Secretome Cues Modulate the Neurogenic Potential of Bone Marrow and Dental Stem Cells. Mol. Neurobiol. (2016). doi:10.1007/s12035-016-0011-3 128. Conover, J. C. & Yancopoulos, G. D. Neurotrophin regulation of the developing nervous system: analyses of knockout mice. Rev. Neurosci. 8, 13–27 129. Boyd, J. G. & Gordon, T. Neurotrophic Factors and Their Receptors in Axonal Regeneration and Functional Recovery After Peripheral Nerve Injury. Mol. Neurobiol. 27, 277–324 (2003). 130. Richner, M. et al. Peripheral nerve injury modulates neurotrophin signaling in the peripheral and central nervous system. Mol Neurobiol. 50, 945–70 (2014). 131. Gordon, T. The role of neurotrophic factors in nerve regeneration. Neurosurgical focus 26, E3 (2009). 132. Yano, H. & Chao, M. V. Neurotrophin receptor structure and interactions. Pharm. Acta Helv. 74, 253–60 (2000). 133. Heumann, R. et al. Differential regulation of mRNA encoding nerve growth factor and its receptor in rat sciatic nerve during development, degeneration, and regeneration: role of macrophages. Proc. Natl. Acad. Sci. U. S. A. 84, 8735–9 (1987). 134. Heumann, R., Korsching, S., Bandtlow, C. & Thoenen, H. Changes of nerve growth factor synthesis in nonneuronal cells in response to sciatic nerve transection. J. Cell Biol. 104, 1623–31 (1987). 135. Funakoshi, H. et al. Differential expression of mRNAs for neurotrophins and their receptors after axotomy of the sciatic nerve. J. Cell Biol. 123, 455–65 (1993). 136. Escandón, E. et al. Regulation of neurotrophin receptor expression during embryonic and postnatal development. J. Neurosci. 14, 2054–68 (1994). 137. Rich, K. M., Alexander, T. D., Pryor, J. C. & Hollowell, J. P. Nerve growth factor enhances regeneration through silicone chambers. Exp. Neurol. 105, 162–70 (1989). 138. Wang, Z. et al. Improved peripheral nerve regeneration with sustained release nerve growth factor microspheres in small gap tubulization. Am. J. Transl. Res. 6, 413–21 (2014).

146

CHAPTER 1: References

139. Hsieh, S.-C. et al. Comparison between two different methods of immobilizing NGF in poly(DL-lactic acid-co-glycolic acid) conduit for peripheral nerve regeneration by 1 EDC/NHS/MES and genipin. J. Biomed. Mater. Res. Part A 99A, 576–585 (2011). 140. Chang, C.-J. Effects of nerve growth factor from genipin-crosslinked gelatin in polycaprolactone conduit on peripheral nerve regeneration- In vitro and in vivo. J. Biomed. Mater. Res. Part A 91A, 586–596 (2009). 141. Kobayashi, N. R., Bedard, A. M., Hincke, M. T. & Tetzlaff, W. Increased expression of BDNF and trkB mRNA in rat facial motoneurons after axotomy. Eur. J. Neurosci. 8, 1018–29 (1996). 142. Meyer, M., Matsuoka, I., Wetmore, C., Olson, L. & Thoenen, H. Enhanced synthesis of brain-derived neurotrophic factor in the lesioned peripheral nerve: different mechanisms are responsible for the regulation of BDNF and NGF mRNA. J. Cell Biol. 119, 45–54 (1992). 143. Sendtner, M., Holtmann, B., Kolbeck, R., Thoenen, H. & Barde, Y. A. Brain- derived neurotrophic factor prevents the death of motoneurons in newborn rats after nerve section. Nature 360, 757–9 (1992). 144. Koliatsos, V. E., Clatterbuck, R. E., Winslow, J. W., Cayouette, M. H. & Price, D. L. Evidence that brain-derived neurotrophic factor is a trophic factor for motor neurons in vivo. Neuron 10, 359–67 (1993). 145. Sendtner, M., Holtmann, B. & Hughes, R. A. The response of motoneurons to neurotrophins. Neurochem. Res. 21, 831–41 (1996). 146. Atwal, J. K., Massie, B., Miller, F. D. & Kaplan, D. R. The TrkB-Shc site signals neuronal survival and local axon growth via MEK and P13-kinase. Neuron 27, 265– 77 (2000). 147. Dolcet, X., Egea, J., Soler, R. M., Martin-Zanca, D. & Comella, J. X. Activation of phosphatidylinositol 3-kinase, but not extracellular-regulated kinases, is necessary to mediate brain-derived neurotrophic factor-induced motoneuron survival. J. Neurochem. 73, 521–31 (1999). 148. Zhang, J. Y., Luo, X. G., Xian, C. J., Liu, Z. H. & Zhou, X. F. Endogenous BDNF is required for myelination and regeneration of injured sciatic nerve in rodents. Eur. J. Neurosci. 12, 4171–80 (2000). 149. Kishino, A., Ishige, Y., Tatsuno, T., Nakayama, C. & Noguchi, H. BDNF Prevents and Reverses Adult Rat Motor Neuron Degeneration and Induces Axonal Outgrowth. Exp. Neurol. 144, 273–286 (1997). 150. Boyd, J. G. & Gordon, T. A dose-dependent facilitation and inhibition of peripheral nerve regeneration by brain-derived neurotrophic factor. Eur. J. Neurosci. 15, 613– 26 (2002). 151. Midha, R., Munro, C. A., Dalton, P. D., Tator, C. H. & Shoichet, M. S. Growth factor enhancement of peripheral nerve regeneration through a novel synthetic hydrogel tube. J. Neurosurg. 99, 555–565 (2003). 152. Wang, B. et al. Functional regeneration of the transected recurrent laryngeal nerve using a collagen scaffold loaded with laminin and laminin-binding BDNF and GDNF. Sci. Rep. 6, 32292 (2016). 153. Santos, D. et al. Preferential Enhancement of Sensory and Motor Axon

147

CHAPTER 1: References

Regeneration by Combining Extracellular Matrix Components with Neurotrophic Factors. Int. J. Mol. Sci. 18, 65 (2016). 154. Hohn, A., Leibrock, J., Bailey, K. & Barde, Y. A. Identification and characterization of a novel member of the nerve growth factor/brain-derived neurotrophic factor family. Nature 344, 339–41 (1990). 155. Kohn, J., Aloyz, R. S., Toma, J. G., Haak-Frendscho, M. & Miller, F. D. Functionally antagonistic interactions between the TrkA and p75 neurotrophin receptors regulate sympathetic neuron growth and target innervation. J. Neurosci. 19, 5393–408 (1999). 156. Yao, B. et al. Phosphorylation of Raf by ceramide-activated protein kinase. Nature 378, 307–310 (1995). 157. Tessarollo, L., Vogel, K. S., Palko, M. E., Reid, S. W. & Parada, L. F. Targeted mutation in the neurotrophin-3 gene results in loss of muscle sensory neurons. Proc. Natl. Acad. Sci. U. S. A. 91, 11844–8 (1994). 158. Klein, R. et al. Disruption of the neurotrophin-3 receptor gene trkC eliminates la muscle afferents and results in abnormal movements. Nature 368, 249–51 (1994). 159. Fariñas, I., Jones, K. R., Backus, C., Wang, X. Y. & Reichardt, L. F. Severe sensory and sympathetic deficits in mice lacking neurotrophin-3. Nature 369, 658–61 (1994). 160. Yamauchi, J., Chan, J. R. & Shooter, E. M. Neurotrophin 3 activation of TrkC induces Schwann cell migration through the c-Jun N-terminal kinase pathway. Proc. Natl. Acad. Sci. U. S. A. 100, 14421–6 (2003). 161. Chan, J. R., Cosgaya, J. M., Wu, Y. J. & Shooter, E. M. Neurotrophins are key mediators of the myelination program in the peripheral nervous system. Proc. Natl. Acad. Sci. U. S. A. 98, 14661–8 (2001). 162. Sterne, G. D., Brown, R. A., Green, C. J. & Terenghi, G. Neurotrophin-3 delivered locally via fibronectin mats enhances peripheral nerve regeneration. Eur. J. Neurosci. 9, 1388–1396 (1997). 163. Sterne, G. D., Coulton, G. R., Brown, R. A., Green, C. J. & Terenghi, G. Neurotrophin-3-enhanced nerve regeneration selectively improves recovery of muscle fibers expressing myosin heavy chains 2b. J. Cell Biol. 139, 709–715 (1997). 164. Simon, M., Terenghi, G., Green, C. J. & Coulton, G. R. Differential effects of NT-3 on reinnervation of the fast extensor digitorum longus (EDL) and the slow soleus muscle of rat. Eur. J. Neurosci. 12, 863–71 (2000). 165. Fan, J. et al. Neural regrowth induced by PLGA nerve conduits and neurotrophin-3 in rats with complete spinal cord transection. J. Biomed. Mater. Res. B. Appl. Biomater. 97, 271–7 (2011). 166. Tang, S. et al. The Effects of Controlled Release of Neurotrophin-3 from PCLA Scaffolds on the Survival and Neuronal Differentiation of Transplanted Neural Stem Cells in a Rat Spinal Cord Injury Model. PLoS One 9, e107517 (2014). 167. Yin, Q., Kemp, G. J., Yu, L. G., Wagstaff, S. C. & Frostick, S. P. Neurotrophin-4 delivered by fibrin glue promotes peripheral nerve regeneration. Muscle Nerve 24, 345–51 (2001).

148

CHAPTER 1: References

168. Simon, M., Porter, R., Brown, R., Coulton, G. R. & Terenghi, G. Effect of NT-4 and BDNF delivery to damaged sciatic nerves on phenotypic recovery of fast and slow 1 muscles fibres. Eur. J. Neurosci. 18, 2460–6 (2003). 169. English, A. W., Cucoranu, D., Mulligan, A., Rodriguez, J. A. & Sabatier, M. J. Neurotrophin-4/5 is implicated in the enhancement of axon regeneration produced by treadmill training following peripheral nerve injury. Eur. J. Neurosci. 33, 2265– 71 (2011). 170. Grothe, C., Meisinger, C. & Claus, P. In vivo expression and localization of the fibroblast growth factor system in the intact and lesioned rat peripheral nerve and spinal ganglia. J. Comp. Neurol. 434, 342–57 (2001). 171. Grothe, C. & Nikkhah, G. The role of basic fibroblast growth factor in peripheral nerve regeneration. Anat. Embryol. (Berl). 204, 171–7 (2001). 172. Davis, J. B. & Stroobant, P. Platelet-derived growth factors and fibroblast growth factors are mitogens for rat Schwann cells. J. Cell Biol. 110, 1353–60 (1990). 173. Haastert, K., Lipokatic´, E., Fischer, M., Timmer, M. & Grothe, C. Differentially promoted peripheral nerve regeneration by grafted Schwann cells over-expressing different FGF-2 isoforms. Neurobiol. Dis. 21, 138–153 (2006). 174. Cui, Y. et al. Collagen scaffolds modified with CNTF and bFGF promote facial nerve regeneration in minipigs. Biomaterials 35, 7819–27 (2014). 175. Nikolaev, S. I., Gallyamov, A. R., Mamin, G. V. & Chelyshev, Y. A. Poly(ε- caprolactone) nerve conduit and local delivery of vegf and fgf2 genes stimulate neuroregeneration. Bull Exp Biol Med. 157, (2014). 176. Lin, L. F., Doherty, D. H., Lile, J. D., Bektesh, S. & Collins, F. GDNF: a glial cell line-derived neurotrophic factor for midbrain dopaminergic neurons. Science 260, 1130–2 (1993). 177. Lin, L. F., Zhang, T. J., Collins, F. & Armes, L. G. Purification and initial characterization of rat B49 glial cell line-derived neurotrophic factor. J. Neurochem. 63, 758–768 (1994). 178. Oppenheim, R. W. et al. Developing motor neurons rescued from programmed and axotomy-induced cell death by GDNF. Nature 378, 287–292 (1995). 179. Henderson, C. E. et al. GDNF: a potent survival factor for motoneurons present in peripheral nerve and muscle. Science 266, 1062–4 (1994). 180. Trupp, M., Belluardo, N., Funakoshi, H. & Ibáñez, C. F. Complementary and overlapping expression of glial cell line-derived neurotrophic factor (GDNF), c-ret proto-oncogene, and GDNF receptor-alpha indicates multiple mechanisms of trophic actions in the adult rat CNS. J. Neurosci. 17, 3554–3567 (1997). 181. Naveilhan, P., ElShamy, W. M. & Ernfors, P. Differential regulation of mRNAs for GDNF and its receptors Ret and GDNFR alpha after sciatic nerve lesion in the mouse. Eur. J. Neurosci. 9, 1450–60 (1997). 182. Burazin, T. C. D. & Gundlach, A. L. Up-regulation of GDNFR-α and c-ret mRNA in facial motor neurons following facial nerve injury in the rat. Mol. Brain Res. 55, 331–336 (1998). 183. Yan, Q., Matheson, C. & Lopez, O. T. In vivo neurotrophic effects of GDNF on neonatal and adult facial motor neurons. Nature 373, 341–4 (1995).

149

CHAPTER 1: References

184. Tajdaran, K., Gordon, T., Wood, M. D., Shoichet, M. S. & Borschel, G. H. A glial cell line-derived neurotrophic factor delivery system enhances nerve regeneration across acellular nerve allografts. Acta Biomater. 29, 62–70 (2016). 185. Tannemaat, M. R. et al. Differential effects of lentiviral vector-mediated overexpression of nerve growth factor and glial cell line-derived neurotrophic factor on regenerating sensory and motor axons in the transected peripheral nerve. Eur. J. Neurosci. 28, 1467–1479 (2008). 186. Santosa, K. B. et al. Nerve allografts supplemented with schwann cells overexpressing glial-cell-line-derived neurotrophic factor. Muscle and Nerve 47, 213–223 (2013). 187. Watabe, K. et al. Rescue of lesioned adult rat spinal motoneurons by adenoviral gene transfer of glial cell line-derived neurotrophic factor. J. Neurosci. Res. 60, 511–9 (2000). 188. Sakamoto, T. et al. Adenoviral vector-mediated GDNF gene transfer prevents death of adult facial motoneurons. Neuroreport 11, 1857–1860 (2000). 189. Eggers, R. et al. Lentiviral vector-mediated gradients of GDNF in the injured peripheral nerve: effects on nerve coil formation, Schwann cell maturation and myelination. PLoS One 8, e71076 (2013). 190. Shakhbazau, A. et al. Doxycycline-regulated GDNF expression promotes axonal regeneration and functional recovery in transected peripheral nerve. J. Control. Release 172, 841–851 (2013). 191. Kokai, L. E., Ghaznavi, A. M. & Marra, K. G. Incorporation of double-walled microspheres into polymer nerve guides for the sustained delivery of glial cell line- derived neurotrophic factor. Biomaterials 31, 2313–2322 (2010). 192. Wood, M. D. et al. Functional motor recovery is improved due to local placement of GDNF microspheres after delayed nerve repair. 110, (2013). 193. Zhang, L. et al. GDNF-enhanced axonal regeneration and myelination following spinal cord injury is mediated by primary effects on neurons. Glia 57, 1178–91 (2009). 194. Chen, B. K. et al. GDNF Schwann cells in hydrogel scaffolds promote regional axon regeneration, remyelination and functional improvement after spinal cord transection in rats. J. Tissue Eng. Regen. Med. (2017). doi:10.1002/term.2431 195. Höke, A. et al. Glial cell line-derived neurotrophic factor alters axon schwann cell units and promotes myelination in unmyelinated nerve fibers. J. Neurosci. 23, 561– 567 (2003). 196. Fricker, F. R. & Bennett, D. L. The role of neuregulin-1 in the response to nerve injury. Future Neurol. 6, 809–822 (2011). 197. Stassart, R. M. et al. A role for Schwann cell-derived neuregulin-1 in remyelination. Nat. Neurosci. 16, 48–54 (2013). 198. Birchmeier, C. & Nave, K.-A. Neuregulin-1, a key axonal signal that drives Schwann cell growth and differentiation. Glia 56, 1491–1497 (2008). 199. Fricker, F. R. et al. Axonally derived neuregulin-1 is required for remyelination and regeneration after nerve injury in adulthood. 31, (2011). 200. Nave, K.-A. & Salzer, J. L. Axonal regulation of myelination by neuregulin 1. Curr.

150

CHAPTER 1: References

Opin. Neurobiol. 16, 492–500 (2006). 201. Ronchi, G. et al. The Neuregulin1/ErbB system is selectively regulated during 1 peripheral nerve degeneration and regeneration. Eur. J. Neurosci. 43, 351–364 (2016). 202. Ronchi, G. et al. ErbB2 receptor over-expression improves post-traumatic peripheral nerve regeneration in adult mice. PLoS One 8, e56282 (2013). 203. Carroll, S. L., Miller, M. L., Frohnert, P. W., Kim, S. S. & Corbett, J. A. Expression of neuregulins and their putative receptors, ErbB2 and ErbB3, is induced during Wallerian degeneration. J. Neurosci. 17, 1642–59 (1997). 204. Kwon, Y. K. et al. Activation of ErbB2 during wallerian degeneration of sciatic nerve. J. Neurosci. 17, 8293–9 (1997). 205. Kerber, G., Streif, R., Schwaiger, F.-W., Kreutzberg, G. W. & Hager, G. Neuregulin-1 Isoforms Are Differentially Expressed in the Intact and Regenerating Adult Rat Nervous System. J. Mol. Neurosci. 21, 149–166 (2003). 206. Gambarotta, G., Fregnan, F., Gnavi, S. & Perroteau, I. Neuregulin 1 role in schwann cell regulation and potential applications to promote peripheral nerve regeneration. Int. Rev. Neurobiol. 108, 223–256 (2013). 207. Mohanna, P.-N., Terenghi, G. & Wiberg, M. Composite PHB-GGF conduit for long nerve gap repair: a long-term evaluation. Scand. J. Plast. Reconstr. Surg. Hand Surg. 39, 129–37 (2005). 208. Joung, I. et al. Secretion of EGF-like domain of heregulinβ promotes axonal growth and functional recovery of injured sciatic nerve. Mol. Cells 30, 477–84 (2010). 209. Suganuma, S. et al. Uncultured adipose-derived regenerative cells promote peripheral nerve regeneration. J. Orthop. Sci. 18, 145–151 (2013). 210. Zhang, J., Zhao, F., Wu, G., Li, Y. & Jin, X. Functional and histological improvement of the injured spinal cord following transplantation of Schwann cells transfected with NRG1 gene. Anat. Rec. (Hoboken). 293, 1933–46 (2010). 211. Carmeliet, P. & Tessier-Lavigne, M. Common mechanisms of nerve and blood vessel wiring. Nature 436, 193–200 (2005). 212. Ferrara, N. & Davis-Smyth, T. The biology of vascular endothelial growth factor. Endocr. Rev. 19, 61–9 (1997). 213. Nowacka, M. M. & Obuchowicz, E. Vascular endothelial growth factor (VEGF) and its role in the central nervous system: a new element in the neurotrophic hypothesis of antidepressant drug action. Neuropeptides 46, 1–10 (2012). 214. Hulse, R. P. et al. Regulation of alternative VEGF-A mRNA splicing is a therapeutic target for analgesia. Neurobiol. Dis. 71, 245–59 (2014). 215. Ruiz de Almodovar, C., Lambrechts, D., Mazzone, M. & Carmeliet, P. Role and Therapeutic Potential of VEGF in the Nervous System. Physiol. Rev. 89, 607–648 (2009). 216. Lindholm, T. et al. Expression of Semaphorins, Neuropilins, VEGF, and Tenascins in Rat and Human Primary Sensory Neurons after a Dorsal Root Injury. Front. Neurol. 8, 49 (2017). 217. Sköld, M. et al. Induction of VEGF and VEGF receptors in the spinal cord after mechanical spinal injury and prostaglandin administration. Eur. J. Neurosci. 12,

151

CHAPTER 1: References

3675–86 (2000). 218. Tang, J. et al. Expression of VEGF and neural repair after alprostadil treatment in a rat model of sciatic nerve crush injury. Neurol. India 57, (2009). 219. Hobson, M. I., Green, C. J. & Terenghi, G. VEGF enhances intraneural angiogenesis and improves nerve regeneration after axotomy. J. Anat. 197 Pt 4, 591–605 (2000). 220. Kempton, L. B. et al. Assessment of axonal growth into collagen nerve guides containing VEGF-transfected stem cells in matrigel. Anat. Rec. 292, 214–224 (2009). 221. Pereira Lopes, F. R. et al. Enhancement of sciatic nerve regeneration after vascular endothelial growth factor (VEGF) gene therapy. Neuropathol. Appl. Neurobiol. 37, 600–612 (2011). 222. Nicolino, S. et al. Expression of alpha2a-2b neuregulin-1 is associated with early peripheral nerve repair along muscle-enriched tubes. Neuroreport 14, 1541–5 (2003). 223. Veikkolainen, V. et al. Function of ERBB4 is determined by alternative splicing. Cell Cycle 10, 2647–2657 (2011). 224. Canetta, S. E., Luca, E., Pertot, E., Role, L. W. & Talmage, D. A. Type III Nrg1 back signaling enhances functional TRPV1 along sensory axons contributing to basal and inflammatory thermal pain sensation. PLoS One 6, e25108 (2011). 225. Bao, J., Wolpowitz, D., Role, L. W. & Talmage, D. A. Back signaling by the Nrg-1 intracellular domain. J. Cell Biol. 161, 1133–1141 (2003). 226. Wang, J. Y., Frenzel, K. E., Wen, D. & Falls, D. L. Transmembrane neuregulins interact with LIM kinase 1, a cytoplasmic protein kinase implicated in development of visuospatial cognition. J. Biol. Chem. 273, 20525–34 (1998). 227. Wang, J. Y., Wigston, D. J., Rees, H. D., Levey, A. I. & Falls, D. L. LIM kinase 1 accumulates in presynaptic terminals during synapse maturation. J. Comp. Neurol. 416, 319–34 (2000). 228. Aizawa, H. et al. Phosphorylation of cofilin by LIM-kinase is necessary for semaphorin 3A-induced growth cone collapse. Nat. Neurosci. 4, 367–73 (2001). 229. Takahashi, M. et al. Suppression of heregulin β signaling by the single N-glycan deletion mutant of soluble ErbB3 protein. J. Biol. Chem. 288, 32910–32921 (2013). 230. Syed, N. et al. Soluble neuregulin-1 has bifunctional, concentration-dependent effects on Schwann cell myelination. J. Neurosci. 30, 6122–31 (2010). 231. Newbern, J. & Birchmeier, C. Nrg1/ErbB signaling networks in Schwann cell development and myelination. Semin. Cell Dev. Biol. 21, 922–928 (2010). 232. Taveggia, C. et al. Neuregulin-1 Type III Determines the Ensheathment Fate of Axons. Neuron 47, 681–694 (2005). 233. Eckert, J. M., Byer, S. J., Clodfelder-Miller, B. J. & Carroll, S. L. Neuregulin-1 beta and neuregulin-1 alpha differentially affect the migration and invasion of malignant peripheral nerve sheath tumor cells. Glia 57, 1501–20 (2009). 234. Ghahramani Seno, M. M., Gwadry, F. G., Hu, P. & Scherer, S. W. Neuregulin 1- alpha regulates phosphorylation, acetylation, and alternative splicing in lymphoblastoid cells. Genome 56, 619–625 (2013).

152

CHAPTER 1: References

235. Hu, X. et al. Bace1 modulates myelination in the central and peripheral nervous system. Nat. Neurosci. 9, 1520–5 (2006). 1 236. Willem, M. et al. Control of Peripheral Nerve Myelination by the -Secretase BACE1. Science (80-. ). 314, 664–666 (2006). 237. Hu, X. et al. Genetic deletion of BACE1 in mice affects remyelination of sciatic nerves. FASEB J. 22, 2970–2980 (2008). 238. La Marca, R. et al. TACE (ADAM17) inhibits Schwann cell myelination. Nat. Neurosci. 14, 857–65 (2011). 239. Huijbregts, R. P. H., Roth, K. A., Schmidt, R. E. & Carroll, S. L. Hypertrophic neuropathies and malignant peripheral nerve sheath tumors in transgenic mice overexpressing glial growth factor beta3 in myelinating Schwann cells. J. Neurosci. 23, 7269–80 (2003). 240. Frohnert, P. W., Stonecypher, M. S. & Carroll, S. L. Constitutive activation of the neuregulin-1/ErbB receptor signaling pathway is essential for the proliferation of a neoplastic Schwann cell line. Glia 43, 104–118 (2003). 241. Fallon, K. B., Havlioglu, N., Hamilton, L. H., Cheng, T. P. H. & Carroll, S. L. Constitutive activation of the neuregulin-1/erbB signaling pathway promotes the proliferation of a human peripheral neuroepithelioma cell line. J. Neurooncol. 66, 273–84 (2004). 242. Marcoccio, I., Ignazio, M., Vigasio, A. & Adolfo, V. Muscle-in-vein nerve guide for secondary reconstruction in digital nerve lesions. J. Hand Surg. Am. 35, 1418–26 (2010). 243. Lisowski, L., Tay, S. S. & Alexander, I. E. Adeno-associated virus serotypes for gene therapeutics. Curr. Opin. Pharmacol. 24, 59–67 (2015). 244. Srivastava, A. In vivo tissue-tropism of adeno-associated viral vectors. Curr. Opin. Virol. 21, 75–80 (2016). 245. Tos, P. et al. Comparison of fresh and predegenerated muscle-vein-combined guides for the repair of rat median nerve. Microsurgery 27, 48–55 (2007). 246. Markus, A., Patel, T. D. & Snider, W. D. Neurotrophic factors and axonal growth. Curr. Opin. Neurobiol. 12, 523–31 (2002). 247. Willerth, S. M. & Sakiyama-Elbert, S. E. Approaches to neural tissue engineering using scaffolds for drug delivery. Adv. Drug Deliv. Rev. 59, 325–338 (2007). 248. Ramburrun, P. et al. A review of bioactive release from nerve conduits as a neurotherapeutic strategy for neuronal growth in peripheral nerve injury. Biomed Res. Int. 2014, 132350 (2014). 249. Ziv-Polat, O., Topaz, M., Brosh, T. & Margel, S. Enhancement of incisional wound healing by thrombin conjugated iron oxide nanoparticles. Biomaterials 31, 741–747 (2010). 250. Marcus, M., Skaat, H., Alon, N., Margel, S. & Shefi, O. NGF-conjugated iron oxide nanoparticles promote differentiation and outgrowth of PC12 cells. Nanoscale 7, 1058–1066 (2015). 251. Skaat, H., Ziv-Polat, O., Shahar, A. & Margel, S. Enhancement of the growth and differentiation of nasal olfactory mucosa cells by the conjugation of growth factors to functional nanoparticles. Bioconjug. Chem. 22, 2600–10 (2011).

153

CHAPTER 1: References

252. Ziv-Polat, O. et al. The role of neurotrophic factors conjugated to iron oxide nanoparticles in peripheral nerve regeneration: in vitro studies. Biomed Res. Int. 2014, 267808 (2014). 253. Leach, J. B., Bivens, K. A., Collins, C. N. & Schmidt, C. E. Development of photocrosslinkable hyaluronic acid-polyethylene glycol-peptide composite hydrogels for soft tissue engineering. J. Biomed. Mater. Res. 70A, 74–82 (2004). 254. Tona, A., Perides, G., Rahemtulla, F. & Dahl, D. Extracellular matrix in regenerating rat sciatic nerve: a comparative study on the localization of laminin, hyaluronic acid, and chondroitin sulfate proteoglycans, including versican. J. Histochem. Cytochem. 41, 593–9 (1993). 255. Adanali, G., Verdi, M., Tuncel, A., Erdogan, B. & Kargi, E. Effects of hyaluronic acid-carboxymethylcellulose membrane on extraneural adhesion formation and peripheral nerve regeneration. J. Reconstr. Microsurg. 19, 29–36 (2003). 256. Zor, F. et al. Effect of VEGF gene therapy and hyaluronic acid film sheath on peripheral nerve regeneration. Microsurgery 34, 209–16 (2014). 257. Wei, Y. T. et al. Hyaluronic acid hydrogels with IKVAV peptides for tissue repair and axonal regeneration in an injured rat brain. Biomed. Mater. 2, S142-6 (2007). 258. Madduri, S., di Summa, P., Papaloïzos, M., Kalbermatten, D. & Gander, B. Effect of controlled co-delivery of synergistic neurotrophic factors on early nerve regeneration in rats. Biomaterials 31, 8402–9 (2010). 259. Chen, J., Chu, Y. F., Chen, J. M. & Li, B. C. Synergistic effects of NGF, CNTF and GDNF on functional recovery following sciatic nerve injury in rats. Adv. Med. Sci. 55, 32–42 (2010). 260. Cao, X. & Shoichet, M. S. Investigating the synergistic effect of combined neurotrophic factor concentration gradients to guide axonal growth. Neuroscience 122, 381–9 (2003). 261. Deister, C. & Schmidt, C. E. Optimizing neurotrophic factor combinations for neurite outgrowth. J. Neural Eng. 3, 172–9 (2006). 262. Pal, I. et al. Enhanced stability and activity of an antimicrobial peptide in conjugation with silver nanoparticle. J. Colloid Interface Sci. 483, 385–93 (2016). 263. Soumen, R. et al. Activation of TrkB receptors by NGFβ mimetic peptide conjugated polymersome nanoparticles. Nanomedicine 8, 271–4 (2012). 264. Meyer, C. et al. Peripheral Nerve Regeneration Through Hydrogel-Enriched Chitosan Conduits Containing Engineered Schwann Cells for Drug Delivery. Cell Transplant. 25, 159–82 (2016). 265. Assunção-Silva, R. C. et al. Induction of neurite outgrowth in 3D hydrogel-based environments. Biomed. Mater. 10, 51001 (2015). 266. Kim, J. A. et al. Enhancement of neurite outgrowth in PC12 cells by iron oxide nanoparticles. Biomaterials 32, 2871–7 (2011). 267. Poggetti, A. et al. How to Direct the Neuronal Growth Process in Peripheral Nerve Regeneration: Future Strategies for Nanosurfaces Scaffold and Magnetic Nanoparticles. Surg. Technol. Int. 30, (2017). 268. Riggio, C. et al. The orientation of the neuronal growth process can be directed via magnetic nanoparticles under an applied magnetic field. Nanomedicine 10, 1549–58

154

CHAPTER 1: References

(2014). 269. Gu, Y. et al. The influence of substrate stiffness on the behavior and functions of 1 Schwann cells in culture. Biomaterials 33, 6672–81 (2012). 270. López-Fagundo, C., Bar-Kochba, E., Livi, L. L., Hoffman-Kim, D. & Franck, C. Three-dimensional traction forces of Schwann cells on compliant substrates. J. R. Soc. Interface 11, 20140247 (2014). 271. Rosso, G., Liashkovich, I., Young, P., Röhr, D. & Shahin, V. Schwann cells and neurite outgrowth from embryonic dorsal root ganglions are highly mechanosensitive. Nanomedicine 13, 493–501 (2017). 272. Pinkernelle, J., Calatayud, P., Goya, G. F., Fansa, H. & Keilhoff, G. Magnetic nanoparticles in primary neural cell cultures are mainly taken up by microglia. BMC Neurosci. 13, 32 (2012). 273. Rivet, C. J., Yuan, Y., Borca-Tasciuc, D.-A. & Gilbert, R. J. Altering iron oxide nanoparticle surface properties induce cortical neuron cytotoxicity. Chem. Res. Toxicol. 25, 153–61 (2012). 274. Nedyalkova, M. et al. Iron oxide nanoparticles - In vivo/in vitro biomedical applications and in silico studies. Adv. Colloid Interface Sci. (2017). doi:10.1016/j.cis.2017.05.003 275. Patel, M. et al. Video-gait analysis of functional recovery of nerve repaired with chitosan nerve guides. Tissue Eng. 12, 3189–99 (2006). 276. Ao, Q. et al. The regeneration of transected sciatic nerves of adult rats using chitosan nerve conduits seeded with bone marrow stromal cell-derived Schwann cells. Biomaterials 32, 787–96 (2011). 277. Zheng, L. & Cui, H.-F. Enhancement of nerve regeneration along a chitosan conduit combined with bone marrow mesenchymal stem cells. J. Mater. Sci. Mater. Med. 23, 2291–302 (2012). 278. Hart, A. M., Wiberg, M. & Terenghi, G. Exogenous leukaemia inhibitory factor enhances nerve regeneration after late secondary repair using a bioartificial nerve conduit. Br. J. Plast. Surg. 56, 444–450 (2003). 279. Pfister, L. A., Papaloïzos, M., Merkle, H. P. & Gander, B. Nerve conduits and growth factor delivery in peripheral nerve repair. J. Peripher. Nerv. Syst. 12, 65–82 (2007). 280. Hato, N. et al. Facial nerve decompression surgery using bFGF-impregnated biodegradable gelatin hydrogel in patients with Bell palsy. Otolaryngol. Head. Neck Surg. 146, 641–6 (2012). 281. Pace, L. A. et al. A Human Hair Keratin Hydrogel Scaffold Enhances Median Nerve Regeneration in Nonhuman Primates: An Electrophysiological and Histological Study. Tissue Eng. Part A 131115063659000 (2013). doi:10.1089/ten.tea.2013.0084 282. Guo, Q. et al. Chitosan conduits filled with simvastatin/Pluronic F-127 hydrogel promote peripheral nerve regeneration in rats. J. Biomed. Mater. Res. B. Appl. Biomater. (2017). doi:10.1002/jbm.b.33890 283. Valentini, R. F., Aebischer, P., Winn, S. R. & Galletti, P. M. Collagen- and laminin- containing gels impede peripheral nerve regeneration through semipermeable nerve guidance channels. Exp. Neurol. 98, 350–6 (1987).

155

CHAPTER 1: References

284. Terris, D. J. et al. Functional recovery following nerve injury and repair by silicon tubulization: comparison of laminin-fibronectin, dialyzed plasma, collagen gel, and phosphate buffered solution. Auris. Nasus. Larynx 26, 117–22 (1999). 285. Zheng, L., Ao, Q., Han, H., Zhang, X. & Gong, Y. Evaluation of the chitosan/glycerol-beta-phosphate disodium salt hydrogel application in peripheral nerve regeneration. Biomed. Mater. 5, 35003 (2010). 286. Zhang, S. et al. A self-assembly pathway to aligned monodomain gels. Nat. Mater. 9, 594–601 (2010). 287. Berns, E. J. et al. Aligned neurite outgrowth and directed cell migration in self- assembled monodomain gels. Biomaterials 35, 185–95 (2014). 288. Rose, J. C. et al. Nerve Cells Decide to Orient inside an Injectable Hydrogel with Minimal Structural Guidance. Nano Lett. 17, 3782–3791 (2017). 289. Dodla, M. C. & Bellamkonda, R. V. Anisotropic scaffolds facilitate enhanced neurite extension in vitro. J. Biomed. Mater. Res. A 78, 213–21 (2006). 290. Yao, C.-C., Yao, P., Wu, H. & Zha, Z.-G. Absorbable collagen sponge combined with recombinant human basic fibroblast growth factor promotes nerve regeneration in rat sciatic nerve. J. Mater. Sci. Mater. Med. 18, 1969–72 (2007). 291. Li, G. et al. Graft of the NT-3 persistent delivery gelatin sponge scaffold promotes axon regeneration, attenuates inflammation, and induces cell migration in rat and canine with spinal cord injury. Biomaterials 83, 233–48 (2016). 292. Yao, L., Billiar, K. L., Windebank, A. J. & Pandit, A. Multichanneled Collagen Conduits for Peripheral Nerve Regeneration: Design, Fabrication, and Characterization. Tissue Eng. Part C Methods 16, 1585–1596 (2010). 293. Yao, L. et al. Controlling dispersion of axonal regeneration using a multichannel collagen nerve conduit. Biomaterials 31, 5789–97 (2010). 294. Barras, F. M., Pasche, P., Bouche, N., Aebischer, P. & Zurn, A. D. Glial cell line- derived neurotrophic factor released by synthetic guidance channels promotes facial nerve regeneration in the rat. J. Neurosci. Res. 70, 746–55 (2002). 295. Dodla, M. C. & Bellamkonda, R. V. Differences between the effect of anisotropic and isotropic laminin and nerve growth factor presenting scaffolds on nerve regeneration across long peripheral nerve gaps. Biomaterials 29, 33–46 (2008). 296. Sakiyama-Elbert, S. E. & Hubbell, J. A. Controlled release of nerve growth factor from a heparin-containing fibrin-based cell ingrowth matrix. J. Control. Release 69, 149–58 (2000). 297. Wood, M. D. et al. Affinity-based release of glial-derived neurotrophic factor from fibrin matrices enhances sciatic nerve regeneration. Acta Biomater. 5, 959–68 (2009). 298. Li, G., Xiao, Q., Zhang, L., Zhao, Y. & Yang, Y. Nerve growth factor loaded heparin/chitosan scaffolds for accelerating peripheral nerve regeneration. Carbohydr. Polym. 171, 39–49 (2017). 299. Lin, C.-C. & Metters, A. T. Hydrogels in controlled release formulations: network design and mathematical modeling. Adv. Drug Deliv. Rev. 58, 1379–408 (2006). 300. Coy, R. H., Evans, O. R., Phillips, J. B. & Shipley, R. J. An integrated theoretical- experimental approach to accelerate translational tissue engineering. J. Tissue Eng.

156

CHAPTER 1: References

Regen. Med. (2016). doi:10.1002/term.2346 1

157

CHAPTER 2 SKELETAL MUSCLE DENERVATION

CONTENT

2.1 INTRODUCTION AND SCIENTIFIC BACKGROUND ...... 161 2.1.1 SKELETAL MUSCLE ANATOMY AT A GLANCE ...... 161 2.1.2 SKELETAL MUSCLE DENERVATION: cellular and molecular events ...... 162 2.1.3 SKELETAL MUSCLE REINNERVATION and CURRENT THERAPIES FOR DENERVATED MUSCLE ...... 164 2.1.4 NEUREGULIN 1 AND ErbB RECEPTORS ...... 166 2.1.4.1 Neuregulin and ErbB receptors structure ...... 167 2.1.4.2 NRG1/ErbB system in skeletal muscle ...... 169 2.2 AIM OF THE RESEARCH ...... 173 2.3 SCIENTIFIC PUBLICATION ...... 175 Modulation of the Neuregulin1/ErbB system after skeletal muscle denervation and reinnervation ...... 175 2.4 DISCUSSION AND FUTURE DIRECTIONS ...... 193

2.5 REFERENCES ...... 197

159

CHAPTER 2: Introduction and Scientific Background

2.1 INTRODUCTION AND SCIENTIFIC BACKGROUND

Skeletal tissue is one of the largest organs of our body, which permits all movements, as well as represents a storage for glucose and proteins. The denervation of muscle occurs in several clinical setting, such as traumatic nerve injury, diabetic neuropathy, alcoholic neuropathy, Charcot-Marie-Tooth disease, amyotrophic lateral sclerosis (ALS), spinal muscular atrophy and some virus infection like polio. The denervation determines muscle 2 atrophy, a condition of great muscle mass reduction and complete loss of muscle function. After a traumatic nerve injury the immediate nerve repair induces a complete rescue in muscle tissue with usually good functional recovery. However when nerve repair is delayed, for chirurgical reasons, post-denervation changes that occurs in muscle tissue aggravates the patient clinical scenario.

Despite its great impact on patients quality life, the muscle atrophy due to nerve traumatic injury is poorly studied and, actually, pharmacological therapies are not available. Understanding the mechanisms guiding the atrophy and the reinnervation process will help us to develop novel therapeutic approaches that act both on nerve and target muscle. These therapies will permit to preserve muscle mass and to achieve correct muscle reinnervation, guaranteeing a complete functional recovery.

2.1.1 SKELETAL MUSCLE ANATOMY AT A GLANCE

Skeletal muscle is a major constituent of the human body, representing the 40% of the total body weight in adult, and it is responsible of voluntary movements and force generation that allows us to interact with the environment. Interesting skeletal muscle is a highly adaptive tissue, able to fast responding to metabolic demand, movement or other body demand. For example skeletal muscle undergo hypertrophy in response to resistance exercise or atrophy in response to fasting, disuse and denervation. The smaller unit of skeletal tissue is given by myocytes or myofibres: a syncytia, with long and cylindrical shape, containing several post-mitotic nuclei localized peripherally close to cell membrane, called sarcolemma. Myonuclei are post-mitotic, thus they cannot divide, and are associated to a constant cytosplasmic area called muscle fibre-myonuclear domain, whose existence is controversial1,2. Myocytes contain the contractive structure given by myofibrils, which consist of repeated units called sarcomeres. Sarcomeres have a diameter of 1 μm and 2-3 μm length, and are composed by myosin and actin filaments held together by structural proteins. Myofibres have been classified into two types accordingly to ATPase activity of myosin, expressing isoforms of myosin heavy chain (MHC) and metabolism: type I fibres (slow twitch) and type II fibres (fast twitch)3. While Type I fibres are thinner, containing many mitochondria and are appropriate for sustained contraction, Type II fibres look thicker, containing less mitochondria, and are glycolytic fibres able to fast contraction, more powerful but less

161

CHAPTER 2: Introduction and Scientific Background sustained. The percentage of each fibre types vary among muscles and can changes under various conditions, such as ageing4. Around sarcolemma, an external basal lamina is visible. Between these two layers are located satellite cells, a stem cell expressing early markers of the myogenic lineage, like Pax7, M-cadherin and Myf5. Satellite cells are usually in a quiescent state and undergo activation only under special conditions like tissue injury and regeneration5. The other component of skeletal muscle is given by the connective tissue, which organization resembles the ones observed in peripheral nerve. In fact, each single myofibre is enveloped by a first connective layer called endomysium. Groups of 20-80 myocytes are surrounded by the perimysium, forming the muscle fascicule. Finally the outmost connective layer is the epimysium which rounded muscle fascicule. Connective tissue is crossed by blood vessels and numerous capillaries that supply oxygen and nutrients to each cellular component. Skeletal muscle contractile activity is controlled by the CNS through action potentials sending from motor neurons to muscle fibres. Each myofibre is connected with a motorneuron through a single neuromuscular junction (NMJ), a highly specialized chemical synapse called also endplate. The nerve terminals are wrapped by specialized glial cells named terminal Schwann cells (SC). A single motorneuron usually contacts more myofibres and all fibres innervated by the same neuron are defined motor unit. Motor unit dimension can vary among muscles. Embedded within muscle, there is the muscle spindles, the sensory receptor specialized to detect muscle stretch, allowing the perception of body position (proprioceptor)6. Muscle spindle sends length information to the CNS thought the associated sensory neurons. It is formed by several specialized muscle fibres, named intrafusal, characterized by specific gene expression pattern and myosin heavy chain, encapsulated by connective tissue.

2.1.2 SKELETAL MUSCLE DENERVATION: cellular and molecular events

Denervation induces several molecular and morphological changes in skeletal muscle. The progressive muscle waste, determined by prolonged denervation, can be summarize in three main stages: i) an immediate loss of contraction function and the progressive loss of muscle mass; ii) a severe muscle atrophy and the complete loss of sarcomeric organization; iii) the muscle fibre degeneration and a visible fibrosis with fibroblast and adipocyte infiltration7,8. At the muscle organ level one of the earliest effect of denervation is the presence of spontaneous fibrillation activity9. This effect is due to spontaneous depolarization of myofibre membrane and correlates with the spread of acetylcholine receptors along fibre membrane, and the later disruption of the sarcoplasmic reticulum8,10. Moreover also the

162

CHAPTER 2: Introduction and Scientific Background vascularisation of the muscle undergoes dramatic remodelling after denervation, with a substantial decrease of capillaries number11. In rodents, the progressive muscle wet weight decrease is measurable within days after denervation, however the velocity and the entity of this process is strictly dependent upon the type of muscle and muscle fibres. In fact type II fibres undergo atrophy quickly respect to fibres type I12,13. Generally in rat models it has been shown an early fast weight 2 decreasing until a stabilization around 10-20% of the original weight after prolonged denervation condition. It has been observed that after denervation the number of fibres does not vary, while fibres type proportion and distribution are altered14. Thus, muscle weight lost is principally related to the decrease of myofibre diameter. This condition is called skeletal muscle atrophy and it is induced by denervation and by other acute or chronic pathological conditions, such as fasting, limb immobilization, sepsis, sarcopenia and neurodegenerative diseases15. After denervation, the NMJ is destabilized, since the axon terminal disappears and only muscle fibres and terminal SC remain. The acetylcholine receptor (AChR), normally present only at NMJ level, diffuses to the whole muscle fibre. Further to changes in density and distribution, also the expression of AChR is modified in denervated muscle: various subtypes are upregulated thanks to muscle regulation factors, bringing to the expression of the embryonic type of AChR. However the turnover of AChR is faster than normal condition16,17. Terminal SC change morphology and become mitotic. Their proliferation is regulated by several signals from motor neurons and denervated tissue18. In single muscle fibre, the initial preservation of sarcomeric structures is followed by prominent ultrastructural changes due to atrophic condition. Muscle atrophy can be described as an unbalance between protein synthesis and protein degradation. Protein catabolism is coordinated by both the ubiquitin-proteasome system (UPS) and the lysosomal/autophagy-related processes19. In particular, UPS plays a predominant role in the removal of myofibrillar protein degradation (as myosin and actin) whereas the lysosomal/autophagy system is responsible for the degradation of protein aggregates, mitochondria and endoplasmic reticulum membranes19,20. Moreover other catabolic systems are activated after denervation, as the calpain system, which disassembles the sarcomeric proteins21. Principal actors of protein degradation by UPS system are Muscle- specific RING-finger 1 (Murf1) and Muscle Atrophy F-Box (MAFbx/Atrogin 1), two E3 ubiquitin ligases. The upregulation of these two proteins is due to the activation of myogenin transcription factor and mTOR-FOXO (forkhead box O) pathways. Also the release of myostatin, a TGF family growth factor, increases after denervation, with subsequent inhibition of the IGF-1/Akt/mTOR pathway, responsible of muscle growth, and the activation of Atrogin 1. Contrasting data have been collected about the changes of the myonuclei numbers after denervation, ranging from great reduction to no changes2,22–24. Myonuclei number is

163

CHAPTER 2: Introduction and Scientific Background strictly related to the concept of muscle fibre-myonuclear domain, the cytoplasmatic volume under the control of a single nucleus, whose modifications under atrophic or hypertrophic condition are still under debate. In concomitant with muscle waste, a regeneration program is also activated, triggered principally by satellite cells. The satellite cells became activated within days after denervation and proliferate until they triplicate the cell population respect to normal condition25. The satellite cells activation, proliferation and final differentiation process are regulated by the expression of several transcription factors: quiescent satellite normally are positive for Pax7 marker; during activation they start to express the myogenic markers MyoD and Myf5 which drive cells in active proliferation, at this stage satellite cells are indicated as myoblasts; after the proliferative phase, myoblasts differentiate into mature myocytes and this transition is accompanied by a decrease of Pax7 and Myf5 expression and an increase of myogenin (MyoG) and MRF45. Satellite cells differentiate and fuse in existing myofibres to replace the lost muscle fibres. Moreover in denervated muscle is possible to observe fibres with central nuclei, which suggest that de novo formation of myofibres occurs thanks to satellite cells differentiation26. Long-term denervation is characterized by a strong decline in satellite cells number, due to the increased ratio of satellite cells apoptosis and the reduction of mitotic ability27. All the described processes are influenced and promoted by the release of growth factors, whose production substantially changes after denervation. NGF, BDNF, NT-3 and CNTF are some of these growth factors regulating the neuromuscular connection remodelling28. Other factors regulate satellite cells proliferation and differentiation, such as hepatocyte growth factor (HGF), insulin-like growth factor-1 (IGF-1), vascular endothelial growth factor (VEGF), fibroblast growth factor-2 (FGF-2), and platelet-derived growth factor (PDGF)29,30.

2.1.3 SKELETAL MUSCLE REINNERVATION and CURRENT THERAPIES FOR DENERVATED MUSCLE

When nerve regeneration occurs property the novel axons elongate until reaching target organ, guided by growth factors produced by SC and muscle tissue. During reinnervation process terminal SC extend process that guide the sprouting of regenerated axons to the correct fibres for reinnervation31,32. Moreover satellite cells have been indicated as a key source of Sema3A after nerve injury; this protein is involved in axon guidance, suggesting that also satellite cells may controls myofibres reinnervation33. The neuron-muscle reconnection results in the restoration of electrical impulses and neurotrophic factors signalling, which determine the regression of the atrophic condition. However the successful NMJ formation and muscle function restoration is related to the morphological changes occurred in muscle tissue and often an optimal nerve regeneration does not correspond to a total functional recovery. In rat extensor digitorum longus muscle

164

CHAPTER 2: Introduction and Scientific Background it has been demonstrated the presence of three defined stages, each characterized by a specific rate of successful reinnervation and functional recovery: i) in the early phase, corresponding to two months after denervation, the muscle tissue can achieve a complete functional recovery, even when the atrophy gives the loss of 90% of the muscle mass; ii) a second phase is characterized by a less restorative ability; iii) in the last phase (from 7 months after nerve injury) the restorative ability is minimal and muscle tissue has been 2 substituted by fat and fibrotic tissue34. These events have been described also for humans, where the time is dilated and the process takes years35,36. The progressive decline of denervated muscle is also associated with the loss of satellite cells and SC, with a reduction of the amount of growth factors release and the ability to guide regenerated axons to the correct fibres. Nevertheless, in human it has been shown that two years after denervation a small amount of SC survives and produces some neurotrophic factors34. In severe nerve injuries or when the nerve repair is delayed, the main goal during the denervation stage is to maintain the trophic state of muscle tissue and to promote terminal SC and satellite cells survival. As muscle tissue function and maintenance is related to the electrical impulses received by innervating nerves, a therapeutic approach proposed is the functional electrical stimulation (FES) of denervated muscle. FES results in muscle mass improvements, restoration of fibre diameter, myosin heavy chain composition and metabolic enzyme37–39, but the effects is dependent on the time between the injury and the beginning of the training: low results are obtained when FES is performed in the acute phase of the injury40,41. However FES might be invasive and implantable devices can be hulking, with a potential electrical injury. The use of exogenous growth factor treatments for denervated muscle has also been explored. Insulin-like growth factor (IGF) has been extensively studied for promoting muscle regenerations, as it is a positive regulator of muscle mass42. Direct injection of IGF in muscle at 1, 3 and 7 days after denervation induces an increase in fibre diameters respect to untreated muscle, even if after 8 weeks the muscle mass is still lower than normal muscle43. Another study showed that the combined injection of a mix of factors (NGF, CNTF and GDNF) after sciatic nerve injury not only increases muscle weight but promotes also the functional recovery44. Also gene therapy approach is a suitable tool for muscle treatment, as AAVs have already been successfully used for gene delivery in muscle and nerve in different animal models45–47. Moimas et al. demonstrated that treatment with AAV2-VEGF of flexor digitorum sublimis muscles results in higher muscle mass after one month of median nerve denervation, with a specific preservation of fast fibres48. Cell therapy has also been proposed as a method for delaying muscle atrophy and fasting nerve regeneration49–51. For example exogenous satellite cells can actively take part to the regenerative process, compensating the loss of local satellite cells52. Transplanted satellite cells or stem cells, moreover, can release growth factors, produce extracellular matrix molecule and give microenviromental stabilization, which in turn sustain the reinnervation process. Shen et al. transplanted satellite cells overexpressing myogenin in gastrocnemius

165

CHAPTER 2: Introduction and Scientific Background muscle at 1 week after denervation and observed a delay in muscle atrophy, with increased rate of muscle mass, however this effect was transient53. Halum and colleagues demonstrated that engineered muscle stem cells expressing CNTF injected in denervated thyroarytenoid muscle significantly promotes reinnervation54. It was reported that the injection of rat mesenchymal stem cells (MSC) in denervated gastrocnemius muscle induces the increase of muscle mass and muscle strength respect to untreated muscle 12 week after surgery, and these effects are not dependent on CNTF release by cells, but, as suggested by authors, it is probably due to a mix of delivered factors49. The clinical translation of these approaches remains to be investigated, however recently it was demonstrated, in a pilot study, that the intramuscular injection of autologous bone marrow- MSC produces significant increase in myofibre diameter and satellite cell and capillary number in patient with partial brachial plexus denervation55. After prolonged denervation, when the endogenous regenerative capacity is exhausted, a therapy aimed to replace the muscle is preferable. Several types of scaffold has been proposed, generally made by absorbable materials and mimicking the natural ECM structure56–58. An advantage of hydrogels is the possibility of sustained release of growth factors, as VEGF and IGF, with successful promotion of myogenesis56. Also the topographic features of the used scaffold seems to be important, for example for the alignment of myoblast and the organization of mature myofibres59. Future advances in growth factors engineering, stem cell biology and biomaterial technologies will be guide us towards a clinically applicable engineered muscle reconstruction.

2.1.4 NEUREGULIN 1 AND ErbB RECEPTORS

As reported in Chapter 1 of this thesis, neuregulins (NRG) comprise a large family of transmembrane and soluble growth factors, derived by the alternative splicing of four gene. The most studied is the Neuregulin 1 (NRG1), an ubiquitous protein involved in the development of several organs such as heart, brain and nerve. The other three gene encode for the corresponding proteins NRG2, NRG3 and NRG4, whose function remain largely obscure. NRG were first identified independently by several groups: two groups found a ligand for the oncogene ErbB261, a group was searching for a factor that stimulates SC proliferation62 and the last group identified a factor that stimulates the synthesis of AChR in muscle63. All these groups isolated a protein that was later referred as NRG1. In the following paragraph the structure of NRG1 and its receptors are described, meanwhile the last paragraph focus on the role of NRG1/ErbB system in the skeletal muscle tissue.

166

CHAPTER 2: Introduction and Scientific Background

2.1.4.1 Neuregulin and ErbB receptors structure

In the last decades different isoforms of NRG1 were isolated and differently named depending on the context of their discovery. Here, all isoforms are named depending on the exon compositions, as reported in several reviews64–66. Common characteristic of NRG1 is the presence of an epidermal growth factor (EGF)-like 2 domain, located in the extracellular portion, close to the transmembrane region. The EGF domain alone is able to activate the ErbB receptors. Alternative splicing and different promoters give rise to six different types of NRG1 (I-VI), which differ for the N terminal exon (Figure 2.1A). NRG1 type I (often referred as ARIA/Heregulin/NDF), type II (identified as GGF) and type III (called also SMDF) have been found in several vertebrates, whereas NRG type IV is encoded only in mammalian and NRG V-VI are typical of primates67. The NRG1 type I, II, IV and V possess an immunoglobulin (Ig)-like domain, absent in NRG1 type III and VI. This domain binds heparin and other glycosaminoglycans and can be a mechanism for limiting soluble NRG1 diffusion or creating a reservoir of the protein trapped in ECM. Moreover NRG1 type III presents an unique folding, in fact both N- and C- terminals are located inside cells (with the exception of NRG1 type III-γ) due to a cystein-rich domain (CRD) and a second transmembrane domain. The EGF-like domain is followed by various alternative domains that can be included or excluded giving rise to a several NRG1 isoforms (greater than 30). The first domain closer to EGF-like domain determines the formation of NRG1 alpha or beta isoforms. These domains seem influence protein activity: isoforms have higher affinity to ErbB receptors and it has been reported to be 10-fold more active than α isoforms. However, recently, it has been proposed that the differences between NRG1 beta and alpha might be not only quantitative but also qualitative68. Down to the alpha or beta domain, the following exon can be 1, 2 (exon missing), 4 or 3, which contain a stop codon (Figure 2.1A). Despite most of the NRG1 are produced as transmembrane isoforms, to be later cleaved and release as active soluble proteins (except for type III that remain an active transmembrane protein), isoforms containing exon 3 are directly released and do not possess an intracellular domain. Also the cytoplasmatic tail of NRG1 can differ depending on the exon included: exon a, b or a stop codon (type c). Most of NRG1 proteins are synthesized as transmembrane precursors that undergo proteolytic cleavage resulting in the release of soluble mature isoform (Figure 2.1 B). The cleavage is given by the action of three type I transmembrane proteases: -site of amyloid precursor protein cleaving enzyme (BACE); tumour necrosis factor-αconverting enzyme (TACE, also referred as ADAM17) and meltrin beta (called also ADAM19). The cleavage of different enzymes can result in different cellular signalling, as observed for NRG1 type III69–71 (Figure 2.1 C).

167

CHAPTER 2: Introduction and Scientific Background

NRG1 action is exerted thanks to a family of single-transmembrane receptor tyrosine kinases, named ErbB proteins. Three ErbB receptors are known, ErbB2, ErbB3 and ErbB4, that have homology with EGF receptor (known also as ErbB1). Only ErbB3 and ErbB4

Figure 2.1. Schematic representation of NRG1 isoforms66. (A) The six types of NRG1 differ from the N- terminal sequences. Moreover the inclusion of several exons give rise to various NRG1 isoforms. (B) Most of the NRG1 isoforms are produced as transmembrane proteins and after proteolytic cleavage are released as soluble mature isoforms, except for NRG1 type III that remains transmembrane. Some isoforms do not possess an intracellular domain and are directly released. (C) Example of different effects of the NRG1 processing by various secretases.

can bind directly NRG1. After ligand binding, ErbB receptors dimerize and the intracellular tyrosine residues are phosphorylated, becoming a docking site for adaptor proteins involved in several transduction pathways. While ErbB4 is the only autonomous receptors, able to bind NRG1 and former homodimer or heterodimers; ErbB3 kinase function is impaired and it can form only active heterodimers; ErbB2 does not bind NRG1, working as co-receptor. Thus NRG1 can act on ErbB2-ErbB3, ErbB2-ErbB4 and ErbB4- ErbB4 dimers. The canonical signals include the activation of Raf–MEK–ERK and PI3K– Akt–S6K pathways, but also c-Abl, jNK, CDK5, Fyn and Pyk264. Moreover ErbB4 can form dimers together with ErbB1, and so additional pathways can be activated by NRG1, such as Src family kinases, PlC, jNK and Abl64. Moreover was also described a non canonical forward signal, given by intracellular fragment of ErbB4 cleavable isoforms. After a -secretase cleavage the ErbB4 intracellular domain (ICD) is released and can translocate into the nucleus where it regulates gene transcription72.

168

CHAPTER 2: Introduction and Scientific Background

In addition to the canonical signal NRG1 can act also through a back-signalling: similarly to ErbB4, after extracellular cleavage NRG1 can be further processed by -secretase cleavage to release NRG1-ICD. This fragment can translocate into the nucleus and regulate gene transcription. It has been demonstrated that NRG1-ICD can bind the zinc-finger transcription factor Eos73. Thus, the peculiarity of NRG1 is that it might function as a soluble or transmembrane ligand but at the same time it might be a receptor for the putative 2 soluble isoform of ErbB4 and ErbB3 activating a back-signalling.

2.1.4.2 NRG1/ErbB system in skeletal muscle

The major component of the peripheral neuromuscular system are motoneurons, terminal SCs, muscle fibres, proprioceptive sensory neurons and fibroblasts. It is now evident that NRG1/ErbB system exerts a crucial role in the development and/or maintenance of all these cell types, except for fibroblasts. General information about the importance of this system in skeletal muscle tissue have been achieved with the study of knockouts mice. Targeted mutation that eliminates all NRG1 alpha expression results in abnormalities breast development during pregnancy but does not reveal any defect in nervous system development, suggesting that NRG1 beta isoforms are sufficient to guide a correct neuromuscolar development74. The inactivation of all the soluble isoforms of NRG1 (NRG1 type I and II) determines mice death in uterus, at E10.5, prior to the development of neuromuscular synapses, whereas SC precursor are normal75,76. Curiously the heterozygous mice look apparently normal, but have a strong reduction of AChR number at NMJ and concomitant synaptic “weakness”77, indicating that soluble NRG1 are necessary for NMJ maintenance. Mice in which the expression of all transmembrane isoforms (NRG type III) are disrupted, have a reduction of SC precursor and SC along axons, the number of motor and sensory neurons is smaller and is evident a failure of nerve terminals to find the endplate78. The inactivation of ErbB2 in muscle results in abnormal NMJ and muscle spindle development79. In adult life mature myofibres express ErbB2, ErbB3 and ErbB4, which are predominantly localized at NMJ levels, specially ErbB4, but also present in the transverse tubular (T- tubules) system80–82. Moreover myocytes produce soluble NRG1 which can bind the ErbB receptors on muscle, in an autocrine loop, or can mediate the motor neurons- or terminal SC- crosstalk. Moreover also terminal SC and motor neurons express soluble NRG1 that can act on muscle fibres. As observed with transgenic mice, NRG1 principally regulates NMJ formation and maintenance. In myocytes soluble NRG1 stimulates the expression and the translocation of AChR in the postsynaptic membrane83. The development and function of the neuromuscular system may depend on NRG1 signals among motorneurons, terminal SC and muscle fibres, a combination of paracrine and justacrine signals, as shown in Figure 2.2. In particular neuron-derived NRG1 seems critical for terminal SC survival, proliferation and differentiation84. Moreover application of NRG1 type II has been shown

169

CHAPTER 2: Introduction and Scientific Background to induce terminal SC process elongation and migration from synapses, with nerve terminal retraction and subsequent regrowth85. However it was also reported that in chicken embryo NRG1 administration, at later phase of reinnervation process, induces terminal SC ingrowth and nerve elongation to the correct endplate85. In addition, recently a study reported that the transmembrane NRG 1 type III expressed by motorneurons in adult mouse plays a role in synapse elimination and NMJ modulation through its action on terminal SC86.

Figure 2.2. Schematic representation of potential NRG1 signalling at NMJ level87. (A-C) Potential NRG1-mediated crosstalk between terminal Schwann cells and motorneurons. (D-G) Possible mechanisms of NRG1 signals between muscle-axon or muscle-muscle. The labelled 1 arrows indicate canonical signalling, where NRG1 activates ErbB. The arrows labelled "2" indicates the involvement of non-NRG/ErbB ligand and receptors. The arrows labelled "3" represent the reverse signalling, where ErbB activates NRG1 that function both as receptor and ligand.

Muscle exercise has been demonstrated to have beneficial effects on NMJ, increasing the synapse size in rats and mice and preventing denervation88. Curiously it was demonstrated that NRG1 proteolytic cleavage and subsequent NRG1 release is increased during exercise89, suggesting an NRG1 involvement in the exercise effects on NMJs. As already mentioned, the ablation of NRG1 or ErbB2 receptors led to abnormal muscle spindle formation, however the NRG1/ErbB system is critical also for its maintenance. It

170

CHAPTER 2: Introduction and Scientific Background has been demonstrated that muscle spindle formation, maturation and maintenance is dependent to the activity of the protease BACE90. Cheret et al. demonstrated that a transgenic mice without BACE expression shows a reduction of number and impaired maturation of fibres in muscle spindle90. The same authors reported that the ablation of soluble NRG1 give similar effects observed for BACE suppression. In cultured muscle cells, NRG1 induces the expression of several genes associated with intrafusal fibres91, in 2 particular a crucial component activated by the intracellular pathways downstream to ErbB receptors is Egr3, a transcription factors that activates several genes involved in muscle spindle formation92. Satellite cells in quiescent state do not express any ErbB receptors. However within 6 h after activation it was detected the expression of ErbB1, ErbB2 and ErbB3, whereas ErbB4 expression was found only later. Golding and colleagues demonstrated that the expression of ErbB2 receptor is involved in the activation of anti-apoptotic pathways, very important during this early phase of satellite cells activation, when cells are particularly susceptible to oxidative stress93. NRG1 plays a crucial role in skeletal myogenesis. However the effect of NRG1 is strictly related to the differentiation state of its target cells94. In vitro NRG1 induces the differentiation of L6 myotubes and mature muscle cells63,94–96; whereas it has been demonstrated that NRG1 inhibits early myoblast differentiation, blocking myogenin expression and promoting myoblast proliferation97. In myofibres ErbB receptors signalling regulates also some aspects of muscle metabolism. In particular NRG1 stimulates muscle uptake of glucose, and this effect is additive to IGF, indicating an independent pathway. NRG1 induces the translocation of GLUT4 glucose transporter in membrane, thus promoting glucose uptake98. This effect is evidence during exercise, where muscle contraction let to an increase of cytosolic Ca2+ concentration that activates metalloproteinase. These enzymes can cleave and release NRG1 soluble isoforms that can act in autocrine way activating ErbB4/ErbB2 dimers, which in turn activates an intracellular signalling, involving the PI3K-PDK1-PKCz pathway, resulting in the GLUT4 vesicles translocation and fusion with membrane99.

171

CHAPTER 2: Aim of The Research

2.2 AIM OF THE RESEARCH

Injuries to peripheral nerve determine the total or partial loss of motor and sensory functions mediated by the compromised nerve. When a severe nerve injury occurs, the nerve regeneration takes long time and the target skeletal muscle undergoes morphological changes that can afterwards compromise the reinnervation and the functional recovery. 2 Thus the faster target reinnervation and the maintenance of muscle optimum trophic state are crucial to obtain optimal restoration of motor and sensory ability. The detailed knowledge of the mechanisms regulating muscle changes after denervation are clinically relevant because of the possibility to predict the successful of reinnervation and to eventually prevent muscle waste in patients.

In this work I focus my attention on NRG1/ErbB system in skeletal muscle after nerve injury. Here is described the expression analysis of all the components of this system in skeletal muscle after denervation and reinnervation, using a well established rat median nerve injury models. This work attempts to collect exhaustive data about NRG1/ErbB system modulation in denervated and reinnervated skeletal muscle, giving its contribution to our knowledge about signalling pathways involved in muscle atrophy and muscle-nerve crosstalk after traumatic nerve injury.

173

CHAPTER 2: Scientific Publication

2.3 SCIENTIFIC PUBLICATION

Modulation of the Neuregulin1/ErbB system after 2 skeletal muscle denervation and reinnervation

Michela Morano1,2, Giulia Ronchi1,2, Valentina Nicolò1, Alessandro Crosio3, Isabelle Perroteau1, Stefano Geuna1,2, Giovanna Gambarotta1, Stefania Raimondo1,2.

1 Department of Clinical and Biological Sciences, University of Torino, 10043 Orbassano, Italy 2 Neuroscience Institute Cavalieri Ottolenghi (NICO), University of Torino, 10043 Orbassano, Italy 3 Microsurgery Unit, AOU Città della Salute e della Scienza, PO CTO, 10126 Torino, Italy

Article in preparation

ABSTRACT

Neuregulin 1 (NRG1) is a growth factor, produced by both peripheral nerve and skeletal muscle. In the muscle, it regulates neuromuscolar junction gene expression, acetylcholine receptor numbers, as well as muscle homeostasis and satellite cell survival. NRG1 signal is mediated by the tyrosine kinase receptors ErbB3 and ErbB4, and their co-receptor ErbB2. The NRG1/ErbB system is well studied in nerve tissue after injury, but little is known about this system in skeletal muscle after denervation/reinnervation processes. Here we performed a detailed time course expression analysis of several NRG1 isoforms and ErbB receptors in rat superficial flexor muscle after three types of median nerve injury, characterized by different severity. We found that ErbB receptors expression is related to the innervated state of the muscle, with an upregulation of ErbB2 clearly associated with the denervation state. Interestingly, NRG1 isoforms are differently regulated depending on nerve injury type. In vitro experiments with C2C12 atrophic myotubes reveals that both NRG1 alpha and beta administration influences the most known atrophic pathways, suggesting that NRG1 may play an anti-atrophic role.

175

CHAPTER 2: Scientific Publication

INTRODUCTION

The innervation of skeletal muscle has a predominant role in muscle structure and function maintenance. Therefore, nerve damage implicates wide molecular and structural modifications of target muscle fibres. Denervated myofibres undergo atrophy, a condition of loss of balance between protein synthesis and protein degradation, with consequently loss of muscle strength, mass and myofibre diameter, activation of apoptosis program and autophagy-lysosome machinery1,2. Atrophic condition is given by the activation of genes known as "atrogenes" and genes like myostatin, FoxO1, FoxO3 and NF-kB play a critical role in this process3. The more studied atrogenes are atrogin and murf1, encoding for two E3 ubiquitin ligases responsible of the degradation of transcription factors related to protein synthesis activation and structural proteins respectively4,5. After denervation, the stem cell pool in muscle, the satellite cells, exit from their quiescent state and become active. Satellite cells proliferate and then differentiate into multi-nucleated myofibres. This mechanism tries to maintain the muscle mass until reinnervation. When reinnervation occurs, the muscle receives again growth factors from motoneurons and all the features of atrophic state recede completely or partially. However, a change in fibre composition, an atrophic condition prolonged in time, an abnormal size and morphology of motor endplates or immature neuromuscolar junctions often compromise the functional recovery after muscle reinnervation. Neuregulin 1 (NRG1) is a trophic factor produced by the nerve, as well as by the muscle. It is a member of the Epidermal Growth Factor (EGF) family and it mediates the crosstalk between motor axon and muscle, terminal Schwann cells and motor axon, and muscle and muscle6. Alternative splicing gives rise to several NRG1 isoforms that can be roughly divided in soluble (Type I, II) or transmembrane (Type III) isoforms. Soluble isoforms exist in a form anchored to the cell membrane and only after a proteolytic cleavage the mature form of the protein is released for paracrine or autocrine signalling. A more detailed classification divides NRG1 in alpha or beta isoforms on the base of a small domain proximal to the EGF-like domain, or in type a, b, c isoforms depending on the exon included at the C-terminal tail7. NRG1 signal is mediated by the ErbB tyrosine kinase receptors ErbB3 and ErbB4, and their co-receptor ErbB2. All the receptors are mainly present at the neuromuscolar junction8. NRG1 regulates myogenesis9, muscle spindle development and maintenance10, glucose transport11,12, mitochondrial oxidative capacity13 and acetylcholine receptor expression8,14. It has been reported that satellite cells do not express ErbB receptors in the quiescent state but some cells are positive for NRG1. However, satellite cells start to express ErbB2 and ErbB3 and later also ErbB4 after activation, and this correlates with the activation of anti- apoptotic signalling15. A role of the NRG1/ErbB system in muscle regeneration has been postulated. Hirata et al. demonstrated that after toxin-induced muscle damage, NRG1 expression increases in satellite cells and in motoneurons16. Anyway, the role of the NRG1/ErbB system during

176

CHAPTER 2: Scientific Publication muscle denervation and reinnervation after nerve injury is far from being understood, since only uncompleted and discordant data are available. Here we performed a detailed time course analysis of NRG1 and ErbB receptors expression in the rat superficial flexor muscle after median nerve injury, comparing three types of injury. Our data reveal a fine regulation of the system strictly correlated with the type of injury and the reinnervation stage. In vitro experiments, using C2C12 cells treated with dexamethasone as a muscle 2 atrophy model, were performed to investigate the effect of NRG1 alpha and beta on atrophic myotubes.

MATERIAL AND METHODS

Animals. All procedures were approved by the Bioethics Committee of the University of Torino, by the Institutional Animal Care and Use Committee of the University of Torino, and by the Italian Ministry of Health, in accordance with the European Communities Council Directive European Communities Council (2010/63/EU), the National Institutes of Health guidelines and the Italian Law for Care and Use of Experimental Animals (DL26/14). Surgery. In this study a total of 73 adult female Wistar rats (Harlan Laboratories), weighing approximately 200 g, were used. Animals were housed in plastic cages with free access to food and water in a constant temperature and humidity, under 12–12 h light/dark cycles. Animals were randomly divided in four experimental groups according to the different types of nerve injury: (i) crush group (axonotmesis): the median nerve was crushed at the mid-humerus level with a non-serrated clamp, as previously described 17; (ii) end-to-end repair (neurotmesis): the median nerve was transected at the same position as in (i) and the proximal and distal nerve stumps were immediately sutured with 2 ETHILON 9/0 stitches; (iii) denervated muscle group: the median nerve was transected and the proximal stump was sutured to the major pectoralis muscle to avoid regeneration, while the distal nerve stump was allowed to degenerate over time; (iv) control group: the median nerve was not injured. Rat were sacrificed at different time points until 28 days after the surgery. To better appreciate the biomolecular differences between reinnervation and denervation process, some animals were sacrificed at a longer time point (12 weeks) for end-to-end and denervated groups. The superficial digitorum flexor muscle was collected, immediately frozen on dry ice and stored at -80°C for molecular analysis, or embedded in optimal cutting temperature (OCT) compound, and then frozen for morphological analysis. Morphological and morphometrical analysis. Morphological analysis was performed after median nerve crush or denervation. Muscles were harvested at different post-operative time-points: 7, 14 and 28 days. After collection, muscles were immediately embedded in OCT, frozen in cold isopentane (-50°C) and then stored at -80°C until use. All samples were cut with a cryostat to obtain transverse sections (10 μm of thickness) that were stored

177

CHAPTER 2: Scientific Publication at -20°C until use. Hematoxylin-Eosin staining was then performed on muscle section, after a quick passage in 70% ethanol to fix and lay the section on glass slide. Morphological analysis was performed using DM4000B microscope and DFC320 digital camera, using IM50 image manager system (Leica Microsystems). Samples were processed as follow: adjacent non-overlapped fields were acquired at 10X magnification to cover all section area and these images were used to measure fibre diameter. We used the systematic protocol previously described for nerve tissue 18. For each group, at least four animals were analyzed, for each muscle one section was analyzed. RNA isolation, cDNA preparation and quantitative real-time PCR (qRT-PCR) analysis. Frozen pieces of superficial flexorum muscle were used for total RNA extraction, analysing three animals for each time point. After mechanical dissociation, tissue was dissolved in TRIzol (Thermo Fisher Scientific, Waltham, MA, USA) and processed according to manufacturer's instructions. For each sample 1 μg total RNA was reverse transcripted (RT). The reaction was carried out in 25 μl containing: 1× RT-Buffer (Fermentas, Thermo Fisher Scientific), 0.1 μg/μl bovine serum albumin (BSA, Promega, Madison, WI, USA), 1 mM dNTPs, 7.5 μM random primers, 0.05% Triton, 33U RNAse Out Inhibitor (Fermentas, Thermo Fisher Scientific) and RT enzyme. The reaction was performed for 10 min at 25°C, 90 min at 42°C, 10 min at 90°C. The obtained cDNA was diluted 1:10 with water and stored at -20°C. Quantitative real-time PCR analysis was performed using a ABI Prism 7300 (Applied Biosystems, Thermo Fisher Scientific). The reaction was done in 25 µl containing 5 µl diluted cDNA (corresponding to 25 ng of starting RNA), 1x Sybr Green PCR Master Mix (BioRad, Hercules, CA, USA) and 300 nM forward and reverse primers. All primers were manually designed to amplify specific isoforms of Neuregulin 1 or ErbB receptors, the sequences are reported in Table 1 (for rat tissue) and Table 2 (for mouse cell line). The reaction protocol is the following: 30 s at 95°C, 40 cycles of denaturation at 95°C for 15 s followed by primer annealing and elongation at 60°C for 1 min. Dissociation curve was always analysed to check the quality of the reaction. Data were analyzed by the ΔΔCt relative quantification method normalizing to the housekeeping gene TATA-box Binding Protein (TBP). We determined the difference between Ct values of target and housekeeping gene (ΔCt), the difference between the ΔCt values of the samples and the ΔCt mean value of control sample was then calculated (ΔΔCt). Data about the time course analysis on muscle tissues are reported as -ΔΔCt instead of 2-ΔΔCt (RQ) because this way is more suitable to appreciate both up- and down-regulation. C2C12 cells culture. C2C12 are a mouse cell line (ATCC® CRL-1772™) derived from a subclone of myoblasts established by D. Yaffe and O. Saxel. Cells were growth in Dulbecco's Modified Eagle's Medium, (#30-2002, DMEM, Invitrogen, UK) supplemented with 100 U/mL penicillin, 100 g/mL streptomycin, 1 mM sodium pyruvate, 2 mM l- glutamine and 10% heat-inactivated fetal bovine serum (FBS; Invitrogen, Thermo Fisher

178

CHAPTER 2: Scientific Publication

Scientific). Cells were grown until 80% confluence and then the medium was changed to differentiation medium (DMEM supplemented with 100 U/mL penicillin, 100 g/mL streptomycin, 1 mM sodium pyruvate, 2 mM l-glutamine and 2% horse serum). The medium was replaced every 2 days and experiments were performed on the fifth day of differentiation. For the molecular analysis differentiated cells were treated as follow: 1 µM dexamethasone (water soluble, D2915, SIGMA, St.Louis, MO, USA) was added to the 2 differentiation medium, alone or in combination with recombinant peptides corresponding to the EGF-like domain of NRG1 alpha (30 nM; #296-HR) or NRG1 beta (5 nM, #396- HB) purchased from R&D systems (Minneapolis, MN, USA). RNA was extracted with TRIzol reagent and proceeded as explained before. For myotube diameter analysis, cells were seeded on a 24 multiwell plate and treated for 24 h with 1 µM dexamethasone, followed by further 48 h treatment with dexamethasone alone or in combination with NRG1 alpha or NRG1 beta. Cells were then fixed in 2% glutaraldehyde and stained with 1% Toluidine Blue. Five random pictures were taken for each well using Nikon eclipse TS100 microscope. Mean myotube diameter was calculated using ImageJ program. At least 40 diameters were manually counted for each well, collecting three measurements for each myotube. All experiments were carried out in biological and technical triplicate. Total protein extraction and Western Blot analysis. Protein extraction was performed using boiling Laemmli buffer (2.5% SDS, 0.125 M Tris–HCl pH 6.8): frozen pieces of superficial flexorum muscle were mechanically dissociated in Laemmli buffer and incubated at 100°C for 3 min. The same buffer was used for C2C12 cell line protein extraction. The BCA (Bicinchoninic Acid, SIGMA) method was used to quantify proteins. Primary antibodies used are: anti ErbB2 (working dilution, w.d. 1:1000, #sc-284), anti ErbB3 (w.d. 1:1000, #sc-285), anti β-actin (w.d. 1:4000, #sc-2228), anti Murf1 (w.d. 1:4000 , #sc-32920) all purchased from Santa Cruz (Dallas, TX, USA); anti p-FoxO3a (Thr32) (w.d. 1:1000, #9464), anti FoxO3a (w.d. 1:1000, #2497), anti AKT (w.d. 1:1000, #9272), anti p-AKT (w.d. 1:1000, #4051), anti ERK (w.d. 1:1000, #9102), anti p-ERK (w.d. 1:1000, #9106) all purchased from Cell Signalling (Danvers, MA, USA); anti Atrogin (anti-Fbx32 w.d. 1:3000, #ab168372, Abcam, Cambridge, MA, USA). The secondary antibodies used are: ECLTM anti-rabbit IgG (w.d. 1:40000, #NA934) and ECLTM anti-mouse IgG (w.d. 1:40000, #NA931) purchased from GE Healthcare (Milano, Italy). Image J program (National Institutes of Health) was used for a quantitative analysis of the western blot bands obtained. Statistical method. The statistical analysis was performed using IBM SPSS Statistics 22.0 software. Gene expression analysis for muscle tissue was performed in biological and technical triplicate. Data are reported as mean ± SEM, and the Anova One Way followed by Bonferroni Post Hoc Test was used for statistical analysis. We carried out two analyses: the first considering each single experimental group at different time points, the second

179

CHAPTER 2: Scientific Publication considering single time points and comparing the three experimental groups (only this analysis is reported in figures). Cell experiments were performed in biological triplicates; all data are presented as mean ± SD. Anova One Way followed by Bonferroni Post Hoc Test was used to assess statistical significance. The direct comparison between two groups was performed using Student's T test. A probability value lower than 0.05 was considered as statistically significant.

RESULTS

Short term analysis of skeletal muscle after denervation and reinnervation

Morphological and morphometrical analysis. A morphological analysis of the superficial digitorum flexor muscle was performed 7, 14 and 28 days after median nerve injury. Macroscopic analysis reveals muscle atrophy in denervated muscle with loss of muscle mass from 7 days to 28 days. In the crush group the mass loss is rescued 28 days after nerve injury (Fig. 1A). Morphometrical analysis carried out on transverse sections of healthy or injured muscle (Fig. 1B), shows that the average of fibre diameters is reduced respect to un-injured muscle for all the time points investigated in both crush and denervated groups (Fig. 1C). In the denervated group, the fibre diameter decreases over time, while in the crush group the trend is different and at 28 days the fibre diameter stops to decrease (Fig. 1C). Indeed, 28 days after nerve injury in the crush group the fibre size distribution returns similar to un-injured muscle, while in denervated group the curve is shifted to the left with a high number of small fibres (Fig. 1D). Regulation of NRG1/ErbB system. A mRNA time course expression analysis of NRG1 isoforms, the NRG1 receptors ErbB3 and ErbB4, and their co-receptors ErbB2 and ErbB1, was performed on superficial digitorum flexor muscle after median nerve injury at different time points. Results show that ErbB1 receptor mRNA expression is not regulated after denervation or denervation/reinnervation process (Fig. 2). Meanwhile, ErbB2 mRNA expression changes significantly in the three experimental models investigated. In denervated and end-to-end repair groups ErbB2 mRNA is upregulated compared to un-injured muscle from day 7 or 14 (denervated: 4.11 ± 1.62-fold, p = 0.003; end-to-end 14 days: 7.92 ± 3.33-fold, p = 0.001) until day 28 (denervated: 9.95 ± 2.25-fold, p = 0.000; end-to-end: 6.88 ± 3.00-fold, p = 0.002). In the crush group ErbB2 mRNA shows an early upregulation with a peak 7 days after injury (3.28 ± 0.73-fold, p = 0.011) and returns to basal level around day-14. Statistical analysis revealed significant differences among groups in the later phase (Fig. 2). ErbB3 expression is similar in the denervated and in the end-to-end repair groups: in the denervated group ErbB3 mRNA is upregulated 3 days (2.45 ± 0.1-fold, p = 0.008), 7 days (2.14 ± 0.37-fold, p = 0.029) and 28 days (2.92 ± 0.75-fold, p = 0.002) after injury, as well as in the end-to-end repair group the upregulation is detectable 3 days (3.75 ± 1.37- fold, p = 0.009); 7 days (3.46 ± 1.29-fold, p=0.007) and 28 days (2.81 ± 0.39-fold, p =

180

CHAPTER 2: Scientific Publication

0.021) after injury. In the crush group ErbB3 mRNA expression significantly increases 3 and 14 days (2.67 ± 0.83-fold, p = 0.031; 2.71 ± 0.09-fold, p = 0.022 respectively) after injury and returns to basal level at 28 days, when significant differences among groups are appreciable (Fig. 2). Finally, ErbB4 mRNA expression does not change significantly in the denervated and in the end-to-end repair groups, whereas in the crush group is detectable an upregulation 7 and 14 days after injury (2.57 ± 0.8-fold 1, p = 0.047; 3.32 ± 0.39-fold, p = 2 0.006 respectively) followed by a return to expression levels comparable to the control (Fig. 2).

Figure 1. Morphometrical analysis of superficial flexor muscle after nerve injury. A. Representative superficial digitorum flexor muscles as they appear after nerve injury. B. Representative transverse sections of muscle, stained with Hematoxylin-Eosin, in control condition and 7, 14 and 28 days after median nerve injury (crush injury or denervation). C. Graphs representing results of morphometrical analysis performed on muscle sections. The mean fibre diameter is reported for each group at different time points after nerve injury. The analysis was performed on four animals per group; data are expressed as a mean ± SD. Statistical analysis: Student's T-test: each value versus control muscle-value. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001 D. The graphs show the muscle fibre size distribution in injured muscle (gray: crush nerve injury; black: denervation) together with the distribution observed in un-injured muscle (white); un-injured n= 732; denervation 7 days n= 855, 14 days n= 838, 28 days n= 973; crush 7 days n= 882, 14 days n= 724, 28 days n= 701.

181

CHAPTER 2: Scientific Publication

The expression of the soluble isoforms of NRG1 (NRG1 type I and II) was assessed in muscle during denervation and reinnervation processes. In crush and end-to-end groups soluble NRG1 mRNA increases early, with a peak at 1 day (crush: 8.35 ± 3.95 -fold, p = 0.002; end-to-end: 5.08± 2.78-fold, p = 0.05) and returns to basal conditions already 3 days after injury. In denervated muscle this peak is not visible and soluble NRG1 expression is not perturbed. Statistical analysis among groups reveals differences at 1 day (Fig. 2). Soluble (and transmembrane) NRG1 is present in different isoforms depending on the exon at the C-terminal region of the EGF-like domain (alpha or beta isoforms) or the exon at the C-terminal of the protein (type a, b, c isoforms). NRG1 alpha and beta are differently modulated after nerve injury: alpha isoform expression increases at day 1 in crush and end- to-end repair groups (crush: 4.66 ± 1.65-fold; p = 0.001; end-to-end: 6.87 ± 2.26-fold, p = 0.085) and 14 days after injury in crush and denervated groups (crush 1.63 ± 0.35-fold; p = 0.015; denervated: 1.43 ± 1.63-fold; p = 0.09); NRG1 beta is upregulated at 1 day in both crush (crush: 12.44 ± 3.34-fold; p = 0.045 ) whereas in E-E group the increase of m-RNA is not statistically relevant; moreover in the crush group the expression return to basal level at 28 days with statistically relevant differences with end-to-end group. In denervated muscle the expression of NRG1 beta is not changed (Fig. 2). Also the expression of NRG1 type a, b, c isoforms is differently modulated by nerve injury (Fig. 2). NRG1 a expression pattern changes depending on nerve injury: in the crush group is strongly upregulated at 1 day (13.7 ± 12.5-fold; p = 0.014) and then returns to basal level of expression; in the end- to-end repair group the expression increases and reaches a peak at 28 days, whereas in the denervated group NRG1 a mRNA is upregulated at 3 days (2.73± 0.35-fold, p = 0.039) and 14 days (10.08 ± 4.15 -fold, p = 0.000) (Fig. 3). The expression of NRG1 b isoform is strongly upregulated 1 day after surgery in crush (13.9 ± 7.85-fold, p = 0.005) and end-to- end repair (5.30 ± 1.8-fold, p = 0.05) groups. At 14 days an increase of NRG1 b mRNA is detectable in the crush group (10.7 ± 6.37-fold, p = 0.011) and in the denervated group (12.12 ± 2.75 -fold, p = 0.007) (Fig. 2). On the contrary, the isoform NRG1 c shows a downregulation statistically relevant at 1 day in denervated muscle (-3.03 ± 1.02-fold, p = 0.004) and at 3, 7 and 28 days for the crush group (-1.44 ± 0.58-fold, p = 0.012; -5.64 ± 0.13-fold, p = 0.000; -2.87 ± 1.6-fold, p = 0.014 respectively) (Fig. 2). Long term analysis of skeletal muscle after denervation and reinnervation

We quantified the expression of ErbB receptors and NRG1 isoforms after long term denervation and reinnervation (12 weeks after denervation or end-to end repair, respectively). The expression of all ErbB receptors is strongly upregulated in the denervated muscle, while in the end-to-end repair group 12 weeks after injury the expression is similar to the control group (Fig. 3). In the end-to-end repair group the expression of soluble NRG1 (I-II) at 12 weeks is comparable to the control muscle, whereas in the denervated muscle a decrease in expression (0.37 ± 0.04-fold; p = 0.000) is detectable (Fig. 3). Finally, the analysis of all NRG1 isoforms reveals that only NRG1 beta

182

CHAPTER 2: Scientific Publication and NRG1 type c are highly downregulated after 12 weeks of denervation (0.11 ± 0.03- fold; p = 0.01; 0.25 ± 0.08-fold, p = 0.027 respectively) (Fig. 3).

Figure 2 The expression of ErbB receptors and NRG1 isoforms is modulated in muscle after nerve injury. Three different types of nerve injury and injury/repair were performed on median nerve

(denervation, crush and end-to-end repair). At different time points, superficial digitorum flexor muscle was collected and used for expression analysis. Graphs show the result of quantitative real-time PCR analysis. The analysis was performed on three animals per group; data are expressed as a mean ± SEM. Statistical analysis (referred to comparison among groups for each single time point): Anova One Way plus Bonferroni Post Hoc Test: *p≤0.05; **p ≤ 0.01; ***p ≤ 0.001. E-E: end-to-end repair.

183

CHAPTER 2: Scientific Publication

Figure 3. NRG1/ErbB expression after 12 weeks of muscle denervation and reinnervation. 12 weeks after denervation or end- to-end repair, the expression of ErbB receptors and NRG1 was analysed by quantitative real-time PCR in superficial digitorum flexor muscle. The analysis was performed on three animals per group; data are expressed as a mean ± SEM. Statistical analysis: Anova One Way plus Bonferroni Post Hoc Test: *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001.

In vitro analysis of NRG1 effects on atrophic myotubes C2C12 cell line was used as a model to study in vitro the NRG1/ErbB system during muscle atrophy. The lack of a model for studying muscle denervation in vitro was overcome using a toxin-induced atrophy model: C2C12 myotubes were treated for 24 h with dexamethasone to induce atrophy; then, mRNA was extracted to study NRG1/ErbB system modulation. First, we evaluated the expression level of two well known atrophic genes, atrogin and murf1. As expected, both were upregulated after dexamethasone treatment (Fig. 4). Also FoxO3 mRNA, the major atrogene activator, was found to be upregulated after treatment (Fig. 4). C2C12 myoblasts and myotubes express ErbB1, ErbB2 and ErbB3, whereas ErbB4 mRNA is barely detectable (data not shown). ErbB2 and ErbB3 expression in C2C12 myotubes increases in dexamethasone-induced atrophy. Furthermore, a strong inhibition of soluble NRG1 I-II mRNA was observed in atrophic condition (Fig. 4).

184

CHAPTER 2: Scientific Publication

Figure 4. Dexamethasone-treated C2C12 myotubes as in vitro model to study the NRG1/ErbB system during muscle atrophy. After 24 h of dexamethasone treatment (1 µM) alone or in combination with NRG1 alpha (30 nM) or NRG1 beta (5 nM), C2C12 mRNA was extracted for quantitative real-time PCR analysis. Data refer to a biological triplicate and are expressed as a mean ± SD. Statistical analysis: Anova One Way plus Bonferroni Post Hoc Test: *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001.

Since NRG1 alpha and beta are upregulated early after denervation/reinnervation and down-regulated at early and late stage of complete denervation in vivo, we hypothesized that exogenous administration of NRG1 could rescue the atrophic condition. C2C12 myotubes were incubated for 24 h in the presence of dexamethasone alone or together with NRG1 alpha or beta (the two isoforms commercially available). Results show that NRG1 beta induces a strong downregulation of murf1 mRNA and a slight reduction of atrogin and FoxO3 expression. On the contrary, NRG1 alpha administration does not affect atrogin, murf1 and FoxO3 expression (Fig. 4). Moreover, ErbB2 expression is decreased in NRG1 beta- group respect to dexamethasone or NRG1 alpha- group. ErbB3 expression in cells treated with NRG1 alpha and beta does not differ from control or dexamethasone- group, due to high standard deviation (Fig. 4). However, the expression of ErbB3 in C2C12 myotubes is already high and correlates with the differentiated stage. Protein analysis confirms the presence of higher level of atrogin and FoxO3 protein in atrophic condition (Fig.5 A, B). Treatment with NRG1 beta reduces both atrogin and FoxO3 levels, while NRG1 alpha downregulates only FoxO3a protein. The phosphorylation of FoxO3, corresponding to the inactivation of the protein, is significantly higher in NRG1 alpha treatment respect to dexamethasone alone. Surprisingly, murf1 protein levels were unchanged in all conditions. Under atrophic condition ErbB2 protein

185

CHAPTER 2: Scientific Publication increases and the administration of NRG1, both alpha and beta, rescues the upregulation (Fig.5 A, B). ErbB3 protein levels do not change in all condition.

Figure 5. Molecular and morphological analysis of dexamethasone-induced atrophy in C2C12 myotubes treated with NRG1. A. Representative pictures of western blot assays performed with proteins extracted from C2C12 myotubes treated for 24 h with 1 µM dexamethasone alone or in combination with NRG1 alpha (30 nM) or NRG1 beta (5 nM). B. Graphs show the results of C2C12 myotube protein quantification. C. The graph shows the results of myotube diameter measurement. C2C12 cells were treated for 24 h with dexamethasone, and then for 48 h with dexamethasone alone or in combination with NRG1 alpha or NRG1 beta. D. Representative images used for myotube diameter measurement. Data refer to a biological triplicate and are expressed as a mean ± SD. Statistical analysis: Anova One Way plus Bonferroni Post Hoc Test: *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001. DEXA: dexamethasone.

186

CHAPTER 2: Scientific Publication

We analysed also the main signalling pathways activated by ErbB2/ErbB3: AKT and ERK. The AKT phosphorylation is strongly induced by NRG1 beta treatment, while ERK phosphorylation does not change significantly among groups (Fig.5 A, B). In addition to the molecular analysis, we investigated morphological changes in cells after NRG1 treatment. The mean myotube diameter was used as atrophic index. Since after 24 h of dexamethasone treatment the diameter differences were small, we decided to treat cells 2 24 h with dexamethasone followed by 48 h in the presence of dexamethasone alone or in combination with NRG1 alpha or beta. After dexamethasone treatment, myotube diameter is lower respect to control, as expected; this difference is lost in NRG1 alpha and beta- groups even if the rescue is not complete (Fig. 5 C, D), however a direct comparison between NRG1 treated group and dexamethasone group revealed a statistical relevant difference (p=0.006 for NRG1 alpha, p=0.027 for NRG1 beta).

DISCUSSION

The aim of this study was to describe the modulation of the NRG1/ErbB system in skeletal muscle after peripheral nerve injury. To investigate the molecular changes occurring in muscle during denervation and reinnervation, we took advantage of three nerve injury models characterized by different severity, namely denervation, crush injury and end-to- end repair. We analyzed muscle quantitative morphology after the most severe nerve injury (the irreversible denervation) and after the faster reinnervation model (crush injury) to follow histological changes in these two opposite conditions. Denervation and crush are well characterized nerve injury models. In both models nerve fibres undergo Wallerian degeneration early after nerve injury. After denervation, nerve fibres degenerate over time19; in contrast, in the crush injury it is already possible to observe a regeneration phase at 14 days, and a nerve fibre maturation phase at 28 days. So, crush injury represents a good model to observe the effects of denervation and reinnervation processes in a short time-window: 12 days after median nerve crush injury, a functional recovery of muscle is already detectable and a rescue of 75% of muscle activity is reached already after 28 days17. Accordingly, in this study we found that after denervation the muscle atrophy worsens over time, while the muscle mass reduction detectable at 7 days after crush injury is restored at 14 and 28 days. Moreover, the muscle fibre size distribution confirms these data: 28 days after crush injury the distribution is similar to the un-injured muscle; on the contrary, in the denervated muscle histograms are shifted to the left, showing a higher number of smaller fibres, coherently with literature data20. As a model of slower muscle reinnervation, we used the end-to-end repair of the median nerve, a more severe nerve injury than crush, characterized by slower nerve regeneration19. At molecular level we focused our attention on the NRG1/ErbB system. NRG1 and its ErbB receptors are well known regulators of nerve growth and maintenance. Their

187

CHAPTER 2: Scientific Publication implication in the peripheral nerve regeneration process is deeply studied19,21,22. However, little is known about the modulation of NRG1 and ErbB receptors in muscle tissue after nerve injury. Since in skeletal muscle NRG1 regulates neuromuscolar junction gene expression, acetylcholine receptor numbers, as well as muscle homeostasis and satellite cells survival6,14,23, we can postulate that alterations of nerve-muscle crosstalk modify also the NRG1/ErbB system. In our previous work and in other studies, the expression of NRG1 was analyzed in the denervated muscle, using RT-PCR, but results were partially discordant, while ErbB2, ErbB3 and ErbB4 receptors were found upregulated11,24,25. Here we extended the expression analysis of different isoforms of NRG1 and ErbB receptors in the three models of nerve injury to differentiate among muscle denervation and slow and fast muscle reinnervation, using the quantitative real-time PCR technique for a time course analysis. In coherence with our previous observation, we found that ErbB2 and ErbB3 receptors are upregulated in denervated muscle over time (indeed, at 12 weeks they are still upregulated). On the contrary, ErbB2 and ErbB3 expression in the reinnervated muscle seems to be inversely correlated with muscle functional recovery. In the crush group their expression returns to basal levels 28 days after injury, whereas in the end-to-end repair group their level decreases later, at 12 weeks, in accordance with the muscle functional (grasping test) recovery17. Also ErbB4 expression is upregulated in denervated muscle over time, while in both reinnervated groups it returns to the basal level 28 days after injury, indicating that this receptor might be mainly involved in the early phases of muscle reinnervation. Indeed, ErbB4 is specifically localized at the neuromuscular junction15; we can therefore speculate that its return to basal expression level might be correlated to nerve-muscle reconnection. The upregulation of ErbB receptors might be correlated to satellite cell activation. It is known that quiescent satellite cells do not express ErbB receptors, but after activation in vitro, they express all the ErbB receptors15. An increase of ErbB3 expression was observed in C2C12 and L6 cell lines after myotube differentiation, correlated with a fast and early expression of NRG1 alpha8,26. The high level of ErbB2 and ErbB3 in the denervation phase might be necessary to the response to NRG1 alpha, which, in L6 cell line, stimulates cell fusion and myotube formation in a dose-dependent manner 26. However, in the denervated muscle, the high level of ErbB receptors might be related to continuous cycles of new fibre generation and fibre atrophy. As far as we know, this is the first study showing the modulation of several NRG1 isoforms (not only alpha and beta, but also type a, b and c) in skeletal muscle after nerve injury. We confirmed the upregulation of NRG1 alpha 14 days after denervation24. Intriguingly, 1 day after nerve injury NRG1 alpha and beta are differentially modulated according to the injury severity, suggesting that their expression might be influenced by the integrity of the nerve: where only axon continuity is lost (crush injury), both NRG1 alpha

188

CHAPTER 2: Scientific Publication and beta are upregulated; where nerve injury is more severe and nerve continuity is lost but immediately repaired (end-to-end repair), only NRG1 alpha is significantly upregulated; finally, where nerve injury is highly severe and nerve continuity is completely lost (denervation), both NRG1 isoforms are not upregulated. It has been demonstrated that after toxin-induced injury, NRG1 is upregulated in muscle during regeneration, although the authors did not discriminate between alpha or beta isoforms16. Taken together, these data 2 indicate that nerve signals might promote muscle regeneration after injury influencing NRG1 expression and that NRG1 alpha might be the principal player in the muscle response to nerve injury. If in the past the difference between NRG1 alpha and beta was described to be only quantitative, with alpha isoforms 10-fold less active than beta isoforms, now literature data demonstrate that their activity can be qualitatively different27,28. We also analyzed NRG1 type a, b and c, which differ for their intracellular domain (ICD) and they have never been studied in the muscle. We observed a similar expression pattern for NRG1 type a and b, where we found different expression peaks in all the experimental models, while NRG1 type c seems to be always downregulated. These data suggest that NRG1 type a and b might be involved in the response to the injury, as previously observed in nerve tissue19. Although these NRG1 isoforms are poorly studied, we know that they are involved in back signalling: the NRG1-ICD can be cleaved and transported in the nucleus where it can modulate gene expression; in neurons, for example, it represses several apoptosis regulators 29. However, further studies are needed to understand the role of these isoforms in denervated and reinnervated muscle. In order to explain the different roles of NRG1 isoforms observed in denervated and reinnervated muscle in vivo, we investigated the effect of NRG1 in an in vitro atrophy model, represented by C2C12 cells treated with dexamethasone. We stimulated cells with the two commercially available recombinant isoforms of NRG1, corresponding to the extracellular domain of NRG1 alpha and beta. In the muscle, the action of NRG1 beta has been shown to depend on the differentiation stage of muscle cells: it induces myogenesis and blocks differentiation in myoblasts9, while it stimulates differentiation in myotube culture30. Moreover, in the L6 myoblast cell lines it has been demonstrated that NRG1 alpha stimulates cell fusion and myotube formation26. Here, we demonstrated that both proteins have an effect on atrophic C2C12 in terms of myotube diameter, although only NRG1 beta is able to reduce atrogin, murf1 and FoxO3 expression levels. Interestingly, NRG1 alpha induces FoxO3 phosphorylation, and reduces FoxO3 and ErbB2 protein levels. We speculate that NRG1 alpha effect is delayed in time respect to NRG1 beta as both restore myotube diameter in a longer experiment, supporting the thesis of quantitative differences between the two isoforms. Nevertheless, we must keep in mind that only myotubes and not myofibres are present in our in vitro culture, thus limiting the effect (or the detection of the effect) exerted by NRG1.

189

CHAPTER 2: Scientific Publication

In summary, our results show a regulation of the NRG1/ErbB system after nerve injury, correlated with the denervation and the reinnervation phases in muscle, and the different expression pattern observed for NRG1 isoforms suggests a specific role of each isoform. Thus, our data highlight the importance to study the NRG1/ErbB system not only in a peripheral nerve perspective, but also in the perspective of the target muscle. Our in vitro experiments demonstrated an active role of NRG1 alpha and beta to rescue atrophic conditions induced by dexamethasone. Future experiments will be needed to address the in vivo efficacy of NRG1 and the challenge will be to identify the exact role exerted by each single NRG1 isoform. ACKNOWLEDGEMENTS. This study was supported by Compagnia di San Paolo (InTheCure project).

AUTHOR CONTRIBUTIONS. S. Raimondo and S. Geuna designed research with contribution from G. Gambarotta and I. Perroteau. G. Ronchi and A. Crosio performed the animal surgery. M. Morano and V. Nicolò performed the experiments. Data analysis and interpretation was carried out by M. Morano, G. Gambarotta and S. Raimondo. M. Morano, G. Ronchi and G. Gambarotta wrote the first draft of the manuscript that was critically revised by S. Raimondo and S. Geuna.

CONFLICT OF INTEREST. The authors declare that they have no conflict of interest.

REFERENCES

1. Bonaldo, P. & Sandri, M. Cellular and molecular mechanisms of muscle atrophy. Dis. Model. Mech. 6, 25–39 (2013). 2. Fanzani, A., Conraads, V. M., Penna, F. & Martinet, W. Molecular and cellular mechanisms of skeletal muscle atrophy: an update. J. Cachexia. Sarcopenia Muscle 3, 163–179 (2012). 3. Jackman, R. W., Cornwell, E. W., Wu, C.-L. & Kandarian, S. C. Nuclear factor-κB signalling and transcriptional regulation in skeletal muscle atrophy. Exp. Physiol. 98, 19–24 (2013). 4. Glass, D. J. Molecular mechanisms modulating muscle mass. Trends Mol. Med. 9, 344–50 (2003). 5. Bodine, S. C. & Baehr, L. M. Skeletal muscle atrophy and the E3 ubiquitin ligases MuRF1 and MAFbx/atrogin-1. AJP Endocrinol. Metab. 307, E469–E484 (2014). 6. Falls, D. L. Neuregulins and the neuromuscular system: 10 years of answers and questions. J. Neurocytol. 32, 619–647 (2003). 7. Falls, D. L. Neuregulins: functions, forms, and signaling strategies. Exp. Cell Res. 284, 14–30 (2003). 8. Jo, S. a, Zhu, X., Marchionni, M. a & Burden, S. J. Neuregulins are concentrated at nerve-muscle synapses and activate ACh-receptor gene expression. Nature 373, 158–161 (1995). 9. Ford, B. D., Han, B. & Fischbach, G. D. Differentiation-dependent regulation of skeletal myogenesis by neuregulin-1. Biochem. Biophys. Res. Commun. 306, 276– 281 (2003). 10. Cheret, C. et al. Bace1 and Neuregulin-1 cooperate to control formation and

190

CHAPTER 2: Scientific Publication

maintenance of muscle spindles. EMBO J. 32, 2015–2028 (2013). 11. Suárez, E. et al. A novel role of neuregulin in skeletal muscle: Neuregulin stimulates glucose uptake, glucose transporter translocation, and transporter expression in muscle cells. J. Biol. Chem. 276, 18257–18264 (2001). 12. Cantó, C. et al. Neuregulin Signaling on Glucose Transport in Muscle Cells. J. Biol. Chem. 279, 12260–12268 (2004). 13. Cantó, C. et al. Neuregulins increase mitochondrial oxidative capacity and insulin 2 sensitivity in skeletal muscle cells. Diabetes 56, 2185–2193 (2007). 14. Fromm, L. & Rhode, M. Neuregulin-1 induces expression of Egr-1 and activates acetylcholine receptor transcription through an Egr-1-binding site. J. Mol. Biol. 339, 483–494 (2004). 15. Golding, J. P., Calderbank, E., Partridge, T. A. & Beauchamp, J. R. Skeletal muscle stem cells express anti-apoptotic ErbB receptors during activation from quiescence. Exp. Cell Res. 313, 341–356 (2007). 16. Hirata, M. et al. Increased expression of neuregulin-1 in differentiating muscle satellite cells and in motoneurons during muscle regeneration. Acta Neuropathol. 113, 451–459 (2007). 17. Ronchi, G. et al. Functional and morphological assessment of a standardized crush injury of the rat median nerve. J. Neurosci. Methods 179, 51–7 (2009). 18. Geuna, S., Tos, P., Battiston, B. & Guglielmone, R. Verification of the two- dimensional disector, a method for the unbiased estimation of density and number of myelinated nerve fibers in peripheral nerves. Annals of Anatomy 182, 23–34 (2000). 19. Ronchi, G. et al. The Neuregulin1/ErbB system is selectively regulated during peripheral nerve degeneration and regeneration. Eur. J. Neurosci. 43, 351–364 (2016). 20. Dedkov, E. I., Borisov, A. B. & Carlson, B. M. Dynamics of postdenervation atrophy of young and old skeletal muscles: differential responses of fiber types and muscle types. J. Gerontol. A. Biol. Sci. Med. Sci. 58, 984–91 (2003). 21. Gambarotta, G., Fregnan, F., Gnavi, S. & Perroteau, I. Neuregulin 1 role in schwann cell regulation and potential applications to promote peripheral nerve regeneration. Int. Rev. Neurobiol. 108, 223–256 (2013). 22. Salzer, J. L. Axonal regulation of Schwann cell ensheathment and myelination. Journal of the peripheral nervous system : JPNS 17 Suppl 3, 14–19 (2012). 23. Rimer, M. Neuregulins at the neuromuscular synapse: Past, present, and future. Journal of Neuroscience Research 85, 1827–1833 (2007). 24. Nicolino, S. et al. Denervation and reinnervation of adult skeletal muscle modulate mRNA expression of neuregulin-1 and ERBB receptors. Microsurgery 29, 464–472 (2009). 25. Rimer, M., Cohen, I., Lømo, T., Burden, S. J. & McMahan, U. J. Neuregulins and erbB receptors at neuromuscular junctions and at agrin-induced postsynaptic-like apparatus in skeletal muscle. Mol. Cell. Neurosci. 12, 1–15 (1998). 26. Kim, D. et al. Neuregulin stimulates myogenic differentiation in an autocrine manner. J. Biol. Chem. 274, 15395–15400 (1999). 27. Eckert, J. M., Byer, S. J., Clodfelder-Miller, B. J. & Carroll, S. L. Neuregulin-1 beta and neuregulin-1 alpha differentially affect the migration and invasion of malignant peripheral nerve sheath tumor cells. Glia 57, 1501–20 (2009). 28. Ghahramani Seno, M. M., Gwadry, F. G., Hu, P. & Scherer, S. W. Neuregulin 1- alpha regulates phosphorylation, acetylation, and alternative splicing in lymphoblastoid cells. Genome 56, 619–25 (2013). 29. Bao, J., Wolpowitz, D., Role, L. W. & Talmage, D. A. Back signaling by the Nrg-1

191

CHAPTER 2: Scientific Publication

intracellular domain. J. Cell Biol. 161, 1133–1141 (2003). 30. Florini, J. R., Samuel, D. S., Ewton, D. Z., Kirk, C. & Sklar, R. M. Stimulation of myogenic differentiation by a neuregulin, glial growth factor 2: Are neuregulins the long-sought muscle trophic factors secreted by nerves? J. Biol. Chem. 271, 12699– 12702 (1996).

192

CHAPTER 2: Discussion and Future Directions

2.4 DISCUSSION AND FUTURE DIRECTIONS

The maintenance of muscle tissue morphology and homeostasis depends on signalling provided by the nervous system, both electrical and chemical. The loss of contact with neurons determines molecular and morphological changes in muscle, due to sequential events that trigger muscle fibres atrophy and loss, determining, after prolonged time 2 without neuron contact, fibrotic infiltrations with great impact on muscle function recovery. Moreover sustained denervation bring to the loss of terminal Schwann cells and satellite cells, which compromise the reinnervation process. However, it has been recently described a sustained myogenesis in human muscle after years of denervation, indicating that the surviving satellite pool is still able to generates new fibres8,100. This finding suggest that a rescue from severe atrophy is possible100. Actually, the proposed therapeutic approach to counteract the progressive muscle changes is represented by the functional electrical stimulation (FES) technique. These approach led to a maintenance and a quantifiable improvement of muscle fibres morphology and function in human muscle within two years after denervation8. However there are no other therapies able to maintain muscle mass and, simultaneously, guide the reinnervation for optimal functional recovery.

Despite muscle tissue engineering has developed novel approaches, based on new biomaterials and protein delivery, for the treatment of muscle traumatic injury, the handling of denervated muscle is little examined. New insights in growth factors modulation and cellular signalling which orchestrate the tissue changes after denervation and control the reinnervation process, will provide the bases for the development of novel therapies for muscle denervation. Thus, we decide to focus our attention on NRG1/ErbB system and study its regulation in denervated and reinnervated muscle.

NRG1/ErbB system has been extensively studied in peripheral nerve, with special regard to its role in Schwann cell differentiation, proliferation, motility, as well as myelination; whereas only few works focus on the involvement of this signalling pathway in muscle tissue response to nerve damage. Our expression analysis revealed that NRG1/ErbB system is widely modulated after nerve injury. Moreover, the direct comparison of three different nerve injuries permitted to asses that the severity of the injury influences the modulation of our target system. For our analysis, we took advantage of three well known nerve injury models characterized by different severity: complete rat median nerve denervation; denervation followed by immediate repair, the end-to-end repair (E-E); nerve injury with fast reinnervation, the crush injury. Main observations are that ErbB2, ErbB3 and ErbB4 are upregulated after nerve injury, but return to basal level of expression after reinnervation or functional recovery in crush and E-E repair groups. On the contrary in denervated muscle the upregulation of these receptors is still visible after 12 weeks. ErbB1 expression does not changes, indicating a less involvement in skeletal muscle changes after denervation. The high level of ErbB2 and ErbB3 detected might be related to the activation

193

CHAPTER 2: Discussion and Future Directions of Satellite cells. Whereas their expression was not detected in quiescent satellite cells, all the ErbB receptors have been found in activated satellite cells in vitro, and a pro-survival role was proposed93. Moreover the differentiation of C2c12 and L6 myoblast cells correlates to ErbB3 expression increase and a fast NRG1 alpha expression83,95.

We evaluated also the expression of several isoforms of NRG1 in denervated/reinnervated muscle. We confirmed that NRG1 alpha is upregulated after 14 days of denervation, as previously observed101. The NRG1 alpha and beta mRNA level increases after 1 day only in crush and E-E group, but not in denervated group, suggesting that nerve continuity preservation (as occurs in crush injury and in E-E repair) might influence NRG1 modulation. The upregulation of NRG1 was detected also after toxin-induced muscle atrophy102. We focus our attention also on NRG type a,b and c isoforms, which role remains generally obscure. Our principal finding was that NRG1 type c seems not affected by the injury, whereas both NRG1 a and b are regulated, suggesting their involvement in the response to the injury, as observed in nerve tissue68.

Despite our work highlights the involvement of NRG1/ErbB system in muscle response to nerve damage, it does not give a complete view of target system modulation, since the analysis remains on the expression level and protein analysis is missing. We are actually work on ErbB proteins expression analysis through western blot technique. Since protein localization might be useful to better understand its actions, we are also working on immunohistochemical analysis. Preliminary data show that ErbB2 is expressed both by satellite cells and myofibres, while ErbB3 seems specifically localized in satellite cells. The localization of NRG1 isoforms by antibody staining is limited by the absence of antibodies that recognize specific isoforms. Thus, the immunofluorescence analysis could give poor information. An alternative approach might be the in situ hybridization that permits the localization of the NRG1 mRNA, thus giving the cellular source of each specific NRG1 isoform. It has been demonstrated that skeletal muscle shows variable susceptibly to several insults, depending to their fibre type composition. In a delayed nerve repair study it was highlighted that fast type fibres shows atrophic condition after one month of delayed repair whereas the reduction on size of slow type fibres is visible only after six months13. Thus, it could be interesting analyse the expression of NRG1 and ErbB receptors in several muscle type or muscle fibres, to see any eventual correlations with atrophy susceptibility. Other authors revealed differences in protein expression and miRNA network in gastrocnemius (containing mainly fast fibres) and soleus muscles (containing mainly slow fibres), underlining how different expression of specific molecular pathways modulates the degree of atrophy after denervation103.

We set up an in vitro experiment to observe the effects of NRG1 alpha/beta administration on C2c12 cells in dexamethasone-induced atrophic condition. We observed that the administration of both alpha and beta isoforms increases myotube diameter respect to

194

CHAPTER 2: Discussion and Future Directions dexamethasone treatment alone, partially rescuing the atrophic condition. The effect of the NRG1 isoforms need to be further explored in vitro to understand the therapeutic potentiality of these molecules. Since C2c12 express only ErbB2 and ERbB3 but not ErbB4, and the population after differentiation protocol is composed by myotubes and not finally differentiate myofibres, this cellular model is limited. Moreover a single cell population culture cannot resemble the in vivo complexity, specially the neuron-muscle 2 fibre interaction. Thus, new in vitro models are needed to better mimic the environment of muscle tissue and to give the opportunity to study the reinnervation process. Mimic denervation or reinnervation in vitro results difficult, however some new technologies offer singular approaches. For example different co-culture systems have been set up in which muscle fibres and motor neurons are seeded together to study NMJ formation and eventually disruption104–106. However these methods show some limitations, like the formation of immature myofibres. Microfluidic chambers have also been proposed for multi-compartment culture where neurons processes and muscle cells are physically isolated from neuron cell bodies107. Actually novel 3D co-culture methods have being investigated to obtain a culture microenvironment as similar as possible to those observed in vivo. The integration of hydrogel or matrix has been shown to improve the maturation of cultured myoblast and the formation of mature NMJ108–111.These innovative culture systems will provide a good model to study the effects of NRG1 administrations on muscle maintenance and reinnervation.

The regulation of the NRG1/ErbB system observed after denervation and reinnervation process, and the literature data about the NRG1 effects on myoblast and myotube cultures, suggest that the manipulation of NRG1 might represent a promising novel approach to treat denervated muscle. Ideal treatment for denervated muscle must exert multiple effects, such as controlling fibres atrophy promoting protein production and reducing protein catabolism, preventing satellite cells apoptosis and supporting myogenesis, promoting terminal Schwann cell survival and guiding the correct reassembly of NMJ and neuron- fibres contact. It has been demonstrated that in vitro both NRG1 beta and alpha are able to induce myogenesis in L6 cells95,97, moreover the signalling of ErbB receptors has been associated to anti-apoptotic signals in activated satellite cells93. NRG1 is also involved in NMJ assembly and plasticity, in fact in vitro NRG1 induces transcription and membrane translocation of acetylcholine receptors (AChRs) in muscle87. Moreover, recently it was shown that direct injection of NRG1 in muscle stabilize postsynaptic AChRs and increase the size of AChR cluster at NMJ112,113. It was also demonstrated that soluble NRG1 administration has different effect on multiple stages of NMJ formation in chicken embryo: in a early phase when neuron-muscle interaction is absent, NRG1 induces AChR cluster formation; during a middle stage, when nerve-muscle contact is made, NRG1 guides the differentiation of pre- and postsynaptic components and their proper apposition; whereas in a later phase NRG1 acts on the terminal SC morphology85. Collectively this data provide evidences of therapeutic potentiality of NRG1. In this respect, Mancuso and colleagues

195

CHAPTER 2: Discussion and Future Directions have recently demonstrated that overexpression of NRG1 by AAV vector injection in muscles prevent muscle denervation inducing motor axon collateral sprouting in SOD1G93A mice (model for the amyotrophic lateral sclerosis neurodegenerative disease)114. Moreover the same authors showed that in wild-type mice after partial or total denervation of the gastrocnemius muscle, the overexpression of NRG1 accelerates the collateral sprouting of regenerating nerve, and the treatment with an inhibitors of ErbB receptors blocks this effect and increases the motor units size114. Indeed NRG1 has been proposed as therapeutic approach also for other muscular diseases, as Duchenne’s muscular dystrophy115 and sarcopenia116.

NRG1 based therapies look promising for muscle reinnervation in chronic neuronal diseases and after traumatic injuries, specially for its ability to control NMJ assembly. Nerve repair therapy should be focus on nerve regeneration as well as on the muscle reinnervation process. A spatial and temporal control of NRG1 delivery might be essential to obtain the desired effect in a specific step of the regenerative process, since it is known from in vitro experiments that NRG1 effect might be different depending on concentration and target: in myoblast cells NRG1 induces proliferation, whereas in myotubes promotes myogenesis. Moreover, determining the precise role exercised by each single isoform remains a challenge. Future molecular and cellular biological in vitro approaches will probably help to define the differences among NRG1 isoforms actions, with special attention on the poor studied intracellular signalling of NRG1 type a and b in muscle cells.

196

CHAPTER 2: References

2.5 REFERENCES

1. Liu, J.-X. et al. Myonuclear domain size and myosin isoform expression in muscle fibres from mammals representing a 100,000-fold difference in body size. Exp. Physiol. 94, 117–29 (2009). 2. Gundersen, K. & Bruusgaard, J. C. Nuclear domains during muscle atrophy: nuclei 2 lost or paradigm lost? J. Physiol. 586, 2675–2681 (2008). 3. Punkt, K., Naupert, A. & Asmussen, G. Differentiation of rat skeletal muscle fibres during development and ageing. Acta Histochem. 106, 145–154 (2004). 4. Ciciliot, S., Rossi, A. C., Dyar, K. A., Blaauw, B. & Schiaffino, S. Muscle type and fiber type specificity in muscle wasting. Int. J. Biochem. Cell Biol. 45, 2191–2199 (2013). 5. Dumont, N. A., Bentzinger, C. F., Sincennes, M.-C. & Rudnicki, M. A. in Comprehensive Physiology 5, 1027–1059 (John Wiley & Sons, Inc., 2015). 6. Maier, A. Development and regeneration of muscle spindles in mammals and birds. Int. J. Dev. Biol. 41, 1–17 (1997). 7. Lapalombella, R. et al. Persistence of regenerative myogenesis in spite of downregulation of activity-dependent genes in long-term denervated rat muscle. Neurol. Res. 30, 197–206 (2008). 8. Carlson, B. M. The biology of long-term denervated skeletal muscle. Eur. J. Transl. Myol. 24, (2014). 9. Midrio, M., Caldesi-Valeri, V., Princi, T., Ruzzier, F. & Velussi, C. Differential effects of disuse preceding denervation on the onset and development of fibrillation in fast and slow muscles. Experientia 33, 209–11 (1977). 10. Bandi, E. et al. Autocrine activation of nicotinic acetylcholine receptors contributes to Ca2+ spikes in mouse myotubes during myogenesis. J. Physiol. 568, 171–80 (2005). 11. Borisov, A. B., Huang, S. K. & Carlson, B. M. Remodeling of the vascular bed and progressive loss of capillaries in denervated skeletal muscle. Anat. Rec. 258, 292– 304 (2000). 12. Borisov, A. B., Dedkov, E. I. & Carlson, B. M. Interrelations of myogenic response, progressive atrophy of muscle fibers, and cell death in denervated skeletal muscle. Anat. Rec. 264, 203–18 (2001). 13. Jonsson, S. et al. Effect of delayed peripheral nerve repair on nerve regeneration, Schwann cell function and target muscle recovery. PLoS One 8, e56484 (2013). 14. Ashley, Z. et al. Atrophy, but not necrosis, in rabbit skeletal muscle denervated for periods up to one year. AJP Cell Physiol. 292, C440–C451 (2006). 15. Sacheck, J. M. et al. Rapid disuse and denervation atrophy involve transcriptional changes similar to those of muscle wasting during systemic diseases. FASEB J. 21, 140–55 (2007). 16. Ma, J. et al. Gene expression of myogenic regulatory factors, nicotinic acetylcholine receptor subunits, and GAP-43 in skeletal muscle following denervation in a rat model. J. Orthop. Res. 25, 1498–505 (2007).

197

CHAPTER 2: References

17. Wu, P. et al. Key changes in denervated muscles and their impact on regeneration and reinnervation. Neural Regen. Res. 9, 1796–809 (2014). 18. Love, F. M. & Thompson, W. J. Schwann cells proliferate at rat neuromuscular junctions during development and regeneration. J. Neurosci. 18, 9376–85 (1998). 19. Bonaldo, P. & Sandri, M. Cellular and molecular mechanisms of muscle atrophy. Dis. Model. Mech. 6, 25–39 (2013). 20. Masiero, E. et al. Autophagy Is Required to Maintain Muscle Mass. Cell Metab. 10, 507–515 (2009). 21. Talbert, E. E., Smuder, A. J., Min, K., Kwon, O. S. & Powers, S. K. Calpain and caspase-3 play required roles in immobilization-induced limb muscle atrophy. J. Appl. Physiol. 114, 1482–1489 (2013). 22. Schmalbruch, H. & Lewis, D. M. Dynamics of nuclei of muscle fibers and connective tissue cells in normal and denervated rat muscles. Muscle Nerve 23, 617– 26 (2000). 23. Aravamudan, B., Mantilla, C. B., Zhan, W.-Z. & Sieck, G. C. Denervation effects on myonuclear domain size of rat diaphragm fibers. J. Appl. Physiol. 100, 1617–22 (2006). 24. Wada, K.-I., Takahashi, H., Katsuta, S. & Soya, H. No decrease in myonuclear number after long-term denervation in mature mice. Am. J. Physiol. Cell Physiol. 283, C484-8 (2002). 25. Viguie, C. A., Lu, D. X., Huang, S. K., Rengen, H. & Carlson, B. M. Quantitative study of the effects of long-term denervation on the extensor digitorum longus muscle of the rat. Anat. Rec. 248, 346–54 (1997). 26. de Castro Rodrigues, A., Andreo, J. C. & de Mattos Rodrigues, S. P. Myonuclei and satellite cells in denervated rat muscles: An electron microscopy study. Microsurgery 26, 396–398 (2006). 27. de Castro Rodrigues, A. & Schmalbruch, H. Satellite cells and myonuclei in long- term denervated rat muscles. Anat. Rec. 243, 430–437 (1995). 28. Kingham, P. J. & Terenghi, G. Bioengineered nerve regeneration and muscle reinnervation. J. Anat. 209, 511–526 (2006). 29. Ten Broek, R. W., Grefte, S. & Von den Hoff, J. W. Regulatory factors and cell populations involved in skeletal muscle regeneration. J. Cell. Physiol. 224, n/a-n/a (2010). 30. Yin, H., Price, F. & Rudnicki, M. A. Satellite Cells and the Muscle Stem Cell Niche. Physiol. Rev. 93, 23–67 (2013). 31. Tam, S. L. & Gordon, T. Neuromuscular activity impairs axonal sprouting in partially denervated muscles by inhibiting bridge formation of perisynaptic Schwann cells. J. Neurobiol. 57, 221–34 (2003). 32. Kang, H., Tian, L. & Thompson, W. Terminal Schwann cells guide the reinnervation of muscle after nerve injury. J. Neurocytol. 32, 975–985 (2003). 33. Tatsumi, R. et al. Possible implication of satellite cells in regenerative motoneuritogenesis: HGF upregulates neural chemorepellent Sema3A during myogenic differentiation. AJP Cell Physiol. 297, C238–C252 (2009). 34. Carlson, B. M. The biology of long-term denervated skeletal muscle. Eur. J. Transl.

198

CHAPTER 2: References

Myol. 24, 3293 (2014). 35. Kern, H. et al. Long-term denervation in humans causes degeneration of both contractile and excitation-contraction coupling apparatus, which is reversible by functional electrical stimulation (FES): a role for myofiber regeneration? J. Neuropathol. Exp. Neurol. 63, 919–31 (2004). 36. Carraro, U., Rossini, K., Mayr, W. & Kern, H. Muscle fiber regeneration in human 2 permanent lower motoneuron denervation: relevance to safety and effectiveness of FES-training, which induces muscle recovery in SCI subjects. Artif. Organs 29, 187–91 (2005). 37. Dow, D. E. et al. Number of contractions to maintain mass and force of a denervated rat muscle. Muscle Nerve 30, 77–86 (2004). 38. Davis, G. M., Hamzaid, N. A. & Fornusek, C. Cardiorespiratory, Metabolic, and Biomechanical Responses During Functional Electrical Stimulation Leg Exercise: Health and Fitness Benefits. Artif. Organs 32, 625–629 (2008). 39. Andersen, J. L., Mohr, T., Biering-Sørensen, F., Galbo, H. & Kjaer, M. Myosin heavy chain isoform transformation in single fibres from m. vastus lateralis in spinal cord injured individuals: effects of long-term functional electrical stimulation (FES). Pflugers Arch. 431, 513–8 (1996). 40. Chilibeck, P. D., Jeon, J., Weiss, C., Bell, G. & Burnham, R. Histochemical changes in muscle of individuals with spinal cord injury following functional electrical stimulated exercise training. Spinal Cord 37, 264–8 (1999). 41. Mohr, T. et al. Long-term adaptation to electrically induced cycle training in severe spinal cord injured individuals. Spinal Cord 35, 1–16 (1997). 42. Schiaffino, S., Dyar, K. A., Ciciliot, S., Blaauw, B. & Sandri, M. Mechanisms regulating skeletal muscle growth and atrophy. FEBS J. 280, 4294–4314 (2013). 43. Day, C. S. et al. Insulin growth factor-1 decreases muscle atrophy following denervation. Microsurgery 22, 144–51 (2002). 44. Chen, J., Chu, Y. F., Chen, J. M. & Li, B. C. Synergistic effects of NGF, CNTF and GDNF on functional recovery following sciatic nerve injury in rats. Adv. Med. Sci. 55, 32–42 (2010). 45. Homs, J. et al. Schwann cell targeting via intrasciatic injection of AAV8 as gene therapy strategy for peripheral nerve regeneration. Gene Ther. 18, 622–30 (2011). 46. Abadie, J., Blouin, V., Guigand, L., Wyers, M. & Cherel, Y. Recombinant adeno- associated virus type 2 mediates highly efficient gene transfer in regenerating rat skeletal muscle. Gene Ther. 9, 1037–43 (2002). 47. Messina, S. et al. VEGF overexpression via adeno-associated virus gene transfer promotes skeletal muscle regeneration and enhances muscle function in mdx mice. FASEB J. 21, 3737–46 (2007). 48. Moimas, S. et al. Effect of vascular endothelial growth factor gene therapy on post- traumatic peripheral nerve regeneration and denervation-related muscle atrophy. Gene Ther. 20, 1014–21 (2013). 49. Jiang, J. et al. Adult rat mesenchymal stem cells delay denervated muscle atrophy. Cell. Mol. Neurobiol. 32, 1287–98 (2012). 50. Schaakxs, D., Kalbermatten, D. F., Raffoul, W., Wiberg, M. & Kingham, P. J.

199

CHAPTER 2: References

Regenerative cell injection in denervated muscle reduces atrophy and enhances recovery following nerve repair. Muscle Nerve 47, 691–701 (2013). 51. Grumbles, R. M., Liu, Y., Thomas, C. M., Wood, P. M. & Thomas, C. K. Acute stimulation of transplanted neurons improves motoneuron survival, axon growth, and muscle reinnervation. J. Neurotrauma 30, 1062–9 (2013). 52. Lazerges, C. et al. Transplantation of primary satellite cells improves properties of reinnervated skeletal muscles. Muscle Nerve 29, 218–26 (2004). 53. Shen, H. et al. Implantation of muscle satellite cells overexpressing myogenin improves denervated muscle atrophy in rats. Brazilian J. Med. Biol. Res. = Rev. Bras. Pesqui. medicas e Biol. 49, e5124 (2016). 54. Halum, S. L., McRae, B., Bijangi-Vishehsaraei, K. & Hiatt, K. Neurotrophic factor- secreting autologous muscle stem cell therapy for the treatment of laryngeal denervation injury. Laryngoscope 122, 2482–96 (2012). 55. Hogendoorn, S. et al. Local injection of autologous bone marrow cells to regenerate muscle in patients with traumatic brachial plexus injury: a pilot study. Bone Joint Res. 3, 38–47 (2014). 56. Koning, M., Harmsen, M. C., van Luyn, M. J. A. & Werker, P. M. N. Current opportunities and challenges in skeletal muscle tissue engineering. J. Tissue Eng. Regen. Med. 3, 407–415 (2009). 57. Cezar, C. A. & Mooney, D. J. Biomaterial-based delivery for skeletal muscle repair. Adv. Drug Deliv. Rev. 84, 188–197 (2015). 58. Kwee, B. J. & Mooney, D. J. Biomaterials for skeletal muscle tissue engineering. Curr. Opin. Biotechnol. 47, 16–22 (2017). 59. Kumbar, S. G., James, R., Nukavarapu, S. P. & Laurencin, C. T. Electrospun nanofiber scaffolds: engineering soft tissues. Biomed. Mater. 3, 34002 (2008). 60. Yan, L. et al. Neuregulin-2 ablation results in dopamine dysregulation and severe behavioral phenotypes relevant to psychiatric disorders. Mol. Psychiatry (2017). doi:10.1038/mp.2017.22 61. Peles, E. et al. Isolation of the neu/HER-2 stimulatory ligand: a 44 kd glycoprotein that induces differentiation of mammary tumor cells. Cell 69, 205–16 (1992). 62. Goodearl, A. D. et al. Purification of multiple forms of glial growth factor. J. Biol. Chem. 268, 18095–102 (1993). 63. Falls, D. L., Rosen, K. M., Corfas, G., Lane, W. S. & Fischbach, G. D. ARIA, a protein that stimulates acetylcholine receptor synthesis, is a member of the neu ligand family. Cell 72, 801–15 (1993). 64. Falls, D. L. Neuregulins: functions, forms, and signaling strategies. Exp. Cell Res. 284, 14–30 (2003). 65. Mei, L. & Xiong, W.-C. Neuregulin 1 in neural development, synaptic plasticity and schizophrenia. Nat. Rev. Neurosci. 9, 437–52 (2008). 66. Gambarotta, G., Fregnan, F., Gnavi, S. & Perroteau, I. Neuregulin 1 role in Schwann cell regulation and potential applications to promote peripheral nerve regeneration. Int. Rev. Neurobiol. 108, 223–56 (2013). 67. Chou, C.-F. & Ozaki, M. In silico analysis of neuregulin 1 evolution in vertebrates. Biosci. Rep. 30, 267–75 (2010).

200

CHAPTER 2: References

68. Ronchi, G. et al. The Neuregulin1/ErbB system is selectively regulated during peripheral nerve degeneration and regeneration. Eur. J. Neurosci. 43, 351–364 (2016). 69. Hu, X. et al. Bace1 modulates myelination in the central and peripheral nervous system. Nat. Neurosci. 9, 1520–5 (2006). 70. Willem, M. et al. Control of Peripheral Nerve Myelination by the -Secretase 2 BACE1. Science (80-. ). 314, 664–666 (2006). 71. La Marca, R. et al. TACE (ADAM17) inhibits Schwann cell myelination. Nat. Neurosci. 14, 857–65 (2011). 72. Sardi, S. P., Murtie, J., Koirala, S., Patten, B. A. & Corfas, G. Presenilin-dependent ErbB4 nuclear signaling regulates the timing of astrogenesis in the developing brain. Cell 127, 185–97 (2006). 73. Bao, J. et al. Activity-dependent transcription regulation of PSD-95 by neuregulin-1 and Eos. Nat. Neurosci. 7, 1250–8 (2004). 74. Li, L. et al. The breast proto-oncogene, HRGalpha regulates epithelial proliferation and lobuloalveolar development in the mouse mammary gland. Oncogene 21, 4900– 7 (2002). 75. Kramer, R. et al. Neuregulins with an Ig-like domain are essential for mouse myocardial and neuronal development. Proc. Natl. Acad. Sci. U. S. A. 93, 4833–8 (1996). 76. Meyer, D. et al. Isoform-specific expression and function of neuregulin. Development 124, 3575–86 (1997). 77. Sandrock, A. W. et al. Maintenance of acetylcholine receptor number by neuregulins at the neuromuscular junction in vivo. Science 276, 599–603 (1997). 78. Wolpowitz, D. et al. Cysteine-rich domain isoforms of the neuregulin-1 gene are required for maintenance of peripheral synapses. Neuron 25, 79–91 (2000). 79. Leu, M. et al. Erbb2 regulates neuromuscular synapse formation and is essential for muscle spindle development. Development 130, 2291–301 (2003). 80. Song, X. M. et al. Muscle fiber type specificity in insulin signal transduction. Am. J. Physiol. 277, R1690-6 (1999). 81. Rumsey, J. W. et al. Tissue engineering intrafusal fibers: dose- and time-dependent differentiation of nuclear bag fibers in a defined in vitro system using neuregulin 1- beta-1. Biomaterials 29, 994–1004 (2008). 82. Cantó, C. et al. Neuregulins mediate calcium-induced glucose transport during muscle contraction. J. Biol. Chem. 281, 21690–7 (2006). 83. Jo, S. a, Zhu, X., Marchionni, M. a & Burden, S. J. Neuregulins are concentrated at nerve-muscle synapses and activate ACh-receptor gene expression. Nature 373, 158–161 (1995). 84. Ma, Z., Wang, J., Song, F. & Loeb, J. A. Critical period of axoglial signaling between neuregulin-1 and brain-derived neurotrophic factor required for early Schwann cell survival and differentiation. J. Neurosci. 31, 9630–40 (2011). 85. Wang, J., Song, F. & Loeb, J. A. Neuregulin1 fine-tunes pre-, post-, and perisynaptic neuromuscular junction development. Dev. Dyn. 246, 368–380 (2017). 86. Lee, Y. Il et al. Neuregulin1 displayed on motor axons regulates terminal Schwann

201

CHAPTER 2: References

cell-mediated synapse elimination at developing neuromuscular junctions. Proc. Natl. Acad. Sci. U. S. A. 113, E479-87 (2016). 87. Falls, D. L. Neuregulins and the neuromuscular system: 10 years of answers and questions. J. Neurocytol. 32, 619–647 (2003). 88. Nishimune, H., Stanford, J. A. & Mori, Y. ROLE of exercise in maintaining the integrity of the neuromuscular junction. Muscle Nerve 49, 315–324 (2014). 89. Lebrasseur, N. K., Coté, G. M., Miller, T. A., Fielding, R. A. & Sawyer, D. B. Regulation of neuregulin/ErbB signaling by contractile activity in skeletal muscle. Am. J. Physiol. Cell Physiol. 284, C1149-55 (2003). 90. Cheret, C. et al. Bace1 and Neuregulin-1 cooperate to control formation and maintenance of muscle spindles. EMBO J. 32, 2015–2028 (2013). 91. Jacobson, C., Duggan, D. & Fischbach, G. Neuregulin induces the expression of transcription factors and myosin heavy chains typical of muscle spindles in cultured human muscle. Proc. Natl. Acad. Sci. U. S. A. 101, 12218–23 (2004). 92. Herndon, C. A., Ankenbruck, N., Lester, B., Bailey, J. & Fromm, L. Neuregulin1 signaling targets SRF and CREB and activates the muscle spindle-specific gene Egr3 through a composite SRF-CREB-binding site. Exp. Cell Res. 319, 718–30 (2013). 93. Golding, J. P., Calderbank, E., Partridge, T. A. & Beauchamp, J. R. Skeletal muscle stem cells express anti-apoptotic ErbB receptors during activation from quiescence. Exp. Cell Res. 313, 341–356 (2007). 94. Florini, J. R., Samuel, D. S., Ewton, D. Z., Kirk, C. & Sklar, R. M. Stimulation of myogenic differentiation by a neuregulin, glial growth factor 2: Are neuregulins the long-sought muscle trophic factors secreted by nerves? J. Biol. Chem. 271, 12699– 12702 (1996). 95. Kim, D. et al. Neuregulin stimulates myogenic differentiation in an autocrine manner. J. Biol. Chem. 274, 15395–15400 (1999). 96. Sandrock, A. W., Goodearl, A. D., Yin, Q. W., Chang, D. & Fischbach, G. D. ARIA is concentrated in nerve terminals at neuromuscular junctions and at other synapses. J. Neurosci. 15, 6124–36 (1995). 97. Ford, B. D., Han, B. & Fischbach, G. D. Differentiation-dependent regulation of skeletal myogenesis by neuregulin-1. Biochem. Biophys. Res. Commun. 306, 276– 281 (2003). 98. Suárez, E. et al. A novel role of neuregulin in skeletal muscle: Neuregulin stimulates glucose uptake, glucose transporter translocation, and transporter expression in muscle cells. J. Biol. Chem. 276, 18257–18264 (2001). 99. Gumà, A., Martínez-Redondo, V., López-Soldado, I., Cantó, C. & Zorzano, A. Emerging role of neuregulin as a modulator of muscle metabolism. Am. J. Physiol. Endocrinol. Metab. 298, E742-50 (2010). 100. Carraro, U. et al. Persistent Muscle Fiber Regeneration in Long Term Denervation. Past, Present, Future. Eur. J. Transl. Myol. 25, 4832 (2015). 101. Nicolino, S. et al. Denervation and reinnervation of adult skeletal muscle modulate mRNA expression of neuregulin-1 and ERBB receptors. Microsurgery 29, 464–472 (2009).

202

CHAPTER 2: References

102. Hirata, M. et al. Increased expression of neuregulin-1 in differentiating muscle satellite cells and in motoneurons during muscle regeneration. Acta Neuropathol. 113, 451–459 (2007). 103. Wiberg, R., Jonsson, S., Novikova, L. N. & Kingham, P. J. Investigation of the Expression of Myogenic Transcription Factors, microRNAs and Muscle-Specific E3 Ubiquitin Ligases in the Medial Gastrocnemius and Soleus Muscles following 2 Peripheral Nerve Injury. PLoS One 10, e0142699 (2015). 104. Das, M., Rumsey, J. W., Bhargava, N., Stancescu, M. & Hickman, J. J. A defined long-term in vitro tissue engineered model of neuromuscular junctions. Biomaterials 31, 4880–4888 (2010). 105. Southam, K. A., King, A. E., Blizzard, C. A., McCormack, G. H. & Dickson, T. C. Microfluidic primary culture model of the lower motor neuron-neuromuscular junction circuit. J. Neurosci. Methods 218, 164–9 (2013). 106. Zahavi, E. E. et al. A compartmentalized microfluidic neuromuscular co-culture system reveals spatial aspects of GDNF functions. J. Cell Sci. 128, 1241–1252 (2015). 107. Taylor, A. M., Rhee, S. W. & Jeon, N. L. in Microfluidic Techniques 321, 167–178 (Humana Press, 2006). 108. Bach, A. D., Beier, J. P. & Stark, G. B. Expression of Trisk 51, agrin and nicotinic- acetycholine receptor epsilon-subunit during muscle development in a novel three- dimensional muscle-neuronal co-culture system. Cell Tissue Res. 314, 263–74 (2003). 109. Morimoto, Y., Kato-Negishi, M., Onoe, H. & Takeuchi, S. Three-dimensional neuron–muscle constructs with neuromuscular junctions. Biomaterials 34, 9413– 9419 (2013). 110. Juhas, M. & Bursac, N. Roles of adherent myogenic cells and dynamic culture in engineered muscle function and maintenance of satellite cells. Biomaterials 35, 9438–9446 (2014). 111. Vilmont, V., Cadot, B., Ouanounou, G. & Gomes, E. R. A system for studying mechanisms of neuromuscular junction development and maintenance. Development 143, 2464–2477 (2016). 112. Schmidt, N. et al. Neuregulin/ErbB regulate neuromuscular junction development by phosphorylation of α-dystrobrevin. J. Cell Biol. 195, 1171–84 (2011). 113. Ngo, S. T., Cole, R. N., Sunn, N., Phillips, W. D. & Noakes, P. G. Neuregulin-1 potentiates agrin-induced acetylcholine receptor clustering through muscle-specific kinase phosphorylation. J. Cell Sci. 125, 1531–43 (2012). 114. Mancuso, R. et al. Neuregulin-1 promotes functional improvement by enhancing collateral sprouting in SOD1G93A ALS mice and after partial muscle denervation. Neurobiol. Dis. 95, 168–178 (2016). 115. Krag, T. O. B. et al. Heregulin ameliorates the dystrophic phenotype in mdx mice. Proc. Natl. Acad. Sci. U. S. A. 101, 13856–60 (2004). 116. Lynch, G. S. Emerging drugs for sarcopenia: age-related muscle wasting. Expert Opin. Emerg. Drugs 9, 345–61 (2004).

203

CHAPTER 3 CARDIAC MUSCLE: ISCHEMIA/REPERFUSION INJURY

3 CONTENT

3.1 INTRODUCTION AND SCIENTIFIC BACKGROUND ...... 207 3.1.1 ISCHEMIA/ REPERFUSION INJURY IN HEART ...... 207 3.1.2 THERAPIES FOR CARDIAC I/R INJURY ...... 209 Conditioning ...... 210 Pharmacological cardioprotection ...... 210 Innovative approaches ...... 211 3.1.3 NRG1/ERBB SYSTEM IN CARDIAC TISSUE ...... 211 NRG1 and ErbB expression and function in heart development ...... 211 NRG1/ ErbB system in adult heart ...... 213 NRG1/ ErbB signalling in injured heart and clinical role of nrg1 ...... 213 3.2 AIM OF THE RESEARCH ...... 217 3.3 SCIENTIFIC PUBLICATION ...... 219 Myocardial ischemia/reperfusion upregulates the transcription of the Neuregulin1 receptor ErbB3, but only postconditioning preserves protein translation: role in oxidative stress...... 219 3.4 DISCUSSION AND FUTURE DIRECTIONS ...... 237 3.5 REFERENCES ...... 240

205

CHAPTER 3: Introduction and Scientific Background

3.1 INTRODUCTION AND SCIENTIFIC BACKGROUND

The World Health Organization declared that cardiovascular disease are the leading worldwide cause of death, representing 31% of all global deaths (referring to 2012). The major cause of these mortality is due to acute myocardial infarction (AMI), which, despite the incessant improvement of therapies, causes the death of 10.000 men and 9500 woman each day1. Moreover the post-infarction reduced left ventricular ejection fraction, an AMI consequent, represents the main cause of chronic heart failure worldwide2. AMI develops when coronary flow is totally or partially interrupted giving myocardial ischemic injury. After AMI, the only way to reduce infarct size is the prompt and successful restore of the blood flow in myocardium; indeed this reperfusion induces myocardial tissue damages, 3 activating a cascade of events that led to cell death and tissue necrosis compromising heart function3,4. These events are indicated as myocardial reperfusion injury. Therefore AMI is also called ischemia/reperfusion (I/R) injury. The mechanisms of I/R - induced cell death are still not completely understood, however they are clearly related to mechanical process and the activation of specific intracellular pathways.

The high incident of I/R injuries and the high mortality associated account for the intense researcher activity on this field. A deeper knowledge of the basal mechanisms and the cellular responses involved in I/R injury will let us to improve the pharmacological approaches proposed in clinics, hopefully reducing mortality related to AMI. Moreover new insight in cardiac function will permit to potentiate also the preventive cardiovascular medicine.

3.1.1 ISCHEMIA/ REPERFUSION INJURY IN HEART

When the abrupt occlusion of epicardial coronary artery occurs, the myocardial tissue located distally to the site of injury becomes ischemic. Prolonged ischemia gives rise to permanent changes in myocardium where cardiac tissue is destroyed and replaced by fibrous scar tissue. When scar tissue interested a large area, the ventricular contractile function is impaired giving a progressive chronic heart failure. The hypoperfused myocardial zone is indicated as area of risk and is characterized by necrotic tissue. The timely restoration of blood flow (reperfusion) limits ischemic damage but induces additional damages, as clearly supported by experimental and clinical evidences. In fact the oxygen and nutrients privation generates a condition in which the re-establishment of blood flow results in inflammation and oxidative damage, rather than restoration of normal functions. Moreover karyolysis, membrane disruption, mitochondrial swelling and disruption are observed in cardiomyocytes, together with interstitial haemorrhage, microvasculature destruction and inflammation reaction5,6. After flow restoration the heart is still not able to supply the required cardiac output, so various morphological modifications of heart tissue occur, collectively named heart remodelling.

207

CHAPTER 3: Introduction and Scientific Background

Figure 3.1 Schematic illustration of cell pathways activated in ischemia/reperfusion injury leading to cell death7

The mechanisms under I/R injury are not completely clear, however several pathways have been detected as possible players in long term damage to cardiomyocytes (Figure 3.1). During ischemia cardiomyocytes undergo to biochemical changes, switching the metabolism to anaerobic form. As the consequence of acidosis due to anaerobic glycolysis, the influx of Na+ increases and intracellular ion accumulation is favoured. The Na+\Ca2+ exchanger reverses its function, determining the entry and the accumulation of Ca2+ at cytoplasm and mitochondrial levels and the consequently loose of intramitochondrial potential difference and the inactivation of electron transport. The rise of Ca2+ results in contracture state maintenance, with risk of rupture of the cell7. Upon reperfusion the mitochondrial membrane potential is re-established determining the uncoupling of oxidative phosphorylation, the increase of Ca2+ uptake into mitochondria and the generation of reactive oxygen species (ROS)8. Moreover the suddenly pH normalization and the increase of Ca2+ activates calpain which digests sarcolemma and the cytoskeleton components9.

The high amount of ROS contributes to the sarcolemma disruption10. Reperfusion is associated with an oxidant and antioxidant imbalance (oxidative stress), due to increased production of ROS and decreased ROS scavenging ability11. ROS are defined as oxygen- containing molecules, unstable and highly reactive for the presence of an unpaired electron

208

CHAPTER 3: Introduction and Scientific Background

(such as superoxide anion radical, O2-•; hydroxyl radical, OH• and hydrogen peroxide, H2O2). Together with ROS, reactive nitrogen species (RNS) are produced during I/R injury. RNS are a group of molecules derived from nitric oxide (NO•), an important bioactive substance produced from L-arginine by nitric oxide synthases (NOS). Examples are dioxide radical (NO2•), peroxynitrite (ONOO-), and nitroxyl (HNO). Both ROS and RNS are physiological produced at low levels by endothelial and myocardial cells and have a role as secondary messengers for cellular homeostasis, mitosis, differentiation and various cell signalling12. During I/R injury three are the main sources of radical generation: i) the mitochondrial electron transport chain, in cardiomyocyte ii) xanthine oxidase (flavoenzyme involved in purine catabolism), in endothelial cells; iii) NADPH oxidase 3 principally in leukocytes11,12. Moreover, cellular defences against oxidative stress is reduced due to a lower activities of enzymes such as superoxide dismutase (SOD), catalase and glutathione peroxidase. At higher uncontrolled concentration ROS and RNS react with lipids, protein and DNA, damaging cells components and activating pathways of stress response. These events trigger the activation of cytokine-mediated cascade and the finally excessive production of the tumor necrosis factor alpha (TNFα), which gives to contractile dysfunctions, hypertrophy and cell death13. Indeed myocardial infarction has been usually associated to necrotic cardiomyocyte death, recently other mechanisms of cell death have been described during I/R injury contributing to infarct zone size, as apoptosis and autophagy. Apoptosis is activated both by extrinsically mechanisms (activation of the sarcolemma receptors FAS or tumor necrosis factor a receptor)13 and by intrinsically pathways given by mitochondrial release of cytochrome c and subsequently activation of caspase signalling14. A pivotal event is the opening of the mitochondrial permeability transition pore (mPTP), a megachannel whose opening is induced by the high concentration of ROS and Ca2+, the restore of neutral pH in cytosol during reperfusion. The consequences are the loss of ATP, the mitochondrial membrane swelling and the activation of mitochondria-mediated apoptosis15,16. Autophagy controls lysosomal degradation and protein recycle; it is activated during I/R injury and its role is controversial. Some studied indicated autophagy as a protective mechanism, however its role in human remain unclear17,18.

3.1.2 THERAPIES FOR CARDIAC I/R INJURY

The standard procedure for patients presenting acute myocardial infarction is the timely restoration of blood flow in the myocardial tissue (reperfusion)1,19. Both mechanical and pharmacological approaches are adopted to perform the reperfusion mainly to reduce the ischemic area. Widely used approaches are angioplasty, thrombolytic therapy and coronary artery bypass graft. Anyway developing new strategies to reduce I/R injury is currently one of the main goals of cardioprotection. The currently investigated approaches are briefly described in the next paragraphs.

209

CHAPTER 3: Introduction and Scientific Background

CONDITIONING A major breakthrough in cardioprotection research was the description of ischemic- preconditioning by Murry et al.20 in 1986. Murry and colleagues reported that brief episodes of ischemia and reperfusion performed before a prolonged ischemic event considerably reduce infarct area in dogs. Almost a decade later, the ischemic postconditioning was described by Zhao et al.21, which demonstrated that brief episodes of ischemia and reperfusion immediately after reflow following sustained ischemia can decrease the infarct area by 30-40% in canine heart. In addition to reduce infarct area, postconditioning gives several protective effects, such as decreasing oxidative stress, levels of myocardial oedema and neutrophil accumulation, as well as preservation of endothelial functions22. This action is exerted through the activation of surface receptors (like bradykinin, adenosine and opioids) in cardiomyocytes and the trigger of pro-survival pathways (as SAFE, RISK and cGMP), which result in cardioprotection mainly by preserving mitochondrial function. Indeed, postconditioning has been shown to inhibit mPTP opening, reduce calcium overload and attenuate oxidative stress23,24. Two years after postconditioning discovery, the first clinical study reported successful effects of this procedure in human heart25. However data obtained from other clinical studies are controversial and high variability was observed. A largest clinical study, with a cohort of 700 patients, showed no beneficial effects of postconditioning procedure22,26. Despite a lot of assumptions have been made, the reasons of variability in postconditioning effects remain to be elucidated, leaving the cardioprotective potential of postconditioning procedure into question.

PHARMACOLOGICAL CARDIOPROTECTION The history of pharmacological cardioprotection has been troubled and disappointing. Current medical therapy may include beta blockers, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, mineralocorticoid receptor antagonists, diuretics, and vasodilators. However significant numbers of patients do not adequately response to cure. Numerous drugs have been investigated, as well as antioxidants, anti‑inflammatory agents, calcium‑channel blockers and erythropoietin; together with molecules acting on mitochondria or modulator of NO signalling. However many of these molecules resulted ineffective in reducing infarct area in AMI or in improving clinical outcomes, highlighting the challenge of translating cardioprotection into clinical advantage22. Among drugs, tested in phase II clinical trials, cyclosporine A, an inhibitors of mPTP opening, looked particularly promising, indeed a later multi centre clinical study showed that no improved clinical outcomes were achieved27. At the moment other compounds obtained good results in proof of concept studies, such as atrial natriuretic peptide, metoprolol or exenatide, which all were able to reduce the infarct area22. Larger studies are needed to confirm the beneficial effects of these pharmacological strategies.

210

CHAPTER 3: Introduction and Scientific Background

INNOVATIVE APPROACHES Various innovative approaches are actually investigated for cardiac disease, such as cell therapy, gene therapy and combination of biomaterial for drugs release. The cardiac regenerative medicine goal is repopulate the injured site with stem cells able to replacing death cells and functional integrating with the rest of heart tissue. Various stem cells were investigated for heart injury and tested in cardiac clinical trials, such as bone marrow- derived mononuclear cells, skeletal myoblasts, hematopoietic stem cells, mesenchymal stem cells, endothelial progenitor cells and cardiac stem cells28. Moreover preclinical investigations were performed with embryonic stem cells and induced pluripotent stem cells. The majority of clinical trials were conducted with bone marrow derived stem cells (BMSC), which in vitro showed high rate of cardiomyocyte differentiation. However the 3 primary mechanism of action of BMSC remain unclear as well as their effective therapeutic function in cardiac injury. Actually the main challenges are reducing cell death rate after transplantation and obtaining the functional integration with local cardiomyocytes28. Recently stem cell therapy has been integrated with other approaches like gene-based therapies, molecular medicine and tissue engineering technology28. Various biomaterials were used to prepare hydrogel or other scaffold for cell easily release in injured heart. Fibrin, methylcellulose, alginate and derivates, hyaluronic acid and decellularized heart matrices are some of the tested materials29. On the other hands, cell-free approaches appeared enticing for their simplicity. Many materials were used to prepare particles as drug carrier for growth factors, cytokines and other molecules involved in cardioprotection, cell recruitment, vasculogenesis and cell proliferation30,31. Nevertheless, several clinical trials failed to reproduce the benefits described in animal models32, raising again the challenges of clinical translation.

3.1.3 NRG1/ERBB SYSTEM IN CARDIAC TISSUE Since its discovery, it was clear that NRG/ErbB system is fundamental for cardiac development, meanwhile the first demonstration of NRG1/ErbB system role in adult heart came from the observation of cardiotoxic side effects of chemotherapeutic agents targeting ErbB receptors. This discover opens a Pandora's box and bring to special attention pointed towards NRG1/ErbB system for the clinical relevance in cancer and cardioprotection medicine. Among chemotherapeutic agents Trastuzumab, a monoclonal anti-ErbB2 antibody used in breast cancer therapy in combination with anthracyclines, is the mainly studied and it has been shown that Trastuzumab-treated patient have a higher incidence of left ventricular systolic dysfunction and heart failure33,34.

NRG1 AND ERBB EXPRESSION AND FUNCTION IN HEART DEVELOPMENT NRG1/ErbB system has been extensively studied in heart development, where it plays an indispensable role in heart tissue organization and zone specification. The study of specific KO mice reveals that ErbB2 and ErbB4 are necessary for ventricular trabeculation, an

211

CHAPTER 3: Introduction and Scientific Background

Figure 3.1 NRG1/ErbB signalling in cardiomyocytes35

essential process for a full-thickness ventricular wall formation, and their depletion bring to premature death during embryogenesis36. Mice with disrupted ErbB3 showed defects in valve formation leading also to embryonic lethality37. ErbB4 resulted important for cardiomyocyte differentiation38. NRG1-null mice (disruption of all NRG1 isoforms) die during midembryogenesis (E 10.5) for absence of normal ventricles formation and valve formation39. Curiously the depletion of NRG1α isoforms is compatible with life, indicating that these isoforms are not necessary for heart development. In the foetal heart soluble NRG1 is released by endocardial endothelium, while ventricular myocytes express ErbB2 and ErbB4 receptors. ErbB3 is present in mesenchymal cells of the endocardial cushion (structure dividing atrium and ventricle) and also in prenatal cardiac myocytes, suggesting a possible additional roles in myocardial development beyond valve formation40,41. NRG1β/ErbB signalling drives the embryonic stem cell differentiation into cardiomyocytes, inducing the cardiac gene expression in both nontrabecular and trabecular myocardium42. Moreover NRG1β plays a role in the development of the cardiac conduction system and the differentiation of foetal cardiomyocytes into cardiac pacemaker-like cells43.

212

CHAPTER 3: Introduction and Scientific Background

NRG1/ ERBB SYSTEM IN ADULT HEART In adult heart various NRG1 isoforms are expressed by cardiac microvascular endothelial cells: at least 3 NRG1α isoforms, whose role is yet unclear, and 8 NRG1β isoforms. Most of NRG1 isoforms expressed are type I, so they require a proteolytic cleavage to be released as active ligands. The expression of ErbB2 and ErbB4 is maintained in adult cardiomyocytes. Regarding ErbB3, recently it has been shown to be expressed in cardiomyocytes44, but its role has not been fully examined. Wide information about NRG1 signalling in adult heart were achieved by the study of conditional knockout mice. Animals without ErbB2 or ErbB4 signalling in heart tissue develop spontaneous dilated cardiomyopathy and are more susceptible to stress like 3 anthracycline antibiotics or pressure overload45,46. Moreover cardiomyocytes of these mice showed defect in cell ultrastructure, like excess mitochondria and thin myofilaments47. It is now clear that NRG1/ErbB system is involved also in physiological adaptation of the heart as response to change in cardiac demand, like in pregnancy heart adaptation. This concept was supported by the observation that the disruption of ErbB signalling in pregnant mice let to increased ventricular dilation and decreased systolic function, with premature death of the animals48. Also the response to physiological stress seem to be related to NRG1 signalling in cardiomyocytes. Indeed physiological hypertrophy induced by exercise in rat corresponds to the upregulation of NRG1, both mRNA and protein, in cardiomyocytes, following by cell hypertrophy and proliferation49. Cardiac NRG1/ErbB signalling regulates the crosstalk between microvascular endothelial cells and cardiac myocytes. In vitro endothelial cells release NRG1 after oxidative stress induced by H2O2 treatment and this action results protective for adult rat ventricular myocytes, as shown by reduced apoptosis50.

Recombinant NRG1β is able to activate ErbB2/ErbB4 in isolated cardiomyocytes with consequent activation of several intracellular pathways, as protein kinase B/PI3K, Src/focal adhesion kinase, MEK/ extracellular signal-regulated kinase and NO synthase (Figure 3.2). These pathways are related to various cellular responses, as cell survival, proliferation, glucose uptake, mitochondrial function, sarcoplasmic reticulum calcium uptake or focal adhesion formation51. NRG1/ErbB system regulates myocardial metabolism through multiple pathways. Similarly to insulin, NRG1β induces glucose uptake in cardiac myocytes through PI3K activation52, as shown for skeletal muscle. Moreover NRG1β regulates the expression and the activity of muscarinic cholinergic receptor in cardiomyocytes, and is able to protects cells from β1-adrenergic receptor–induced death53,54.

NRG1/ ERBB SIGNALLING IN INJURED HEART AND CLINICAL ROLE OF NRG1 Beside in vitro data about NRG1/ErbB signalling in cardiomyocytes, several subsequent in vivo studies reinforced the hypothesis that the NRG1/ErbB system is implicated in the pathophysiology of different cardiac acute and chronic diseases. The ischemia/reperfusion

213

CHAPTER 3: Introduction and Scientific Background injury is a strong activator of NRG1/ErbB signalling in heart tissue inducing an increase in ventricular levels of NRG1 mRNA and protein, together with ErbB4 phosphorylation50,55. In animal model was observed, at early stage of chronic heart disease, an upregulation of NRG1 and ErbB receptors in cardiac tissue; indeed in the later stage of the disease their expression decline, in coincidence with the pump failure and the development of ventricular hypertrophy56. In patients with advanced heart failure increased levels of NRG1 but decreased expression of ErbB2/B4 has been observed in left ventricular tissue57. Further, circulating NRG1β, whose level normally increases after muscular exercise, has been found increased in patients with chronic heart failure and high NRG1 levels correlate with disease severity and risk of death or cardiac transplantation58. These finding have let to the hypothesis that NRG1β may be a potential biomarker in the diagnosis of MI, acute coronary syndrome and general heart failure. However a recent study on a cohort of 319 patients reported that a single serum NRG-1 evaluation is not predictive of a diagnosis of MI or acute coronary syndrome59.

Several animal studies have shown the therapeutic action of NRG1 in heart failure. The observed beneficial effects are not completely clear but might be related to the following mechanisms: 1) cardiomyocyte proliferation60; ii) organization of sarcomeric structure61; iii) limitation of myocardial damage62; iv) promotion of angiogenesis63; v) attenuation of mitochondrial dysfunction64; vi) prevention of apoptosis63; vii) reduction of oxidative stress64; viii) reversal of ventricular remodeling65. In rat MI model, 4 week after injury intravenous injection of NRG1 reduced left ventricular remodeling inhibiting mitochondrial dysfunction and myocyte apoptosis64. Further lentivirus carrying NRG1 gene was injected in rat infarcted myocardium and results showed high angiogenesis and reduction of apoptosis63. Additionally, the prolonged release of NRG1 by microparticles has been shown to induce proliferation of cardiomyocytes, improvement in left ventricular ejection fraction, increased number of arteries and higher progenitor cell recruitment in a mouse model of MI66.

To date, Clinical Trial.gov has listed seven clinical trials evaluating the efficacy of NRG1β administration as cardioprotective agent for the treatment of chronic heart failure. Two isoforms of NRG1 are actually under development for therapeutic use. Neucardin is the short isoform of NRG1β (epidermal growth factor domain fragment). From preliminary clinical data, Neucardin administration resulted in improvement of cardiac structure and function in patients with chronic heart failure67. A phase II and III clinical trials have been recently closed and collected data will show the efficacy of Neucardin68,69. A second NRG1 isoform tested is Cimaglermin alpha, which corresponds to a large full-length recombinant NRG1 type II β3. It was demonstrated that Cimaglermin intravenous infusion (twice for week) improves cardiac function in rat and swine model of myocardial infarction70–72. The tolerability and safety of cimaglermin alpha single intravenous infusions in patient with chronic heart failure has been investigated in a

214

CHAPTER 3: Introduction and Scientific Background clinical trials recently closed. Patients treated with cimaglermin presented a trend towards improvement of left ventricular function. Adverse effects recorded are headache (33%) and nausea (27%), associated mainly at higher dose of the cimaglermin73, as reported for Neucardin.

215

CHAPTER 3: Aim of The Research

3.2 AIM OF THE RESEARCH

The high mortality associated with AMI and the elevated number of patients that do not respond positively to classical therapies underline the necessity to study the intrinsic cardioprotective mechanisms guiding heart tissue regeneration after I/R injury. The basic science research provides increasing information about cardioprotection, on whose bases new clinical approaches can be designed and current therapies can be potentiated.

PostC treatment is a interesting cardioprotective strategy which consists in brief cycles of ischemia/reperfusion carried out after a sustained ischemia. This treatment reduces the infarct size, decreases oxidative stress and preserves endothelial functions. During my 3 doctoral research I studied whether the protective PostC procedure can influence NRG1/ErbB system, which recently has been pointed as a cardioprotective pathway. The protective action of NRG1 is actually studying in clinical trials, however the molecular basis of its action are still unclear. Thus, secondary aim of this research is achieving new insights into NRG1 story of cardioprotection.

217

CHAPTER 3: Scientific Publication

3.3 SCIENTIFIC PUBLICATION

Myocardial ischemia/reperfusion upregulates the transcription of the Neuregulin1 receptor ErbB3, but only postconditioning preserves protein translation: role in oxidative stress.

Michela Moranoa, Carmelina Angottia, Francesca Tullioa, Giovanna Gambarottaa, Claudia Pennaa, 3 Pasquale Pagliaroa and Stefano Geunaa,b. a Department of Clinical and Biological Sciences, University of Turin, Torino, Italy b Neuroscience Institute Cavalieri Ottolenghi (NICO),Torino, Italy

International Journal of Cardiology 2017 Apr 15;233:73-79. doi: 10.1016/j.ijcard.2017.01.122.

ABSTRACT

Neuregulin1 (Nrg1) and its receptors ErbB are crucial for heart development and for adult heart structural maintenance and function and Nrg1 has been proposed for heart failure treatment. Infarct size is the major determinant of heart failure and the mechanism of action and the role of each ErbB receptor remain obscure, especially in the post-ischemic myocardium. We hypothesized that Nrg1 and ErbB are affected at transcriptional level early after ischemia/reperfusion (I/R) injury, and that the protective postconditioning procedure (PostC, brief cycles of ischemia/reperfusion carried out after a sustained ischemia) can influence this pathway. The Langendorff’s heart was used as an ex-vivo model to mimic an I/R injury in the whole rat heart; after 30 min of ischemia and two hours of reperfusion, with or without PostC, Nrg1 and ErbB expression were analyzed by quantitative real-time PCR and Western blot. While no changes occur for ErbB2, ErbB4 and Nrg1, an increase of ErbB3 expression occurs after I/R injury, with and without PostC. However, I/R reduces ErbB3 protein, whereas PostC preserves it. An in vitro analysis with H9c2 cells exposed to redox-stress indicated that the transient over-expression of ErbB3 alone is able to increase cell survival (MTT assay), limiting mitochondrial dysfunction (JC-1 probe) and apoptotic signals (Bax/Bcl-2 ratio). This study suggests ErbB3 as a protective factor against death pathways activated by redox stress and supports an involvement of this receptor in the pro-survival responses.

219

CHAPTER 3: Scientific Publication

INTRODUCTION

Neuregulin1 (Nrg1) is a signalling protein belonging to the Epidermal Growth Factor (EGF) gene family that mediates various cellular processes, such as cell growth, survival, migration, in different cell types, through the ErbB receptors. Nrg1/ErbB system is essential for a correct cardiac development, as demonstrated by the lethal effects of its depletion at embryonic stages 5,17. Furthermore, it is now clear that also in adult heart this signalling plays a critical role in the normal function as well as in ischemia or other pathological conditions 13,18. In the NRG1 gene, alternative splicing gives rise to different isoforms and, in adult heart, cardiac microvascular endothelial cells (EC) express soluble Nrg1 isoforms (type I and type II), both alpha and beta variants 14, which stimulate cell survival and growth 19, glucose up-take 14, protein synthesis and “hypertrophic” gene expression 20. Nrg1 mediated the cross-talk between EC and cardiomyocytes that express ErbB receptors 21, and the deletion of Nrg1 from EC increases the infarct area and the number of TUNEL positive cells after ischemia and reperfusion (I/R) injury 22. ErbB receptors belong to the tyrosine kinase receptor family, and work as dimers. Among the four ErbB receptors only ErbB3 and ErbB4 can bind directly Nrg1; ErbB3 signals only as heterodimer and ErbB2 is the preferred partner for heterodimerization, while ErbB4 can form both homo and heterodimers. Initially only ErbB1, ErbB2 and ErbB4 were thought to be expressed in adult heart. In 2011, Camprecios and colleagues demonstrated that mouse post-natal cardiomyocytes express a functional ErbB3 protein that is localized mainly in the outer areas of T-Tubules with a non-uniform distribution 23. Although a methylation of ERBB3 gene has been observed in human dilated cardiomyopathy 24, the function of ErbB3 in adult heart remains still unknown. Recently, it has been demonstrated that a E3 ligase known as “neuregulin receptor degradation protein-1” (Nrdp1), which targets specifically ErbB3 25, is upregulated after I/R injury. The mouse model over-expressing Nrdp1 is characterized by higher infarct size, increased TUNEL-positive nuclei and inflammatory cells 26. It was postulated that Nrdp1 is a pro-apoptotic signal in heart during I/R injury and that its action is mediated principally by the degradation of ErbB3. For its pro-survival effect Nrg1 has been proposed as a potential drug for heart failure treatment. Several pre-clinical studies in rat or mouse models of heart failure and two clinical trial 27,28 demonstrated that intravenous administration of recombinant soluble Nrg1 improved cardiac contractility and relaxation 29,30, left ventricular remodelling 31, decreased apoptosis 32 and attenuated mitochondrial dysfunction 31. However, the molecular bases of this beneficial effect remain unclear. It has not yet been investigated whether Nrg1/ErbB system is modulated by existing therapeutic strategies aimed to protect the heart against ischemia and reperfusion injury. One of the most interesting cardioprotective strategy is the so-called “Postconditioning” (PostC, i.e., brief cycles of ischemia/reperfusion carried out after a sustained ischemia) 33– 36, which can reduce the infarct size in animal models 37,38 and humans 39–41. The

220

CHAPTER 3: Scientific Publication cardioprotective mechanisms activated by PostC comprise the activation of multiple pathways, including the so-called RISK (Akt/ERK/GSK3-β) and SAFE (JNK/STAT3) pathways 42–44. Since these pathways are downstream of the Nrg1/ErbB system 29, 30, 31, which in PostC context is scarcely investigated, we studied how this system is influenced by I/R and the PostC procedures. We analysed the early changes in Nrg1/ErbB system following ischemia/reperfusion injury in isolated rat heart, with or without the PostC treatment. Moreover, we investigated in a cardiomyocyte cell line whether the ErbB3 receptor plays a role in oxidative stress. Overall results suggest that this receptor can be involved in the myocardial response to ischemia and reperfusion challenge. In particular, we confirmed that ErbB3 is expressed in 3 rat adult heart and we demonstrated, for the first time, that its expression is upregulated in post-ischemic heart, suggesting that ErbB3 protein might play a role in PostC and oxidative injury limitation.

MATERIAL AND METHODS Animals. Male Wistar rats (n=18, 5–6 month old, body weight 450–550 g) were purchased from Harlan (Bresso, MI, Italy). Animals received care in compliance with the Italian law (DL-116, January 27, 1992) and with the European Directive 2010/63/EU on the protection of animals used for scientific purposes. These laws are in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85–23, revised 1996). Langendorff’s heart model. Methods for Langendorff’s isolated rat hearts were similar to those previously described 45,46. In brief, each animal was weighed and treated with heparin (800 U/100 g b.w., i.m.). Then, 10 min afterwards, animal was sacrificed, the heart was rapidly excised and placed in ice-cold buffer solution and weighed. Isolated hearts were retrogradely perfused at constant flow (9 ± 1 ml/min/g) with oxygenated Krebs–Henseleit buffer (127 mM NaCl, 17.7 mM NaHCO3, 5.1 mM KCl, 1.5 mM CaCl2, 1.26 mM MgCl2,

11 mM D-glucose) (Sigma-Aldrich, St. Louis, MO, USA) gassed with 95% O2 and 5%

CO2, paced at 280 bpm and kept in a temperature-controlled chamber (37°C).

Experimental Protocols. After 20 min of stabilization, hearts were randomly divided in three groups: (1) Control group (Sham), hearts were subjected to 150 min perfusion only; (2) I/R group, hearts underwent 30 min of global ischemia and then a period of 120 min full reperfusion; (3) PostC group, after 30 min ischemia, hearts underwent a PostC protocol (5 cycles of 10 sec reperfusion and 10 sec of global ischemia) then a period of 120 min full reperfusion 37.

Myocardial Injury Infarct size. Immediately after reperfusion each heart was quickly removed from the perfusion apparatus and the apical part of the heart was collected and frozen for

221

CHAPTER 3: Scientific Publication biomolecular analysis. The remaining basal part of the left ventricle was dissected into 2-3 mm circumferential slices. Following 20 min of incubation at 37°C in 0.1% solution of nitro-blue tetrazolium in phosphate buffer, unstained tissue was carefully separated from stained tissue by an independent observer. The unstained tissue represents the amount of death cells, the stained tissue represents the viable cells. A gravimetric method was used: the unstained mass was weighed and then expressed as a percentage of left ventricular mass 37,45,46. Reagents necessary to assess myocardial infarction were purchased from Sigma (USA). Lactate dehydrogenase analysis. Since in isolated rat hearts PostC is known to reduce the production of lactate dehydrogenase (LDH) during reperfusion, the release of this enzyme was tested. Samples of coronary effluent (2 ml) were collected with a catheter inserted into the right ventricle via the pulmonary artery. Samples were collected immediately before ischemia during reperfusion. Thereafter samples were collected every 20 min until the end of reperfusion. LDH release was measured as previously described and data are expressed as cumulative values for the entire reperfusion period 37,45. RNA isolation and cDNA preparation. Total RNA was extracted from half apex from each heart, previously dissociated by pestles, with TRIzol (Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions. Glycogen (5 μg) was added as a carrier to facilitate RNA precipitation. For each sample 1μg total RNA was reverse transcripted (RT): samples were warmed at 65 °C for 5 minutes and the reaction was carried out in 20 μl containing: 1 × RT-Buffer (Life Technologies), 0.1 μg/μl bovine serum albumin (BSA, Promega, Fitchburg, WI, USA), 0.5 mM dNTPs, 7.5μM random decamers (Life Technologies), 5mM DTT (Life Technologies), 40 U RNAse Out Inhibitor and 200U SUPERSCRIPT III Rev transcript (Life Technologies,18080044). The reaction was performed for 10 min at 25 °C, 90 min at 50 °C, 15 min at 70 °C. The obtained cDNA was diluted 10 folds with water and stored at -20 °C. Quantitative real-time PCR (qRT-PCR) analysis. Quantitative real-time PCR analysis was performed in a 7300 real-time PCR system (Applied Biosystems). The reaction was carried out in 20 µl containing 5 µl diluted cDNA (corresponding to 25 ng of starting RNA), 1x Sybr Green PCR Master Mix (Bio-Rad, Hercules, CA, USA) and 300nM forward and reverse primers. For each sample a technical triplicate was performed. All primers were designed to amplify specific isoforms of ErbB receptors or Neuregulin-1 and Nrdp1 (Table 1). The reaction was carried out with the following protocol: 30 seconds (sec) at 95°C, 40 cycles of denaturation at 95 °C for 15 sec followed by primer annealing and elongation at 60 °C for 1 minute. Dissociation curve was always analysed to control the quality of the reaction.

Data were analyzed by the ΔΔCt relative quantification method normalizing to the geometric average of two housekeeping genes, Ubiquitin C (UbC) and Hypoxanthine guanine phosphoribosyltransferase (HPRT). We determined the difference between Ct

222

CHAPTER 3: Scientific Publication values of target and housekeeping gene (ΔCt), the difference between the ΔCt values of the samples and the ΔCt mean value of control sample was then calculated (ΔΔCt). The normalized relative quantity (NRQ) was determined using the formula: NRQ=2-(ΔΔCt). NRQ value >1 reflects an increasing of target gene expression, NRQ<1 reflects a decreasing of target gene expression 47. The analysis was carried out in biological quintuplicate (for each experimental group n=5) and results were expressed as mean ± standard deviation. Total protein extraction and Western Blot analysis. Total proteins were extracted from heart apexes with boiling Laemmli buffer (2.5% SDS, 0.125M Tris–HCl pH 6.8) dissociating the tissue with pestles and incubating the protein extract at 100°C for 3 3 minutes. The BCA method was used to quantify proteins and the same amount of protein for each sample was loaded for SDS-PAGE analysis. Proteins were transferred to a nitrocellulose membrane (Bio-Rad, 162-0093). Primary antibodies used are: rabbit polyclonal anti-ErbB3 (working dilution, w.d., 1:1000, sc-285, Santa Cruz, Santa Cruz, CA, USA), anti-ErbB2 (w.d. 1: 1000, sc-284, Santa Cruz), anti-alpha sarcomeric actinin (w.d. 1:2500, A7811, SIGMA), anti-phospho S6 ribosomal protein (w.d. 1:2000, 4858, Cell Signaling). The secondary antibodies used are: ECLTM anti-rabbit IgG (w.d. 1:40000, NA934, GE Healthcare, Little Chalfont, Buckinghamshire, UK) and ECLTM anti-mouse IgG (w.d. 1:40000, NA931, GE Healthcare). The western blot quantitative analysis was performed using the Image J program. Cell culture and in vitro experiments. For the in vitro analysis we used H9C2 cells, a rat myoblast cell line provided by American Collection of Cell Cultures (ATCC® CRL- 1446™ Milan Italy). Cells were cultured in Dulbecco’s Modified Eagle’s Medium Nutrient mixture F-12 HAM (DMEM, D8437, SIGMA) supplemented with 10% fetal bovine serum (FBS, GIBCO, Life Technologies) and 1% (v/v) streptomycin/penicillin (Wisent Inc, Quebec, Canada). To express the ErbB3 protein, the expression vector pcDNA3-B3, kindly provided by Dr. John G. Koland (Department of Pharmacology, University of Iowa, College of Medicine, Iowa City, IA) was used 48. For transient transfections, Lipofectamine 2000 (Life Technologies) was used following manufacturer’s instruction. Briefly, cells were grown in 20cm2 dishes, then medium was changed to Optimem (Life Technologies) and cells were transfected with 4µg plasmid DNA and 4µl Lipofectamine 2000. Control mock transfections were carried out with the empty vector pcDNA3. After 16 hours the medium was changed to normal culture medium and after other 24 hours cells were detached and left to grow in 60 mm dish. For the oxidative stress we treated H9C2, H9C2-ErbB3 and H9C2-mock cells with hydrogen peroxide solution (H202, H1009, SIGMA), used for 2 hours, in a range of concentration 100 µM-300 µM to obtain 40-50% of cell mortality. Untreated H9c2 were

223

CHAPTER 3: Scientific Publication compared to H9c2 exposed to 50 ng/ml Nrg1-β1 (recombinant human NRG1-β1/HRG1-β1 extracellular domain, # 377-HB-050, R&D Systems, Minneapolis, MN). A pre-treatment of the cells with 50 ng/ml Nrg1 for 2 h was also performed, followed by the oxidative stress with hydrogen peroxide solution. Cell survival rate was obtained using the thiazolyl blue tetrazolium bromide (MTT, M2128, SIGMA) assay, following producer’s instruction. Cell survival rate of each cell group were analysed separately. For each independent experiment, data for each condition were calibrated to the control (“no stress”) sample. For Bax and Bcl-2 analysis, cells were seeded on 60 mm dish and exposed to oxidative stress as described above. After 3h of reperfusion (cells in medium with 2% of fetal bovin serum) protein were extracted with boiling Laemmli buffer. Primary antibodies used are: anti-Bax (w.d. 1:300, sc-23959, Santa Cruz, CA, USA); anti-Bcl-2 (w.d. 1:500, sc-492 , Santa Cruz, CA, USA); anti-β actin (w.d. 1:4000, #A5316, Sigma, Germany). JC-1 fluorescent probe (Molecular Probes, Eugene, OR, USA) was used to measure loss of mitochondrial transmembrane potential after the oxidative stress protocol. In brief, subgroup of cells after 1 hour of reperfusion were incubated with medium containing JC-1 (10 μg/ml) at 37°C for 20 min and then washed twice with PBS. Fluorescent values were acquired using a GloMax-Multi Detection System (Promega Corporation, Madison, WI, USA) with 485 nm and 530 nm as the green excitation and emission wavelengths, respectively, and 535 nm and 590 nm as red excitation and emission wavelengths, respectively. The ratio between red and green JC-1 florescence was taken as an index of mitochondrial membrane potential 49. Statistical analysis. Statistical analysis was performed using IBM SPSS program. For real- time PCR data and in vitro data Student’s t-Test was used. For protein quantitative analysis One Way Anova plus Fisher’s LSD (Least Significant Difference) or plus Bonferroni Post Hoc test was used. Data are presented as mean ± SD. A probability value lower than 0.05 was considered as statistically significant.

RESULTS

Taking advantage of the ex vivo Langendorff’s heart model, we investigated the changes that occur in Nrg1 and ErbB expression in heart tissue following a global ischemia/reperfusion challenge (I/R group), or following the postconditioning procedure (PostC group). Infarct mass analysis confirmed the effectiveness of PostC procedure in reducing ischemia/reperfusion injury (48±16% vs 29±3.5% of risk area, in I/R and PostC respectively; p<0.05). LDH analysis corroborated the infarct size data (LDH release was 381±110 in I/R group and 188±80 IU in PostC hearts; p<0.05). The molecular analysis was carried out on five animals for each condition. The expression of Nrg1and ErbB receptors in the left ventricle was analysed using qRT- PCR; the mRNA level in I/R hearts was compared with the mRNA level observed in PostC

224

CHAPTER 3: Scientific Publication hearts. Data show that the four ErbB receptor genes are transcripted in adult rat heart (Fig. 1). After 30 min ischemia and 2 hours reperfusion, the mRNA levels of ErbB1 and ErbB2 did not differ significantly from the Sham (Fig. 1a), regardless of PostC procedure. Similar results were obtained for the four ErbB4 isoforms (Fig. 1b). Conversely, ErbB3 mRNA (Fig. 1a) displayed a 2.8 fold increase after I/R (p<0.05compared to the Sham) and a 2.6 fold increase after PostC (P=0.081 with respect to Sham). Nrg1 transcription was also investigated, discriminating between the two isoforms alpha and beta. qRT-PCR analysis did not reveal changes in transcription among samples, as shown in Fig. 1c. The analysis was then shifted from the mRNA level to the protein level, to investigate 3 ErbB3 protein expression. The western blot analysis confirmed the expression of ErbB3 protein in rat adult heart (Fig. 2a). A quantitative analysis revealed an appreciable downregulation (p=0.063) of ErbB3 in I/R samples compared to the PostC (Fig. 2a), in partial contrast with mRNA analysis. We then analysed a marker of protein translation, namely phospho-S6 protein (p-S6), a subunit of ribosome S40 (when S6 subunit is phosphorylated the translation is permitted). The analysis revealed that I/R samples have a lower level of p-S6 when compared to PostC (p=0.012, Fig. 2a). Yet in PostC sample p-S6 was upregulated respect to the Sham (p<0.05, Fig. 2a). Furthermore, we checked the expression of Nrdp1, the E3 ligase targeting ErbB3, and we found that in I/R group the Nrdp1 mRNA level is significantly higher compared to Sham (2.6± 0.9 fold increase, p=<0.05) but not in PostC group (Fig. 2b). Nrdp1 protein levels were analyzed in a western blot assay: the protein is downregulated in PostC group (p=0.041), but not in I/R group, respect to the Sham group (Fig 2c). A Student's t-test between PostC and I/R group reveals a statistically relevant difference (p< 0.05). These data together could explain the lower amount of ErbB3 protein in I/R samples despite the upregulation of ErbB3 mRNA expression. Intriguingly, PostC preserves ErbB3 protein. Data regarding Nrdp1 suggest that ErbB3 protein absence contributes to the I/R damage and that the presence and stability of ErbB3 protein could protect the tissue. So we examined in vitro the ability of ErbB3 receptor to defend cells from oxidative damage. H9c2 cells were used as a model; because they express ErbB2 and soluble Nrg1, and a barely detectable ErbB3 and ErbB4 (data not shown), they were transiently transfected for the expression of ErbB3 receptor, and were subjected to oxidative stress with hydrogen peroxide. In non-transfected cells the administration of Nrg1, before or simultaneously to the oxidative stress, had no effect on cell survival rate in a MTT test (Fig 3a). The expression of ErbB3 was confirmed with a western blot, together with ErbB2 protein constitutively expressed by H9c2 (Fig. 3b). Results show that ErbB3-expressing cells tolerate oxidative stress better than control cells (H9c2-mock) in each treatment evaluated. Nevertheless, the addition of Nrg1 to the medium did not increase cell survival rate in H9c2-ErbB3 and H9c2-mock cells (Fig. 3c). We analysed the Bax/Bcl-2 ratio as an indicator of cell apoptosis. We found that in H9c2-ErbB3 cells the ratio, after oxidative 225

CHAPTER 3: Scientific Publication stress, remains lower respect to H9c2-mock cells (Fig. 3d,e). Moreover, the analysis of JC- 1 fluorescent probe confirms that ErbB3 expression in H9c2 cells significantly limits the reduction of mitochondrial membrane potential with respect to H9c2-mock cells in all treatment performed (Fig 3f).

Figure 1. ErbB3 mRNA is upregulated in Langendorff's model of heart after ischemia/ reperfusion. The graphs show the results of quantitative real-time PCR analysis on heart apexes after ischemia/reperfusion injury performed in Langendorff's model. (a) ErbB1 and ErbB2 transcription is not affected by the injury or by the postconditioning treatment at the time-point evaluated. ErbB3 transcription is upregulated in I/R. (b) All theErbB4isoforms are not perturbed by theischemia/reperfusion injury. (c) No changes in transcription were detected for Neuregulin1, alpha or beta isoforms. Statistical analysis: One- way ANOVA was not signiﬁcant. Student's t-Test analysis (comparison between Sham andI/R or Sham and PostC).*:p b 0.05.Sham =control group; I/R=ischemia/reperfusion group; PostC = postconditioning group.

226

CHAPTER 3: Scientific Publication

DISCUSSION To determine whether ErbB receptors and Nrg1 are early affected by the ischemia- reperfusion injury and PostC treatment, we analyzed mRNA and protein expression levels in samples collected from the rat heart injured by global I/R, in a Langendorff’s model which, like all experimental paradigms, has advantages and disadvantages50. For instance, Langendorff’s model allows to perform a global ischemia, subjecting the heart to a homogeneous stress insult, and to remove unwanted blood cells, protein and mRNA from the myocardium, thus allowing to perform a “clean” molecular analysis. The expression analysis of ErbB receptors reveals an upregulation of ErbB3 mRNA after ischemic injury, while the expression of the other ErbB receptors is not perturbed. As far 3 as we know, this is the first time that ErbB3 mRNA is shown to be regulated in post- ischemic heart. Since no statistic differences in ErbB3 expression were detectable between PostC and I/R samples, it is likely that this upregulation is mainly a direct early response to the I/R, not modifiable with PostC. Yet, PostC can positively affect the translation machinery, thus preserving ErbB3 protein synthesis. It is known that ErbB2 and ErbB4 mRNA expression is reduced in human and rat heart after heart failure 13,51,52. This was observed also in diabetic animals, showing cardiac dysfunctions 53. It was proposed that the downregulation of ErbB2 and ErbB4 is a late characteristic of the chronic heart injury. Here we focused on the early changes occurring in gene expression after I/R challenge and PostC, suggesting that in the early-phase ErbB2 and ErbB4 are not regulated at transcriptional level. ErbB3 mRNA is upregulated after I/R, while for ErbB3 protein only a tendency to downregulation in I/R group, but not in PostC samples, can be observed. This tendency is considered the effect of a balance between protein synthesis and degradation in the protected and non-protected heart. Our data regarding the level of P-S6 provide evidences that after I/R injury the cell damage affects also the translation machinery, with a reduction of protein production, and that the PostC procedure can rescue this damage. These data support the idea that ribosomal S6 subunit is a convergence point of protective signaling 54. Moreover, Nrdp1 mRNA was upregulated after I/R injury, in line with literature data about Nrdp1 26. It has been shown that mouse model overexpressing Nrdp1 exhibit a downregulation of ErbB3 protein but only after I/R injury 26. Nrdp1 degrades ErbB3 and our data suggest that this mechanism is perturbed after PostC treatment. In fact, we detected lower amount of Nrdp1 protein in PostC samples respect to I/R samples. Directly or indirectly mechanical PostC can influence ErbB3 protein levels, so we can assume that ErbB3 signalling pathway can be part of the protective signalling recruited by PostC treatment. It is well known that PostC activates PI3K-Akt pathway 55. Actually, ErbB3 receptor has six docking sites for the binding of PI3K-p85 and this can result in a higher activation of PI3K-Akt pathway respect to the other ErbB receptors 56,57. Zhang et al.,

227

CHAPTER 3: Scientific Publication observed in mice after coronary artery ligation that the phosphorylation of ErbB3 increases respect to the sham, meanwhile ErbB3 protein is unchanged 26. At the time point analyzed we did not observe ErbB3 or AKT phosphorylation (data not shown), however we cannot rule out that it occurs in an earlier or later time. In literature, data about Nrg1 expression in damaged heart are conflicting: in diabetic animals Nrg1 protein is reduced 53, while in human failing myocardium Nrg1 expression increases 52. In animal models of heart failure, after an initial upregulation, Nrg1 level decreases, coincidently with the development of ventricular hypertrophy and pump failure 58.

Figure 2. Analysis of ErbB3, pS6 protein and Nrdp1 mRNA and protein after ischemia/reperfusion injury with or without post conditioning treatment. (a)The graph shows the results of the protein quantiﬁcation (technical triplicate) performed on heart apexes after ischemia/reperfusion injury; ErbB3 andp-S6proteinswere normalized to alpha sarcomeric actinin protein, using ImageJ program. About ErbB3 a tendency to downregulation can be detected for the I/R samples but not for PostC samples. p-S6 protein results upregulated in PostC samples compared to Sham (p = 0.033) or I/R samples (p = 0.012). The image shows a representative Western blots. (b) The graph shows the results of quantitative real-time PCR analysis. Nrdp1 mRNA is upregulated in I/R group (p = 0.011) respect to the control. (c) The results of a quantitative analysis (technical triplicate) of Nrdp1 protein are visible in the graph. Nrdp1 is downregulated in PostC group (p = 0.041), but not in I/R group. The image shows representative Western blot bands of Nrdp1 protein. Three animal's apexes were analysed for each group. Statistical analysis: One Way ANOVA plus Fisher's LSD test. * p b 0.05.Sham= control group; I/R = ischemia/reperfusion group; PostC = postconditioning group.

228

CHAPTER 3: Scientific Publication

Figure 3. Oxidative stress in H9c2 cells over-expressing ErbB3. (a) The graph shows the results of the thiazolyl blue tetrazolium bromide (MTT) assay in H9c2 cells exposed to oxidative stress (100 μMH 2O2 for 2 h). The addition of Nrg1 (50 ng/ml)in the medium immediately before (for 2 h) or concurrently to the oxidative stress does not affect cell survival. The assay was performed as technical seven-fold and biological triplicate. Statistical analysis refers to One Way ANOVA plus Bonferroni post hoc test. (b) Representative Western blot bands, conﬁrming the expression of ErbB3 protein in transfected H9c2 cells. (c) MTT analysis was performed to assess cell survival rate of H9c2 expressing ErbB3 exposed to oxidative stress. Four conditions were assayed (no stress, H2O2,H 2O2 with Nrg1, pre-treatment with Nrg1). H9c2-ErbB3 tolerate oxidative stress better than control cells (H9c2-mock). (d) The graph shows the ratio between Bax and Bcl-2 protein in H9c2-ErbB3 and -mock cells exposed to oxidative stress. (e) Representative Western blot of 2 independent experiments is shown. (f) The graph illustrates the effects of oxidative stress on mitochondrial membrane potential in H9c2-mock and H9c2-ErbB3 cells, analysed by JC-1 assay. All treatment values were normalized to mean value of no-stress groupeither inH9c2- ErbB3and H9c2-mockcells.Statisticalanalysis: Student'st-Test analysis. *: p b 0.05;**: p b 0.01.

In I/R scenario, cardioprotection by Nrg1 induced pharmacological conditioning has been described: in vivo Nrg1 pre-conditioning activates the protective PI3K/Akt pathway 59. Similarly, post-conditioning with Nrg1, in both in situ and in isolated murine hearts, exerted a cardioprotective effect via PI3K/Akt pathway 60. The same authors evidenced that Nrg1 activates this pathway only after I/R injury but not in sham samples. Therefore, we can argue that these protective effects involved also ErbB3 post-ischemic upregulation.

Recently D’Uva et al., demonstrated that transient expression of ErbB2 in adult mice heart after I/R injury results in a good functional and anatomical regeneration, and the authors postulated that probably treatment with Nrg1 avoids the loss of ErbB2 in cardiomyocytes or activates ErbB2 pathway 61. The pivotal role of ErbB receptors in cell survival after hypoxia/reoxygenation has also been observed using zinc pyrithione, which restores the

229

CHAPTER 3: Scientific Publication basal zinc levels during I/R and prevents apoptosis by activating PI3K/Akt 62. Since in our data ErbB3 mRNA is upregulated after I/R, we can argue that the cellular response to Nrg1 and its beneficial effects are mediated by ErbB2-ErbB3 heterodimer 63. In fact, among ErbB receptors, only ErbB3 and ErbB4 can bind directly Nrg1. Our analysis revealed that Nrg1 mRNA does not change significantly after I/R or PostC, at the time point investigated. Nevertheless, in PostC ErbB3 protein is upregulated and a protective autocrine loop, involving physiological levels of Nrg1 and upregulated ErbB3, can occur. To investigate the role of ErbB3 receptor in the context of redox stress, H9c2 cell model was used. The in vitro assays demonstrated that the expression of ErbB3 receptor in H9c2 cells can increase cell survival rate and can ameliorate mitochondrial resistance to oxidative stress. Nrg1 should stimulate an increase of cell survival rate in H9c2-ErbB3 cells, however H9c2-ErbB3 cells are protected also without exogenous Nrg1 addition, suggesting that ErbB3 is activated by the overexpression or by the endogenous Nrg1. Similarly to what seen for ErbB4, our data suggest that oxidative stress activated ErbB3 signalling and that this receptor is involved in the protective adaptation to cardiac oxidative stress. It is becoming more and more clear that Nrg1/ErbB system is a potential target for therapy in heart failure as highlighted by the promising results obtained in the clinical studies with Nrg1 administration27,28. However, the intrinsic mechanism of Nrg1 action is not completely understood; a better knowledge of Nrg1 and ErbB regulation in ischemia/reperfusion injury is now required to ameliorate the cure and increase the heart performance after acute myocardial infarction.

CONCLUSIONS This is the first study emphasizing ErbB3 mRNA upregulation in the heart subjected to acute I/R challenge. Here, we show that PostC may counterbalance the effects of I/R on ErbB3 protein downregulation and that ErbB3 receptor transfection in cardiomyocytes inhibits death mechanisms activated by redox stress. These novel evidences of a pivotal role for ErbB3 in the context of I/R and redox biology highlights the importance of studies on these receptors not only for the optimization of Nrg1 treatment for heart diseases, but also for the development of new cancer therapies with no side effects on heart. ACKNOWLEDGEMENTS. The study was supported by grant from University of Turin (ex 60% fund).

CONFLICT OF INTEREST. The authors report no relationships that could be construed as a conflict of interest.

230

CHAPTER 3: Scientific Publication

REFERENCES

1. Verma, S. et al. Fundamentals of reperfusion injury for the clinical cardiologist. Circulation 105, 2332–6 (2002). 2. Lee, K. F. et al. Requirement for neuregulin receptor erbB2 in neural and cardiac development. Nature 378, 394–8 (1995). 3. Camenisch, T. D., Schroeder, J. A., Bradley, J., Klewer, S. E. & McDonald, J. A. Heart-valve mesenchyme formation is dependent on hyaluronan-augmented activation of ErbB2–ErbB3 receptors. Nat. Med. 8, 850–5 (2002). 4. Gassmann, M. et al. Aberrant neural and cardiac development in mice lacking the ErbB4 neuregulin receptor. Nature 378, 390–394 (1995). 3 5. Meyer, D. & Birchmeier, C. Multiple essential functions of neuregulin in development. Nature 378, 386–390 (1995). 6. Zhao, Y. Y. et al. Neuregulins promote survival and growth of cardiac myocytes. Persistence of ErbB2 and ErbB4 expression in neonatal and adult ventricular myocytes. J. Biol. Chem. 273, 10261–10269 (1998). 7. Erickson, S. L. et al. ErbB3 is required for normal cerebellar and cardiac development: a comparison with ErbB2-and heregulin-deficient mice. Development 124, 4999–5011 (1997). 8. Lai, D. et al. Neuregulin 1 sustains the gene regulatory network in both trabecular and nontrabecular myocardium. Circ. Res. 107, 715–27 (2010). 9. Rentschler, S. et al. Neuregulin-1 promotes formation of the murine cardiac conduction system. Proc. Natl. Acad. Sci. 99, 10464–10469 (2002). 10. Ozcelik, C. et al. Conditional mutation of the ErbB2 (HER2) receptor in cardiomyocytes leads to dilated cardiomyopathy. Proc. Natl. Acad. Sci. U. S. A. 99, 8880–5 (2002). 11. García-Rivello, H. et al. Dilated cardiomyopathy in Erb-b4-deficient ventricular muscle. Am. J. Physiol. Heart Circ. Physiol. 289, H1153-60 (2005). 12. Lemmens, K., Doggen, K. & De Keulenaer, G. W. Activation of the neuregulin/ErbB system during physiological ventricular remodeling in pregnancy. Am. J. Physiol. Heart Circ. Physiol. 300, H931-42 (2011). 13. Odiete, O., Hill, M. F. & Sawyer, D. B. Neuregulin in cardiovascular development and disease. Circ. Res. 111, 1376–85 (2012). 14. Cote, G. M., Miller, T. A., Lebrasseur, N. K., Kuramochi, Y. & Sawyer, D. B. Neuregulin-1alpha and beta isoform expression in cardiac microvascular endothelial cells and function in cardiac myocytes in vitro. Exp. Cell Res. 311, 135–146 (2005). 15. Kuramochi, Y. et al. Myocyte contractile activity modulates norepinephrine cytotoxicity and survival effects of neuregulin-1beta. Am. J. Physiol. Cell Physiol. 286, C222-9 (2004). 16. Okoshi, K. et al. Neuregulins Regulate Cardiac Parasympathetic Activity: Muscarinic Modulation of -Adrenergic Activity in Myocytes From Mice With 231

CHAPTER 3: Scientific Publication

Neuregulin-1 Gene Deletion. Circulation 110, 713–717 (2004). 17. Negro, A., Brar, B. K. & Lee, K.-F. Essential roles of Her2/erbB2 in cardiac development and function. Recent Prog. Horm. Res. 59, 1–12 (2004). 18. Rupert, C. E. & Coulombe, K. L. The roles of neuregulin-1 in cardiac development, homeostasis, and disease. Biomark. Insights 10, 1–9 (2015). 19. Wadugu, B. & Kuhn, B. The role of neuregulin/ErbB2/ErbB4 signaling in the heart with special focus on effects on cardiomyocyte proliferation. AJP: Heart and Circulatory Physiology 302, H2139–H2147 (2012). 20. Baliga, R. R. et al. NRG-1-induced cardiomyocyte hypertrophy. Role of PI-3- kinase, p70(S6K), and MEK-MAPK-RSK. Am. J. Physiol. 277, H2026–H2037 (1999). 21. Pentassuglia, L. & Sawyer, D. B. ErbB/integrin signaling interactions in regulation of myocardial cell-cell and cell-matrix interactions. Biochim. Biophys. Acta 1833, 909–16 (2013). 22. Hedhli, N. et al. Endothelium-derived neuregulin protects the heart against ischemic injury. Circulation 123, 2254–2262 (2011). 23. Camprecios, G. et al. Expression, localization, and regulation of the neuregulin receptor ErbB3 in mouse heart. J Cell Physiol 226, 450–455 (2011). 24. Haas, J. et al. Alterations in cardiac DNA methylation in human dilated cardiomyopathy. EMBO Mol. Med. 5, 413–429 (2013). 25. Diamonti, a J. et al. An RBCC protein implicated in maintenance of steady-state neuregulin receptor levels. Proc. Natl. Acad. Sci. U. S. A. 99, 2866–2871 (2002). 26. Zhang, Y. et al. Cardiac-specific overexpression of E3 ligase Nrdp1 increases ischemia and reperfusion-induced cardiac injury. Basic Res. Cardiol. 106, 371–383 (2011). 27. Jabbour, A. et al. Parenteral administration of recombinant human neuregulin-1 to patients with stable chronic heart failure produces favourable acute and chronic haemodynamic responses. Eur. J. Heart Fail. 13, 83–92 (2011). 28. Gao, R. et al. A phase II, randomized, double-blind, multicenter, based on standard therapy, placebo-controlled study of the efficacy and safety of recombinant human neuregulin-1 in patients with chronic heart failure. J. Am. Coll. Cardiol. 55, 1907– 1914 (2010). 29. Liu, X. et al. Neuregulin-1/erbB-Activation Improves Cardiac Function and Survival in Models of Ischemic, Dilated, and Viral Cardiomyopathy. J. Am. Coll. Cardiol. 48, 1438–1447 (2006). 30. Hill, M. F. et al. Intravenous Glial Growth Factor 2 (GGF2) Isoform of Neuregulin- 1 beta Improves Left Ventricular Function, Gene and Protein Expression in Rats after Myocardial Infarction. PLoS One 8, (2013). 31. Guo, Y. F. et al. Neuregulin-1 attenuates mitochondrial dysfunction in a rat model of heart failure. Chin. Med. J. (Engl). 125, 807–814 (2012).

232

CHAPTER 3: Scientific Publication

32. Xiao, J. et al. Therapeutic effects of neuregulin-1 gene transduction in rats with myocardial infarction. Coronary Artery Disease 23, 460–468 (2012). 33. Barsukevich V, G. A. Distinct cardioprotective mechanisms of immediate, early and delayed ischaemic postconditioning. Basic Res Cardiol. 110, 452 (2015). 34. Kin, H. et al. Postconditioning attenuates myocardial ischemia-reperfusion injury by inhibiting events in the early minutes of reperfusion. Cardiovasc. Res. 62, 74–85 (2004). 35. Prunier, F. et al. The RIPOST-MI study, assessing remote ischemic perconditioning alone or in combination with local ischemic postconditioning in ST-segment elevation myocardial infarction. Basic Res. Cardiol. 109, (2014). 3 36. Ovize, M. et al. Postconditioning and protection from reperfusion injury: where do we stand?Position Paper from the Working Group of Cellular Biology of the Heart of the European Society of Cardiology. Cardiovasc. Res. 87, 406–423 (2010). 37. Penna, C. et al. Post-conditioning reduces infarct size in the isolated rat heart: Role of coronary flow and pressure and the nitric oxide/cGMP pathway. Basic Res. Cardiol. 101, 168–179 (2006). 38. Mykytenko, J. et al. Persistent beneficial effect of postconditioning against infarct size: Role of mitochondrial KATP channels during reperfusion. Basic Res. Cardiol. 103, 472–484 (2008). 39. Staat, P. et al. Postconditioning the human heart. Circulation 112, 2143–2148 (2005). 40. Garcia, S. et al. Long-term follow-up of patients undergoing postconditioning during ST-elevation myocardial infarction. J. Cardiovasc. Transl. Res. 4, 92–98 (2011). 41. Yetgin, T. et al. Impact of multiple balloon inflations during primary percutaneous coronary intervention on infarct size and long-term clinical outcomes in ST-segment elevation myocardial infarction: Real-world postconditioning. Basic Res. Cardiol. 109, (2014). 42. Cohen, M. V. & Downey, J. M. Signalling pathways and mechanisms of protection in pre- and postconditioning: Historical perspective and lessons for the future. British Journal of Pharmacology 172, 1913–1932 (2015). 43. Boengler, K., Hilfiker-Kleiner, D., Drexler, H., Heusch, G. & Schulz, R. The myocardial JAK/STAT pathway: from protection to failure. Pharmacol. Ther. 120, 172–85 (2008). 44. Heusch, G., Musiolik, J., Gedik, N. & Skyschally, A. Mitochondrial STAT3 activation and cardioprotection by ischemic postconditioning in pigs with regional myocardial ischemia/reperfusion. Circ. Res. 109, 1302–1308 (2011). 45. Pagliaro, P. et al. Nitroxyl affords thiol-sensitive myocardial protective effects akin to early preconditioning. Free Radic. Biol. Med. 34, 33–43 (2003). 46. Penna, C., Mancardi, D., Rastaldo, R., Losano, G. & Pagliaro, P. Intermittent

233

CHAPTER 3: Scientific Publication

activation of bradykinin B2 receptors and mitochondrial KATP channels trigger cardiac postconditioning through redox signaling. Cardiovasc. Res. 75, 168–177 (2007). 47. Gambarotta, G. et al. Identification and validation of suitable housekeeping genes for normalizing quantitative real-time PCR assays in injured peripheral nerves. PLoS One 9, e105601 (2014). 48. Hellyer, N. J., Kim, H. H., Greaves, C. H., Sierke, S. L. & Koland, J. G. Cloning of the rat ErbB3 cDNA and characterization of the recombinant protein. Gene 165, 279–284 (1995). 49. Perrelli, M. G. et al. Catestatin reduces myocardial ischaemia/reperfusion injury: Involvement of PI3K/Akt, PKCs, mitochondrial KATP channels and ROS signalling. Pflugers Arch. Eur. J. Physiol. 465, 1031–1040 (2013). 50. Bell, R. M., Mocanu, M. M. & Yellon, D. M. Retrograde heart perfusion: The Langendorff technique of isolated heart perfusion. Journal of Molecular and Cellular Cardiology 50, 940–950 (2011). 51. Jiang, Z. & Zhou, M. Neuregulin signaling and heart failure. Curr Hear. Fail Rep 7, 42–47 (2010). 52. Rohrbach, S., Niemann, B., Silber, R. E. & Holtz, J. Neuregulin receptors erbB2 and erbB4in failing human myocardium. Depressed expression and attenuated activation. Basic Res. Cardiol. 100, 240–249 (2005). 53. Gui, C., Zhu, L., Hu, M., Lei, L. & Long, Q. Neuregulin-1/ErbB signaling is impaired in the rat model of diabetic cardiomyopathy. Cardiovasc Pathol 21, 414– 420 (2012). 54. Yano, T. et al. Pivotal role of mTORC2 and involvement of ribosomal protein S6 in cardioprotective signaling. Circ. Res. 114, 1268–1280 (2014). 55. Wagner, C., Tillack, D., Simonis, G., Strasser, R. H. & Weinbrenner, C. Ischemic post-conditioning reduces infarct size of the in vivo rat heart: Role of PI3-K, mTOR, GSK-3??, and apoptosis. Mol. Cell. Biochem. 339, 135–147 (2010). 56. Olayioye, M. A., Neve, R. M., Lane, H. A. & Hynes, N. E. The ErbB signaling network: receptor heterodimerization in development and cancer. EMBO J. 19, 3159–3167 (2000). 57. Hellyer, N. J., Kim, M. S. & Koland, J. G. Heregulin-dependent Activation of Phosphoinositide 3-Kinase and Akt via the ErbB2/ErbB3 Co-receptor. J. Biol. Chem. 276, 42153–42161 (2001). 58. Lemmens, K., Doggen, K. & De Keulenaer, G. W. Role of neuregulin-1/ErbB signaling in cardiovascular physiology and disease: Implications for therapy of heart failure. Circulation 116, 954–960 (2007). 59. Fang, S. J. et al. Neuregulin-1 preconditioning protects the heart against ischemia/reperfusion injury through a PI3K/Akt-dependent mechanism. Chin. Med. J. (Engl). 123, 3597–3604 (2010).

234

CHAPTER 3: Scientific Publication

60. Ebner, Lange S, Hollenbach, Steinbronn, Ebner, Fischaleck, Braun-Dullaeus, Weinbrenner, S. R. In Situ Postconditioning With Neuregulin-1β Is Mediated by a PI3K/Akt-Dependent Pathway. Can. J. Cardiol. 31, 76–83 (2015). 61. D’Uva G, Aharonov A, Lauriola M, Kain D, Yahalom-Ronen Y, Carvalho S, Weisinger K, Bassat E, Rajchman D, Yifa O, Lysenko M, Konfino T, Hegesh J, Brenner O, Neeman M, Yarden Y, Leor J, Sarig R, Harvey RP, T. E. ERBB2 triggers mammalian heart regeneration by promoting cardiomyocyte dedifferentiation and proliferation. Nat Cell Biol. 17, 627–38 (2015). 62. Bodiga, V. L., Thokala, S., Vemuri, P. K. & Bodiga, S. Zinc pyrithione inhibits caspase-3 activity, promotes ErbB1-ErbB2 heterodimerization and suppresses 3 ErbB2 downregulation in cardiomyocytes subjected to ischemia/reperfusion. J. Inorg. Biochem. 153, 49–59 (2015). 63. Roskoski, R. The ErbB/HER family of protein-tyrosine kinases and cancer. Pharmacological Research 79, 34–74 (2014).

235

CHAPTER 3: Discussion and Future Directions

3.4 DISCUSSION AND FUTURE DIRECTIONS

Growing body of evidence demonstrates that myocardial reperfusion injury reduces the positive effects, in term of ischemic area, given by the restoration of blood flow after coronary occlusion. Therefore contrasting I/R damage, in order to reduce morbidity and mortality of coronary ischemic disease, represents one of the main challenges of clinical intervention after AMI. In this field postconditioning procedure has been characterized by promising but highly variable results. Thus several pharmacological approaches are now being studied and tested as valid alternative or supportive therapies. Among investigated compounds NRG1 received great attention by researchers and clinicians for its wide range 3 of actions in adult heart. Both NRG1 and PostC activate specifics intracellular pathways, like SAFE or RISK pathways, that have been associated to survival signalling in cardiomyocytes under ischemic damage. We focus our attention on possible influences of PostC procedure on NRG1 and ErbB receptors expression in ischemic heart. We performed the analysis in a ex vivo Langendorff’s heart model of I/R injury. The main evidence, from our data, is that ErbB3 mRNA levels increases after I/R injury, but only postconditioning preserves ErbB3 protein levels. Indeed in postconditioning-protected heart the expression of Nrdp1, an E3 ligase targeting mainly ErbB3, decreases, suggesting a reduced ErbB3 protein elimination. Moreover the phosphorylation of ribosomal protein S6 increases after PostC, indicating the activity of the translation machinery. For my knowledge, it is the first time that the phosphorylation of ribosomal protein S6 is shown in cardiac tissue after postconditioning treatment. Previously Yano et al.74 reported that preconditioning procedure activates ribosomal protein S6, which was indicated as a "convergence point of cardioprotective signaling". Our results about Nrdp1 are coherent with literature data: it has been demonstrated that Nrdp1 increases after I/R injury and mouse animal overexpressing Nrdp1 shows higher infarct size and mortality after I/R injury75. Our data suggest that PostC influences ErbB system, directly or indirectly through Nrdp1, resulting in higher levels of ErbB3 protein respect not-protected hearts. But does ErbB3 be one of the players of protective response activated by PostC treatment? We shortly investigated in vitro how ErbB3 might play a role in protective responses activated in cardiomyocytes under oxidative stress condition. We found that in H9c2 cell model the overexpression of ErbB3 induces an increase in cell survival rate after oxidative stress and ameliorates mitochondrial resistance. Thus ErbB3/ErbB2 dimer might be able to protect cells from oxidative stress-induced apoptosis. Despite supporting this claim, our data do not provide an exhaustive and complete evidence about putative ErbB3 role in I/R injury. Other authors reported that the overexpression of Nrdp1, negative regulator of ErbB3, in HEK293 cell line increases cell death in stressed condition76, suggesting that ErbB3 elimination might contribute to the higher cell mortality. To further understand the role of ErbB3 after I/R injury, ErbB3 expression need to be confirmed in animal models of

237

CHAPTER 3: Discussion and Future Directions myocardial infarction or chronic heart disease, since all literature data are about ErbB2/ErbB4 dimer. Several data about gene expression analysis of heart tissue after myocardial infarction have been already collected by scientific community thank to microarray or deep sequencing analysis77–80; these data can be used for fast and cheap exploration of ErbB3 expression in vivo, taking advantage of the availability of multi- temporal analysis. Finally, conditional ErbB3 knockout mice will be useful to explore the effects of ErbB3 absence in cardiomyocytes after I/R injury. Our analysis was performed using ex vivo heart model of Langherdorff, investigating a single time point after injury. This model has various advantages, like the possibility to induce global ischemia and obtain a clean tissue for molecular analysis without blood cells and circulating compounds. However, as every model, it shows limitations and is far from resembling human heart condition; thus it is important moves the analysis to carefully chosen in vivo animal model and, preferably, to patients. The discrepancy between preclinical model and human is particularly tricky in cardioprotection field. As already mentioned, most of the studied compounds, which in vitro and in pre-clinical studies obtained promising results, fail to be effective in clinical trials. While comprehensive explanation for this unsuccessful translation of positive pre-clinical findings into clinics is not yet achieved, one crucial point is that pre-clinical models cannot recapitulate a complete clinical scenario. We should mention that reperfusion research studies are predominantly performed on healthy animals, besides it is clear from epidemiological evidences that ischemic episode and final infarct size in human are influenced by several risk factors and comorbidities (like obesity, diabetes, hypertension, cigarette smoking, diet and alcohol consumption)11,81. In addition, anatomical and physiological differences among animal models and human might also compromise the possibility to extrapolate data for human therapy. As underline by Ibáñez. and colleagues, the widely used small animal, as mice, have higher heart rate compared to human and, in relationship to heart dimension, they have different oxygen and nutrient diffusion cardiac tissue, which results in variable tolerance to myocardial infarction19. Moreover differences in collateral blood flow influence the infarct area, as well as animal genetic background. Most of these problems are faced also with large animal models. In addiction it seems that also the day and hours of the experiments influence the response of the heart to I/R injury82. Despite technical limitations, from our analysis we can extrapolate two main issues. First, PostC might influence NRG1/ErbB system, at least ErbB3 expression. While new insights are needed, I can hypothesize that PostC procedure accompanied by NRG1 administration might results in synergistic protective effect in cardiac tissue. The second point is the highlight of ErbB3 expression in adult heart and its possible involvement in cell response to I/R injury. The ErbB3 expression in adult heart is particularly relevant considering ErbB receptors importance in cancer therapy. Targeting ErbB receptors has been an established strategy for tumours treatments and recently ErbB3 is pointed as potential pharmacological target in case of resistance to EGFR and ErbB2 directed therapies83. The cardio-toxicity of

238

CHAPTER 3: Discussion and Future Directions cancer therapy is principally associated with anti-ErbB2 treatments, but since the role of cardiac ErbB3 in adult heart remains obscure we cannot exclude a cardio-toxicity side effects also for ErbB3-targetting therapy. Cardiovascular diseases represent the leading cause of mortality worldwide and, despite advances in therapies, the mortality for AMI or for its complications remains elevated. The number of cases have been estimated to increase in relationship with factors such as stress events or air pollution, which more and more influence our life. Body of evidence supports the assertion that climate changes and pollution trigger of AMI84. In the last decade research activity on cardioprotection field has been increased, nevertheless the molecular basis of I/R injury remain largely obscure. In this scenario NRG1 emerges as one of the 3 promising cardioprotective factors. The large literature on NRG1 showed that NRG1/ErbB system, beside being fundamental for correct heart development, plays a role in adult heart homeostasis, metabolism and adaptation51,85. As already mentioned, the expression of NRG1 and ErbB receptors have shown to be altered in AMI and HF in animal model as well as in patients. However it is not clear how this alteration contributes or causes the progression of the disease. At the moment Clinical Trials.gov reports an open recruitment for a large clinical study about NRG/ErbB signalling in human heart (ClinicalTrials.gov Identifier: NCT02820233), which will provide new insights on NRG1/ErbB system role in heart function and disease progression. Regarding NRG1 clinical trials for treatment of AMI and heart failure, collected data provide evidence for anti-remodelling effects and improved cardiac functions in treated patients73,86. NRG1 acts on several cell targets: besides cardiomyocytes, also cardiac fibroblasts, endothelial cells and inflammatory cells are responsive to NRG187. The understanding of NRG1 effects on all cell types that regulates heart remodelling and tissue regeneration, is predominant to design a therapeutic approach that completely take advantage of the wide action of NRG1. However we should mention that, despite evidences about beneficial effects of NRG1 in the setting of heart disease, there are some concerns related to its long term administration. The potential development of cancer requires attention. Furthermore circulating NRG1 might also give side effects at central nervous system level. In order to avoid "off target effects", alternative NRG1 delivery methods can be considered, like the local injection with a matrix or using nanomaterials carrying the compound and targeting specifically the heart tissue.

239

CHAPTER 3: References

3.5 REFERENCES

1. O’Gara, P. T. et al. 2013 ACCF/AHA Guideline for the Management of ST- Elevation Myocardial Infarction: Executive Summary: A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 127, 529–555 (2013). 2. Callender, T. et al. Heart failure care in low- and middle-income countries: a systematic review and meta-analysis. PLoS Med. 11, e1001699 (2014). 3. Verma, S. et al. Fundamentals of reperfusion injury for the clinical cardiologist. Circulation 105, 2332–6 (2002). 4. Yellon, D. M. & Hausenloy, D. J. Myocardial Reperfusion Injury. N. Engl. J. Med. 357, 1121–1135 (2007). 5. Reimer, K. A., Jennings, R. B. & Tatum, A. H. Pathobiology of acute myocardial ischemia: metabolic, functional and ultrastructural studies. Am. J. Cardiol. 52, 72A– 81A (1983). 6. Kloner, R. A. et al. Ultrastructural evidence of microvascular damage and myocardial cell injury after coronary artery occlusion: which comes first? Circulation 62, 945–52 (1980). 7. Murphy, E. & Steenbergen, C. Mechanisms Underlying Acute Protection From Cardiac Ischemia-Reperfusion Injury. Physiol. Rev. 88, 581–609 (2008). 8. Choi, E., Cha, M.-J. & Hwang, K.-C. Roles of Calcium Regulating MicroRNAs in Cardiac Ischemia-Reperfusion Injury. Cells 3, 899–913 (2014). 9. Inserte, J., Hernando, V. & Garcia-Dorado, D. Contribution of calpains to myocardial ischaemia/reperfusion injury. Cardiovasc. Res. 96, 23–31 (2012). 10. Schlüter, K. D., Jakob, G., Ruiz-Meana, M., Garcia-Dorado, D. & Piper, H. M. Protection of reoxygenated cardiomyocytes against osmotic fragility by nitric oxide donors. Am. J. Physiol. 271, H428-34 (1996). 11. Granger, D. N. & Kvietys, P. R. Reperfusion injury and reactive oxygen species: The evolution of a concept. Redox Biol. 6, 524–551 (2015). 12. ZWEIER, J. & TALUKDER, M. The role of oxidants and free radicals in reperfusion injury. Cardiovasc. Res. 70, 181–190 (2006). 13. Kleinbongard, P., Schulz, R. & Heusch, G. TNFα in myocardial ischemia/reperfusion, remodeling and heart failure. Heart Fail. Rev. 16, 49–69 (2011). 14. Orogo, A. M. & Gustafsson, Å. B. Cell death in the myocardium: my heart won’t go on. IUBMB Life 65, 651–6 (2013). 15. Heusch, G., Boengler, K. & Schulz, R. Inhibition of mitochondrial permeability transition pore opening: the Holy Grail of cardioprotection. Basic Res. Cardiol. 105, 151–4 (2010). 16. Ong, S.-B. & Gustafsson, A. B. New roles for mitochondria in cell death in the reperfused myocardium. Cardiovasc. Res. 94, 190–6 (2012). 17. Przyklenk, K. et al. Acute induction of autophagy as a novel strategy for cardioprotection: getting to the heart of the matter. Autophagy 7, 432–3 (2011).

240

CHAPTER 3: References

18. Singh, K. K. et al. Autophagy gene fingerprint in human ischemia and reperfusion. J. Thorac. Cardiovasc. Surg. 147, 1065–1072.e1 (2014). 19. Ibáñez, B., Heusch, G., Ovize, M. & Van de Werf, F. Evolving Therapies for Myocardial Ischemia/Reperfusion Injury. J. Am. Coll. Cardiol. 65, 1454–1471 (2015). 20. Murry, C. E., Jennings, R. B. & Reimer, K. A. Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation 74, 1124–36 (1986). 21. Zhao, Z.-Q. et al. Inhibition of myocardial injury by ischemic postconditioning during reperfusion: comparison with ischemic preconditioning. Am. J. Physiol. Heart Circ. Physiol. 285, H579–H588 (2003). 22. Hausenloy, D. J. & Yellon, D. M. Ischaemic conditioning and reperfusion injury. 3 Nat. Rev. Cardiol. 13, 193–209 (2016). 23. Najafi, M. et al. Inhibition of mitochondrial permeability transition pore restores the cardioprotection by postconditioning in diabetic hearts. J. Diabetes Metab. Disord. 13, 106 (2014). 24. Vinten-Johansen, J. et al. Postconditioning--A new link in nature’s armor against myocardial ischemia-reperfusion injury. Basic Res. Cardiol. 100, 295–310 (2005). 25. Staat, P. et al. Postconditioning the human heart. Circulation 112, 2143–2148 (2005). 26. Hahn, J.-Y. et al. Ischemic postconditioning during primary percutaneous coronary intervention: the effects of postconditioning on myocardial reperfusion in patients with ST-segment elevation myocardial infarction (POST) randomized trial. Circulation 128, 1889–96 (2013). 27. Cung, T.-T. et al. Cyclosporine before PCI in Patients with Acute Myocardial Infarction. N. Engl. J. Med. 373, 1021–1031 (2015). 28. Huang, P., Tian, X., Li, Q. & Yang, Y. New strategies for improving stem cell therapy in ischemic heart disease. Heart Fail. Rev. 21, 737–752 (2016). 29. Weinberger, F., Mannhardt, I. & Eschenhagen, T. Engineering Cardiac Muscle Tissue: A Maturating Field of Research. Circ. Res. 120, 1487–1500 (2017). 30. Pascual-Gil, S., Garbayo, E., Díaz-Herráez, P., Prosper, F. & Blanco-Prieto, M. J. Heart regeneration after myocardial infarction using synthetic biomaterials. J. Control. Release 203, 23–38 (2015). 31. Simon-Yarza, T., Bataille, I. & Letourneur, D. Cardiovascular Bio-Engineering: Current State of the Art. J. Cardiovasc. Transl. Res. (2017). doi:10.1007/s12265- 017-9740-6 32. Formiga, F. R. et al. Angiogenic therapy for cardiac repair based on protein delivery systems. Heart Fail. Rev. 17, 449–473 (2012). 33. Seidman, A. et al. Cardiac Dysfunction in the Trastuzumab Clinical Trials Experience. J. Clin. Oncol. 20, 1215–1221 (2002). 34. Pondé, N. F., Lambertini, M. & de Azambuja, E. Twenty years of anti-HER2 therapy-associated cardiotoxicity. ESMO Open 1, e000073 (2016). 35. Mendes-Ferreira, P., De Keulenaer, G. W., Leite-Moreira, A. F. & Brás-Silva, C. Therapeutic potential of neuregulin-1 in cardiovascular disease. Drug Discovery Today 18, 836–842 (2013).

241

CHAPTER 3: References

36. Lee, K. F. et al. Requirement for neuregulin receptor erbB2 in neural and cardiac development. Nature 378, 394–8 (1995). 37. Camenisch, T. D., Schroeder, J. A., Bradley, J., Klewer, S. E. & McDonald, J. A. Heart-valve mesenchyme formation is dependent on hyaluronan-augmented activation of ErbB2–ErbB3 receptors. Nat. Med. 8, 850–5 (2002). 38. Gassmann, M. et al. Aberrant neural and cardiac development in mice lacking the ErbB4 neuregulin receptor. Nature 378, 390–394 (1995). 39. Meyer, D. & Birchmeier, C. Multiple essential functions of neuregulin in development. Nature 378, 386–390 (1995). 40. Zhao, Y. Y. et al. Neuregulins promote survival and growth of cardiac myocytes. Persistence of ErbB2 and ErbB4 expression in neonatal and adult ventricular myocytes. J. Biol. Chem. 273, 10261–10269 (1998). 41. Erickson, S. L. et al. ErbB3 is required for normal cerebellar and cardiac development: a comparison with ErbB2-and heregulin-deficient mice. Development 124, 4999–5011 (1997). 42. Lai, D. et al. Neuregulin 1 sustains the gene regulatory network in both trabecular and nontrabecular myocardium. Circ. Res. 107, 715–27 (2010). 43. Rentschler, S. et al. Neuregulin-1 promotes formation of the murine cardiac conduction system. Proc. Natl. Acad. Sci. 99, 10464–10469 (2002). 44. Campreciós, G. et al. Expression, localization, and regulation of the neuregulin receptor ErbB3 in mouse heart. J. Cell. Physiol. 226, 450–455 (2011). 45. Ozcelik, C. et al. Conditional mutation of the ErbB2 (HER2) receptor in cardiomyocytes leads to dilated cardiomyopathy. Proc. Natl. Acad. Sci. U. S. A. 99, 8880–5 (2002). 46. García-Rivello, H. et al. Dilated cardiomyopathy in Erb-b4-deficient ventricular muscle. Am. J. Physiol. Heart Circ. Physiol. 289, H1153-60 (2005). 47. Crone, S. A. et al. ErbB2 is essential in the prevention of dilated cardiomyopathy. Nat. Med. 8, 459–465 (2002). 48. Lemmens, K., Doggen, K. & De Keulenaer, G. W. Activation of the neuregulin/ErbB system during physiological ventricular remodeling in pregnancy. Am. J. Physiol. Heart Circ. Physiol. 300, H931-42 (2011). 49. Waring, C. D. et al. The adult heart responds to increased workload with physiologic hypertrophy, cardiac stem cell activation, and new myocyte formation. Eur. Heart J. 35, 2722–31 (2014). 50. Kuramochi, Y. et al. Cardiac endothelial cells regulate reactive oxygen species- induced cardiomyocyte apoptosis through neuregulin-1beta/erbB4 signaling. J. Biol. Chem. 279, 51141–51147 (2004). 51. Odiete, O., Hill, M. F. & Sawyer, D. B. Neuregulin in cardiovascular development and disease. Circ. Res. 111, 1376–85 (2012). 52. Cote, G. M., Miller, T. A., Lebrasseur, N. K., Kuramochi, Y. & Sawyer, D. B. Neuregulin-1alpha and beta isoform expression in cardiac microvascular endothelial cells and function in cardiac myocytes in vitro. Exp. Cell Res. 311, 135–146 (2005). 53. Kuramochi, Y. et al. Myocyte contractile activity modulates norepinephrine cytotoxicity and survival effects of neuregulin-1beta. Am. J. Physiol. Cell Physiol.

242

CHAPTER 3: References

286, C222-9 (2004). 54. Okoshi, K. et al. Neuregulins Regulate Cardiac Parasympathetic Activity: Muscarinic Modulation of -Adrenergic Activity in Myocytes From Mice With Neuregulin-1 Gene Deletion. Circulation 110, 713–717 (2004). 55. Fang, S.-J. et al. Neuregulin-1 preconditioning protects the heart against ischemia/reperfusion injury through a PI3K/Akt-dependent mechanism. Chin. Med. J. (Engl). 123, 3597–604 (2010). 56. Lemmens, K., Doggen, K. & De Keulenaer, G. W. Role of neuregulin-1/ErbB signaling in cardiovascular physiology and disease: Implications for therapy of heart failure. Circulation 116, 954–960 (2007). 57. Rohrbach, S., Niemann, B., Silber, R.-E. & Holtz, J. Neuregulin receptors erbB2 3 and erbB4 in failing human myocardium -- depressed expression and attenuated activation. Basic Res. Cardiol. 100, 240–9 (2005). 58. Ky, B. et al. Neuregulin-1 beta is associated with disease severity and adverse outcomes in chronic heart failure. Circulation 120, 310–7 (2009). 59. Yiadom, M. Y. A. B. et al. Diagnostic Utility of Neuregulin for Acute Coronary Syndrome. Dis. Markers 2016, 8025271 (2016). 60. Bersell, K., Arab, S., Haring, B. & Kühn, B. Neuregulin1/ErbB4 signaling induces cardiomyocyte proliferation and repair of heart injury. Cell 138, 257–70 (2009). 61. Gu, X. et al. Cardiac functional improvement in rats with myocardial infarction by up-regulating cardiac myosin light chain kinase with neuregulin. Cardiovasc. Res. 88, 334–43 (2010). 62. Liu, X. et al. Neuregulin-1/erbB-Activation Improves Cardiac Function and Survival in Models of Ischemic, Dilated, and Viral Cardiomyopathy. J. Am. Coll. Cardiol. 48, 1438–1447 (2006). 63. Xiao, J. et al. Therapeutic effects of neuregulin-1 gene transduction in rats with myocardial infarction. Coronary Artery Disease 23, 460–468 (2012). 64. Guo, Y. F. et al. Neuregulin-1 attenuates mitochondrial dysfunction in a rat model of heart failure. Chin. Med. J. (Engl). 125, 807–814 (2012). 65. Gui, C., Zhu, L., Hu, M., Lei, L. & Long, Q. Neuregulin-1/ErbB signaling is impaired in the rat model of diabetic cardiomyopathy. Cardiovasc Pathol 21, 414– 420 (2012). 66. Formiga, F. R. et al. Controlled delivery of fibroblast growth factor-1 and neuregulin-1 from biodegradable microparticles promotes cardiac repair in a rat myocardial infarction model through activation of endogenous regeneration. J. Control. Release 173, 132–139 (2014). 67. Gao, R. et al. A phase II, randomized, double-blind, multicenter, based on standard therapy, placebo-controlled study of the efficacy and safety of recombinant human neuregulin-1 in patients with chronic heart failure. J. Am. Coll. Cardiol. 55, 1907– 1914 (2010). 68. Zensun Sci. & Tech. Co., L. A Phase III, Multi-center,Randomized, Double-blind, Based on Standard Therapy, Placebo-controlled Study of the Survival of Recombinant Human Neuregulin-1β in Patients With Chronic Systolic Heart Failure. ClinicalTrialsgov NCT01131637

243

CHAPTER 3: References

69. Zensun Sci. & Tech. Co., L. A Randomized, Parallel, Placebo-controlled, Double- blind Phase IIa Study of Efficacy and Safety of Recombinant Human Neuregulin-1 (Neucardin) in Subjects With Stable Chronic Heart Failure. ClinicalTrialsgov. NCT01251406 70. Hill, M. F. et al. Intravenous Glial Growth Factor 2 (GGF2) Isoform of Neuregulin- 1 beta Improves Left Ventricular Function, Gene and Protein Expression in Rats after Myocardial Infarction. PLoS One 8, (2013). 71. Galindo, C. L. et al. Anti-remodeling and anti-fibrotic effects of the neuregulin-1β glial growth factor 2 in a large animal model of heart failure. J. Am. Heart Assoc. 3, e000773 (2014). 72. Parry, T. J. et al. Effects of neuregulin GGF2 (cimaglermin alfa) dose and treatment frequency on left ventricular function in rats following myocardial infarction. Eur. J. Pharmacol. 796, 76–89 (2017). 73. Lenihan, D. J. et al. A Phase I, Single Ascending Dose Study of Cimaglermin Alfa (Neuregulin 1β3) in Patients With Systolic Dysfunction and Heart Failure. JACC Basic to Transl. Sci. 1, (2016). 74. Yano, T. et al. Pivotal role of mTORC2 and involvement of ribosomal protein S6 in cardioprotective signaling. Circ. Res. 114, 1268–1280 (2014). 75. Zhang, Y. et al. Cardiac-specific overexpression of E3 ligase Nrdp1 increases ischemia and reperfusion-induced cardiac injury. Basic Res. Cardiol. 106, 371–383 (2011). 76. Zhou, A., Pan, D., Yang, X. & Zhou, J. Overexpression of Nrdp1/FLRF sensitizes cells to oxidative stress. Biochem. Biophys. Res. Commun. 410, 771–4 (2011). 77. Tulacz, D. et al. Transcriptional profiling of left ventricle and peripheral blood mononuclear cells in a rat model of postinfarction heart failure. BMC Med. Genomics 6, 49 (2013). 78. Kuster, D. W. D. et al. Left ventricular remodeling in swine after myocardial infarction: a transcriptional genomics approach. Basic Res. Cardiol. 106, 1269–81 (2011). 79. Suresh, R. et al. Transcriptome from circulating cells suggests dysregulated pathways associated with long-term recurrent events following first-time myocardial infarction. J. Mol. Cell. Cardiol. 74, 13–21 (2014). 80. Harpster, M. H. et al. Earliest changes in the left ventricular transcriptome postmyocardial infarction. Mamm. Genome 17, 701–15 (2006). 81. Anderson, J. L. & Morrow, D. A. Acute Myocardial Infarction. N. Engl. J. Med. 376, 2053–2064 (2017). 82. Suárez-Barrientos, A. et al. Circadian variations of infarct size in acute myocardial infarction. Heart 97, 970–6 (2011). 83. Zhang, N., Chang, Y., Rios, A. & An, Z. HER3/ErbB3, an emerging cancer therapeutic target. Acta Biochim. Biophys. Sin. (Shanghai). 48, gmv103 (2015). 84. Claeys, M. J., Rajagopalan, S., Nawrot, T. S. & Brook, R. D. Climate and environmental triggers of acute myocardial infarction. Eur. Heart J. 38, ehw151 (2016). 85. Rupert, C. E. & Coulombe, K. L. The roles of neuregulin-1 in cardiac development,

244

CHAPTER 3: References

homeostasis, and disease. Biomark. Insights 10, 1–9 (2015). 86. Gao, R. et al. A Phase II, randomized, double-blind, multicenter, based on standard therapy, placebo-controlled study of the efficacy and safety of recombinant human neuregulin-1 in patients with chronic heart failure. J. Am. Coll. Cardiol. 55, 1907– 14 (2010). 87. Galindo, C. L., Ryzhov, S. & Sawyer, D. B. Neuregulin as a heart failure therapy and mediator of reverse remodeling. Curr. Heart Fail. Rep. 11, 40–9 (2014).

245

GENERAL CONCLUSIONS

by Michela Morano

247

CONCLUSIONS

During my PhD experience I focused my research on growth factor role in tissue repair and their delivery by biomaterials to improve the regenerative process outcome. The PhD program has been integrated with an international internship at the MHH Institute of Hannover and an interdisciplinary collaboration with the group of Physiologists and with private companies, which provided the tested biomaterials. Overall experience permits to understand the great value of collaboration and interdisciplinary approaches to solve technical problems currently facing in bioengineering field. I showed in this thesis that growth factor-based therapy, with engineered proteins, might be a powerful tool to enhance tissue regeneration. Future technological progresses and biological researches will provide novel insights in the regenerative medicine and will permit the clinical transposition of promising growth factor-therapy.

Regenerative medicine aims to replace or repair damaged tissues and organs to restore the normal function and represents a promising treatment for traumatic injuries and degenerative diseases, circumventing the need for donation. In the last decades regenerative medicine has become highly interdisciplinary, holding the integration of traditional scientific disciplines, as biology and chemistry, with novel emerging research fields like mechanobiology, synthetic biology, mathematical modelling and biomimetic material science. The cellular and molecular complexity and the spatial-temporal features of each well defined step in the regeneration process make the regenerative medicine the perfect target for further developments and applications of novel bioengineering solutions. Thus, in the last decades, the regenerative medicine evolves in more integrated therapeutic approaches based on novel biomaterials and implants. Among investigated materials, chitosan emerged as a promising candidate for medical and pharmaceutical applications, thanks to its biodegradability, biocompatibility and non-toxicity. As described in this thesis, in the peripheral nerve regeneration field, chitosan-based conduits are now accepted for clinical use achieving good results in the reconstruction of digital nerves and, thus, representing an optimal basis for the development of more complex scaffolds supporting nerve regeneration along great distances. The increasing efforts to mimic the natural environment of the regenerating tissue using biomaterials led to novel discoveries about the chemical and physical nature of cell- matrix crosstalk, improving our knowledge about cell interaction with the surrounding environment. Results of these findings give rise to 3D cell culture for more efficient in vitro studies, and to the development of biomaterials with tuneable stiffness which can adapt to specific cell types. In this regard, hydrogels are a suitable tool for 3D cultures, as demonstrated by our in vitro studies with DRG cultures on NVR hydrogel. However, our data suggest that for in vivo application the hydrogel composition and stiffness need to be rethought to avoid side effects as physical obstruction of regeneration. Further studies will permit to define more complex matrices whose features can be modulated to better support the cell behaviour required in a precise regeneration step.

249

CONCLUSIONS

Tissue regeneration is a sequential phase process, which includes inflammation, degradation, tissue formation and maturation, involving the orchestrated action of several molecules and the interaction of various cell types. All these regeneration steps might be guided and enhanced by the scaffold itself, whose definition shifted from a passive and only mechanical support to an active device that takes part to the regenerative process. As described in this thesis, we investigated the enrichment of chitosan tube for the repair of long nerve defect in rat, showing that growth factor delivery can be an optimal strategy to enhance the nerve regeneration inside the scaffold. Growth factors are naturally secreted peptides which act in an autocrine way or are trapped by the extracellular matrix for a paracrine signalling. The growth factor-mediated effects are essential to the regenerative process. In fact, the growth factor binding to cell receptors activates specific intracellular pathways that trigger several cell behaviours, such as cell migration, survival, dedifferentiation, differentiation, adhesion or proliferation. An intricate network of growth factors signalling determines the successfully tissue progression into sequential events that trigger the regeneration. Thus, growth factors represent a good therapeutic target. The basic sciences is a precious research field, which permits the continuous progression of our knowledge about cell signalling and the identification of new therapeutic targets for more effective therapies. As we showed, NRG1 is widely investigated for its therapeutic applications in cardiovascular disease, however the bases of its action and the intracellular mechanisms activated are largely unknown. We demonstrated that ErbB3 receptor is upregulated after ischemia/reperfusion injury, suggesting its role in oxidative response after injury. Thus, even if further investigations on its role are needed, ErbB3 might be considered as a new therapeutic target for acute myocardial infarction.

While some growth factors have been accepted for clinical use, such as human growth hormone (hGH; Humatrope®), administrated to children with short stature, or platelet- derived growth factor-BB (PDGF-BB; Regranex®) used for treatment of lower extremity diabetic neuropathic ulcers, or bone morphogenetic factor-2 (BMP-2) (InFUSE™ Bone Graft/LT-Cage™) for lumbar spine fusion; generally, the clinical application of growth factors is hindered by inherent limitations related to their native protein forms. Main limitations are given by short half-life, low protein stability and rapid cellular up-take. However, further challenges arise from exogenous growth factor delivery for therapeutic use, such as high cost of production, poor expression of recombinant protein in yeast, elevated cost and difficulty of purification. Moreover, also the inadequate methods of growth factor delivery have determined the limited success of the clinical transposition of the growth factor-based therapies. Collectively, these restrictions create a need for new technologies and approaches that will turn growth factors into more amenable tools for clinic use. The poor factor stability can be overtaken for example by the conjugation with nanoparticles, as we demonstrated for GDNF and NRG1 conjugated with iron-oxide nanoparticles. Other strategies might be the release by microspheres or the embedding into

250

CONCLUSIONS hydrogel, moreover recombinant factors can be produced with site specific mutation that increase the stability or removing the protease cleavage sites. Also the production methods can be simplified by using small domain of growth factor instead of the entire protein, as we demonstrated for the extracellular domain of ErbB4 receptors (ecto-ErbB4) which retains its biological activity.

The factor delivery strategy is strictly related to the concept of factor concentration. It remains an open question which is the optimal concentration for each single factor to be effective avoiding side-effects in a defined regenerative process. As highlighted also by our analysis of different concentration of conjugated factors and their effects on neurite outgrowth, from in vitro and in vivo analysis is clear that the effect exerted by specific growth factors might depend on their concentration. Several strategies have been proposed to increase scaffold retention of factors such as the use of ECM components or fragments that are recognized by specific domains in growth factor sequences. Hydrogel mediated release is also a possible strategy, even if the accessibility of the factor depends on hydrogel degradability. However, our inability to sustain a release comparable to physiological factor concentration is probably one of the main causes for unsuccessful results of growth factor therapies. To bypass this problem, it has been proposed the use of stem cell secretome as a suitable tool for growth factor delivery; however, the efficacy and the applicability of this method need to be further investigated. Furthermore, the current proposed therapies are overly simplistic, mostly delivering only one factor. Generally, a mix of factors might recapitulate the biological conditions and act on several cellular types. Thus, increased efforts need to be done to easily test in vitro various factor combinations and factor concentration to find the optimal conditions for in vivo application, because not all the combinations might result in synergistic effects, as we showed in our in vitro testing with conjugated factors.

A key area of future research will be the temporal and spatial control of factor delivery, which can be important to localize factor effects. As we observed for NRG1 in denervated muscle, the expression of growth factors is finely regulated in the regenerative process with specific timing and expression patterns, that probably reflects the action-window of growth factor. Multilayer scaffolds, composed by materials with different degradation kinetics, have been proposed for the temporal control of factors delivery, but the technique needs to be optimized. Some strategies to immobilize factors on scaffold surface have been investigated, such as the use of chemical crosslinkers that covalently bind growth factors to free chemical groups of the biomaterials. However, these methods do not permit a fine control of protein spatial distribution and are not suitable for a mix of factors with different kinetic of action. The novel technology of the 3D printing might be useful to finely decide the scaffold structure and the spatial distribution of growth factors using different bioinks, however further technical improvements are required.

251

CONCLUSIONS

It has to be considered that specific regeneration process might require specific needs. For example, in nerve regeneration inside tubular nerve device the orientation of the matrix and the sequential release of factors have emerged as prominent issues to be addressed for optimal nerve regeneration. In ischemic cardiac tissue one important topic is the choice of the delivery strategy, as direct intramiocardial injection of growth factors can be dangerous for the fragile heart wall integrity and it has been demonstrated to have minimal effects on angiogenesis, but the hydrogel release might physically impede the tissue repair. Moreover, a challenge is obtain a matrix that has mechanical properties similar to heart and can both be used as delivery system and as structural support to the damaged heart. In skeletal muscle regeneration, the release of stabilized growth factors or a gene therapy approach results preferable after denervation, whereas hydrogel injection and cell-based therapy might be useful in case of great loss of muscle mass.

To date regenerative medicine let to the development of new therapeutic approaches, with the FDA approval, to treat several pathologies. Anyway, various improvements are necessary for the further advancement of regenerative medicine field. The current goal is to establish a perfect combination of biomaterials, growth factors and cells. Furthermore, more attention needs to be pointed out on the challenges and need for clinical use of these biomaterials or engineered factors, already at the beginning step of the design stage. The future will see significant efforts to the production of new biomaterials combined with growth factors or cells under safety and efficacy standards with a well defined quality control programme. Moreover, advancements of our knowledge about tissue regeneration in mammals and other species will provide further useful information on biological mechanisms, suggesting novel concepts that might be transferred to the tissue bioengineering field. The knowledge of the influence on the regenerative process exerted by factors such as age, microbiome or disease state of the patients, which can determine the choice of a specific therapy instead of another, will be interesting for future advances. The great advances done in the last few years in regenerative medicine suggest that new breakthroughs will be make in the next years bringing us close to reach the full potential of the growth factors therapy combined with biomaterial and cell approaches.

252

ACKNOWLEDGMENTS

I would like to thank all the people that contributed to the projects presented here. Dear Prof. Geuna, my tutor, I would like to thank you for the possibility to work in your laboratory, to take part to an European Project and to be immersed in an international and stimulating atmosphere. Thanks also for your trust in my personal abilities. Dear Prof. Grothe, my sincere thanks for the five months that I spent in your group and for the excellent science, the defined methods and the great organization that I had the opportunity to know and be part of. Thanks also to Prof. Kirsten Haastert-Talini for her patience and kindly support. Thanks to all the people of Prof. Grothe's group, especially Sandra and Cora, which welcomed me in such a kindly way. Dear Prof. Gambarotta, or better Dear Giò, you have helped me in countless ways. Your patience, great knowledge and brilliant mind gave me the opportunity to growth as a biologist. Your passion and dedication were an inspiration for me. Thanks for all!! Thanks also for our travels by car together! Dear Benedetta, dear Davide, what a wonderful adventure was this PhD with you in the lab, sharing all the disappointments as well as the happiest moments. Thanks Betta for our conversations; your dedication and your organization, but above all your professional growth in these years has been for me a great example. I wish you all the best! Dear Stefania, Marwa, Giulia, Federica, Luisa, Sara and Nicoletta, you taught me that research means collaboration, and that the growth of people that surround us must be always our first caring. I spent more time with you that with my family.. so thank you for being my lab family in these years! Special thanks to Marwa, for your kindness and for being my adventure mate for Schwann cell cultures!!! Dear Valentina, your passion, your never ending motivation and your help in the last year have been for me really precious. I greatly appreciated our discussions about so many topics. Dear all from the group of Prof. Pagliaro, many thanks for our pleasant collaboration. Dear colleagues from the Department of Clinical and Biological Science, it was great to be surrounded by all of you. Especially, thanks to Francesca for our conversations in cell culture room and for having so much funny time on the 43 bus in the morning! To my friends Giulia, Matteo, Samuel and Emily: thank you for every moments of service together in these years, because, even if far from the scientific field, this was an important part of my personal growth, conditioning my life in the lab. To my parents and my sister Elisa, thank you for supporting my past and future choices. To my husband Stefano, thank you for your unconditional trust and for being my compass during so many adventures. Our wedding will be always the best memory of these PhD years!!

255