Demonstrating the Importance of Membrane Repair in Response to Disease and Injury

Demonstrating The Importance Of Membrane Repair In Response To Disease And Injury

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy

in the Graduate School of The Ohio State University

Brian J. Paleo

Graduate Program in Biomedical Sciences Graduate Program

The Ohio State University

2020

Dissertation Committee

Dr. Noah Weisleder, Advisor

Dr. Brandon Biesiadecki

Dr. Anthony Brown

Dr. Federica Accornero

Copyrighted by

Brian J. Paleo

2020

Abstract

Most cells in the human body have the capacity to reseal their cellular membranes following a disruption of the lipid bilayer. This membrane repair response involves a coordinated chain of events that are essential to maintain cellular homeostasis and prevent cell death. Membrane repair initially emerged as an important field in skeletal muscle physiology, however as the field continues to mature the importance of membrane repair has been noted in many cell types and in the context of injuries related to bacterial toxins, ischemic events, and traumatic injury. As the field begins to expand, it is important to understand the value of membrane repair in the tissue studied and if proteins known to be involved in membrane repair appear in specific tissues. Studies from our laboratory group and others demonstrated that mitsugumin 53 (MG53), a muscle-enriched tripartite motif (TRIM) family protein also known as TRIM72, is an essential component of the cell membrane repair machinery in striated muscle. In an effort to demonstrate the value of both MG53/TRIM72 and membrane repair in

Duchenne muscular dystrophy (DMD), I studied the effects of genetic knock out of

MG53/TRIM72 in the mdx mouse model of DMD. We observed muscle pathology consistent with the mdx mouse following the initial phase of pathology at 6 weeks of age.

However, aging of the mice and resulting accumulation of repeated bouts of injury due to

ii the lack of dystrophin protein led to robust fibrosis throughout skeletal and cardiac muscles.

To expand the knowledge of membrane repair in tissues other than skeletal muscle we investigated if increasing membrane repair can have protective effects in the peripheral nervous system. Since many neurons are terminally differentiated, increasing cell survival following injury may minimize the impact of these injuries and provide the translational potential for the treatment of neuronal diseases. While several cell types are known to survive injury through plasma membrane repair mechanisms there has been little investigation of membrane repair in neurons and even fewer efforts to target membrane repair as a therapy in neurons. Interestingly, recombinant human MG53

(rhMG53) can be applied exogenously to increase the membrane repair capacity of various cell types both in vitro and in vivo. Thus, we assessed the therapeutic potential of rhMG53 to increase membrane repair in cultured neurons and in an in vivo mouse model of neurotrauma. We found that a robust repair response exists in various neuronal cells and that rhMG53 can increase neuronal membrane repair both in cultured cells and a mouse model of peripheral nerve injury. These findings provide direct evidence of conserved membrane repair responses in neurons and that these repair mechanisms can be targeted as a potential therapeutic approach for neuronal injury.

The previous study has identified the value of membrane repair in the nervous system, but identifying potential therapeutic proteins involved in membrane repair can provide targeted treatment for diseases that involve cell membrane injuries. We have previously shown that TRIM72/MG53 can increase plasma membrane repair in skeletal

iii and cardiac muscle as well as non-muscle cell types where it is not usually expressed.

This observation led us to screen for novel TRIM family proteins that may be able to mediate membrane repair in neuronal cells. We found that TRIM2 transfected cells show an increased capacity for membrane resealing following multi-photon laser injury, while knock down of TRIM2 decreases membrane repair. Because TRIM2 is highly expressed in the nervous system and was previously shown to regulate neurofilaments (NFL) we tested if TRIM2 protects against membrane damage in neurons through regulation of ubiquitination of NFLs and if disruption of TRIM2 or NFL leads to compromised membrane repair or neuronal cell death. Using confocal microscopy, TRIM2 and NFL were not observed to co-localize and immunoprecipitation only revealed an interaction when the two proteins were overexpressed. Additionally, knockout of NFL in primary neurons did not affect the ability of the cells to repair their membranes. The available data suggest that TRIM2 must use an alternative mechanism to mediate membrane repair in neurons.

Overall, these studies demonstrate the importance of membrane repair in both skeletal muscle, cardiac muscle, and nervous tissue. Future studies of membrane repair will increase the understanding of this essential process and establish that increasing membrane repair is a valuable target for therapeutic treatment.

Dedication

This document is dedicated to my wife, family and friends for their patience and understanding.

Acknowledgments

This document would not have been completed without the mentorship and support of my advisor Dr. Noah Weisleder. I am extremely grateful for the time and effort he has put into helping me get through my time in graduate school. Working in his lab has been an extremely enjoyable experience, and he has taught me how to think and examine problems like a scientist.

I would also like to thank the current and former members of my committee Dr.

Brandon Biesiadecki, Dr. Anthony Brown, Dr. Federica Accornero, and Dr. Jessica

Lerch. I appreciate the guidance you have given me, and the aid the members of your lab have given me when I needed help with my projects.

Lastly, I want to thank all the current and former members of the Weisleder lab.

Eric Beck has been crucial for the completion of my dissertation both through experimental help and editing of the document. I also want to acknowledge the help of my fellow graduate students Dr. Alisa Blazek, Dr. Kevin McElhanon, and Dr. Tom

Kwiatkowski have been by helping me troubleshoot experimental issues. Finally, this work would not have been completed without the aid of the hard-working undergraduate research assistants that have worked with me: Kassidy Banford, Alex Carsel, Allison

Miller, and Francesca Veon.

This work was aided by the Nishikawara fund and the Center For Muscle Health vi and Neuromuscular Disorders.

vii

Vita

2011………………………………………… B.S. Biology, California State University,

Long Beach

2012-2013……………………………………Discovery Post-baccalaureate Research

Education Program, The Ohio State

University

2013-present ………………………………...Graduate Research Associate, The Ohio

State University

Publications

Blazek, A. D.*, Paleo, B. J.* & Weisleder, N. Plasma Membrane Repair: A Central

Process for Maintaining Cellular Homeostasis. Physiology (Bethesda) 30, 438-448, doi:10.1152/physiol.00019.2015 (2015). * Equal Contribution

Talbert E.E., Cuitiño M.C. Ladner K.J., Rajasekerea P. V., Siebert M., Shakya R., Leone

G.W., Ostrowski M.C., Paleo B., Weisleder N,. Reiser P.J.,Webb A., Timmers C.D.,

Eiferman D.S., Evans D.C., Dillhoff M.E., Schmidt C.R., Guttridge D.C. Modeling

viii

Human Cancer-induced Cachexia. Cell reports 28, doi:10.1016/j.celrep.2019.07.016

(2019).

Paleo B.J., Madalena K.M., Mital R. McElhanon K.E., Kwiatkowski T.A. Rose A.L.,

Lerch J.K. Weisleder N; Enhancing Membrane Repair Increases Regeneration in a

Sciatic Injury Model. PloS one 15, doi:10.1371/journal.pone.0231194 (2020).

Fields of Study

Major Field: Biomedical Sciences Graduate Program

Table of Contents

Abstract ...... ii

Dedication ...... v

Acknowledgments...... vi

Vita ...... viii

List of Figures ...... xi

Chapter 1. Introduction ...... 1

Chapter 2. MG53 Deficiency Exacerbates Skeletal Muscle Pathology in the mdx Mouse

...... 27

Chapter 3. Enhancing Membrane Repair Increases Regeneration In A Sciatic Injury

Model ...... 61

Chapter 4 Targeting Neuronal Membrane Repair Through TRIM2 Modulation ...... 87

Chapter 5 Summary, Significance, and Future Work ...... 103

Bibliography ...... 115

List of Figures

Figure 1: Models of the plasma membrane repair process...... 22

Figure 2: Major membrane repair proteins and their hypothesized roles in the repair process...... 24

Figure 3: Proposed roles of MG53/TRIM72 in mediating membrane repair...... 26

Figure 4: Breeding strategy for the generation of MG53/mdx DKO mice...... 45

Figure 5: DKO mice maintain muscle mass consistent with mdx mice...... 47

Figure 6: Histological analysis of skeletal muscle of mdx and DKO mice at 6 weeks. ... 48

Figure 7: Membrane repair and integrity is compromised in DKO mice...... 50

Figure 8: Membrane repair proteins are altered in DKO mice...... 52

Figure 9: Age affects body weight, but does not affect tissue weight ...... 54

Figure 10: Histology and functional analysis of skeletal muscle of mdx and DKO mice at

1.5 years...... 55

Figure 11: Masson’s trichrome stain analysis implicates robust fibrosis of DKO mice at

1.5 years...... 57

Figure 12: Membrane repair and integrity is compromised in aged DKO mice...... 59

Figure 13: MG53 is not expressed in the nervous system...... 80

Figure 14: MG53 increases membrane repair capacity in N2a cells ...... 81 xi

Figure 15: rhMG53 decreases membrane damage in vitro ...... 83

Figure 16: Treatment with rhMG53 significantly increases regeneration past the crush site...... 85

Figure 17: TRIM2 is essential for membrane resealing in vitro...... 98

Figure 18: TRIM2 does not bind to neurofilament light chain...... 100

Figure 19: Lack of neurofilament light chain does not affect membrane resealing in NFL

KO neurons...... 102

xii

Chapter 1. Introduction

This chapter has been adapted from the published article, “Plasma Membrane Repair: A

Central Process for Maintaining Cellular Homeostasis; Physiology Volume 30, Issue 6

November 2015, Pages 438-448” from authors Alisa D. Blazek, Brian J. Paleo and Noah

Weisleder. ADB and BJP were co-first authors and NW was the corresponding author.

The contributions of each were as follows:

a) Writing ADB BJP: Original draft preparation. ADB BJP NW: Review & editing.

i) ADB drafted the following sections: Calpain, Caveolin, Polymerase-1 and

Transcriptase Release Factor, Affixin, AHNAK, S100A10, and Conclusion.

ii) BJP drafted the following sections: Introduction, Motor Proteins, Acid

Sphingomyelinase, ESCRT, Synaptotagmin/SNAREs, Ferlin Family, and

Mg53.

iii) Equal contribution to the draft of the Annexin section.

b) Prepared figures; ADB, BJP, and NW contributed equally. Presentation of figures

provided by the journal Physiology.

Introduction

The plasma membrane of eukaryotic organisms is composed of a phospholipid bilayer with proteins embedded in the bilayer that provide support and allow the

1 regulated exchange of molecules between the intra- and extracellular environment. This bilayer establishes a physical barrier that allows the cell to maintain molecular and ionic gradients that the cell can use to perform signaling and also keep the intracellular environment free of potentially toxic conditions outside the membrane. However, due to disease or injury, the integrity of the barrier function of the plasma membrane can be compromised. During these injury events, toxic levels of extracellular components, including calcium and other molecules, can enter through the membrane and adversely affect the function of the cell and lead to eventual cell death. Disruptions in the lipid bilayer will reseal through the thermodynamic properties of the component phospholipids if the membrane has a simple lipid structure similar to a vesicle or liposome1 (Fig. 1A).

However, eukaryotic cells have cytoskeletal components anchored to the membrane, along with attachments to the extracellular matrix necessary to maintain proper cellular structure and function. These attachments increase membrane rigidity and produce mechanical tension on the membrane that reduces the effectiveness of simple thermodynamic resealing2,3. As a result, larger disruptions to the membrane will remain open unless the cell can use an active process to repair that disruption. Thus, cells have evolved multiple processes of cell membrane repair to maintain the integrity of the membrane when it is damaged or broken. Membrane repair was first reported in the literature as early as the 1950s4 as a calcium-mediated process that actively seals the injury in response to membrane damage.

Studies in the 1950s reported on free floating eggs demonstrating the ability to survive rupture of their membranes. Later in the 1960s, it was reported that cells could

2 react to membrane damage in the presence of physiological levels of calcium and that the cells showed a rapid response to the wound with vesicles appearing at the site of the injury5. After this fundamental discovery, decades passed with little advancement in the field and it wasn’t until the early 1990s that interest began to increase. Dr. Paul McNeil advanced the field by proving that cells experience membrane damage and resealing in vivo in the intestines6. These resealing events occur when induced, as well as with physiological levels of injury associated with the digestion process. Work from his laboratory group also demonstrated that muscle fibers naturally experience membrane disruptions during bouts of exercise and actively reseal their membrane following the damage7. While Dr. McNeil was establishing the physiological relevance of the membrane repair response in mammalian tissues, Dr. Richard Steinhardt began a study that provided a better mechanistic understanding of the cell membrane repair process. He demonstrated that the plasma membrane was not simply passively resealing as shown previously in some red blood cell experiments8, his results showed that following an injury event that there was vesicle trafficking to the injury site9. He also demonstrated through the use of the botulinum and tetanus toxins that members of the SNAREs, a family of exocytotic proteins, are required for membrane repair in an exocytotic process similar to the exocytotic events of neurotransmitter release8,9.

The first proposed mechanism of membrane repair is referred to as the ‘patch hypothesis’10 (Fig. 1B). This proposed mechanism hypothesizes that following an injury event, intracellular vesicles are trafficked to the injury area. Next, due to the high concentration of calcium, the vesicles fuse to form a patch the same size as the injury.

The edges would in turn fuse with the injury site to patch the injured membrane. The evidence for the patch model first came from work done in sea urchin eggs. Seawater was injected into the egg and vesicles were observed at the microneedle site10. Although much of the evidence for the patch hypothesis has been observed in sea urchin and Xenopus eggs, a recent article has found that muscle cells can produce a ‘cap’ to patch their membrane11. Whether the observed cap is, in fact, a similar mechanism to the patch hypothesis remains to be seen, but a cap forming at the site of an injury can certainly be perceived as a patch.

As the field has advanced and membrane repair has been recognized as being important to different diseases and injuries, alternative methods of repair have also been proposed. Similar to the patch method, exocytosis has been proposed to occur when vesicles are trafficked and then fuse to the edge of the injury adding lipids to the membrane until enough lipids are added to close the wound6-9,12,13 (Fig. 1B). Early studies indicated lysosomes as the vesicle population required for membrane repair14, however, there have been conflicting reports on the use of lysosomes in membrane repair15.

While there is evidence that lysosomes are the vesicles responsible for closing the wound, there is some controversy related to exocytosis as a method to reseal wounds.

Some believe that exocytosis aids in resealing the wound but is not the primary cause of the wound closure. They believe that during the exocytotic event an enzyme called acid sphingomyelinase (ASM) is released into the extracellular space16-19. ASM converts sphingomyelin in the membrane to ceramide, which induces an invagination in the

4 membrane allowing for the injury to be endocytosed.

Although the original repair mechanism was thought to be exocytotic, it is clear that cells have a wide variety of mechanisms to maintain the integrity of their membranes. Constriction of the membrane around disruptions can also contribute to membrane repair20 (Fig. 1C). Endocytotic mechanisms seem to be required for larger membrane disruptions, while smaller disruptions of less than 100nm reseal through budding and exocytosis (Fig. 1D). Repair through budding involves pinching the membrane at the injured site and shedding the injured membrane into the extracellular space21-23 (Fig. 1E). Endocytosis is also thought to contribute to membrane repair by internalization of the injured membrane23,24 (Fig. 1F). The specific repair mechanism at work at a given time may depend on several factors, potentially including the size of the injury. There may be instances where a cell could be capable of using multiple methods of membrane repair, or it could be simply a matter of what process the cell has used recently in the area of the injury. It is important to remember that these methods can be cell specific and that certain circumstances such as disease, injury, or ischemic condition could play a role in the type of repair that is reported.

Membrane damage is not necessarily required to recruit membrane repair.

Damage can occur as a result of an ischemia-reperfusion injury, like heart attack or stroke25-28. An ischemic event occurs when blood supply to a tissue is interrupted, leading to a lack of nutrients and oxygen to the tissue. The lack of oxygen does not allow for the tissue to replenish adenosine triphosphate (ATP), thus causing disequilibrium due to

ATP-dependent pumps failing to maintain the sodium potassium gradient as well as the

5 calcium gradient29,30. Calcium influx can be taken up by the mitochondria to produce reactive oxygen species (ROS) and may lead to cell death. ROS can lead to lipid peroxidation which leads to modification of lipids in the plasma membrane31-33. The dysfunction of gradients, production of ROS, and lipid peroxidation all lead to decreased membrane integrity. The decrease in membrane integrity leaves the cell vulnerable to injury, and the damaged lipids must be replaced for normal membrane function.

Cells that are exposed to bacterial pathogens are vulnerable to pore forming toxins

(PFTs)16,34. PFTs can be produced by pathogenic bacteria as a defense mechanism against the host immune system. These toxins are released to induce lysis of the attacking immune cells to protect the bacteria from being destroyed. The holes formed by the toxins cannot spontaneously seal and must be removed by the cell to avoid death. The two common forms of membrane repair used to remove the pores are shedding and endocytosis21-23.

While it is important to understand that membrane repair may be required in many tissues, mechanically active tissue is most vulnerable to injury. These tissues include the epithelium that covers the majority of the body, the inner lumen of the gastrointestinal tract, the endothelia of the blood vessels, and the muscles used for body movement. As the early work by Dr. McNeil showed, muscles regularly experience injuries to their membranes during bouts of exercise. These stretch injuries occur as a result of eccentric contractions which occur when the muscle is required to contract as the muscle is lengthening7,35. Imagine a weightlifter exercising their bicep muscle with a bicep curl. As they are slowly lowering the weight, the bicep is lengthening, but since 6 there is a weight in their hand the muscle must contract to maintain the weight. Interest in muscle membrane repair has driven the field to explore repair mechanisms due to their role in the disease muscular dystrophy36,37. Duchenne muscular dystrophy is a disorder that occurs as a result of a truncated or missing protein called dystrophin that serves as a shock absorber between the contractility machinery and the extracellular matrix38. Loss or deletion of this protein makes the muscle vulnerable to contractile related injuries and is extremely dependent on membrane repair to reseal the frequent membrane injuries39.

Despite the importance of membrane repair in cellular function, the field has only recently begun to expand with the discovery of more proteins linked to resealing damaged cell membranes.

Caveolin

Caveolin family (-1, -2, and -3) proteins are 21-24 kDa integral membrane proteins enriched in invaginations of the plasma membrane involved in membrane transport called caveolae40 (Fig. 1C, Fig. 3). Cav-3 is muscle specific and has been most closely linked with membrane repair. Mutations in Cav-3 cause autosomal dominant limb-girdle muscular dystrophy (LGMD) 1C and autosomal dominant rippling muscle disease (AD-RMD)41. Cav-3 has been shown by co-immunoprecipitation to interact with

MG53/TRIM72 and dysferlin39, and cav-3 overexpression regulates membrane fusion events by downregulating MG53/TRIM72 induced membrane budding and preventing development of filopodia-like structures42. Disruption of the cav-3/dysferlin/MG53 complex can affect the localization and membrane repair function of the other

7 components. For example, dominant negative cav-3 mutations associated with the development of muscular dystrophy have been shown to cause retention of

MG53/TRIM72 39 or dysferlin43 in the Golgi apparatus and loss of membrane repair capacity. However, other studies using ultrastructural analysis of dysferlin trafficking showed that dysferlin can still reach the plasma membrane in the absence of cav-3, but that it is rapidly endocytosed44,45.

Polymerase-1 and transcriptase release factor (PTRF/ Cavin-1/ Cav-p60)

PTRF may aid in the formation of caveolae at the plasma membrane since PTRF localizes to caveolae, and loss of PTRF results in caveolae loss and a dystrophic phenotype46. PTRF is required for formation of caveolae, as lack of PTRF expression leads to loss of caveolae, while expression of PTRF is sufficient for caveolae formation.

Experiments using fluorescence lifetime imaging showed that cholesterol is required for the interaction between PTRF and caveolin, as cholesterol depletion abrogates the PTRF- caveolin interaction, suggesting that PTRF may be involved in stabilizing the membrane curvature of caveolae47. Additionally, it was shown by immunoprecipitation that PTRF binds dysferlin, and in fact, may be required for the correct localization of dysferlin, as

PTRF mutation results in decreased dysferlin at the cell membrane46. Knockdown of

PTRF by shRNA results in decreased membrane repair capacity, while overexpression of

PTRF can rescue dystrophic muscle membrane repair44. During the membrane repair process, PTRF binds to dysferlin and may anchor MG53/TRIM72 to cholesterol, as

MG53/TRIM72 cannot bind cholesterol unless PTRF is present44,48. In this model, PTRF

8 anchors MG53/TRIM72 by binding to exposed cholesterol at the membrane injury site

(Fig. 3).

Motor proteins (Nonmuscle Myosin IIA and IIB/ Kinesin)

Since vesicle trafficking is an important aspect of membrane repair, motor proteins must be involved in this process to allow trafficking to occur. Membrane repair is sensitive to inhibitors of both myosin and kinesin motor proteins8,49. Studies have shown that non-muscle myosin IIA and IIB are important motor proteins that specifically mediate vesicle trafficking during membrane repair49. Antisense knockdown of myosin

IIB suppressed exocytosis and membrane resealing, and knockdown of myosin IIA inhibited the rate of resealing at repeated wound sites. These studies are supported by the observation that nonmuscle myosin IIA facilitates the transport of vesicles containing

MG53/TRIM72 to the site of membrane injury50 (Fig. 3).

Affixin (β-parvin/ integrin-linked kinase (ILK)-binding protein)

Affixin was first discovered to be an integrin-linked kinase binding protein that localizes to focal adhesions and likely contributes to their maturation. In muscle cells, affixin and ILK colocalize at sites where the Z band attaches to the sarcolemma due to an interaction with dysferlin51. Affixin’s interaction with dysferlin and altered immunoreactivity in Myoshi Myopathy (MM), LGMD2B, and LGMD1C suggests a role for this protein in muscle membrane repair, possibly through the organization of F-actin

51 (Fig. 2A). For example, affixin has been shown by immunoprecipitation and pull-down assay to interact with guanine nucleotide exchange factor αPIX (ARHGEF6 or Cool-2),

9 which functions to regulate actin skeleton adhesion51-53, as well as α-actinin which plays a role in the organization of the cytoskeleton54. Although it appears that affixin may participate in cytoskeletal remodeling during membrane repair, the precise role of affixin in this process remains speculative.

Acid sphingomyelinase

Acid sphingomyelinase (ASM) is an enzyme that cleaves the phosphorylcholine head of sphingomyelin to generate ceramide, a molecule that leads to membrane invagination, and contributes to the process of endocytosis during membrane repair55,56.

In response to injury, lysosomes fuse with the injured cell membrane and release ASM from the cell, and the action of ASM causes invagination of the membrane and endocytosis of the injury site 16-19 (Fig. 2B). The evidence for ASM in membrane repair includes that dysferlin deficient C2C12 cells showed less secretion of ASM than control cells16. When ASM was inhibited and cells were permeabilized using a pore forming toxin, the cells could not sufficiently repair16. Treatment with exogenous ASM was sufficient to restore the membrane integrity of ASM depleted cells16 and in a dystrophic patient-derived myoblast cell line16.

ESCRT

Endosomal sorting complex required for transport (ESCRT) is involved in viral budding, cytokinesis, and spontaneous budding of the plasma membrane. ESCRT subunits are classified into five complexes57. The ESCRT III complex recently has been shown to be involved in endocytosis and budding in response to membrane

10 damage23,34,58. After injury, apoptosis-linked gene (ALG)-2 binds Ca2+ and leads to the accumulation of ALG-2-interacting protein X (ALIX), ESCRT III, and vacuolar protein sorting-associated protein during calcium-dependent wound repair. The observed recruitment of ESCRT was followed by blebbing of the membrane and shedding of the wound (Fig. 2B). ESCRT III was involved in endocytosis of the membrane after the insertion of the bacterial pore forming toxin SLO23. The complex appears to be critical for the repair of injuries less than 100nm23.

Calpain

After membrane disruption, rapid reorganization of cytoskeletal elements at the site of injury is necessary. Treatment with the actin-depolymerizing enzyme DNase I, significantly enhanced resealing, demonstrating that disassembly of the actin skeleton is important in membrane resealing59. Calpain 3 is a cysteine protease that cleaves cytoskeletal proteins, such as talin and vimentin, and may aid in early remodeling of various proteins during membrane repair as cleaved fragments of these cytoskeletal proteins could not be recovered in calpain null cells60. Calpain 3 cleaves AHNAK, inhibits AHNAK’s interactions with dysferlin and myoferlin, and regulates AHNAK protein turnover (Fig. 2). Through these actions, calpain regulates cytoskeletal structure and interaction of the cytoskeleton with the cell membrane61,62. In fact, loss of calpain 3 in particular has been shown to cause limb girdle muscular dystrophy 63. Calpain has been shown to be required for calcium-dependent membrane repair, since knockout of the small subunit of the calpain enzyme results in failure to reseal after laser induced membrane disruption60. Calpain could contribute to membrane repair through other

11 mechanisms as well. Studies suggest that calpain can remodel sarcomeres since calpain 3 binds titin and is present in other regions of the sarcomere, such as the Z disk, costameres, and myotendinous junctions 64. Thus, the role of calpain may be to increase local loosening of the sarcomere via proteolysis and facilitate removal of damaged and cleaved proteins by the proteosome65. Other studies suggest a role for calpain 3 in post- membrane repair sarcolemmal remodeling because loss of calpain in limb girdle muscular dystrophy leads to disorganization of myofibers and a lack of organized sarcomeres and sarcomere proteins in myotubes61,64,66. Finally, calpain activity leads to dysferlin cleavage to produce subunits that function in membrane repair67.

AHNAK (Desmoyokin)

AHNAK (translated to giant in Hebrew), or desmoyokin, is a 629.1 kDa tripartite nucleoprotein with potential functions as diverse as fat metabolism, DNA repair, autoantigenicity, cell signaling, calcium channel regulation, and tumor metastasis68. It has been proposed that interaction of AHNAK with annexin 2/S100A10 regulates organization of the actin cytoskeleton and architecture of the cell membrane, as AHNAK- specific siRNA prevents actin cytoskeleton reorganization69. In membrane repair,

AHNAK has been associated with enlargeosomes, vesicles that rapidly exocytose in response to calcium influx; however, its exact role within these vesicles is unknown70.

AHNAK’s ability to bind actin may signify a role for the protein in membrane resealing, although its presence within enlargeosomes during recruitment to the membrane injury suggests earlier involvement in the repair process68. The co-localization of AHNAK with annexin 2 within vesicles is controversial, as studies also have reported that annexin is on 12 the cytosolic face of the vesicle71. In fact, calcium sensitive annexin 2 may be involved in recruitment of AHNAK and S100A10 to the plasma membrane in response to calcium- annexin binding68. Furthermore, co-immunoprecipitation and mass spectrometry studies confirm that AHNAK interacts with dysferlin which may localize and stabilize AHNAK at the sarcolemma through its transmembrane domain, an interaction possibly regulated by calpain 361,72 (Fig. 2).

S100A10

S100A10 is a small, approximately 10kDa EF-hand Ca2+ binding protein also known as annexin 2 light chain or p11. S100A10 forms a heterotetrameric complex with annexin A2 by forming a S100A10 dimer in the middle of two annexin A2 chains. This complex targets to the plasma membrane in a calcium regulated manner that is dependent on its interaction with annexin 2 (Fig. 2C). At the plasma membrane, S100A10 interacts with a number of proteins where it may be essential for surface presentation of proteins such as ion channels73. Further, proteomic and structural analyses have identified that the heterotrimeric complex binds cytosolic proteins AHNAK and dysferlin, and that it is responsible for the recruitment of AHNAK to the cell membrane where the entire complex acts as a scaffold for membrane repair74-77. Although S100A10 is known to be a central player in the membrane repair complex, its exact function in membrane repair has not been identified. One model suggests that membrane fusion may occur due to the ability of S100A10-annexin A2 to bridge adjacent phospholipid membranes78.

Synaptotagmin/SNAREs

Synaptotagmins (Syt) are a family of proteins that contain two C2 domains (C2A and C2B), which some members use to bind Ca2+ or phospholipids79-83. Much of the focus on synaptotagmins has been in synaptic neurotransmitter release in neurons through their interactions with soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNARE)84-90. The general mechanism of this interaction is that SNAREs on the vesicle interact with SNAREs on the membrane to dock the vesicle to the membrane.

Then, synaptotagmins already bound to the vesicle then bind Ca2+, and become able to interact with the SNAREs to fuse the vesicle to the membrane, and expel the contents of the vesicle (Fig. 2D).

The relationship between synaptotagmins and SNAREs is important because of their roles in Ca2+ sensing and vesicle membrane fusion. Syt I is found exclusively in the nervous system91,92, and antibodies against the C2A domain of Syt I cause inhibition of membrane repair in severed axons of squid and crayfish giant axons93. This same study used antibodies against a domain of a SNARE protein, syntaxin, to inhibit interaction with synaptotagmin’s C2A domain93. Similar to inhibition of Syt I, blocking syntaxin inhibited resealing of the severed axon93. Syt VII is ubiquitously expressed and found proximal to the membrane on lysosomal-associated membrane protein 1 (LAMP-1) positive lysosomes. Cells injured in the presence of Syt VII blocking antibodies showed a decreased capacity to reseal membrane disruptions14 and fibroblasts taken from Syt VII deficient mice showed defective lysosomal exocytosis and decreased capacity to reseal their plasma membrane94.

Membrane repair is not solely dependent on synaptotagmins because studies using 14 botulinum toxin A, which cleaves SNARE proteins, showed inhibition of resealing of sea urchin eggs9. Disruption of the formation of the SNARE complex using the cytoplasmic domain of synaptobrevin 2, a SNARE protein, was also able to block membrane resealing95.

Ferlin Family (Dysferlin/ Myoferlin/ Otoferlin)

Dysferlin is a type II transmembrane protein from the ferlin family that localizes to the plasma membrane and T-tubules of muscle fibers 96,97 (Fig. 2). Members of the

Ferlin family contain five-seven C2 domains that facilitate lipid and protein binding.

Dysferlin contains a C2A domain responsible for calcium sensing, and the C2 domains also confer lipid binding capability to the protein, giving it the ability to sense a membrane disruption through calcium release and to bind lipids to help reseal the membrane.

Dysferlin was initially identified as the target gene mutated in Myoshi Myopathy

(MM) and limb-girdle muscular dystrophy type 2B (LGMD2B)37,98-100. These autosomal recessive forms of muscular dystrophy present with weakness in proximal muscles, Limb

Girdle Muscular Dystrophy Type 2B (LGMD2B), or weakness in distal muscles, Myoshi

Myopathy (MM). Although dysferlin is expressed ubiquitously, skeletal and cardiac muscle are the most affected tissue when the protein is lost. Subsequent dysferlin knockout mouse studies showed that muscle fibers101 and hearts102,103 from these animals demonstrated a decreased ability to reseal the plasma membrane36, leading to muscular dystrophy and late onset cardiomyopathy36,102. These results identified dysferlin as the first protein to contribute directly to membrane repair in striated muscles, resulting in 15 dysferlin’s prominent role in the membrane repair literature104. Dysferlin deficient mice exhibit muscular dystrophy symptoms similar to human disease. This is characterized by an impaired response to membrane injury. Control fibers resealed normally, while knockout fibers showed prolonged increased dye influx following laser injury.

Subsequent studies linked dysferlin to roles in membrane receptor recycling, endocytosis, vesicle trafficking, membrane turnover, focal adhesion, and ATP dependent intracellular signaling and modulation of the immune system62,105-111. Recent in vivo studies indicated that dysferlin function may be as important for T-tubule structure as in sarcolemmal membrane repair112, illustrating that much is still unknown about dysferlin function113.

Other members of the ferlin family also have been linked to membrane repair. For example, myoferlin shares a 56% homology to dysferlin and is expressed during muscle development to facilitate myoblast fusion111,114. WT mice show low expression of myoferlin; however, increased expression was observed in mdx and gamma-sarcoglycan null mouse models of muscular dystrophy109,115. Transgenic mouse experiments showed that myoferlin overexpression could compensate for dysferlin in membrane repair; however, myoferlin could not prevent all dystrophy symptoms116. Another dysferlin family member, otoferlin, regulates synaptic vesicle exocytosis in cochlear hair cells and may be involved in membrane repair specifically within these cells117,118.

Annexin

Annexins are a gene family consisting of a C-terminus containing phospholipid and Ca2+ binding sites, and a variable N-terminus75. The N-terminus is considered the unique region of the protein and is responsible for interaction with different binding 16 partners. The C-terminus is considered the conserved region and confers its calcium and phospholipid binding. The annexins are unique in that they can bind lipids while in their

Ca2+-bound conformation and the concentration of the Ca2+ signal determines which particular annexin family member will bind to the phospholipid119. Although the exact mechanism remains unclear, numerous studies suggest roles for annexins in vesicle movement, fusion, and patch formation during membrane repair, 78,120,121. Annexins A1 and A2 bind dysferlin in a Ca2+ dependent manner and may contribute to membrane repair by their ability to assist in the aggregation and fusion of intracellular vesicles by their association with lipid rafts of the plasma membrane122,123. Density gradient centrifugation experiments confirmed the Ca2+ and annexin-dependent association of these rafts. Electron microscopic evidence shows that annexin may be involved in membrane fusion during exocytosis, as well as serving as a scaffold for endosomes 119,124

(Fig. 2C). Annexins also have been shown to mediate microvesicle release and blebbing, an indirect repair mechanism that involves sealing off a damaged membrane segment by accumulation of annexin at the neck of the membrane bleb125. Serum levels of annexins

A1 and A2 also increase in patients with dysferlinopathy, which indicates that the proteins are involved with the progression of the disease. Additionally, annexins A1, A2, and A6 have been shown by live imaging to locate to membrane disruptions and to assemble into a “cap” on the membrane repair patch to assist with membrane resealing. A recent study has shown that recombinant annexin A6 has been shown to aid in membrane repair and to alleviate the progression of the mouse model of Duchenne muscular dystrophy. Annexins A4 and A6 have been shown to induce bending and constriction of

17 the membrane in response to the free edges of membranes. Other studies indicate that annexin A5 (AnxA5) can bind to injured membranes and form a two dimensional array that is thought to contribute to membrane repair, as AnxA5 null cells show defective repair capacity 126. Finally, annexin A6 has been shown by live imaging to locate to membrane disruptions and assemble into a "cap" on the membrane repair patch to assist with membrane resealing127.

MG53/TRIM72

Mitsugumin 53 (MG53)/TRIM72 has been shown to be a vital component of the membrane repair machinery in multiple cell types128-131. It is a member of the tripartite motif family of E3 ubiquitin ligases (TRIM72) that was originally identified in an immunoproteomic library screen for muscle enriched proteins132. Although native

MG53/TRIM72 protein was initially thought to be found only in skeletal and cardiac muscle129, recent studies have shown expression and membrane repair function in other tissues130,131,133. Along with native protein repair capacity, overexpression of

MG53/TRIM72 in cells that do not express MG53/TRIM72 also shows protective effects against membrane injury134. MG53/TRIM72 null mice display defective membrane repair129, progressive myopathy129, and increased susceptibility to injury in the heart128,135, lungs131,133 and kidneys130.

MG53/TRIM72 interacts with phosphatidylserine to associate with intracellular vesicles and the inner leaflet of the plasma membrane129 (Fig. 3). Once a cell is injured,

MG53/TRIM72 is thought to react to the oxidized extracellular environment entering the

18 cell by forming higher molecular weight units136. MG53/TRIM72 tethered vesicles traffic to the membrane disruption, allowing vesicles to fuse and patch the injured membrane129.

This process appears to involve a dysferlin and caveolin-3 (Cav-3) containing complex that regulates repair in the sarcolemma39,137. A study using ballistic injury and super resolution microscopy of human myotubes showed that MG53/TRIM72 is recruited quickly to the membrane (2 seconds), followed by dysferlin about 10 seconds post- injury15. MG53/TRIM72 and dysferlin form a lattice that fills the wound area with vesicles and closes the wound15.

Interestingly, when myotubes are injured in media containing exogenous human

MG53/TRIM72 protein (rhMG53), the protein can be seen to localize to the injury site134.

This mechanism is similar to intracellular MG53/TRIM72 binding to exposed phosphatidylserine129 and does not require endogenous MG53/TRIM72 or dysferlin134.

This association with injury sites can increase membrane resealing134. Application of rhMG53 to animal models of various diseases, such as muscular dystrophy134, myocardial infarct27, lung injury131,133, compartment syndrome138 and acute kidney injury130 can reduce the pathology in these models. While this property of the protein is dependent on the phosphatidylserine binding capacity of rhMG53, it is possible that other aspects of the protein function, such as regulation of intracellular signaling, could also contribute to this ability to increase membrane repair capacity27,139-141. Further studies should determine the rhMG53 extracellular mechanism of action and assess its potential as a therapeutic agent.

Conclusions

While clear evidence for the involvement of the discussed proteins in membrane repair exists, it should be noted that other molecules, such as Amphiphysin 2 (BIN1)142,

Lamp-114, and others, have been speculated to play a role in membrane resealing due to the nature of their actions in the cell or due to their interactions with known members of the membrane repair machinery. For example, a recent study identified ATPase EH- domain containing 2 as a novel membrane repair protein that colocalizes at the site of membrane injury with F-actin and annexin A1143. Moesin (membrane-organizing extension spike protein) has been shown to interact with dysferlin and appears to cross- link plasma membranes and actin cytoskeletons46. As more information about specific proteins and membrane repair in general is obtained, we expect that additional candidates will be discovered.

In considering the molecules known to be important in the membrane repair process, it becomes clear that many of the steps in membrane repair are processes necessary for normal cellular functions. For example, myosin, kinesin, annexin, and

SNAREs are involved in vesicle trafficking and membrane fusion, integral events for numerous cell activities144. Calpain and affixin are important for cytoskeletal remodeling during cell motility, responses to the cell environment, and mitosis145,146. PTRF, ESCRT,

ASM and S100A10 are active at the plasma membrane to form caveolae, membrane buds, and other membrane invaginations, as well as to mediate membrane fusion18,23,46,78.

Dysferlin is integral to multiple activities, such as endocytosis, vesicle trafficking, membrane turnover, and others106-108. Membrane repair can be considered an emergency response in which these cellular processes are used to reseal the membrane and allow cell

20 survival120,147,148. However, it is necessary to move beyond dissection of this supportive cellular machinery in order to gain a full understanding of membrane repair. Our current understanding of membrane repair is limited to a subset of cellular functions and protein interactions leaving compelling questions unanswered. For example, how does the cell differentiate the membrane repair process from normal cell functions and what are the specific signaling pathways that allow this differentiation? Furthermore, what are the signals that direct the repair machinery to the site of injury? While calcium-dependent mechanisms are known signals for the assembly of membrane repair proteins at the damage site and fusion of membrane surfaces, additional mediators may exist. Discovery and characterization of these mediators and pathways is an important next step in understanding the membrane repair process.

Figure 1: Models of the plasma membrane repair process.

A: Thermodynamic resealing occurs spontaneously due to tension produced by the disordered arrangement of the membrane phospholipids at the open edge of the break. B:

Exocytosis can contribute by trafficking intracellular vesicles to the wounded area where they can fuse with each other and the injured membrane to form a repair patch. C: Wound constriction occurs when caveolae cluster and fuse around larger wounds, leading to wound constriction and intracellular fusion of caveolar endosomes. D: Budding/blebbing of the membrane portion containing the wound site with release of the newly formed vesicles into the extracellular space also involves exocytosis. E: Exocytosis of an intracellular patch and

22 fusion to the wound site could result in the extracellular release or “shedding” of the wound site. F: Endocytosis of wounds occurs via invagination of caveolar vesicles and subsequent intracellular fusion of caveolae.

Figure 2: Major membrane repair proteins and their hypothesized roles in the repair process.

A: Affixin, which also binds dysferlin, localizes to focal adhesions and may organize actin. B: ESCRT and acid sphingomyelinase (ASM) facilitate exocytosis and endocytosis. ESCRT has been found to be involved in both endocytosis and budding.

ASM is secreted and cleaves sphingomyelin to generate ceramide, leading to membrane invagination of the injury site. C: The annexin and S100A10 complex binds dysferlin and may recruit AHNAK to the membrane due to annexin’s ability to bind lipid rafts.

Annexin/S100A10 may also bridge adjacent phospholipids to form endosomes. Annexin accumulates at the neck of membrane blebs to mediate microvesicle release. D:

Synaptotagmin and SNARE proteins interact at the plasma membrane via a

24 conformational change in synaptotagmin present on synaptic vesicles to fuse the vesicles with the membrane.

Figure 3: Proposed roles of MG53/TRIM72 in mediating membrane repair.

MG53/TRIM72 interacts with phosphatidylserine in the plasma membrane in a complex containing dysferlin and Cav-3. MG53/TRIM72 and dysferlin close the membrane wound with vesicles. Vesicle transport is facilitated by myosin motor proteins. Cav-3 may regulate MG53/TRIM72-mediated membrane fusion and is enriched in caveolae or plasma membrane invaginations. PTRF may aid in the formation and stabilization of these caveolae by interaction with Cav-3 and MG53/TRIM72 through cholesterol. PTRF also binds and may help localize dysferlin to the plasma membrane.

Chapter 2. MG53 Deficiency Exacerbates Skeletal Muscle Pathology in the mdx Mouse

This chapter has not been adapted from any currently published article and represents work conducted specifically by BJP. Subsequent expected publication of this work will list BJP as a first author. His contributions to the work presented in this chapter include:

a) Writing: BJP Original draft preparation. BJP NW: Review and editing

b) Figure 2.1

i) Panels A, B, C: BJP Tissue processing/sectioning/staining, imaging, data

acquisition, and analysis/presentation

ii) Panel D: BJP Contractility protocol, data acquisition, and

analysis/presentation.

c) Figure 2.2

i) Panel A) BJP Tissue processing/sectioning/staining, imaging, and

analysis/presentation. KEM Data acquisition.

ii) Panel B) BJP Data acquisition, and analysis/presentation. KEM Laser

injury/imaging

d) Figure 2.3

i) Panels A, B) BJP Western blotting, imaging, data acquisition, and

analysis/presentation.

e) Figure 2.4

i) Panel A, B, C: BJP Tissue processing/sectioning/staining, imaging, and

analysis/presentation. KS SR Data acquisition.

ii) Panel D: BJP Contractility protocol, data acquisition, and

analysis/presentation.

f) Figure 2.5

i) BJP Tissue processing/sectioning/staining, imaging, and

analysis/presentation. KB Data acquisition.

g) Figure 2.6

i) Panels A, B, C) BJP Tissue processing/sectioning, analysis/presentation. KB

staining, imaging, and data acquisition.

ii) Panel D) BJP Data acquisition, and analysis/presentation. KEM Laser

injury/imaging.

Introduction

Muscular dystrophy is a group of genetic disorders characterized by loss of muscle mass and muscle weakness. Duchenne muscular dystrophy (DMD) is the most prevalent form of muscular dystrophy affecting approximately 1 in 5000 male births149,150. The disease affects skeletal, respiratory, and cardiac muscle, with initial symptoms presenting at 2-5 years of age. Children with DMD tend to have an abnormal gait, difficulties rising from the floor, and calf pseudohypertrophy. Most patients are wheelchair bound by their teens and succumb to the disease in their 20s due to respiratory or heart failure. 28

DMD is caused by mutations in the dystrophin gene, which is one of the largest genes in the genome containing 79 exons and spans 2.4 megabases of the X chromosome151. Due to the large size of the dystrophin gene, it is susceptible to frameshift errors and point mutations152,153. Interestingly, the protein contains several domain repeats, and if a portion of the protein is deleted it remains somewhat functional.

This was discovered naturally through the milder form of muscular dystrophy called

Becker muscular dystrophy154,155.

The dystrophin protein provides structure and stability to the muscle fiber membrane by linking the extracellular matrix and intracellular cytoskeleton156. This stability is provided by the N-terminus binding intracellular actin cell157, and the C- terminus binds to dystroglycan in the dystrophin-associated protein complex (DAPC)156.

The absence of dystrophin destabilizes the DAPC, leaving the muscle membrane fragile and vulnerable to compression and stretch related injury158. Initial damage to the myofiber is repaired through satellite cell regeneration, but the regenerative capacity of the muscle is exhausted and fibrotic tissue and fatty tissue infiltration eventually impede muscle contractility causing weakness and loss of funtion159. Calcium homeostasis has also been implicated as a contributor to DMD pathophysiology. Repeat damage leads to elevated levels of calcium which can activate proteases, which may exacerbate fiber damage160.

The mdx mouse is commonly used to model human dystrophic deficiency. The mdx mouse contains a spontaneous mutation causing a premature stop codon in exon 23 which leads to undetectable dystrophin protein levels161. The mice developed

29 characteristics similar to muscular dystrophy patients and have become the most used preclinical mouse model for DMD studies. An interesting characteristic of this mouse is that at approximately 3-5 weeks of age the mouse undergoes muscle necrosis followed by stabilization of the muscle phenotype162. While the mdx myofibers are more susceptible to contraction and stretch related muscle damage, this model does not perfectly mimic disease progression in humans163,164.

Mitsugumin 53 (MG53)/TRIM72 is a member of the tripartite motif family of E3 ubiquitin ligases, and has been shown to be a vital component of the membrane repair machinery in several cell types128-131. The tripartite motif (TRIM) proteins comprise a large gene family containing approximately 80 members that all contain a canonical

TRIM domain on the N-terminal end of the protein containing a Really Interesting New

Gene (RING) domain, zinc binding motifs called B-boxes, and coiled coil domains165,166.

The TRIM domain of these proteins provides E3 ubiquitin ligase enzymatic activity. E3 ubiquitin ligases interact with the generic ubiquitin proteasome machinery to provide specificity for the target protein. While there have been many E3 ubiquitin ligase proteins identified in recent years, there are not enough of these proteins present within the genome to account for all of the target proteins. Formation of TRIM family proteins into heterodimers with differential affinity for different targets would provide an effective mechanism for producing a great deal of diversity from a relatively small number of genes167-169.

The TRIM family is segregated into subfamilies by their variable C-terminus region: class I to XI. MG53/TRIM72 is located in the Class IV family that contains

PRY/SPRY domains in the C-terminal region. The SPRY domain is responsible for selectivity and specificity during protein-protein interactions170,171. When cysteine 242 of the SPRY domain was mutated to an alanine, MG53/TRIM72 failed to locate to the injury site indicating that the SPRY domain is essential to membrane repair129.

Endogenous MG53/TRIM72 has also been demonstrated to contribute to myogenesis. IGFR signals through insulin receptor substrate 1 (IRS-1) to activate MyoD- induced myogenesis172. During myogenesis, MyoD drives transcription of genes that lead to skeletal muscle differentiation173. In C2C12 myoblasts, overexpression of

MG53/TRIM72 inhibited myogenesis, while knock down facilitates myogenesis through degradation of insulin receptor substrate 1 (IRS-1)174. Knockout MG53/TRIM72 mice also have high levels of IRS-1, which indicates that MG53/TRIM72 regulates IRS-1.

MG53/TRIM72 is proposed to be a negative feedback regulator of skeletal muscle myogenesis through the degradation of IRS-1 and the shutdown of the MyoD signaling141.

Along with its role in skeletal muscle, MG53/TRIM72 is essential for the maintenance of cardiac function during injury events. Cardiomyocytes have a limited capacity to regenerate and tissue death can lead to impaired function175. The most common form of cardiac injury is an occlusion of a blood vessel leading to the death of the tissue supplied by that vessel. However, challenges also arise if blood flow is restored due to ischemia reperfusion injury.

Although MG53/TRIM72 does have normal cellular processes, much of the work done on the protein is focused on the membrane repair process129,176. Much of the research done on MG53/TRIM72 has been for its ability to increase membrane repair when overexpressed128-131, additionally it has been shown to be involved in metabolism141,177. Knock out of MG53/TRIM72 results in defective membrane repair and progressive myopathy as the mouse ages. Interestingly, the dysferlin knockout mouse and

MG53/TRIM72 knockout mouse both have a membrane repair defect, but the dysferlin knockout mouse develops muscle pathology around six months in age77 while the

MG53/TRIM72 knockout mouse develops pathology around one year of age129.

MG53/TRIM72 has been shown to be critical for membrane repair, but it is unclear why mutations in its membrane repair binding partners lead to muscular dystrophy, while MG53/TRIM72 has not specifically been linked to the disease. The process of membrane repair is thought to maintain homeostasis, both by repairing the damage to the cell and by decreasing energy demand by not having to replace the dead cell. During normal physiology, membrane repair does not have a significant role, however it is extremely important during stress events7. This is elucidated by the fact that when recombinant human MG53/TRIM72 is injected intramuscularly with cobra venom, cardiotoxin VII, there is significantly less muscle damage and inflammation when compared to control134. It has also been shown to be important in the heart where

MG53/TRIM72 knockout mice are more susceptible to ischemia reperfusion injury. In δ- sarcoglycan (δ-SG)-deficient hamsters, overexpression of MG53/TRIM72 in the heart

32 decreased heart failure, indicating that MG53/TRIM72 and membrane repair is relevant in the heart of muscular dystrophy animals178.

MG53/TRIM72 has been shown to be essential to the membrane repair process, however the importance of the protein in a chronic, diseased state has not yet been elucidated. The mdx undergoes a short bout of necrosis at approximately 3-5 weeks of age, followed by a stabilized pathology that is susceptible to membrane injury. In this study we access the value of the membrane repair protein MG53/TRIM72 in the DMD mouse model by the generation of a MG53/mdx double knockout (DKO) mouse. We find that after the 3-5 week critical period of muscle necrosis, the DKO mice have a membrane repair defect in their flexor digitorum brevis (FDB) muscle, and reduced membrane integrity in the tibialis anterior (TA) muscle when compared to the mdx mouse. Additionally, the maximal force generated by the extensor digitorum longus

(EDL) is reduced compared to the mdx mouse. Aging the mice to 1.5 years resulted in increased fibrosis in the DKO animals when compared to the age matched mdx mouse.

The aged DKO mice also have a breakdown of membrane integrity and membrane repair.

These data indicate MG53 is important for the long term maintenance of muscle in mdx mice, and disruption of the protective effect of MG53 will lead to breakdown of muscle membrane integrity and lead to fibrotic muscle pathology.

Results

Deficiency of MG53/TRIM72 does not affect histological differences in skeletal muscles of 6 week old mdx mice.

The mdx mouse has a discernable phenotype at 3-5 weeks in which it undergoes large amounts of muscle damage and inflammation. Double knockout mice sacrificed at 6 weeks of age did not show differences in raw weight in heart, diaphragm, or skeletal muscle at this age, or when normalized to body weight (Fig 5). Histologically, there was no difference in cross sectional area or central nuclei of EDL, soleus, and TA muscles

(Fig. 6A, B, C). However, the DKO mouse showed a significant decrease in maximal force in the EDL when compared to the mdx mouse (Fig. 6D). There was no observed difference between the force outputs of soleus muscles between the two phenotypes (Fig.

6D). These results indicate that although the double knockout mouse EDL did not show a difference in cross sectional area (CSA), MG53 may play a role in maintaining the level of force produced by the muscle.

Membrane repair is compromised in the 6 week old DKO mouse

Membrane repair has been shown to be a crucial process in skeletal muscle and has been shown to be a targeted treatment in the mdx mouse134. Additionally, the membrane repair proteins dysferlin, MG53, and caveolin-3 are upregulated in DMD patients. Since the mdx mouse lacks functional dystrophin, the integrity of the plasma membrane is compromised leading to muscle fibers becoming more reliant on membrane repair to keep fibers alive. mdx and DKO mouse EDL and TA muscles were subjected to

IgG staining at 6 weeks of age to investigate the amount of IgG antibodies trapped in muscle fibers after membrane injury. DKO TA muscles showed a significant increase in

IgG positive areas (Fig. 7A). To directly test membrane repair, whole FDB muscle taken from 6-week-old mice were subjected to infrared laser injury ex vivo using a multiphoton

34 microscope. DKO mice showed a significant increase in dye influx indicating significantly compromised membrane repair (Fig. 7B). These findings indicate that although the mdx mouse has a defect in membrane integrity, adding a defect in membrane repair can exacerbate the membrane injuries.

Membrane repair proteins are increased in the 6 week old DKO mouse

As shown in previous studies, knockout of MG53 in wild type mice leads to membrane repair defects. It is expected to see compensation by other membrane repair proteins to accommodate the stress of increased injury as a result of decreased membrane integrity. In both the EDL and soleus, we observed an increase in caveolin-3 protein (Fig.

8). Interestingly, we did not see an increase in dysferlin expression, but the expression was capricious which may indicate that levels may be changing in response to different stages of injury and regeneration.

DKO mouse in response to age stress

As animals age, their muscles accumulate stress. This is especially relevant in mice that have compromised membrane repair and membrane fragility. The DKO mice developed a hunched posture at approximately 1.5 years of age and showed significantly decreased body mass (Fig. 9). However, the skeletal muscle, heart, and diaphragm mass were not significantly different from the mdx mouse when normalized to body weight

(Fig. 9). Histologically, there was also no difference in cross sectional area or central nuclei of EDL, soleus, and TA muscles (Fig. 10A, B, C). The force generated from both mice decreased when compared to the younger 6 week old mice, and similarly to the 6

35 week old mice, the EDL exhibited significantly less force in the DKO mouse when compared to the mdx mouse (Fig. 10D).

Robust fibrosis occurs in aged DKO mice

Repeated damage to muscle that occurs with muscular dystrophy eventually results in excessive deposition of fibrous tissue which results in impairment of muscle function, muscle fiber regeneration, and susceptibility to re-injury. The mdx mouse undergoes repeated cycles of muscle damage and regeneration, leading to increased fibrotic muscle tissue. At 1.5 years, the DKO mouse has significantly more fibrotic tissue in every tested skeletal muscle (soleus, EDL, and TA), when compared to the mdx mouse

(Fig. 11A-C). Similarly, the hearts of the DKO mice also contained elevated levels of fibrosis (Fig. 11D). However, diaphragm muscle did not show a difference between genotypes (Fig. 11E). The diaphragm has been shown to be the most injury susceptible muscle in the mdx model, and it may be in a more advanced disease state at the time of sampling. These results indicate that the loss of MG53 may not be critical for the initial injury that occurs at 3-5 weeks, but the repeated injury in the dystrophic model requires a complete membrane repair response.

Membrane repair and integrity is compromised in aged DKO mice.

Aged muscle undergoes various molecular and physical changes in response to the stress of aging. This is particularly relevant to dystrophic muscle due to repeated bouts of injury and regeneration. Similarly, to the fibrosis analysis, the DKO EDL, soleus, and TA muscles all showed a break down of membrane integrity when compared to aged mdx mice (Fig. 12A-C). The DKO mice also had decreased membrane repair

36 when tested with a laser injury assay (Fig. 12D). These results confirm the value of membrane repair in the mdx model. The loss of MG53, although not debilitating in young mice, is crucial for the maintenance of myofibers as they undergo repeated bouts of damage.

Discussion

Our studies show that dystrophic muscles are dependent on the membrane repair protein MG53 to survive the naturally occurring repeated injury. Dystrophic muscle is dependent on the membrane repair process due to the lack of the dystrophin protein that provides a force-transferring link from the cytoskeleton to the extracellular matrix. At the six-week timepoint, we observed minimal histological differences between the DKO and mdx mouse. However, the EDL muscle of the DKO mouse was functionally weaker than the mdx age matched control. Our findings indicate that in the absence of MG53, other membrane repair proteins are upregulated to compensate for the deficiency. In response to aging mice 1.5 years, DKO mice contain significantly more fibrotic muscle in all skeletal muscles examined. Fibrosis is a hallmark of muscular dystrophy, and lack of

MG53 exacerbated the phenotype. We interpret these findings to indicate, that while

MG53 may not be critical for the initial dystrophic injury, it is essential for maintaining muscle composition during the repeated bouts of damage caused by dystrophic deficiency. The value of membrane repair in muscular dystrophy is well documented, and a previous study179 illustrated that the absence of the critical membrane repair protein dysferlin in the mdx model, leads to massive muscle necrosis and weakness.

Our results suggest that MG53 is critically important to the function of the EDL muscle and heart of dystrophic mice. As seen in the maximal contractile force generated by the EDL muscle of mice 6 week and 1.5 year old, the EDL produces significantly less force in the DKO mice when compared to mdx age matched controls. Previous studies have shown that the EDL of mdx mice is susceptible to contraction related membrane injuries163. This indicated that the fast twitch nature of the EDL muscle may increase the strain on the membrane caused by the muscle contraction, leading to muscle damage and a drop in force.

An interesting comparison when investigating membrane repair in muscular dystrophy mouse model is the dysferlin mdx double knockout mouse. This mouse model was observed to have severe muscle damage at six months when compared to the mdx mouse179. Although both dysferlin and MG53 play a role in membrane repair, dysferlin is also involved in satellite cell fusion180, whereas MG53 is only expressed in mature myotubes50. The formation of new myotubes is crucially important in mdx mice due to its susceptibility to membrane damage and is likely a cause for differences between both double knockout models.

Although, the histological data do not reveal any differences in central nuclei or cross sectional area, it is important to highlight the low n values for those analysis. With a lower n value, the analysis loses statistical power, and differences in the analyzed data could be obscured as a result. This decreased sample size could account for differences in maximal force produced by the EDL muscle. An observed functional difference between the genotypes was measured (Fig4 D), while this data did not correlate to the measured

CSA. My analysis pointed to the defect in membrane repair compromising the ability of the EDL muscle to contract due to injury, however with the low n value for the muscle histology it cannot be ruled out that they may be a difference in CSA.

Our study focused on the value of membrane repair in the muscular dystrophy mouse model. We knocked out the critical membrane repair protein MG53 to illicit a membrane repair defect in these mice. Minimal histological differences were observed in

6 week old mice, however robust fibrosis and reduced membrane integrity were observed in aged mice. This demonstrates that acute injuries where muscles can regenerate are easily compensated, however continuous bouts of injury tax the regeneration capacity of muscle leading to fibrotic muscle. This model demonstrates the value of membrane repair in repeat injury and cements its role in maintaining muscle fiber survival in dystrophic muscle.

Materials and methods

Mouse model breeding

Mice were bred and maintained in standardized conditions at 22 ± 2C under a 12- hr/12-hr light cycle (lights on at 7 a.m. EST). Mice were provided standard mouse chow and drinking water ad libitum. All experimental procedures were approved by The Ohio

State University Institutional Animal Care and Use Committee. Animals were maintained in accordance with the recommendations of the NIH Guide for the Care and Use of

Laboratory Animals. MG53 knockout mice (MG53-/-) were previously generated129 and crossed with commercially available mdx mice (Jackson Laboratory). Three generations of crosses of male progeny with mdx female mice produced littermates MG53-/- /mdx

39 double knockout (DKO), or MG53+/+ /mdx (mdx) (Fig. 4). Lack of MG53 and dystrophin expression was confirmed by Western blotting using standard techniques as described below.

Tissue preparation for histological procedures

Male mice at various ages (indicated where appropriate) were euthanized by use of CO2 asphyxiation followed by cervical dislocation, and hind limb muscles were extracted. Tissue was then fixed with 10% phosphate buffered formalin, followed by a 24 hour incubation with 70% ethanol. Tissue was then embedded in paraffin (Thermo) and

12μm sections were mounted on SuperFrost Plus slides (Fisher Scientific, Hampton, NH,

USA). Slides were deparaffinized through changes of xylene and rehydrated through decreasing concentration of ethanol. The slides were then utilized for hemotoxylin/eosin

(Thermo Scientific) or Masson’s trichrome (American MasterTech) for immunohistochemical staining.

Muscle Fiber Size Measurements

Muscle fiber cross-sectional area (CSA) was determined from sections of EDL, soleus, and TA muscles stained with hemotoxylin/eosin. Image files were blinded before analysis to preclude any bias. Muscle fiber CSA was determined by outlining individual fibers using ImageJ. Muscle fiber CSA was analyzed for their frequency distribution

(200µm bins), and graphed based on the percentage of fibers corresponding to the binned fiber size.

Central Nuclei Measurements

Centrally located nuclei were counted from fibers measured for CSA. Fibers were counted as positive for centrally located nuclei if nuclei were located inside of the perimeter of the outlined fiber. Image files were blinded before analysis to preclude any bias.

IgG staining and analysis

12μm paraffin embedded sections of EDL, soleus, and TA muscles were deparaffinized, and rehydrated followed by antigen retrieval with Citra Plus Solution

(Biogenex, CA, USA). Sections were then blocked in 2.5% bovine serum albumin

(SIGMA) for 1 hour, incubated overnight with goat anti-mouse IgG antibody conjugated with Alexa Fluor 488 (Life Technologies). Analysis of the % area IgG+ was conducted using ImageJ. The area of an image representing skeletal muscle was quantified by visually setting a threshold which included the entirety of the muscle section to be analyzed an d excluding histological artifacts and the dark background where no tissue was present. A second threshold corresponding to areas of each skeletal muscle section positive for IgG was applied to each image. The % area of IgG+ skeletal muscle was then determined by dividing the area IgG+ by the total area of each image representing skeletal muscle.

Masson’s Trichrome Analysis

Analysis of the % fibrosis was conducted using ImageJ on Masson’s trichrome stained sections. 20x images were stitched together encompassing the entire muscle using the EVOS FL Auto 2 inverted microscope imaging system (ThermoFisher). Using

ImageJ, the area of an image representing tissue was quantified by visually setting a

41 threshold which included the entirety of the muscle section to be analyzed and excluding histological artifacts and background. The ImageJ plug-in, color deconvolution was used to separate the blue channel, and a second threshold corresponding to areas of the tissue positive for fibrosis was measured. The % fibrosis of tissue was then determined by dividing the area positive fibrosis by the total area of each image representing whole tissue.

Ex vivo assessment of skeletal muscle contractility

Isolated mouse muscle contractility was measured as described previously 129, briefly EDL and soleus muscles were dissected and mounted between two electrodes in a chamber filled with Tyrode’s solution supplemented with 2 mM Ca and 12 mM glucose.

Oxygen was bubbled in the Tyrode’s solution to ensure sufficient oxygenation of the muscles during the protocol. Muscles were stretched to ensure maximal force using 80hz pulses and constant stimulatory voltage (one 80hz pulse every minute for 30 minutes) was applied to allow for the muscles to equilibrate. Following equilibration, muscles were stimulated at frequencies from 1-150hz to generate a force vs frequency curve. The force versus frequency curve was generated by stimulating muscles at the following frequencies: 1, 5, 10, 20, 30, 40, 50, 60, 80, 100, 120, 140, and 150 hz. Maximal force (F- max) was taken from the peak force on the resulting force frequency curve.

Western blotting

Tissue was taken from mice and extracted using Radioimmunoprecipitation Assay buffer (RIPA; Cell Signaling Technology, Danvers, MA, USA). Protein concentrations

42 were determined in accordance with the standard Bradford Assay using bovine serum albumin (BSA) standards. Protein samples (20µg/lane) were separated by SDS-PAGE at room temperature on 4%–15% gradient gels (Bio-Rad, Hercules, CA, USA) and were transferred to nitrocellulose membranes (Bio-Rad, Hercules, CA, USA). Blots were probed for TRIM72/MG53 with a custom polyclonal antibody (Pacific Immunology, San

Diego, CA, USA), dysferlin (Leica Biosystems, IL, USA), Caveolin-3 (abcam,

Cambridge, UK), GAPDH (Cell Signaling Technology), anti-mouse horseradish peroxidase (HRP)–conjugated secondary antibodies, and anti-rabbit HRP–conjugated secondary antibodies (Cell Signaling Technology). The blots were developed using enhanced chemiluminescence (ECL) substrate (Bio-Rad). An Azure Biosystems imager was used to visualize chemiluminescent blots. Quantification of immunoreactive bands was performed by densitometric analysis using ImageJ’s integrated density measurement and normalized with GAPDH levels.

Membrane repair assessment following laser injury

Whole flexor digitorum brevis (FDB) muscles were dissected from mice and mounted on a 35mm glass bottom imaging dish with commercially available liquid bandage. Membrane damage was induced in Tyrode’s solution containing 2.0 mM Ca2+, using the Olympus FV1000 multi-photon laser scanning confocal system. For laser injury measurements, injury was induced in the presence of 2.5 μM FM4-64 fluorescent lipophilic dye (Life Technologies). A circular area was selected along the edge of the cell membrane and irradiated at 20-30% laser power for 5 s. Pre- and post-damage images were captured every 3 s, continuing for 57 s. The extent of membrane damage was

43 analyzed using ImageJ software, by measuring the fluorescence intensity encompassing the site of damage. To preclude any potential for bias, all of the experiments were performed in a blinded fashion.

Statistical Analysis

Graphical representation and statistical analysis of data was performed using

GraphPad Prism version 8. All results are presented as mean ±SEM. Cross sectional area measurements were grouped based on frequency distribution in bins of 200µm.

Measurements were analyzed by a two-way ANOVA with a Bonferroni post hoc test.

Central nuclei, F-max, laser injury AUC, percent fibrosis, and protein expression measurements were analyzed by unpaired two-tailed t-test assuming unequal variances,

Welch's T-test for unbalanced designs.

Figure 4: Breeding strategy for the generation of MG53/mdx DKO mice.

DKO mice were generated through three generations of breeding. F0 consisted of breeding MG53KO (MG53 -/-) and mdx (DMD -/-). F1 crossed the male progeny of F0 45

(MG53 -/-), mdx) with a female mdx mouse. F2 crossed MG53+/-, mdx littermates to produce a F3 generation of MG53KO/mdx(MG53 -/-, DMD -/-) and MG53WT/mdx

MG53 +/+, DMD -/-) mice. Male mice were used to account for sex as a variable.

Figure 5: DKO mice maintain muscle mass consistent with mdx mice. mdx and DKO mice were sacrificed at 6 weeks of age, and wet muscle weights were measured immediately after surgical dissection. Muscles effected by mdx pathology were measured to investigate differences in pathologies between genotypes. mdx n=13, DKO n=14. Statistical analysis was performed by unpaired two-tailed t-test assuming unequal variances, Welch's T-test for unbalanced designs. Body weight p=0.4332, heart p=0.2931, diaphragm p= 0.8420, EDL p=0.1302, soleus p=0.3482, TA p=0.0671. Data represented as means ± SEM.

Figure 6: Histological analysis of skeletal muscle of mdx and DKO mice at 6 weeks.

H & E staining of EDL (a), soleus (b), and tibialis anterior (c) sections from mdx and

DKO mice at six weeks of age were analyzed for frequency distribution of myofiber cross-sectional area (CSA), and central nuclei. There was no difference in the histology of any of the muscles (n=3) for all groups tested. (d) Maximal force (F-max) of ex vivo

EDL and soleus muscles from mdx and DKO mice. Force was significantly reduced in the DKO EDL muscles. Differences in F-max were compared by unpaired two-tailed t- test assuming unequal variances, Welch's T-test for unbalanced designs. F-max EDL: mdx n=10, DKO n=13 p=0.0414; Soleus: mdx n=12, DKO n=7, p=0.2532. *= p< 0.05.

Data represented as means ± SEM to indicate the confidence level in the mean at each frequency. Scale bar= 100µm

Figure 7: Membrane repair and integrity is compromised in DKO mice.

(A) Representative images of IgG staining. Paraffin sections of EDL and TA muscles stained with fluorescent anti-mouse-IgG antibodies demonstrate the distribution of IgG- 50 positive and IgG-negative fibers in selected skeletal muscles. Quantification analysis of

IgG-positive fibers for the EDL and TA show a significant increase in positive muscle fiber damage in the DKO group. EDL: n=4, p= 0.8456; TA n=4, p=0.0081. (B)

Representative images of FM4-64 dye in whole FDB muscles from mdx and DKO mice.

The area under the curve (AUC) of FM4-64 fluorescence traces display different membrane resealing in DKO mice. mdx n=22 fibers; DKO n=16 fibers. p=0.0052. **= p< 0.01 Data represented as means ± SEM. (A) scale bar = 50µm, (B) scale bar= 20 µm.

Figure 8: Membrane repair proteins are altered in DKO mice.

(A) Soleus and (B) EDL muscle lysates from 6-week old mice were used for western blotting to detect changes in membrane repair proteins. Cav-3 protein expression appears to increase in both muscles, however statistical significance was only seen in the EDL muscle. Statistical analysis was performed with an unpaired two-tailed t-test assuming unequal variances, Welch's T-test for unbalanced designs. A) (n=4) Dysf p=0.8747, Cav-

3 p=0.0756; B) (n=4) Dysf p= 0.8326, Cav-3 p=0.0206 *= p< 0.05. Data represented as means ± SEM.

Figure 9: Age affects body weight, but does not affect tissue weight mdx and DKO mice were sacrificed at 1.5 years of age, and wet muscle weights were measured immediately after surgical dissection. Muscles affected by mdx pathology were measured to investigate differences in pathologies between genotypes. mdx n=7, DKO n=8. Statistical analysis was performed by unpaired two-tailed t-test assuming unequal variances, Welch's T-test for unbalanced designs. Body weight p=0.0161, heart p=0.1414, diaphragm p= 0.4269, EDL p=0.2225, soleus p=0.0661, TA p=0.2164. Data represented as means ± SEM.

Figure 10: Histology and functional analysis of skeletal muscle of mdx and DKO mice at

1.5 years.

H & E staining of EDL (A), soleus (B), TA (C) sections from mdx and DKO mice at 1.5 years of age were analyzed for frequency distribution of myofiber cross-sectional area

(CSA), and central nuclei. n= 4 per group (d) Maximal force of ex vivo EDL and soleus muscles from mdx and DKO mice. EDL: mdx n=11, DKO n=9 p= 0.0150; Soleus mdx n= 10, DKO n=9 p= 0.8944. CSA was analyzed with two-way ANOVA, Differences in central nuclei counts and maximal force were tested for statistical significance with an unpaired two-tailed t-test assuming unequal variances, Welch's T-test for unbalanced designs. *= p< 0.05. Data represented as means ± SEM. Scale bars=100 µm

Figure 11: Masson’s trichrome stain analysis implicates robust fibrosis of DKO mice at

1.5 years.

DKO mice experience repeat bouts of damage throughout their lifetime leading to increased fibrotic tissue in skeletal (a, b, and c) and heart muscle (d). Levels of fibrosis in the diaphragm were elevated, but no major differences in the fibrotic tissue in mdx and

DKO diaphragms were observed. Data represented as means ± SEM. n=5 for all groups.

EDL: p=0.0216, Soleus p=.0490, TA p= 0.0280, Heart p= 0.0015, Diaphragm p=0.6512.

(A,B) Scale bar=50 µm, (C,D) Scale bar=100 µm, E Scale bar=250 µm

Figure 12: Membrane repair and integrity is compromised in aged DKO mice.

(A) Representative images of IgG staining. Paraffin sections of EDL, soleus and TA muscle stained with fluorescent anti-mouse-IgG antibodies demonstrate the distribution of IgG-positive and IgG-negative fibers in selected skeletal muscles. Quantification analysis of IgG-positive fibers for the EDL, Soleus, and TA show a significant increase in positive muscle fiber damage in the DKO group. Data analyzed by unpaired two-tailed t- test assuming unequal variances, Welch's T-test for unbalanced designs. N=4 for all groups. EDL p=0.0192, Soleus p=0.0221, TA p= 0.0399. (B) Representative images and time- dependent accumulation of FM4-64 dye in whole FDB muscles from mdx and

DKO mice. The area under the curve (AUC) of FM4-64 fluorescence traces display different membrane resealing in DKO mice. AUC was analyzed with t-test. mdx n=10

DKO n=9p=0.0046 **= p< 0.01. Data represented as means ± SEM. (A, B) Scale bar=

Scale bar=50 µm, (C) Scale bar=20 µm, (D) Scale bar=20 µm.

Chapter 3. Enhancing Membrane Repair Increases Regeneration In A Sciatic Injury

Model

This chapter has been adapted from the published article:

Paleo B.J., Madalena K.M., Mital R. McElhanon K.E., Kwiatkowski T.A. Rose A.L.,

Lerch J.K. Weisleder N; Enhancing Membrane Repair Increases Regeneration in a

Sciatic Injury Model. PloS one 15, doi:10.1371/journal.pone.0231194 (2020).

The contributions of each were as follows:

a) Writing: BJP Original draft preparation. BJP KMM JKL NW Review and editing.

b) Fig 3.1

i) Panel A) BJP Western blotting, imaging, data acquisition, and presentation.

ii) Panel B) BJP Tissue processing/sectioning/staining, and imaging.

c) Fig 3.2

i) Panels A, B) BJP Cell culture, transfection, laser injury/imaging, data acquisition,

and analysis/presentation.

ii) Panels C, D) BJP Cell culture, transfection, laser injury/imaging, data acquisition,

and analysis/presentation.

d) Fig 3.3

i) Panels A, B, C) BJP Primary cell culture, laser injury/imaging, data acquisition, and

analysis/presentation.

ii) Panel D) AR TK Data acquisition, and analysis/presentation.

e) Fig 3.4

i) A, B, C) BJP JL Immunohistochemistry. BJP Imaging. JL Surgery. Data acquisition

RM. KMM analysis/presentation.

ii) Panel D) BJP Immunohistochemistry, imaging, and presentation. JL Surgery.

Introduction

A wide variety of neurodegenerative diseases and traumatic injuries can cause irreversible damage to neural tissue and compromise various functions of the nervous system. Due to limited treatment options after neurodegeneration, there are intensive efforts underway to investigate the mechanisms of neuronal regeneration and recovery from injury. Despite such efforts, these processes are not yet fully understood, and this lack of knowledge has hampered the development of therapeutics in this area. Since many neurons are terminally differentiated and axonal regeneration can result in reinnervation of off-target tissue 181-183, emphasis on maintaining cell survival could provide increased translational potential for the treatment of neuronal injury. While several cell types are known to survive injury through plasma membrane repair mechanisms, there has been little investigation of membrane repair in neurons and even fewer efforts to target membrane repair in diseases affecting neurons 128,130,131,133,135

Plasma membrane repair is a conserved mechanism observed in many cells from simple single cell eggs to most adult mammalian cell types 184. Membrane repair mechanisms function to close disruptions in the plasma membrane to restore structural integrity and to maintain barrier function to prevent cell death following injury. These repair mechanisms allow cells to survive an injury, which has great advantages over 62 regeneration of a new cell to replace a large, complex cell like a neuron. While disruptions in the plasma membrane measuring approximately 1 nm or less will reseal through thermodynamic rearrangement of the component phospholipids, larger injuries generally require an active repair response to restore membrane integrity 185.

The mechanical tension on the plasma membrane produced by the connections of the membrane to the cytoskeleton and extracellular matrix can force disruptions to remain open and require compensation by active membrane repair mechanisms 185-187. Several previous studies helped to establish the cellular framework of the plasma membrane repair process. While multiple models of membrane repair have been proposed, most of these models involve exocytotic and/or endocytotic vesicle trafficking to facilitate the resealing of membrane disruptions 11,188,189. Given the importance of plasma membrane repair for cell survival, it is likely that most of these putative pathways contribute to the repair process in a cell-type and injury-type dependent fashion.

The current understanding of the plasma membrane repair response comes mainly from studies in striated muscle cells. Much of this interest comes from studies in muscular dystrophies, such as the dysferlinopathies produced by the lack of a membrane repair protein dysferlin 104. In striated muscle fibers, repair of most sarcolemmal membrane disruptions involves calcium dependent translocation of intracellular vesicles to the injury site where these vesicles then fuse with each other and the plasma membrane, to form a repair patch that restores the integrity of the membrane. This process has several similarities to the release of neurotransmitters from neurons 8. Given this fact, it is not surprising that proteins involved in vesicle fusion and neurotransmitter

63 release have also been shown to be involved in membrane repair. Inhibition of synaptotagmin I and syntaxin have been used to block membrane repair in axons in the past 93,190.

Studies from our laboratory group and others demonstrated that mitsugumin 53

(MG53), a muscle-enriched tripartite motif (TRIM) family protein also known as

TRIM72, is an essential component of the cell membrane repair machinery in multiple cell types, including striated muscle, liver, and alveolar epithelial cells 42,128,129,131,135,191.

TRIM72/MG53 is an essential component of the membrane repair machinery as

TRIM72/MG53 ablation results in defective membrane repair, progressive skeletal myopathy, and vulnerability to ischemia-reperfusion injury 129,135. Interestingly, when

TRIM72/MG53 is expressed in non-muscle cell types it can still function in a similar fashion to increase membrane repair in non-muscle cells 134. Moreover, exogenously applied recombinant human MG53 (rhMG53) can increase membrane repair and the integrity of muscle and non-muscle cells both in vitro 134 and in vivo 130,134. Specific evidence that rhMG53 is effective in treating injuries comes from studies that efficaciously treated mouse models of muscular dystrophies 134,192. Similar to results seen when overexpressing the protein inside various cell types, rhMG53 also increased membrane repair in non-muscle cells in vivo. Mice subjected to acute kidney injury showed decreased effects from ischemia reperfusion injury when treated with rhMG53

130. Interestingly, rhMG53 has been very successful in treating ischemia/reperfusion injury and has been also used in the treatment of skeletal muscle 28,138, cardiac muscle 27, and liver tissue 191. Also, aerosolized rhMG53 has ameliorated the effects of ventilator-

64 induced lung injury 133,193. These studies introduce the possibility that rhMG53 could be used to increase the repair capacity of other non-muscle cell types like neurons.

While these previous studies focused on striated muscles, they identified some of the molecular components of the membrane repair process 194, other studies established that disruption in plasma membrane repair and integrity can occur in many cell types aside from muscle fibers 6,195,196. Changes in membrane repair capacity can lead to a number of diseases including heart failure, Alzheimer's disease and neurodegeneration

36,102,195,197-203. Despite the relevance of membrane repair to these disease states, there has been little investigation of membrane repair specifically in neurons. It is possible that by increasing the membrane repair capacity of neurons, we could potentially minimize the death of these cells and thereby, affect the progression of various neuronal diseases.

While rhMG53 has been shown to increase membrane repair capacity 130,134, the lack of knowledge of the membrane repair process in neurons presents a novel opportunity to explore the potential efficacy of modulating membrane repair in neurons. In this study, we assess the therapeutic potential of rhMG53 in neuronal cells both in vitro and in vivo.

We find that a robust repair response exists in various neuronal cells and that rhMG53 can increase neuronal membrane repair in vitro. Additional experiments found that treatment with rhMG53 significantly increased regeneration in an in vivo mouse model of sciatic nerve injury. These data indicate that neurons have an endogenous membrane repair response that can be targeted with rhMG53, and further indicates that there are potential therapeutic benefits to elevating membrane repair in neurons that could have protective effects against injuries to the nervous system.

Results

TRIM72/MG53 is not expressed in neurons

Initial TRIM72/MG53 studies indicated that it was expressed exclusively in the striated muscle tissues of the skeletal muscle and the heart 129. Further examination showed that TRIM72/MG53 expression might appear exclusively in skeletal muscle in humans 204. While TRIM72/MG53 does appear to be highly enriched in striated muscle, recent studies indicate that under certain conditions, TRIM72/MG53 expression can be found in certain cell populations in non-muscle tissues. TRIM72/MG53 expression has been shown in lung type II alveolar epithelial cells where it is important for resisting mechanical injury to the lung 133. While the liver does not usually express

TRIM72/MG53, it appears that ischemia can induce expression in this tissue 191. Thus, we conducted studies to determine if TRIM72/MG53 was expressed in neurons. Western blot analysis was used to examine mouse tissues from the central nervous system (whole brain and spinal cord lysates) and the peripheral nervous system (sciatic nerve lysate).

TRIM72/MG53 expression was not observed in any of these neural tissues when compared to the positive control of skeletal muscle lysate (Fig. 13 A).

Since previous studies indicate that TRIM72/MG53 may be expressed in subpopulations of cells in a given tissue, we used immunohistochemistry to test for expression of TRIM72/MG53 in the mouse spinal cord (Fig. 13B). No TRIM72/MG53 expression was observed in individual cells seen in these sections when compared to the positive control of mouse skeletal muscle. These results from several different areas of

66 the nervous system establish that TRIM72/MG53 is not detected in the peripheral or central nervous system.

Membrane repair responses are active in cultured neurons

While previous studies examined the cellular response in severed axons that likely makes use of membrane repair responses 205,206, there have been limited studies of the membrane repair response in the cell body of neurons. We examined whether cell membrane disruptions in the neuron body lead to the formation of a typical membrane repair patch as seen in other cell types following membrane repair. To visualize the formation of a membrane repair patch, we transfected Neuro2a (N2a) cells with a GFP tagged TRIM72/MG53 construct (GFP-MG53) that has been shown to translocate to the site of membrane repair patches by multiple laboratory groups 44,129. Transfected N2a cells were injured using a multiphoton infrared laser and the localization of GFP-MG53 or GFP alone (as a control) were monitored using confocal microscopy. We observed that following membrane injury, the GFP-MG53 moves rapidly to the injury site while GFP remains diffusely localized throughout the cytoplasm (Fig. 14A). Quantification of GFP fluorescence at the site of injury from multiple cells (Fig. 14B) shows that GFP is bleached at the injury site and that there is no rapid recovery of GFP fluorescence at that site. In contrast, the GFP-MG53 rapidly translocates to the injury site with the fluorescent signal reaching a plateau at approximately 30 seconds after the cell is injured. These results indicate that a membrane repair patch can effectively form at injury sites in the cell body of neuronal origin cells in a manner to that previously observed in muscle cells.

Cultured neurons display effective membrane repair

While we observed that neural origin cells can form a membrane repair patch, those experiments are not able to resolve if that patch is effective at restoring the integrity of the membrane following injury. To test if the membrane repair patch is functional, we used an FM4-64 dye exclusion assay to determine how effective the membrane repair response is in these cells. The lipophilic dye FM4-64 fluoresces poorly in aqueous solution when outside the cell but provides a strong fluorescent signal when bound to lipids within the cell membrane or the cytosol. When multiphoton confocal laser microscopy is used to injure the membrane, the FM4-64 dye enters the cell and the fluorescent signal increases until the membrane reseals. Less dye influx corresponds to more efficient membrane repair in the cell.

N2a cells were transfected with GFP-MG53 or GFP and then injured by the infrared laser (Fig. 14C). Quantification of the extent of FM4-64 dye entry in multiple cells shows that N2a cells can effectively reseal their membrane as the dye influx will rapidly stabilize within 30 seconds of the injury. We also find that transfection of GFP-

MG53 accelerates membrane repair in N2a cells as it does in other non-muscle cell types

134. These results indicate that neural origin cells intrinsically display membrane repair in their cell bodies and that this membrane repair can be accelerated by expression of

TRIM72/MG53.

MG53 increases the membrane repair capacity of cultured neuronal cells

Given that neural cells display an effective membrane repair response and that this membrane repair response can be accelerated by expressing TRIM72/MG53, we tested if rhMG53 can increase the membrane repair capacity in primary isolated neurons,

68 and cultured neural origin cells. We first tested if treatment with rhMG53 before laser injury could improve membrane repair in isolated mouse dorsal root ganglion (DRG) neurons (Fig. 15A). We find that low dose rhMG53 (1 µM) can increase membrane repair by significantly decreasing the influx of FM4-64 dye (Fig. 15B). We also find that the use of another membrane resealing agent, poloxamer 188 (P188) 207,208, also increases membrane repair when applied to cultured DRG neurons (Fig. 15C), albeit when provided at a high dose (100 µM). These results show that primary neurons have a membrane repair response and that it can be improved through application of known membrane resealing agents.

While laser based injury is widely used to assess membrane repair capacity 129,209-

211 we determined if other methods of neuronal injury could be affected by rhMG53. We used an assay that mechanically injured N2a cells with glass microbeads. In this assay, the amount of intracellular lactate dehydrogenase (LDH) that leaks into the extracellular space gives an indication of the extent of membrane repair in the cell population; less

LDH release indicates more effective membrane repair. Application of rhMG53 (1 µM) can improve membrane repair to the same extent as a high dose of P188 (100 µM) following mechanical injury to N2a cells (Fig. 15D). These findings indicate that neuronal cells can reseal following mechanical injury and that these resealing agents can improve membrane repair resulting from other sources of injury. rhMG53 increases regeneration after sciatic nerve crush injury in vivo

Since we observe that isolated neurons display membrane repair and that this repair response can be improved with the application of rhMG53, we wanted to establish

69 if there could be therapeutic effects in vivo. To test this, we determined if rhMG53 can be effective at mitigating damage to allow for increased nerve regeneration in a mouse model of sciatic nerve crush injury. C57Bl/6J mice were anesthetized, and the sciatic nerve was surgically exposed to allow for a 10 second mechanical crush injury.

Immediately following the crush injury, 1µL of rhMG53 (1mg/mL) or saline vehicle control was injected into the epineurium distal to the crush site. At 3 days post-injury, the sciatic nerve was removed and fixed in paraformaldehyde then immunostained for

SCG10, a marker for regenerating axons 1,212 (Fig. 16A). The intensity of SCG10 staining was quantified along the length of the sciatic nerve. rhMG53 treated nerves were observed to have regenerating axons extending further than the saline treated nerves (Fig.

16B). A regeneration index was generated by calculating the distance from the crush site at which the average intensity of SCG10 staining was half that at the crush site 213-215.

There was a two-fold increase in regeneration of rhMG53 treated nerves when compared to saline treated (Fig. 16C). These findings indicate that rhMG53 treatment significantly increased regeneration in the sciatic nerves compared to saline treated control nerves. The measured effect of rhMG53 was amplified in our analysis due to the great increase of regeneration that occurs with treatment. Many treated nerves failed to drop below the threshold, and thus had a robust regeneration index value. This effect can be directly linked to the presence of rhMG53, because at 3 days post-injury, rhMG53 was observed by immunocytochemistry in the rhMG53 treated nerves while no MG53 immunostaining signal appeared in the saline treated nerves (Fig. 16D). Thus, the presence of rhMG53 in

70 the nerve correlates with an increased survival of neurons that allows for more robust regeneration at the crush injury.

Discussion

Our studies show that neurons have an endogenous membrane repair mechanism that is capable of forming a membrane repair patch in the cell body in response to various forms of cell injury. These membrane repair patches are capable of restoring the barrier function of the plasma membrane 8,216. We also find that overexpression of

TRIM72/MG53 or the application of rhMG53 can increase the membrane repair capacity of neurons in vitro and in vivo. This capacity for rhMG53 to increase membrane repair in neurons translated to increased regeneration in a mouse model of sciatic nerve injury.

Since rhMG53 can compensate for mechanical injury of N2a cells and laser injury of

DRG cell bodies by increasing membrane repair, we interpret this finding to mean that application of rhMG53 to the injured nerve could potentially lead to increased cell survival and/or an overall decrease in the damage done to neurons by the crush injury, which leads to increased regeneration or more effective regeneration. This hypothesis is supported by a growing list of evidence showing the benefits of rhMG53 as a facilitator of membrane repair to treat various cellular injuries 27,130,133,138,217, including ischemia/reperfusion, chemical, and damage resulting from genetic mutations such as in muscular dystrophy mouse models.

Our results suggest that rhMG53 could potentially be used for a variety of neurotrauma events and potentially various neurodegenerative diseases. We show that rhMG53 can be effective in the neuron cell body where it could contribute to the survival of the whole

71 cell and not just to projected axons. This makes rhMG53 a potential treatment for a neurotrauma like traumatic brain injury, spinal cord injury, or disruption of peripheral nerves. Outside the direct increase of membrane repair capacity, there may be additional effects of rhMG53 that could be beneficial in the treatment of neural injury. Injuries to the central nervous system are exacerbated by various inflammatory responses of the immune system 218,219. Previous studies suggest that following lipopolysaccharide- induced neural toxicity, the application of rhMG53 reduced microglial activation 220.

While an exact mechanism for this effect is not known, this capacity may allow rhMG53 to both dampen the immune response, as well as increase membrane repair, which would increase its value as a therapeutic for treating injuries to the nervous system. This concept is supported in a rat stroke model, where rats subjected to a brief ischemia/reperfusion event and then treated with rhMG53, show a smaller infarct area 26. That study and our nerve crush model both showed efficacy when rhMG53 was applied immediately post- injury. This is an important finding because it shows that there is potential for the use of rhMG53 as a therapy since treatment for an injury must be effective after an injury occurs. Glial cells play an important role the regeneration process 221, which could lead to future studies investigating if there are benefits to increasing membrane repair in glial cells or other non-neuron cell types in the nervous system.

We were unable to find TRIM72/MG53 protein expression in any of the neural tissues that we examined. While there is no native TRIM72/MG53 in these tissues, the overexpression of TRIM72/MG53 or delivery of rhMG53 can increase membrane repair.

This suggests that TRIM72/MG53 is interacting with conserved membrane repair

72 machinery present in many different cell types to accelerate the membrane repair process when it is necessary to improve membrane repair and allow for cell survival. It is also possible that there are other proteins that function in membrane repair in neurons that have yet to be discovered. Future studies of the neuron membrane repair process will help to determine if there are other proteins that can be linked to this process in neurons.

While our results and previous studies show that rhMG53 can increase membrane repair capacity and reduce the effects of injury, the long-term effects of treatment with this protein are unknown. Recent studies suggest that circulating levels of native

TRIM72/MG53 protein can affect insulin sensitivity and metabolic homeostasis 222.

These potential issues would be less of a concern for treating neural injuries where there is an acute injury that does not require prolonged application of the protein. In any case, it may prove to be more beneficial to target the endogenous membrane repair mechanism in neurons through different means. Our results, supported by other studies 25, also indicate that P188 can be effective at increasing neural membrane repair, so it may represent another potential therapeutic approach.

Our study focuses on the impact of increasing the membrane repair capacity in the nervous system. We used rhMG53 to increase the repair capacity in neurons and found that increasing membrane repair in a sciatic nerve crush model extended the length of regenerating neurons. This could address multiple unmet needs in treating diseases and injuries to the nervous system so future studies will address the efficacy of such approaches.

Materials and methods

Mouse models

C57Bl/6J were bred and maintained in standardized conditions at 22 ± 2C under a

12-hr/12-hr light cycle (lights on at 7 a.m. EST). Standard mouse chow and drinking water were provided ad libitum. All experimental procedures were approved by The Ohio

State University Institutional Animal Care and Use Committee. Animals were maintained in accordance with the recommendations of the NIH Guide for the Care and Use of

Laboratory Animals.

Western blotting

Tissues were collected from adult mice and were titrated in the presence of

Radioimmunoprecipitation Assay buffer (RIPA; Cell Signaling Technology, Danvers,

MA, USA) to extract protein for isolation. Protein concentrations were determined in accordance with the standard Bradford Assay using bovine serum albumin (BSA) standards. Protein samples (10 µg/lane skeletal muscle; 40µg/lane spinal cord, brain, and sciatic nerve; 5n g/lane rhMG53) were separated by SDS-PAGE at room temperature on

10% gels at 150 V and were transferred to nitrocellulose membranes (Bio-Rad, Hercules,

CA, USA). Blots were stained with Ponceau S (Boston Bioproducts, Ashland, MA, USA) to visualize total protein. Blots were probed for TRIM72/MG53 with a custom polyclonal antibody (Pacific Immunology, San Diego, CA, USA), and anti-rabbit horseradish peroxidase (HRP)–conjugated secondary antibodies (Cell Signaling Technology). The blots were developed using enhanced chemiluminescence (ECL) substrate (Bio-Rad). An

Azure Biosystems imager was used to visualize chemiluminescent blots per manufacturer’s directions.

Transient transfection

Neuro2a (American Type Culture Collection (ATCC), Manassas, VA, USA) cells were cultured in DMEM (SIGMA, St. Louis, MO, USA) containing 10% FBS (VWR

International, Radnor, PA, USA), 1x penicillin-streptomycin-glutamine (Life

Technologies, Carlsbad, CA, USA). ~1 x 106 cells were seeded onto 35mm culture dishes and allowed to adhere for ~12 hours. Cells were transfected with GFP, MG53-GFP, or

MG53-MBP (maltose-binding protein) and an empty GFP vector at a ratio of 3:1 using

Lipofectamine 3000(Life Technologies, Grand Island, NY). After 6-8 hours, the media containing transfection reagent was aspirated off the cells and replaced with DMEM supplemented with 10% FBS and 1x penicillin-streptomycin-glutamine. Transfection efficiency of ~60-70% was viewed using florescent microscopy, and laser injury was performed 2 days post transfection.

Membrane damage assay following laser injury

Neuro2a (American Type Culture Collection (ATCC), Manassas, VA, USA) cells were cultured in DMEM (SIGMA, St. Louis, MO, USA) containing 10% FBS (VWR

International, Radnor, PA, USA), 1x penicillin-streptomycin-glutamine (Life

Technologies, Carlsbad, CA, USA). Membrane damage was induced in Tyrode’s solution with 2.0 mM Ca2+, using the Olympus FV1000 multi-photon laser scanning confocal system. For laser injury measurements, injury was induced in the presence of 2.5 μM

FM4-64 fluorescent lipophilic dye (Life Technologies). rhMG53 protein was dissolved in saline solution and used at 1μM concentration. A circular area was selected along the edge of the cell membrane and irradiated at 20% laser power for 5 s. Pre- and post-

75 damage images were captured every 3 s, continuing for 57 s. The extent of membrane damage was analyzed using ImageJ software, by measuring the fluorescence intensity encompassing the site of damage. To preclude any potential for bias, all of the experiments were performed in a blinded fashion.

Mouse cervical, thoracic, and lumbar DRG neurons were dissected from CO2 asphyxiated adult mice and dissociated using Dispase ll (10mg/ml; SIGMA) and

Collagenase type 1 (500 g/ml; Worthington Biochemical Corp., Lakewood, NJ, USA) for three separate 45min incubations at 37°C. Next, DRG were triturated in 0.5 ml of

HBSS media, and centrifuged at 3000 rpm for 3 min. The neuron-enriched pellet was resuspended in 0.1 ml of primary neuronal growth media (LONZA, Basel, Switzerland).

Neurons were plated onto coverslips pre-coated with poly-L-lysine (0.1 mg/ml; SIGMA) and laminin (10 µg/ml; SIGMA). DRG neurons grew for 5 days at 37°C in a 5% CO2 humidified incubator. After 5 days the neurons were treated as described above for laser injury.

Membrane damage assays following glass bead rotation damage

Using aseptic techniques, 1x105 Neuro2a cells were plated into 2mL microcentrifuge (VWR International) tubes and incubated at 37º and 5% CO2 for 18 hours in

500µL complete medium. After the cells adhered to the bottom of the tube, the cells were washed with 200µL PBS. Following the initial wash, 200µL of 2mM Ca2+ Tyrode’s solution and 20µL of ≤106µm glass beads (~4.20 mg) (SIGMA) were added to the culture tubes. All tubes were sealed with parafilm and rotated 360º, 15 times, ~4 seconds per revolution, using a Fisher Scientific Hematology/Chemistry Mixer 346. Following

76 these rotations, 10µL of the supernatant from each sample was transferred to a 96 well plate with technical duplicates. Lactate dehydrogenase (LDH) levels in these samples were determined using a commercial colorimetric assay kit per manufacturer’s instructions (Takara, Japan; MK401). A set of tubes were lysed using 2% Triton X-100

(Sigma) to provide a maximum LDH release level that was used to normalize the data.

All experimental conditions were performed in triplicate. For a no damage control, glass beads were not added to the sample and the tubes were not inverted.

Immunostaining procedures

Mice were euthanized by use of CO2 asphyxiation followed by cervical dislocation, and spinal cord and tibialis anterior muscles were extracted. Tissue was then fixed with 10% phosphate buffered formalin, followed by a 24 hour incubation with 70% ethanol. Tissue was then embedded in paraffin (Thermo) and 12μm sections were mounted on SuperFrost Plus slides (Fisher Scientific, Hampton, NH, USA). Slides were deparaffinized through changes of xylene and rehydrated through decreasing concentration of ethanol. Antigen retrieval was then performed using Citra Plus Solution

(Biogenex, CA, USA). Slides were rinsed with PBS and incubated with a 2.5% BSA

(SIGMA) blocking solution for 1 h. Sections were then washed and incubated overnight with TRIM72/MG53 (custom rabbit polyclonal antibody generated Pacific Immunology)

192, dysferlin (Leica Biosystems, IL, USA), beta-III Tubulin (Novus Biologicals,

Littleton, CO, USA) and a 488-conjugated goat anti-rabbit IgG antibody, 568-conjugated goat anti-chicken IgG antibody , or 568-conjugated goat anti-mouse IgG antibody (Life

Technologies).

Mice were euthanized and perfused by transcardial perfusion with PBS followed by 4% paraformaldehyde in PBS. Nerves were post-fixed in paraformaldehyde for two hours, rinsed in PBS, and then stored in 30% sucrose in PBS at 4ºC. Sciatic nerves were embedded in optimal cutting temperature compound (OCT; VWR International) and frozen at -80°C; 12μm sections were cut using a cryostat and thaw-mounted on

SuperFrost Plus slides (Fisher Scientific), then stored at -20°C until use. After drying at

RT⁰, slides were rinsed with PBS and incubated with a 2.5% BSA (SIGMA) blocking solution for 1 h. Sections were then washed and incubated overnight with

TRIM72/MG53 (Pacific Immunology), SCG10 (Novus Biologicals), and a 488- conjugated goat anti-rabbit IgG antibody, or 568-conjugated goat anti-rabbit IgG antibody (Life Technologies).

Surgical nerve crush procedures

Four month old male C57Bl/6 mice were anesthetized (80 mg/kg ketamine; 10 mg/kg xylazine) then, following shaving and aseptic preparation of the right hind leg, a

Dumont #5 forceps was used to crush (10 s duration) the sciatic nerve. The site of nerve injury was marked by charcoal. 1µL of rhMG53 (1mg/mL) or saline vehicle control was injected into the epineurium, distal to the crush site. The 1mg/ml dose was selected to produce these therapeutic levels in the target tissue134 by accounting for dilution in the endoneurial fluid, leakage from the crush site, and quick diffusion at the crush site. The muscle/fascia layer was pulled together and the skin was sealed with a single wound clip.

Three days after crush injury, mice were perfused, and the sciatic nerves were harvested and prepared for histology as described above.

Statistical Analysis

Graphical representation and statistical analysis of data was performed using

GraphPad Prism version 8. All results are presented as mean ±SEM. Figure 14E and 15C:

Area under the curve values were generated from individual traces of dye influx over time. Data were analyzed by unpaired two-tailed t-test assuming unequal variances,

Welch's T-test for unbalanced designs. Figure 16B was analyzed using two-way ANOVA to test for comparisons of effect of distance and treatment.

Figure 13: MG53 is not expressed in the nervous system.

(A) Lysates from an adult wild type mouse were used for western blotting to detect levels of MG53. One fourth the amount of muscle lysate was used compared to nervous tissue lysate, and no band was detected in the nervous system lysate. Ponceau S stain was used to visualize total protein in lysates (B) Sections from wild type mice were analyzed for the presence of MG53. Cross sections of tibialis anterior muscle was positive for MG53 expression, and dysferlin was used as a counterstain. Longitudinal sections of spinal cord immunostained for MG53 show no MG53 expression, with beta III tubulin used as a counterstain.

Figure 14: MG53 increases membrane repair capacity in N2a cells

(A)A multiphoton microscope was used to injure transfected Neuro2a cells transfected with GFP or GFP tagged MG53. After injury, the GFP signal can be seen moving to the injury site in the GFP-MG53 cells, whereas no signal was detected in the GFP control cells (B). Graphical representation of change of fluorescence at the injury site over time 81 indicates that GFP-MG53 accumulates at the injury site over the span of 1 minute, while the GFP injured cells become photobleached and fail to recover their signal at the site of injury (n=8 per group). (C) In the presence of lipophilic dye (FM 4-64), Neuro2a cells transfected with GFP or co-transfected with GFP and MG53 –MBP were injured with a multiphoton microscope. Representative images show dye accumulating at the injury site and the change in fluorescence graph indicated that MG53 transfected cells show reduced dye intake when compared to GFP control. (D, E) Traces were analyzed using the area under the curve (AUC), and analyzed by unpaired two-tailed t-test assuming unequal variances, Welch's T-test for unbalanced designs. (n=12 per group, **=p>.001, p=0.0041 by t-test). Data represented as means ± SEM.

Figure 15: rhMG53 decreases membrane damage in vitro

(A-C) Dorsal root ganglion neurons were cultured for 5 days, and then incubated with a lipophilic dye (FM 4-64) and injured in the presence of 1µM rhMG53, 100µM P188, or control using a multiphoton microscope. (A) Representative images before and after the injury event visualize dye influx at the site of injury. (B, C) Traces marking the change in fluorescence over time graph indicated that rhMG53 and P188 treated cells show reduced dye intake when compared to control. Traces were analyzed using the area under the curve (AUC), and compared using a t-test (*p<0.01, ***p<0.001, n=8 per group, Tyrodes v rhMG53 p=0.0477, Tyrodes vs P188 p= 0.0008). (D) Rotation damage assay using

100µm glass beads to damage Neuro2a cells. rhMG53 treated cells have reduced LDH leakage, and thus greater resealing and mitigation of injury, (n= 12 cells per group

*p<0.01, ***p<0.001 by one-way ANOVA using Bonferroni’s multiple comparisons test) (Tyrodes v rhMG53 p=0.0004, Tyrodes v P188 p=0.0123). Data represented as 83 means ± SEM.

Figure 16: Treatment with rhMG53 significantly increases regeneration past the crush site.

A) Longitudinal sections of sciatic nerves from mice that received a crush injury and saline or rhMG53 were immunostained with SCG10. Treatment with rhMG53 increased

SCG10 intensity distal to the crush site by 3 dpi. Results are expressed as mean ± SEM.

Significant for treatment (***p=0.0006) and distance (**p=0.009) by two-way ANOVA.

B, C) Treatment with rhMG53 increased the regeneration index. Results are expressed as mean ± SEM, ***p=0.001, unpaired t test (n=8 per group p=0.0002). D) Longitudinal sections of sciatic nerves from mice that received a crush injury and saline or rhMG53 were immunostained with MG53. rhMG53 treated nerves were observed to stain positive for MG53 distal to the crush site where injection occurred.

Chapter 4 Targeting Neuronal Membrane Repair Through TRIM2 Modulation

This chapter has not been adapted from any currently published article and represents work conducted specifically by BJP. Subsequent expected publication of this work will list BJP as a first author.

Introduction

Plasma membrane repair is a highly conserved mechanism that appears in nearly every eukaryotic organism from single cell amoebas to most cell types in the human body, illustrating that repair of plasma membrane damage is an important aspect of normal cellular physiology184. While a simple lipid bilayer will reseal through thermodynamic principles, the presence of the cytoskeleton produces mechanical tension on the lipid bilayer thus preventing spontaneous resealing of disruptions185. Increased mechanical tension on the membrane can force disruptions to stay open and prevent resealing of the membrane186,187, while decreased membrane tension can lead to increased membrane disruption due to an unstable membrane. After injury, membrane tension has been shown to decrease. Cytoskeletal remodeling can increase tension and endo- and exocytosis have been reported to be preceded by a drop in membrane tension. This is relevant in the muscle disease Duchenne muscular dystrophy (DMD). DMD is caused by

87 a mutation in the dystrophin gene, which acts as a shock absorber between the internal cytoskeleton and the extracellular matrix. This leaves patients with muscles that are more prone to membrane tears and rely more heavily on membrane repair to keep muscle fibers alive. Also, mutations in the intermediate filaments keratin, and desmin alter the mechanical properties of the membrane.

Most cells in the human body reseal their cellular membranes following disruption of the membrane through trafficking of intracellular vesicles to the injury site where they fuse to form a membrane repair patch6-9,12,13. Disruption of these membrane repair processes can result in a number of diseases including muscular dystrophy,

Alzheimer's, and neurodegeneration36,197-199. Some molecular components of the membrane repair process have been identified, including the discovery that tripartite motif (TRIM) family proteins are involved in trafficking of vesicles during membrane repair in striated muscle39,129,136,137. Previous discoveries from our laboratory group demonstrated that mitsugumin 53 (MG53), a muscle-enriched tripartite motif (TRIM72) family protein42, is an essential component of the cell membrane repair machinery in skeletal and cardiac muscle128,129,135. Interestingly, when TRIM72/MG53 is delivered to non-muscle cell types, it can still function in a similar fashion to increase membrane repair in non-muscle cells. I demonstrated that membrane repair is present in neuronal cells and that it can be enhanced by using rhMG53 (Ch. 3). Previous work in our lab has led to the discovery that other members of the TRIM family of proteins are able to modulate the membrane repair process.

The TRIM family of proteins are multifunctional proteins with a conserved domain structure consisting of a RING finger with zinc-binding motifs and a coiled-coil domain and are divided into classes based on their C-terminal region. TRIM72 contains a

PRY-SPRY domain, and TRIM2, a screened modulator of membrane repair, contains an

NHL domain. TRIM2 shows high expression levels in the mouse cerebellum, the retina, and the hippocampus 223. TRIM2 is also expressed as 2 isoforms of similar length, with no observed difference between them. TRIM2 has been shown to bind to neurofilament light subunit (NFL) and regulates NFL ubiquitinylation via its RING-finger domain223.

Additionally, TRIM2 participates in neuronal plasticity, and axon initialization224.

TRIM2-GT mice have greater levels of the neurofilament light subunit in axons and resulting axonal swellings in the CNS, neurodegeneration with juvenile-onset tremor, and ataxia. CRISPR-Cas9 mice generated by deleting the ubiquitin ligase domain of TRIM2 showed similar pathology to the TRIM2-GT mice225. However, in contrast to the TRIM2-

GT mice, the ubiquitin ligase deficient TRIM2 mice generated using CRISPR-Cas9 did not have changes in NFL levels in the brain225.

In this study, we address the ability of TRIM2 to contribute to the membrane repair process in neurons though its interaction with NFL. We show that overexpression of TRIM2 increases the membrane repair response in Neuro2a (N2a) cells, while knockdown with shRNA impairs the membrane repair response. We also discovered that when both proteins were overexpressed, TRIM2 was shown to interact with NFL.

However, TRIM2 did not interact with NFL in spinal cord lysates. Lastly, knockout NFL mice did not show a difference in membrane repair when compared to WT littermates.

Results

TRIM2 contributes to membrane repair in vitro

Based on previous work, we established that neurons have a robust membrane repair response, and that this response can be elevated by the use of rhMG53 (Ch3). We further sought to examine the native response to membrane injury. To examine this process, we transfected N2a cells with TRIM2-GFP, RNAi targeting endogenous TRIM2, and control plasmids. Transfected N2a cells were incubated with the lipophilic dye FM4-

64 and injured using a multiphoton infrared laser injury monitored by confocal microscopy. When multiphoton confocal laser microscopy is used to injure the membrane, the FM4-64 dye enters the cell and the fluorescent signal increases until the membrane reseals. Less dye influx corresponds to more efficient membrane repair in the cell. Images were analyzed pre- and post-injury, and results were quantified over time.

We find that overexpression of TRIM2 provides protective effects in N2a cells, while the knockdown of expression decreases membrane repair (Fig. 17A). Post-injury GFP tagged

TRIM2 was visualized moving from its cytoskeletal localization to a more punctate localization (Fig. 17B). These results indicate that TRIM2 is essential to the membrane repair process and is able to respond post-injury.

TRIM2 does not bind to neurofilament light chain.

TRIM2 has previously been shown to interact with and ubiquitinate neurofilament light chain of the neurofilament heteropolymer223. Previous to this work, a screen of

TRIM protein localization in U-2 OS cells revealed TRIM2 localizing to cytoplasmic filaments, but the authors concluded that it did not co-localize to vimentin, β-tubulin or

90 actin226. To confirm the TRIM2-NFL interaction, we co-transfected N2a cells with three plasmids: GFP, NFL, and NFM, or GFP-TRIM2, NFL, and NFM. These lysates along with spinal cord lysates were probed for either TRIM2 or NFM. TRIM2 immunoprecipitation was able to pull down the neurofilament complex in TRIM2 transfected cells and spinal cord lysate. However, TRIM2 was not detected in the spinal cord pull down (Fig. 15A). NFM co-immunoprecipitation was able to pull down the neurofilament complex in all three lysates. TRIM2 was not detected in the spinal cord pull down (Fig. 18B). To visualize the interaction, TRIM2 was co-transfected into SW13

(vim-) and N2a cells with NFM, TRIM2-GFP, and NFL-mCherry plasmids show no colocalization with TRIM2 and NFL (Fig. 18C). These results indicate that if TRIM2 is being overexpressed in cells, there may be an interaction between TRIM2 and NFL, but there was not a detectable level of interaction in the spinal cord.

Lack of neurofilament light chain does not affect membrane resealing in NFL KO neurons.

Based on previous literature, TRIM2 and NFL should interact223. However,

TRIM2 was not shown to interact with NFL in vivo, and only interacted when both proteins were co-expressed (Fig. 18A). To test if NFL is involved in membrane repair, we cultured primary dorsal root ganglion (DRG) neurons for 5 days and then exposed them to a fluorescent lipophilic membrane impenetrable dye (FM4-64) before injury by a

250nm diameter IR laser pulse. We saw no difference in the membrane repair response between NFL KO and WT littermate DRG neurons (Fig. 19 ). These results indicate that the presence of NFL does not influence the membrane repair process.

Discussion

Our study addresses TRIM2’s membrane repair function through interaction with the neurofilament protein NFL. We demonstrate that overexpression of TRIM2 leads to an increased membrane repair response when compared to GFP transfected control cells.

We also show that knockdown of TRIM2 leads to a decreased membrane repair response.

Although the majority of work done with TRIM2 focuses on the interaction of TRIM2 and NFL, the initial result from the Balastik et al. paper has not been repeated. In an attempt to replicate their results, TRIM2 and NFL were co-expressed along with neurofilament medium chain (NFM), to form neurofilament polymers, and TRIM2 or

NFM was immunoprecipitated from the lysates. Similar to Balastik et al., overexpression yielded TRIM2 and NFL interaction, however spinal cord lysates did not reveal an interaction between the proteins. To further investigate whether NFL is involved in the mechanism of membrane repair, NFL KO and WT littermate DRG neurons were cultured for 5 days and subjected to laser injury. We observed no difference between the neurons of either genotype.

Membrane repair is a field that is just beginning to draw the focus of neuroscientists. Recently, membrane disruption has been studied in multiple sclerosis227,

Alzheimer's199,228, and injury to the central26,229 and peripheral nervous system (Ch3).

Finding a mechanism to increase the repair capacity of neurons could provide an attractive therapeutic target for these diseases. TRIM2 is expressed in neurons and has been shown to be able to increase neuron survival after injury230,231.

The interaction between TRIM2 and NFL has been the subject of many studies, however few have actually performed the experiment that looks at their interaction. We examined this interaction, and found that when TRIM2 and NFL were overexpressed an interaction occurred, but this interaction was absent in spinal cord lysates. Interestingly,

GFP tagged TRIM2 appear to bind the cytoskeleton of the N2a cells, but with co- transfection of the NFL containing polymers the two proteins were not seen to co- localize. These results could indicate that TRIM2 could bind to NFL while it is in its non- filamentous form, and the neurofilament polymer obscures TRIM2’s NFL binding site.

Another caveat of the experiment could be the addition of mCherry to NFL and GFP to

TRIM2. Although these proteins help with visualization of the interaction, they may interfere with the protein-protein interaction, and their molecular weights may appear similar on a western. To rectify this issue, immunoprecipitation should be performed without the fluorescent tags. There is still the issue of the lack of interaction in the spinal cord. There is also an issue that can be seen in the 20% input lanes of the western where insufficient stripping can lead to bands remaining present after probing for a second antibody. There was an attempt to avoid this issue by using antibodies raised in different animal species, however the best course of action would be to use fluorescent secondaries and different animal species.

A possible explanation for differences in the results obtained here and those in the literature could have to do with the two isoforms of TRIM2 expressed. The antibody used should recognize both isoforms based on the data sheet, but in the blots it appears to preferentially target the lower molecular weight isoform. The antibody is able to target

93 the higher molecular weight isoform when overexpressed, but the preference of one isoform over the other could potentially explain the capricious results observed with the

TRIM2 NFL interaction. Additionally, one study looking into the rare TRIM2 mutations in humans indicated accumulation of neurofilament in a sural nerve biopsy232. These results were consistent with the TRIM2 gene trap mouse223. However, mice generated by knocking out the RNG domain of TRIM2 did not show similar results and, in fact showed normal levels of neurofilament. Future studies should investigate the difference between the two TRIM2 isoforms because the different results from human disease and mouse experimentation could be explained by alternate function of the isoforms.

Our study focuses on a possible mechanism of membrane repair in neurons. We show that TRIM2 overexpression and knockdown alters the membrane repair response.

We also show that the interaction between TRIM2 and NFL may be artificially forced by overexpression but may not be strong in vitro. Although the initial hypothesis of

TRIM2’s membrane repair mechanism proved to be false, greater study of the cytoskeletal protein TRIM2 binds to will likely lead to promising therapeutic results.

Materials and Methods

Mouse models

Mice were bred and maintained in standardized conditions at 22 ± 2C under a 12- hr/12-hr light cycle (lights on at 7 a.m. EST). All animals had ad libitum access to standard mouse chow and drinking water. All experimental procedures were approved by

The Ohio State University Institutional Animal Care and Use Committee. Animals were maintained in accordance with the recommendations of the NIH Guide for the Care and

Use of Laboratory Animals. NFL knock out mice were provided by Dr. Anthony Brown, and were previously described233.

Transient transfection

Transient transfection was performed as described in Ch. 3. Briefly, N2a cells were transfected with GFP, TRIM2-GFP, or shRNA targeting a knockdown of TRIM2 using Lipofectamine 3000(Life Technologies, Grand Island, NY). After 6-8 hours, the media containing transfection reagent was aspirated off the cells and replaced with

DMEM supplemented with 10% FBS and 1x penicillin-streptomycin-glutamine.

Transfection efficiency of ~60-70% was viewed using fluorescent microscopy, and laser injury was performed 4 days post transfection.

For immunoprecipitation and microscopy, N2a cells were transfected with GFP

NFL-mCherry, and NFM, or TRIM2-GFP NFL-mCherry, and NFM, using

Lipofectamine 3000(Life Technologies, Grand Island, NY) at a ratio of 3:1:1. After 6-8 hours, the media containing transfection reagent was aspirated off the cells and replaced with DMEM supplemented with 10% FBS and 1x penicillin-streptomycin-glutamine.

Transfection efficiency of ~60-70% was viewed using fluorescent microscopy, and cells were lysed or fixed with 4% paraformaldehyde 2 days post transfection.

Laser Injury Assay

Laser injury was performed as described in Ch 3.

Immunoprecipitation

Spinal cord lysate was surgically removed from 6-month-old mice and homogenized in Tris-Triton buffer (50mM Tris pH 8, 150mM NaCl, 1% Triton X-100,

2mM EGTA)234. GFP NFL-mCherry, and NFM, or TRIM2-GFP NFL-mCherry, and

NFM transfected N2a cells were also lysed in the Tris-Triton Buffer. Protein concentrations were determined using a standard Bradford assay using BSA as standards.

20 g of protein lysate was saved as a 20% input control. Equal volume of 100 g of lysate was added to microcentrifuge tubes, and pre-cleared with 10 l Dynabeads Protein

G magnetic beads (Invitrogen) for 1 hour at 4C. Beads were then pelleted, and the supernatant was moved to a new tube. Lysates were then incubated with either NFM, or

TRIM2 primary antibodies overnight at 4C with rotation. The next morning 20 l of beads were added to the tubes and incubated for four hours at 4C with rotation. Beads were then pelleted and washed five times with PBS. Samples were then eluted from the beads by boiling at 90C for 5 min in laemmli buffer containing β-mercaptoethanol (Bio-

Rad).

Precipitate samples were separated by SDS-PAGE at room temperature on 10% gel (Bio-Rad, Hercules, CA, USA) and were transferred to nitrocellulose membranes

(Bio-Rad, Hercules, CA, USA). Blots were probed for TRIM2, NFM, and NFL and anti- mouse horseradish peroxidase (HRP)–conjugated secondary antibodies, anti-goat HRP– conjugated secondary antibodies (Cell Signaling Technology), and anti-rabbit HRP– conjugated secondary antibodies (Cell Signaling Technology). The blots were developed using enhanced chemiluminescence (ECL) substrate (Bio-Rad). An Azure Biosystems imager was used to visualize chemiluminescent blots. Blots were stripped of antibodies following probing using harsh stripping buffer (100 mM 2-Mercaptoethanol, 2% SDS,

62.5 mM Tris-HCl pH 6.8) then washed four times before use for additional western blotting.

Statistical Analysis

Graphical representation and statistical analysis of data was performed using

GraphPad Prism version 8. All results are presented as mean ±SEM. Data were analyzed by unpaired two-tailed t-test assuming unequal variances, Welch's T-test for unbalanced designs. Figure 17B and 19 were analyzed with a one-way ANOVA using Bonferroni’s multiple comparisons test.

Figure 17: TRIM2 is essential for membrane resealing in vitro.

Cultured N2a cells were transfected with plasmids containing GFP, TRIM2, or shRNA targeting TRIM2. 60hrs post-transfection, the cells were exposed to a fluorescent lipophilic membrane impenetrable dye (FM4-64) before injury by a 250nm diameter IR laser pulse. (A) Representative images of neurons pre- and post-injury.(B) Area under the curve (AUC) generated from change in fluorescence at injury site from 0-50 sec. One- way ANOVA, Bonferronis post hoc test, GFP v TRIM2 p=0.0393, GFP v shRNA- 98

TRIM2 p=0.0209. GFP, TRIM2, and shRNA n=11; TRIM2 0 Ca n=10.(C)

Representative western blot showing consistent knock down of TRIM2 expression. (D)

Cultured N2a cells were transfected with GFP tagged TRIM2 and injured with a 250nm

IR laser pulse. GFP tagged TRIM2 localizes to the cytosol and cytoskeletal filaments in

N2a cells before injury. 10 minutes post-injury, TRIM2 localizes to cytoskeletal filaments and shows punctate distribution. Data represented as means ± SEM.

Figure 18: TRIM2 does not bind to neurofilament light chain.

Lysates from Neuro2a neuroblasts co-transfected with GFP, NFL, and NFM, or GFP-

TRIM2, NFL, and NFM, or spinal cord lysates were immunoprecipitated with (A) NFM or (B) TRIM2 antibodies. (A) NFM co-immunoprecipitation was able to pull down the neurofilament complex in all three lysates. TRIM2 was only detected in the overexpressed lysate pull down. (B)TRIM2 co-immunoprecipitation was able to pull

100 down the neurofilament complex only in the spinal cord lysate, but TRIM2 was not detected in the spinal cord pull down. (C) TRIM2 co-transfected into SW13 (vim-) and

Neuro2a cells with NFM, TRIM2-GFP, and NFL-mCherry plasmids show no colocalization with TRIM2 and NFL. (Molecular Weights: NFM=~150kDa,

TRIM2=~80kDa, NFL=~68kDA, GFP-TRIM2= ~107kDa, NFL-mCherry=~95kDa)

101

Figure 19: Lack of neurofilament light chain does not affect membrane resealing in NFL

KO neurons.

Primary dorsal root ganglion (DRG) were dissected and dissociated to a single cell suspension. Cells were plated on coverslips and grown for 5 days. On day 5, cells were exposed to a fluorescent lipophilic membrane impenetrable dye (FM4-64) before injury by a 250nm diameter IR laser pulse. (A) Representative images of neurons pre- and post- injury. Arrow indicates injury location. (B) Graph represents change in fluorescence normalized to baseline over time. Area under the curve (AUC) generated from change in fluorescence at injury site from 0-55sec.Statistical analysis performed by Student’s t test

NFL KO (n=15) vs WT (n=14) p= 0.6112. Data represented as mean ± SEM

102

Chapter 5 Summary, Significance, and Future Work

The overarching goal of these studies was to increase the knowledge of the membrane repair process and the importance of this cellular response to compensate for injury and disease. These findings should help to determine the potential value for targeting membrane repair as a treatment for various pathologies in different tissues throughout the body. One way to consider the value of membrane repair as a native protective mechanism and as a potential therapeutic target is to consider why the membrane repair mechanism is present in nearly every cell type of the body, but can have some variation in how extensively membrane repair occurs in the cells of different tissue235,236. It is important for all cells in the body to have some capacity for membrane repair due the higher energetic cost of regenerating new cells as opposed to ensuring the survival of the current population of cells. The balance between these costs varies from one cell type to another where large, structurally complex cells, such as muscle fibers or neurons, require a more robust cell membrane repair response to avoid the loss of these costly cells in relevant tissues. This increased need for cell membrane repair in such cells is further required by the lower regenerative capacity of the cells. While skeletal muscle fibers can be effectively regenerated through the action of satellite cells, cardiomyocytes and most neurons have a relatively limited, or non-existent, regenerative capacity within 103 the tissue. In contrast, tissues with smaller, less complicated cells, such as the skin or the liver have a robust regenerative capacity. One could hypothesize that the cell membrane repair response is not as vital because damaged cells are more easily replaced through regeneration of the injured tissue. Comparison of the repair response in multiple cell types in this work provided an opportunity to both better understand the membrane repair systems in these cells and how these could be targeted as therapeutic interventions.

In the context of these studies, I examined the membrane repair response in two general cell types: skeletal muscle fibers and neurons. These provided a good model system to better understand the role of membrane repair in different cell types. While skeletal muscle fibers are the most commonly studied in cell type dependent membrane repair systems, much less is known about the membrane repair response in neurons.

Membrane repair in the context of muscular dystrophies

In general, previous efforts at studying membrane repair occurred in tissues most prone to injury such as skin, muscle, and gastrointestinal tissue. These efforts led to the identification of many proteins involved in the membrane repair process. Additionally, diseases like Duchenne (DMD) and limb girdle muscular dystrophies (LGMD), have provided clues as to the consequences of altered membrane repair. These diseases have made it clear that membrane repair is an important process in preventing pathology in skeletal muscle, however the role of specific proteins in mediating membrane repair in specific disease contexts have not been fully elucidated. One approach to test the impact of specific proteins is to cross a knockout mouse for a specific gene that contributes to membrane repair into a mouse model of the disease of interest. Another approach

104 assesses the impact of a membrane repair protein on a stressor that increases the extent of membrane injury and makes the tissue more reliant on membrane repair to avoid loss of cells. The specific methodology used is dependent on the particular protein of interest and the disease that is being studied.

While there have been previous studies on the function of MG53 in membrane repair, there are several outstanding questions that remain. One of these questions was how native MG53 functions within the context of DMD. To study this question, I chose to cross the knockout mouse with a DMD mouse model, known as the mdx mouse. The mdx mouse is susceptible to membrane injury due to the lack of the stabilizing protein dystrophin. While the muscles of the mdx and MG53/mdx DKO mice were histologically similar at the early timepoint of 6 weeks, the EDL muscles showed significant weakness, and FDB muscles showed compromised resealing when subjected to the laser injury assay. The histological results from the DKO mice were interesting and seemed to raise the issue of whether membrane repair played a crucial role in DMD. Fiber cross sectional area and percentage of centrally located nuclei can be used as markers for damage and regeneration of muscle fibers. Damage to the myofibers leads to greater satellite cell recruitment and increased centrally located nuclei counts. If fibers were experiencing more damage due to the absence of effective membrane repair, it is expected that the centrally located fiber count would be increased, however that result was not seen at the 6 week timepoint. The DMD field has questioned if the disease is an issue of membrane integrity or of dysfunctional membrane repair. The results obtained using the laser injury assay certainly point to the DKO mice having a cumulative effect of the loss of both

105 proteins, meaning the issue of maintaining the permeability of the membrane can be made worse by creating a defect in membrane repair. In fact, this initial bout of injury at

6 weeks may be so damaging to the muscle that the defect in membrane repair does not affect the pathology in a significant way. We also saw an upregulation in the membrane repair protein caveolin-3, so there is certainly also a membrane repair compensation occurring. Cav-3 is responsible for vesicular trafficking and mutation in the protein causes LGMD1C. These mutations can cause depletion of dysferlin from the plasma membrane and improper trafficking of MG53 and have also been linked to defective membrane repair. Cav-3 compensation may be able to offset the loss of MG53 by increasing the ability of the muscle to repair itself. However, after the mice were aged, they developed severe skeletal and cardiac fibrosis.

Previous researchers crossed the mdx mouse with the dysferlin knockout mouse.

This creates a similar defect to the mouse we created. The authors of the paper did not look at the 6-week timepoint. The authors did look at a 6-month timepoint, which is the timepoint that pathology begins in dysferlin knockout mice. 6-month old dysferlin/mdx

DKO mice exhibit necrotic muscle, greater central nuclei counts, and fibrotic deposition.

In both the dysferlin and mdx DKO mice, pathology is observed over time indicating that a deficiency in membrane repair manifests into pathology over time. One can envision a pathologic mechanism where the compromised membrane repair leads to an increase in the number of muscle fibers that succumb to cell membrane injuries during the progression of muscular dystrophy. The accumulation of the increased rate of muscle

106 fiber death would lead to the acceleration of DMD pathology observed when membrane repair is disrupted by knockout of either MG53 or dysferlin.

MG53 and dysferlin are able to interact to form a complex39, so it would be expected that knockout of either protein would yield similar results. However, skeletal muscle pathology presents between 6-8 months in the dysferlin null mouse36, while pathology doesn’t appear until approximately 12 months in the MG53 knockout mouse129. These results are similar to those seen in the mdx double knockout mice for both genotypes179. These results lead to the question of whether one protein is more important to membrane repair, or is this disparity due to the normal function of the proteins? Dysferlin has been extensively studied and has been linked to essential processes like transverse tubule stabilization and satellite fusion112,180. These two functions are essential to both normal physiology and chronic stress events. A proposed future study can create a MG53/dysferlin DKO mouse and a MG53/dysferlin/mdx triple knockout mouse to observe if deletion of both membrane repair proteins creates a summation in pathology. These experiments will create useful insights into how the function of either of these proteins perturbs the pathology in the MG53/mdx or dysferlin/mdx mouse skeletal muscle.

Much of this work focused on the effects of muscular dystrophy on skeletal muscle pathology, however cardiomyopathy is an important component in the progression of the DMD and a major contributor to mortality in DMD patients237-239.

Another interesting finding of the study is the pathology observed in the heart of the

DKO animals. Normal mdx mice exhibit cardiomyopathy as they begin to age past one

107 year old, and MG53KO hearts are more susceptible to cardiac stress events240,241. Aged

DKO mice develop approximately double the cardiac fibrosis than the mdx mouse. Of all the muscles tested for fibrotic tissue, the heart showed the most significant pathology.

Future studies should investigate functional changes in the heart to determine the time points at which cardiac dysfunction occurs. A valuable experiment would be the use of echocardiograms to measure the heart function of these animals as they age. An echocardiogram is a nonlethal measurement that can measure how much blood the heart is pumping with each beat, ejection fraction, and fractional shortening, a measure of left ventricular contractility which can be used to assess heart function.

Membrane repair in the context of the nervous system

Although skeletal and cardiac striated muscles are subject to naturally occurring stress events from stretching and contraction of muscles, most other tissues are exposed to mechanical stress less often and presumably would rely less on membrane repair as they are damaged less frequently. However, peripheral nerves are interconnected with the skeletal muscle system and thus are subjected to stretch, compression, or transection in the course of normal physiology and to a much greater extent following traumatic injuries to the body. I hypothesize that such events produce injuries to the cell membrane of neurons that require membrane repair to compensate for these cellular injuries, maintain the integrity of the cell membrane, assure effective conduction by the nerves, and prevent the death of individual neurons. Preventing the loss of neurons would not only help maintain the function of the nervous system it would also reduce the energetic expense of regenerating an axon.

108

Membrane repair has not been studied extensively in the mammalian nervous system, but invertebrate studies revealed that injuries to the axon constrict both sides but do not collapse due to tightly packed vesicles at the transection sites93,242. These experiments involved unmyelinated axons, and although the mammalian nervous system contains some unmyelinated axons, the majority have some form of myelin around their axon. The myelin provides an extra membrane to the axon and adds an interesting dynamic when looking at axonal damage in the mammalian nervous system. It also leads to the questions of, if neuronal membrane repair can be targeted and can function improve after injury?

To address the value of membrane repair in neurons, we wanted to answer several key questions. First, is there an active membrane repair response in mammalian neurons?

Are key membrane repair proteins, such as MG53, expressed in neurons? Can we increase the membrane repair response in neurons using methods that have increased membrane repair in other cell types? Does increasing membrane repair minimize the impact of traumatic injuries to neurons?

For these studies we used purified recombinant human MG53 to elevate membrane repair in the target neuronal origin cells. We first showed that injuries to the cell body are able to be repaired and that with rhMG53 treatment, repair can be enhanced.

Interestingly, overexpression of MG53 in the cells also conferred protection from injury.

MG53 expression is primarily limited to skeletal and cardiac muscle, however similar to what we have shown in neuronal cells, the protein is effective at enhancing membrane repair in many different cell types. In addition to examining the effects of membrane

109 repair on the cell body, we examined rhMG53’s effect on axons in the sciatic nerve injury model. Sciatic nerves were crushed and treated with saline or rhMG53, and 3 days post- injury we observed robust regeneration in the treatment group. The rate of regeneration is crucial for reinnervation of the target tissue and speeding recovery post-injury. The results of this study suggest that enhancing membrane repair could be a viable therapeutic avenue to explore.

The rhMG53 study revealed that membrane repair is functional in neurons and axons, and can be enhanced to increase regeneration. However, it is not clear how to best target the process. Previous literature showed that rhMG53 can cross the blood brain barrier in response to an ischemic episode26. This could be a viable therapeutic option for treating injuries to the central nervous system, but it is not clear if rhMG53 is able to cross the blood brain barrier in the absence of severe trauma and thus, may severely limit treatment options. Future studies need to investigate targetable proteins to enhance membrane repair. One such study could investigate membrane repair proteins that are present in muscle and neuronal tissue. Dysferlin is not exclusively expressed in skeletal muscle243-245 and one study has linked the protein to neuritic plaques in Alzheimer’s brains246. Investigation of dysferlin in the nervous system is severely lacking, and investigating if it is able to function in the brain similarly to muscle could be therapeutically beneficial. Another aspect that will be explored in the following sections is investigating other members of the TRIM family for similar function and homology to

MG53.

TRIM proteins as mediators of membrane repair

110

The studies presented here and previous studies from multiple investigators link the tripartite motif protein MG53 with membrane repair. The tripartite motif family of proteins is a large family of E3 ubiquitin ligases with a high degree of homology between many proteins in the family. Although not every member of the family is studied extensively, the presence of the canonical TRIM motif in all of these proteins means they have the potential to function as E3 ubiquitin ligases. It is not clear if the E3 ubiquitin ligase activity of any TRIM protein is crucial to membrane repair or if other aspects of the protein are the key contributors to membrane repair. Previous work shows that deletion of the E3 ligase domain of MG53 leads to cytosolic localization of MG53, while deleting the SPRY domain, which is responsible for protein interaction specificity, leads to punctate localization42. Both deletion constructs lead to mislocalization, which leads to the conclusion that although it is not known if ubiquitination is important in membrane repair, the loss of some of the protein structure has the potential to alter the ability of

MG53 to localize to the site of membrane injuries.

Previously unpublished work from our laboratory group has elucidated that several members of the family are able to modulate membrane integrity in response to injury, and that many of the proteins are expressed throughout various tissues besides striated muscle. TRIM proteins in general are expressed in many different tissue types so they are attractive targets to study membrane repair. Since MG53 is predominantly expressed in striated muscles, other TRIM family proteins could function in a similar capacity in other cell types. Previous studies from our group showed that MG53 could function in membrane repair even when expressed in many different cell types,

111 suggesting that there could be other TRIM proteins that function similarly in other cell types. Although our laboratory has identified these potential TRIM family proteins that could contribute to membrane repair in some cells, MG53 remains the most widely studied membrane repair protein of the family. Current studies in our laboratory are underway to identify the specific regions in the protein that confer membrane repair activity. This work will help identify homologous regions in other members of the large family that may be responsible for membrane repair.

TRIM2 as a mediator of membrane integrity in the nervous system

Although targeting membrane repair could be helpful to treat many different diseases, the nervous system is particularly interesting because neurons are terminally differentiated, and membrane repair could keep them alive following injury. As expected with a large family of proteins, many TRIM proteins are expressed in the nervous system223,247,248, however one member shares many similarities to MG53. TRIM2 is expressed exclusively in neural tissues and has been shown to interact with motor proteins248 and has been screened as a protein of interest that is able to bind to acidic phospholipids249. Interacting with acidic phospholipids, especially phosphatidylserine, appears to be an important aspect of membrane repair because several prominent membrane repair proteins are able to interact with them129,250,251. Additionally, TRIM2 was previously shown to interact with the cytoskeletal protein neurofilament light chain223. Moreover, mutations in TRIM2 have been linked to the development of a recessive form of Charcot-Marie-Tooth disease named CMT2R232,252. Only two cases of

112 this disease have been reported, one 18-year old and an infant that died at 3 years old232,252. These similarities to MG53 led us to hypothesize that TRIM2 could potentially function in membrane repair in neurons as MG53 functions in striated muscle.

To examine TRIM2’s role in membrane repair, we subjected overexpressing or knockdown TRIM2 transfected cells to a laser injury assay. We observed that TRIM2 overexpression aided membrane repair while knock down impaired the repair process.

Interestingly, GFP tagged TRIM2 was observed as having predominantly filamentous localization. TRIM2 is not unique in its ability to bind to cytoskeletal proteins, members of the C-I subfamily of TRIM proteins are able to interact with microtubules253. The filamentous localization of TRIM2 was considered to be most likely with neurofilament

(NFL), however NFL and TRIM2 were not observed to co-localize when tagged protein was overexpressed and immunoprecipitation did not yield significant interaction between the two proteins. Furthermore, knockout of NFL did not affect membrane repair when tested in the laser injury assay. Although TRIM2 does not appear to regulate membrane repair through effects on NFL, it does appear that TRIM2 can modulate membrane repair in neural origin cells.

There are several future studies that could determine the mechanisms that mediate

TRIM2 effects on membrane repair. One approach is to identify additional binding partners of TRIM2 in neurons. This would involve pulling down TRIM2 from lysates and analyzing the pull down with mass spectrometry. Additionally, knockout cells can be generated using CRISPR-Cas9 and the cells could be screened for abnormalities in the cytoskeletal architecture.

113

Conclusion

In these studies, we examined membrane repair in the skeletal and cardiac muscle along with neurons. These studies were able to increase the knowledge of the membrane repair in DMD and emphasize the importance of membrane repair in the nervous system.

These findings suggest that robust membrane repair is observed in different cell types and that neurons are a targetable strategy to elevate membrane repair. The ability to highlight membrane repair will provide an important therapeutic tool to both keep cells alive and maintain normal tissue function.

114

Bibliography

1 Parsegian, V. A., Rand, R. P. & Gingell, D. Lessons for the study of membrane fusion from membrane interactions in phospholipid systems. Ciba Found Symp 103, 9-27 (1984). 2 Raucher, D. & Sheetz, M. P. Characteristics of a membrane reservoir buffering membrane tension. Biophys J 77, 1992-2002, doi:10.1016/s0006-3495(99)77040- 2 (1999). 3 Chernomordik, L. V., Melikyan, G. B. & Chizmadzhev, Y. A. Biomembrane fusion: a new concept derived from model studies using two interacting planar lipid bilayers. Biochim Biophys Acta 906, 309-352 (1987). 4 Heilbrunn, L. THE DYNAMICS OF LIVING PROTOPLASM. Vol. 41 (Academic Press Inc, 1956). 5 Chambers, R. & Chambers, E. L. Explorations into the Nature of the Living Cell. Academic Medicine 36, 966 (1961). 6 McNeil, P. L. & Ito, S. Gastrointestinal cell plasma membrane wounding and resealing in vivo. Gastroenterology 96, 1238-1248 (1989). 7 McNeil, P. L. & Khakee, R. Disruptions of muscle fiber plasma membranes. Role in exercise-induced damage. Am J Pathol 140, 1097-1109 (1992). 8 Steinhardt, R. A., Bi, G. & Alderton, J. M. Cell membrane resealing by a vesicular mechanism similar to neurotransmitter release. Science 263, 390-393 (1994). 9 Bi, G. Q., Alderton, J. M. & Steinhardt, R. A. Calcium-regulated exocytosis is required for cell membrane resealing. J Cell Biol 131, 1747-1758 (1995). 10 Terasaki, M., Miyake, K. & McNeil, P. L. Large plasma membrane disruptions are rapidly resealed by Ca2+-dependent vesicle-vesicle fusion events. J Cell Biol 139, 63-74 (1997). 11 Demonbreun, A. R. et al. An actin-dependent annexin complex mediates plasma membrane repair in muscle. J Cell Biol 213, 705-718, doi:10.1083/jcb.201512022 (2016). 12 Miyake, K. & McNeil, P. L. Vesicle accumulation and exocytosis at sites of plasma membrane disruption. J Cell Biol 131, 1737-1745 (1995). 13 McNeil, P. L. & Ito, S. Molecular traffic through plasma membrane disruptions of cells in vivo. J Cell Sci 96 ( Pt 3), 549-556 (1990). 14 Reddy, A., Caler, E. V. & Andrews, N. W. Plasma membrane repair is mediated by Ca(2+)-regulated exocytosis of lysosomes. Cell 106, 157-169 (2001). 115

15 Lek, A. et al. Calpains, cleaved mini-dysferlinC72, and L-type channels underpin calcium-dependent muscle membrane repair. J Neurosci 33, 5085-5094, doi:10.1523/jneurosci.3560-12.2013 (2013). 16 Defour, A. et al. Dysferlin regulates cell membrane repair by facilitating injury- triggered acid sphingomyelinase secretion. Cell Death Dis 5, e1306, doi:10.1038/cddis.2014.272 (2014). 17 Corrotte, M. et al. Caveolae internalization repairs wounded cells and muscle fibers. Elife 2, e00926, doi:10.7554/eLife.00926 (2013). 18 Tam, C. et al. Exocytosis of acid sphingomyelinase by wounded cells promotes endocytosis and plasma membrane repair. J Cell Biol 189, 1027-1038, doi:10.1083/jcb.201003053 (2010). 19 Draeger, A. & Babiychuk, E. B. Ceramide in plasma membrane repair. Handb Exp Pharmacol, 341-353, doi:10.1007/978-3-7091-1511-4_17 (2013). 20 Bement, W. M., Forscher, P. & Mooseker, M. S. A novel cytoskeletal structure involved in purse string wound closure and cell polarity maintenance. J Cell Biol 121, 565-578 (1993). 21 Keyel, P. A. et al. Streptolysin O clearance through sequestration into blebs that bud passively from the plasma membrane. J Cell Sci 124, 2414-2423, doi:10.1242/jcs.076182 (2011). 22 Babiychuk, E. B., Monastyrskaya, K., Potez, S. & Draeger, A. Intracellular Ca(2+) operates a switch between repair and lysis of streptolysin O-perforated cells. Cell Death Differ 16, 1126-1134, doi:10.1038/cdd.2009.30 (2009). 23 Jimenez, A. J. et al. ESCRT machinery is required for plasma membrane repair. Science 343, 1247136, doi:10.1126/science.1247136 (2014). 24 Idone, V. et al. Repair of injured plasma membrane by rapid Ca2+-dependent endocytosis. J Cell Biol 180, 905-914, doi:10.1083/jcb.200708010 (2008). 25 Gu, J. H. et al. Poloxamer 188 protects neurons against ischemia/reperfusion injury through preserving integrity of cell membranes and blood brain barrier. PLoS One 8, e61641, doi:10.1371/journal.pone.0061641 (2013). 26 Yao, Y. et al. MG53 permeates through blood-brain barrier to protect ischemic brain injury. Oncotarget 7, 22474-22485, doi:10.18632/oncotarget.7965 (2016). 27 Liu, J. et al. Cardioprotection of recombinant human MG53 protein in a porcine model of ischemia and reperfusion injury. J Mol Cell Cardiol 80, 10-19, doi:10.1016/j.yjmcc.2014.12.010 (2015). 28 Zhu, H. et al. Amelioration of ischemia-reperfusion-induced muscle injury by the recombinant human MG53 protein. Muscle Nerve 52, 852-858, doi:10.1002/mus.24619 (2015). 29 Murphy, E., Cross, H., & Steenbergen, C. Sodium Regulation During Ischemia Versus Reperfusion and Its Role in Injury. Circulation research 84, doi:10.1161/01.res.84.12.1469 (1999). 30 Murphy, E., Cross, H., & Steenbergen, C. Is Na/Ca Exchange During Ischemia and Reperfusion Beneficial or Detrimental? Annals of the New York Academy of Sciences 976, doi:10.1111/j.1749-6632.2002.tb04771.x (2002).

116

31 Arroyo, CM, Kramer, JH., Dikens, BF & Wegliki, WB. Identification of Free Radicals in Myocardial Ischemia/Reperfusion by Spin Trapping With Nitrone DMPO. FEBS letters 221, doi:10.1016/0014-5793(87)80360-5 (1987). 32 Garlick, PB, Davies, MJ, Hearse, DJ, & Slater TF. Direct Detection of Free Radicals in the Reperfused Rat Heart Using Electron Spin Resonance Spectroscopy. Circulation research 61, doi:10.1161/01.res.61.5.757 (1987). 33 Zweier, JL, Flaherty, JT, & Weisfeldt, ML. Direct Measurement of Free Radical Generation Following Reperfusion of Ischemic Myocardium. Proceedings of the National Academy of Sciences of the United States of America 84, doi:10.1073/pnas.84.5.1404 (1987). 34 Corrotte, M., Fernandes, M. C., Tam, C. & Andrews, N. W. Toxin pores endocytosed during plasma membrane repair traffic into the lumen of MVBs for degradation. Traffic 13, 483-494, doi:10.1111/j.1600-0854.2011.01323.x (2012). 35 Friden, J., Sjostrom, M. & Ekblom, B. Myofibrillar damage following intense eccentric exercise in man. Int J Sports Med 4, 170-176, doi:10.1055/s-2008- 1026030 (1983). 36 Bansal, D. et al. Defective membrane repair in dysferlin-deficient muscular dystrophy. Nature 423, 168-172, doi:10.1038/nature01573 (2003). 37 Liu, J. et al. Dysferlin, a novel skeletal muscle gene, is mutated in Miyoshi myopathy and limb girdle muscular dystrophy. Nat Genet 20, 31-36, doi:10.1038/1682 (1998). 38 Blake, D. J., Weir, A., Newey, S. E. & Davies, K. E. Function and genetics of dystrophin and dystrophin-related proteins in muscle. Physiol Rev 82, 291-329, doi:10.1152/physrev.00028.2001 (2002). 39 Cai, C. et al. Membrane repair defects in muscular dystrophy are linked to altered interaction between MG53, caveolin-3, and dysferlin. J Biol Chem 284, 15894- 15902, doi:10.1074/jbc.M109.009589 (2009). 40 Kovtun, O., Tillu, V. A., Ariotti, N., Parton, R. G. & Collins, B. M. Cavin family proteins and the assembly of caveolae. J Cell Sci 128, 1269-1278, doi:10.1242/jcs.167866 (2015). 41 Ohsawa, Y. et al. Caveolin-3 regulates myostatin signaling. Mini-review. Acta Myol 27, 19-24 (2008). 42 Cai, C. et al. MG53 regulates membrane budding and exocytosis in muscle cells. J Biol Chem 284, 3314-3322, doi:10.1074/jbc.M808866200 (2009). 43 Hernández-Deviez, D. J. et al. Aberrant dysferlin trafficking in cells lacking caveolin or expressing dystrophy mutants of caveolin-3. Hum Mol Genet 15, 129- 142, doi:10.1093/hmg/ddi434 (2006). 44 Zhu, H. et al. Polymerase transcriptase release factor (PTRF) anchors MG53 protein to cell injury site for initiation of membrane repair. J Biol Chem 286, 12820-12824, doi:10.1074/jbc.C111.221440 (2011). 45 Hernandez-Deviez, D. J. et al. Caveolin regulates endocytosis of the muscle repair protein, dysferlin. J Biol Chem 283, 6476-6488, doi:10.1074/jbc.M708776200 (2008).

117

46 Cacciottolo, M. et al. Reverse engineering gene network identifies new dysferlin- interacting proteins. J Biol Chem 286, 5404-5413, doi:10.1074/jbc.M110.173559 (2011). 47 Hill, M. M. et al. PTRF-Cavin, a conserved cytoplasmic protein required for caveola formation and function. Cell 132, 113-124, doi:10.1016/j.cell.2007.11.042 (2008). 48 Mariano, A., Henning, A. & Han, R. Dysferlin-deficient muscular dystrophy and innate immune activation. FEBS J 280, 4165-4176, doi:10.1111/febs.12261 (2013). 49 Togo, T. & Steinhardt, R. A. Nonmuscle myosin IIA and IIB have distinct functions in the exocytosis-dependent process of cell membrane repair. Mol Biol Cell 15, 688-695, doi:10.1091/mbc.E03-06-0430 (2004). 50 Lin, P. et al. Nonmuscle myosin IIA facilitates vesicle trafficking for MG53- mediated cell membrane repair. FASEB J 26, 1875-1883, doi:10.1096/fj.11- 188599 (2012). 51 Matsuda, C. et al. Dysferlin interacts with affixin (beta-parvin) at the sarcolemma. J Neuropathol Exp Neurol 64, 334-340 (2005). 52 Matsuda, C. et al. Affixin activates Rac1 via betaPIX in C2C12 myoblast. FEBS Lett 582, 1189-1196, doi:10.1016/j.febslet.2008.01.064 (2008). 53 Mishima, W. et al. The first CH domain of affixin activates Cdc42 and Rac1 through alphaPIX, a Cdc42/Rac1-specific guanine nucleotide exchanging factor. Genes Cells 9, 193-204 (2004). 54 Yamaji, S. et al. Affixin interacts with alpha-actinin and mediates integrin signaling for reorganization of F-actin induced by initial cell-substrate interaction. J Cell Biol 165, 539-551, doi:10.1083/jcb.200308141 (2004). 55 Schissel, S. L., Schuchman, E. H., Williams, K. J. & Tabas, I. Zn2+-stimulated sphingomyelinase is secreted by many cell types and is a product of the acid sphingomyelinase gene. J Biol Chem 271, 18431-18436 (1996). 56 Holopainen, J. M., Angelova, M. I. & Kinnunen, P. K. Vectorial budding of vesicles by asymmetrical enzymatic formation of ceramide in giant liposomes. Biophys J 78, 830-838, doi:10.1016/s0006-3495(00)76640-9 (2000). 57 Nabhan, J. F., Hu, R., Oh, R. S., Cohen, S. N. & Lu, Q. Formation and release of arrestin domain-containing protein 1-mediated microvesicles (ARMMs) at plasma membrane by recruitment of TSG101 protein. Proc Natl Acad Sci U S A 109, 4146-4151, doi:10.1073/pnas.1200448109 (2012). 58 Scheffer, L. L. et al. Mechanism of Ca²⁺-triggered ESCRT assembly and regulation of cell membrane repair. Nat Commun 5, 5646, doi:10.1038/ncomms6646 (2014). 59 Miyake, K., McNeil, P. L., Suzuki, K., Tsunoda, R. & Sugai, N. An actin barrier to resealing. J Cell Sci 114, 3487-3494 (2001). 60 Mellgren, R. L., Zhang, W., Miyake, K. & McNeil, P. L. Calpain is required for the rapid, calcium-dependent repair of wounded plasma membrane. J Biol Chem 282, 2567-2575, doi:10.1074/jbc.M604560200 (2007).

118

61 Huang, Y. et al. Calpain 3 is a modulator of the dysferlin protein complex in skeletal muscle. Hum Mol Genet 17, 1855-1866, doi:10.1093/hmg/ddn081 (2008). 62 de Morrée, A. et al. Proteomic analysis of the dysferlin protein complex unveils its importance for sarcolemmal maintenance and integrity. PLoS One 5, e13854, doi:10.1371/journal.pone.0013854 (2010). 63 Richard, I. et al. Loss of calpain 3 proteolytic activity leads to muscular dystrophy and to apoptosis-associated IkappaBalpha/nuclear factor kappaB pathway perturbation in mice. J Cell Biol 151, 1583-1590 (2000). 64 Duguez, S., Bartoli, M. & Richard, I. Calpain 3: a key regulator of the sarcomere? FEBS J 273, 3427-3436, doi:10.1111/j.1742-4658.2006.05351.x (2006). 65 Beckmann, J. S. & Spencer, M. Calpain 3, the "gatekeeper" of proper sarcomere assembly, turnover and maintenance. Neuromuscul Disord 18, 913-921, doi:10.1016/j.nmd.2008.08.005 (2008). 66 Mellgren, R. L. et al. Calcium-dependent plasma membrane repair requires m- or mu-calpain, but not calpain-3, the proteasome, or caspases. Biochim Biophys Acta 1793, 1886-1893, doi:10.1016/j.bbamcr.2009.09.013 (2009). 67 Redpath, G. M. et al. Calpain cleavage within dysferlin exon 40a releases a synaptotagmin-like module for membrane repair. Mol Biol Cell 25, 3037-3048, doi:10.1091/mbc.E14-04-0947 (2014). 68 Davis, T. A., Loos, B. & Engelbrecht, A. M. AHNAK: the giant jack of all trades. Cell Signal 26, 2683-2693, doi:10.1016/j.cellsig.2014.08.017 (2014). 69 Benaud, C. et al. AHNAK interaction with the annexin 2/S100A10 complex regulates cell membrane cytoarchitecture. J Cell Biol 164, 133-144, doi:10.1083/jcb.200307098 (2004). 70 Borgonovo, B. et al. Regulated exocytosis: a novel, widely expressed system. Nat Cell Biol 4, 955-962, doi:10.1038/ncb888 (2002). 71 Lorusso, A. et al. Annexin2 coating the surface of enlargeosomes is needed for their regulated exocytosis. EMBO J 25, 5443-5456, doi:10.1038/sj.emboj.7601419 (2006). 72 Huang, Y. et al. AHNAK, a novel component of the dysferlin protein complex, redistributes to the cytoplasm with dysferlin during skeletal muscle regeneration. FASEB J 21, 732-742, doi:10.1096/fj.06-6628com (2007). 73 Rescher, U. & Gerke, V. S100A10/p11: family, friends and functions. Pflugers Arch 455, 575-582, doi:10.1007/s00424-007-0313-4 (2008). 74 Rezvanpour, A., Santamaria-Kisiel, L. & Shaw, G. S. The S100A10-annexin A2 complex provides a novel asymmetric platform for membrane repair. J Biol Chem 286, 40174-40183, doi:10.1074/jbc.M111.244038 (2011). 75 Ozorowski, G., Milton, S. & Luecke, H. Structure of a C-terminal AHNAK peptide in a 1:2:2 complex with S100A10 and an acetylated N-terminal peptide of annexin A2. Acta Crystallogr D Biol Crystallogr 69, 92-104, doi:10.1107/s0907444912043429 (2013). 76 Dempsey, B. R. et al. Structure of an asymmetric ternary protein complex provides insight for membrane interaction. Structure 20, 1737-1745, doi:10.1016/j.str.2012.08.004 (2012). 119

77 Kobayashi, K., Izawa, T., Kuwamura, M. & Yamate, J. in J Toxicol Pathol Vol. 25 135-147 (2012). 78 Gerke, V. & Moss, S. E. Annexins: from structure to function. Physiol Rev 82, 331-371, doi:10.1152/physrev.00030.2001 (2002). 79 Brose, N., Petrenko, A. G., Südhof, T. C. & Jahn, R. Synaptotagmin: a calcium sensor on the synaptic vesicle surface. Science 256, 1021-1025 (1992). 80 Sutton, R. B., Davletov, B. A., Berghuis, A. M., Südhof, T. C. & Sprang, S. R. Structure of the first C2 domain of synaptotagmin I: a novel Ca2+/phospholipid- binding fold. Cell 80, 929-938 (1995). 81 Schiavo, G., Gu, Q. M., Prestwich, G. D., Söllner, T. H. & Rothman, J. E. Calcium-dependent switching of the specificity of phosphoinositide binding to synaptotagmin. Proc Natl Acad Sci U S A 93, 13327-13332 (1996). 82 Micheva, K. D., Holz, R. W. & Smith, S. J. Regulation of presynaptic phosphatidylinositol 4,5-biphosphate by neuronal activity. J Cell Biol 154, 355- 368 (2001). 83 Holz, R. W. et al. A pleckstrin homology domain specific for phosphatidylinositol 4, 5-bisphosphate (PtdIns-4,5-P2) and fused to green fluorescent protein identifies plasma membrane PtdIns-4,5-P2 as being important in exocytosis. J Biol Chem 275, 17878-17885, doi:10.1074/jbc.M000925200 (2000). 84 Davis, A. F. et al. Kinetics of synaptotagmin responses to Ca2+ and assembly with the core SNARE complex onto membranes. Neuron 24, 363-376 (1999). 85 Earles, C. A., Bai, J., Wang, P. & Chapman, E. R. The tandem C2 domains of synaptotagmin contain redundant Ca2+ binding sites that cooperate to engage t- SNAREs and trigger exocytosis. J Cell Biol 154, 1117-1123, doi:10.1083/jcb.200105020 (2001). 86 Fukuda, M. et al. Role of the C2B domain of synaptotagmin in vesicular release and recycling as determined by specific antibody injection into the squid giant synapse preterminal. Proc Natl Acad Sci U S A 92, 10708-10712 (1995). 87 Mikoshiba, K. et al. Role of the C2A domain of synaptotagmin in transmitter release as determined by specific antibody injection into the squid giant synapse preterminal. Proc Natl Acad Sci U S A 92, 10703-10707 (1995). 88 Bommert, K. et al. Inhibition of neurotransmitter release by C2-domain peptides implicates synaptotagmin in exocytosis. Nature 363, 163-165, doi:10.1038/363163a0 (1993). 89 Geppert, M. et al. Synaptotagmin I: a major Ca2+ sensor for transmitter release at a central synapse. Cell 79, 717-727 (1994). 90 Geppert, M., Goda, Y., Stevens, C. F. & Südhof, T. C. The small GTP-binding protein Rab3A regulates a late step in synaptic vesicle fusion. Nature 387, 810- 814, doi:10.1038/42954 (1997). 91 Fox, M. A. & Sanes, J. R. Synaptotagmin I and II are present in distinct subsets of central synapses. J Comp Neurol 503, 280-296, doi:10.1002/cne.21381 (2007). 92 Wendland, B., Miller, K. G., Schilling, J. & Scheller, R. H. Differential expression of the p65 gene family. Neuron 6, 993-1007 (1991).

120

93 Detrait, E. et al. Axolemmal repair requires proteins that mediate synaptic vesicle fusion. J Neurobiol 44, 382-391 (2000). 94 Chakrabarti, S. et al. Impaired membrane resealing and autoimmune myositis in synaptotagmin VII-deficient mice. J Cell Biol 162, 543-549, doi:10.1083/jcb.200305131 (2003). 95 Shen, S. S., Tucker, W. C., Chapman, E. R. & Steinhardt, R. A. Molecular regulation of membrane resealing in 3T3 fibroblasts. J Biol Chem 280, 1652- 1660, doi:10.1074/jbc.M410136200 (2005). 96 Anderson, L. V. et al. Dysferlin is a plasma membrane protein and is expressed early in human development. Hum Mol Genet 8, 855-861 (1999). 97 Ampong, B. N., Imamura, M., Matsumiya, T., Yoshida, M. & Takeda, S. Intracellular localization of dysferlin and its association with the dihydropyridine receptor. Acta Myol 24, 134-144 (2005). 98 Bashir, R. et al. A gene related to Caenorhabditis elegans spermatogenesis factor fer-1 is mutated in limb-girdle muscular dystrophy type 2B. Nat Genet 20, 37-42, doi:10.1038/1689 (1998). 99 Illa, I. et al. Distal anterior compartment myopathy: a dysferlin mutation causing a new muscular dystrophy phenotype. Ann Neurol 49, 130-134 (2001). 100 Ueyama, H. et al. A new dysferlin gene mutation in two Japanese families with limb-girdle muscular dystrophy 2B and Miyoshi myopathy. Neuromuscul Disord 11, 139-145 (2001). 101 Ho, M. et al. Disruption of muscle membrane and phenotype divergence in two novel mouse models of dysferlin deficiency. Hum Mol Genet 13, 1999-2010 (2004). 102 Chase, T. H., Cox, G. A., Burzenski, L., Foreman, O. & Shultz, L. D. Dysferlin deficiency and the development of cardiomyopathy in a mouse model of limb- girdle muscular dystrophy 2B. Am J Pathol 175, 2299-2308, doi:10.2353/ajpath.2009.080930 (2009). 103 Han, R. et al. Dysferlin-mediated membrane repair protects the heart from stress- induced left ventricular injury. J Clin Invest 117, 1805-1813, doi:10.1172/JCI30848 (2007). 104 Cooper, S. T. & Head, S. I. Membrane Injury and Repair in the Muscular Dystrophies. Neuroscientist, doi:10.1177/1073858414558336 (2014). 105 Demonbreun, A. R. et al. Impaired muscle growth and response to insulin-like growth factor 1 in dysferlin-mediated muscular dystrophy. Hum Mol Genet 20, 779-789, doi:10.1093/hmg/ddq522 (2011). 106 Evesson, F. J. et al. Reduced plasma membrane expression of dysferlin mutants is attributed to accelerated endocytosis via a syntaxin-4-associated pathway. J Biol Chem 285, 28529-28539, doi:10.1074/jbc.M110.111120 (2010). 107 Covian-Nares, J. F., Koushik, S. V., Puhl, H. L. & Vogel, S. S. Membrane wounding triggers ATP release and dysferlin-mediated intercellular calcium signaling. J Cell Sci 123, 1884-1893, doi:10.1242/jcs.066084 (2010).

121

108 Han, R. et al. Genetic ablation of complement C3 attenuates muscle pathology in dysferlin-deficient mice. J Clin Invest 120, 4366-4374, doi:10.1172/jci42390 (2010). 109 Wenzel, K. et al. Increased susceptibility to complement attack due to down- regulation of decay-accelerating factor/CD55 in dysferlin-deficient muscular dystrophy. J Immunol 175, 6219-6225 (2005). 110 Nagaraju, K. et al. Dysferlin deficiency enhances monocyte phagocytosis: a model for the inflammatory onset of limb-girdle muscular dystrophy 2B. Am J Pathol 172, 774-785, doi:10.2353/ajpath.2008.070327 (2008). 111 Wenzel, K. et al. Novel sequence variants in dysferlin-deficient muscular dystrophy leading to mRNA decay and possible C2-domain misfolding. Human mutation 27, 599-600, doi:10.1002/humu.9424 (2006). 112 Kerr, J. P. et al. Dysferlin stabilizes stress-induced Ca2+ signaling in the transverse tubule membrane. Proc Natl Acad Sci U S A 110, 20831-20836, doi:10.1073/pnas.1307960110 (2013). 113 Kerr, J. P., Ward, C. W. & Bloch, R. J. Dysferlin at transverse tubules regulates Ca(2+) homeostasis in skeletal muscle. Frontiers in physiology 5, 89, doi:10.3389/fphys.2014.00089 (2014). 114 Doherty, K. R. et al. Normal myoblast fusion requires myoferlin. Development 132, 5565-5575, doi:10.1242/dev.02155 (2005). 115 Demonbreun, A. R. et al. Myoferlin regulation by NFAT in muscle injury, regeneration and repair. J Cell Sci 123, 2413-2422, doi:10.1242/jcs.065375 (2010). 116 Lostal, W. et al. Lack of correlation between outcomes of membrane repair assay and correction of dystrophic changes in experimental therapeutic strategy in dysferlinopathy. PLoS One 7, e38036, doi:10.1371/journal.pone.0038036 (2012). 117 Johnson, C. P. & Chapman, E. R. Otoferlin is a calcium sensor that directly regulates SNARE-mediated membrane fusion. J Cell Biol 191, 187-197, doi:10.1083/jcb.201002089 (2010). 118 Roux, I. et al. Otoferlin, defective in a human deafness form, is essential for exocytosis at the auditory ribbon synapse. Cell 127, 277-289, doi:10.1016/j.cell.2006.08.040 (2006). 119 Gerke, V., Creutz, C. E. & Moss, S. E. Annexins: linking Ca2+ signalling to membrane dynamics. Nat Rev Mol Cell Biol 6, 449-461, doi:10.1038/nrm1661 (2005). 120 Han, R. & Campbell, K. P. Dysferlin and muscle membrane repair. Curr Opin Cell Biol 19, 409-416, doi:10.1016/j.ceb.2007.07.001 (2007). 121 McNeil, A. K., Rescher, U., Gerke, V. & McNeil, P. L. Requirement for annexin A1 in plasma membrane repair. J Biol Chem 281, 35202-35207, doi:10.1074/jbc.M606406200 (2006). 122 Lennon, N. J. et al. Dysferlin interacts with annexins A1 and A2 and mediates sarcolemmal wound-healing. J Biol Chem 278, 50466-50473 (2003). 123 Babiychuk, E. B. & Draeger, A. Annexins in cell membrane dynamics. Ca(2+)- regulated association of lipid microdomains. J Cell Biol 150, 1113-1124 (2000). 122

124 Bharadwaj, A., Bydoun, M., Holloway, R. & Waisman, D. Annexin A2 heterotetramer: structure and function. Int J Mol Sci 14, 6259-6305, doi:10.3390/ijms14036259 (2013). 125 Draeger, A., Monastyrskaya, K. & Babiychuk, E. B. Plasma membrane repair and cellular damage control: the annexin survival kit. Biochem Pharmacol 81, 703- 712, doi:10.1016/j.bcp.2010.12.027 (2011). 126 Bouter, A. et al. Annexin-A5 assembled into two-dimensional arrays promotes cell membrane repair. Nat Commun 2, 270, doi:10.1038/ncomms1270 (2011). 127 Swaggart, K. A. et al. Annexin A6 modifies muscular dystrophy by mediating sarcolemmal repair. Proc Natl Acad Sci U S A 111, 6004-6009, doi:10.1073/pnas.1324242111 (2014). 128 Cao, C. M. et al. MG53 constitutes a primary determinant of cardiac ischemic preconditioning. Circulation 121, 2565-2574, doi:10.1161/circulationaha.110.954628 (2010). 129 Cai, C. et al. MG53 nucleates assembly of cell membrane repair machinery. Nat Cell Biol 11, 56-64, doi:10.1038/ncb1812 (2009). 130 Duann, P. et al. MG53-mediated cell membrane repair protects against acute kidney injury. Sci Transl Med 7, 279ra236, doi:10.1126/scitranslmed.3010755 (2015). 131 Kim, S. C. et al. TRIM72 is required for effective repair of alveolar epithelial cell wounding. Am J Physiol Lung Cell Mol Physiol 307, L449-459, doi:10.1152/ajplung.00172.2014 (2014). 132 Weisleder, N., Takeshima, H. & Ma, J. Immuno-proteomic approach to excitation--contraction coupling in skeletal and cardiac muscle: molecular insights revealed by the mitsugumins. Cell Calcium 43, 1-8, doi:10.1016/j.ceca.2007.10.006 (2008). 133 Jia, Y. et al. Treatment of acute lung injury by targeting MG53-mediated cell membrane repair. Nat Commun 5, 4387, doi:10.1038/ncomms5387 (2014). 134 Weisleder, N. et al. Recombinant MG53 protein modulates therapeutic cell membrane repair in treatment of muscular dystrophy. Sci Transl Med 4, 139ra185, doi:10.1126/scitranslmed.3003921 (2012). 135 Wang, X. et al. Cardioprotection of ischemia/reperfusion injury by cholesterol- dependent MG53-mediated membrane repair. Circulation research 107, 76-83, doi:10.1161/CIRCRESAHA.109.215822 (2010). 136 Hwang, M., Ko, J. K., Weisleder, N., Takeshima, H. & Ma, J. Redox-dependent oligomerization through a leucine zipper motif is essential for MG53-mediated cell membrane repair. Am J Physiol Cell Physiol 301, C106-114, doi:10.1152/ajpcell.00382.2010 (2011). 137 Waddell, L. B. et al. Dysferlin, annexin A1, and mitsugumin 53 are upregulated in muscular dystrophy and localize to longitudinal tubules of the T-system with stretch. J Neuropathol Exp Neurol 70, 302-313, doi:10.1097/NEN.0b013e31821350b0 (2011).

123

138 Corona, B. T. et al. Effect of recombinant human MG53 protein on tourniquet- induced ischemia-reperfusion injury in rat muscle. Muscle Nerve 49, 919-921, doi:10.1002/mus.24160 (2014). 139 Ham, Y. M. & Mahoney, S. J. Compensation of the AKT signaling by ERK signaling in transgenic mice hearts overexpressing TRIM72. Exp Cell Res 319, 1451-1462, doi:10.1016/j.yexcr.2013.02.016 (2013). 140 Jung, S. Y. & Ko, Y. G. TRIM72, a novel negative feedback regulator of myogenesis, is transcriptionally activated by the synergism of MyoD (or myogenin) and MEF2. Biochemical and biophysical research communications 396, 238-245, doi:10.1016/j.bbrc.2010.04.072 (2010). 141 Lee, C. S. et al. TRIM72 negatively regulates myogenesis via targeting insulin receptor substrate-1. Cell Death Differ 17, 1254-1265, doi:10.1038/cdd.2010.1 (2010). 142 Prokic, I., Cowling, B. S. & Laporte, J. Amphiphysin 2 (BIN1) in physiology and diseases. J Mol Med (Berl) 92, 453-463, doi:10.1007/s00109-014-1138-1 (2014). 143 Marg, A. et al. Sarcolemmal repair is a slow process and includes EHD2. Traffic 13, 1286-1294, doi:10.1111/j.1600-0854.2012.01386.x (2012). 144 Derby, M. C. & Gleeson, P. A. New insights into membrane trafficking and protein sorting. International review of cytology 261, 47-116, doi:10.1016/S0074- 7696(07)61002-X (2007). 145 Rosenberger, G., Gal, A. & Kutsche, K. AlphaPIX associates with calpain 4, the small subunit of calpain, and has a dual role in integrin-mediated cell spreading. J Biol Chem 280, 6879-6889, doi:10.1074/jbc.M412119200 (2005). 146 Honda, S. et al. Activation of m-calpain is required for chromosome alignment on the metaphase plate during mitosis. J Biol Chem 279, 10615-10623, doi:10.1074/jbc.M308841200 (2004). 147 Miyake, K. & McNeil, P. L. Mechanical injury and repair of cells. Critical care medicine 31, S496-501, doi:10.1097/01.CCM.0000081432.72812.16 (2003). 148 McNeil, P. L. & Kirchhausen, T. An emergency response team for membrane repair. Nat Rev Mol Cell Biol 6, 499-505, doi:10.1038/nrm1665 (2005). 149 Mendell, JR, & Lloyd-Puryear, M. Report of MDA Muscle Disease Symposium on Newborn Screening for Duchenne Muscular Dystrophy. Muscle & nerve 48, doi:10.1002/mus.23810 (2013). 150 Ryder, S. et al. The Burden, Epidemiology, Costs and Treatment for Duchenne Muscular Dystrophy: An Evidence Review. Orphanet journal of rare diseases 12, doi:10.1186/s13023-017-0631-3 (2017). 151 Hoffman, E. P., Brown, R. H., Jr. & Kunkel, L. M. Dystrophin: the protein product of the Duchenne muscular dystrophy locus. Cell 51, 919-928, doi:10.1016/0092-8674(87)90579-4 (1987). 152 Muntoni, F., Torelli, S. & Ferlini, A. Dystrophin and mutations: one gene, several proteins, multiple phenotypes. Lancet Neurol 2, 731-740, doi:10.1016/s1474- 4422(03)00585-4 (2003).

124

153 Bushby, K. et al. Diagnosis and management of Duchenne muscular dystrophy, part 2: implementation of multidisciplinary care. Lancet Neurol 9, 177-189, doi:10.1016/s1474-4422(09)70272-8 (2010). 154 Wein, N., Alfano, L. & Flanigan, K. M. Genetics and emerging treatments for Duchenne and Becker muscular dystrophy. Pediatr Clin North Am 62, 723-742, doi:10.1016/j.pcl.2015.03.008 (2015). 155 Heydemann, A. Skeletal Muscle Metabolism in Duchenne and Becker Muscular Dystrophy-Implications for Therapies. Nutrients 10, doi:10.3390/nu10060796 (2018). 156 Gao, Q. Q. & McNally, E. M. The Dystrophin Complex: Structure, Function, and Implications for Therapy. Compr Physiol 5, 1223-1239, doi:10.1002/cphy.c140048 (2015). 157 Way, M., Pope, B., Cross, R. A., Kendrick-Jones, J. & Weeds, A. G. Expression of the N-terminal domain of dystrophin in E. coli and demonstration of binding to F-actin. FEBS Lett 301, 243-245, doi:10.1016/0014-5793(92)80249-g (1992). 158 Straub, V. & Campbell, K. P. Muscular dystrophies and the dystrophin- glycoprotein complex. Curr Opin Neurol 10, 168-175, doi:10.1097/00019052- 199704000-00016 (1997). 159 Klingler, W., Jurkat-Rott, K., Lehmann-Horn, F. & Schleip, R. in Acta Myol Vol. 31 184-195 (2012). 160 Hopf, F. W., Turner, P. R. & Steinhardt, R. A. Calcium misregulation and the pathogenesis of muscular dystrophy. Subcell Biochem 45, 429-464, doi:10.1007/978-1-4020-6191-2_16 (2007). 161 Bulfield, G, Siller, WG, Wight, PA, & Moore KJ. X chromosome-linked muscular dystrophy (mdx) in the mouse. Proceedings of the National Academy of Sciences of the United States of America 81, doi:10.1073/pnas.81.4.1189 (1984). 162 Radley-Crabb, H. G. et al. in PLoS One Vol. 9 (2014). 163 Moens, P., Baatsen, P. H. & Marechal, G. Increased susceptibility of EDL muscles from mdx mice to damage induced by contractions with stretch. J Muscle Res Cell Motil 14, 446-451, doi:10.1007/bf00121296 (1993). 164 Yucel, N., Chang, A. C., Day, J. W., Rosenthal, N. & Blau, H. M. Humanizing the mdx mouse model of DMD: the long and the short of it. npj Regenerative Medicine 3, 1-11, doi:doi:10.1038/s41536-018-0045-4 (2018). 165 Hatakeyama, S. TRIM proteins and cancer. Nat Rev Cancer 11, 792-804, doi:10.1038/nrc3139 (2011). 166 Ozato, K., Shin, D. M., Chang, T. H. & Morse, H. C. TRIM family proteins and their emerging roles in innate immunity. Nat Rev Immunol 8, 849-860, doi:10.1038/nri2413 (2008). 167 Sanchez, J. G. et al. The tripartite motif coiled-coil is an elongated antiparallel hairpin dimer. doi:10.1073/pnas.1318962111 (2014). 168 Napolitano, L. M. & Meroni, G. TRIM family: Pleiotropy and diversification through homomultimer and heteromultimer formation. IUBMB Life 64, 64-71, doi:10.1002/iub.580 (2012).

125

169 Koliopoulos, M. G., Esposito, D., Christodoulou, E., Taylor, I. A. & Rittinger, K. Functional role of TRIM E3 ligase oligomerization and regulation of catalytic activity. Embo j 35, 1204-1218, doi:10.15252/embj.201593741 (2016). 170 D'Cruz, A. A., Babon, J. J., Norton, R. S., Nicola, N. A. & Nicholson, S. E. Structure and function of the SPRY/B30.2 domain proteins involved in innate immunity. Protein Sci 22, 1-10, doi:10.1002/pro.2185 (2013). 171 James, L. C., Keeble, A. H., Khan, Z., Rhodes, D. A. & Trowsdale, J. Structural basis for PRYSPRY-mediated tripartite motif (TRIM) protein function. doi:10.1073/pnas.0609174104 (2007). 172 Motohashi, N. et al. Regulation of IRS1/Akt insulin signaling by microRNA-128a during myogenesis. J Cell Sci 126, 2678-2691, doi:10.1242/jcs.119966 (2013). 173 Berkes, C. A. & Tapscott, S. J. MyoD and the transcriptional control of myogenesis. Semin Cell Dev Biol 16, 585-595, doi:10.1016/j.semcdb.2005.07.006 (2005). 174 Yi, J.-S. et al. MG53-induced IRS-1 ubiquitination negatively regulates skeletal myogenesis and insulin signalling. Nature Communications 4, 1-12, doi:doi:10.1038/ncomms3354 (2013). 175 Senyo, S. E., Lee, R. T. & Kuhn, B. Cardiac regeneration based on mechanisms of cardiomyocyte proliferation and differentiation. Stem Cell Res 13, 532-541, doi:10.1016/j.scr.2014.09.003 (2014). 176 Weisleder, N., Takeshima, H. & Ma, J. Mitsugumin 53 (MG53) facilitates vesicle trafficking in striated muscle to contribute to cell membrane repair. Commun Integr Biol 2, 225-226 (2009). 177 Lee, H. et al. MG53-IRS-1 (Mitsugumin 53-Insulin Receptor Substrate-1) Interaction Disruptor Sensitizes Insulin Signaling in Skeletal Muscle. J Biol Chem 291, 26627-26635, doi:10.1074/jbc.M116.754424 (2016). 178 He, B. et al. Enhancing muscle membrane repair by gene delivery of MG53 ameliorates muscular dystrophy and heart failure in delta-Sarcoglycan-deficient hamsters. Mol Ther 20, 727-735, doi:10.1038/mt.2012.5 (2012). 179 Han, R., Rader, E. P., Levy, J. R., Bansal, D. & Campbell, K. P. Dystrophin deficiency exacerbates skeletal muscle pathology in dysferlin-null mice. Skelet Muscle 1, 35, doi:10.1186/2044-5040-1-35 (2011). 180 Cohen, T. V., Cohen, J. E. & Partridge, T. A. Myogenesis in dysferlin-deficient myoblasts is inhibited by an intrinsic inflammatory response. Neuromuscul Disord 22, 648-658, doi:10.1016/j.nmd.2012.03.002 (2012). 181 Xie, W., Strong, J. A. & Zhang, J. M. Active Nerve Regeneration with Failed Target Reinnervation Drives Persistent Neuropathic Pain. in eNeuro Vol. 4 (2017). 182 Brushart, TM. Motor Axons Preferentially Reinnervate Motor Pathways. The Journal of neuroscience : the official journal of the Society for Neuroscience 13, doi:10.1523/JNEUROSCI.13-06-02730.1993 (1993). 183 Brushart, TM. & Marsel Mesulam, M. Transganglionic Demonstration of Central Sensory Projections From Skin and Muscle With HRP-lectin Conjugates. Neuroscience letters 17, doi:10.1016/0304-3940(80)90051-8 (1980). 126

184 McNeil, P. L., Miyake, K. & Vogel, S. S. The endomembrane requirement for cell surface repair. Proc Natl Acad Sci U S A 100, 4592-4597, doi:10.1073/pnas.0736739100 (2003). 185 Togo, T., Krasieva, T. B. & Steinhardt, R. A. in Mol Biol Cell Vol. 11 4339- 4346 (2000). 186 McNeil, P. L. & Steinhardt, R. A. Loss, Restoration, and Maintenance of Plasma Membrane Integrity. J Cell Biol 137, 1-4 (1997). 187 Togo, T., Alderton, J. M., Bi, G. Q. & Steinhardt, R. A. The mechanism of facilitated cell membrane resealing. J Cell Sci 112 ( Pt 5), 719-731 (1999). 188 Castro-Gomes, T., Koushik, A. B. & Andrews, N. W. ESCRT: nipping the wound in the bud? Trends Biochem Sci 39, 307-309, doi:10.1016/j.tibs.2014.06.001 (2014). 189 Castro-Gomes, T., Corrotte, M., Tam, C. & Andrews, N. W. Plasma Membrane Repair Is Regulated Extracellularly by Proteases Released from Lysosomes. PLoS One 11, e0152583, doi:10.1371/journal.pone.0152583 (2016). 190 Yoo, S., Nguyen, M. P., Fukuda, M., Bittner, G. D. & Fishman, H. M. Plasmalemmal sealing of transected mammalian neurites is a gradual process mediated by Ca(2+)-regulated proteins. J Neurosci Res 74, 541-551, doi:10.1002/jnr.10771 (2003). 191 Yao, W. et al. MG53 anchored by dysferlin to cell membrane reduces hepatocyte apoptosis which induced by ischaemia/reperfusion injury in vivo and in vitro. J Cell Mol Med 21, 2503-2513, doi:10.1111/jcmm.13171 (2017). 192 Gushchina, L. V. et al. Treatment with Recombinant Human MG53 Protein Increases Membrane Integrity in a Mouse Model of Limb Girdle Muscular Dystrophy 2B. Mol Ther 25, 2360-2371, doi:10.1016/j.ymthe.2017.06.025 (2017). 193 Nagre, N. et al. Inhaled TRIM72 Protein Protects Ventilation Injury to the Lung through Injury-guided Cell Repair. Am J Respir Cell Mol Biol, doi:10.1165/rcmb.2017-0364OC (2018). 194 Towler, M. C., Kaufman, S. J. & Brodsky, F. M. Membrane traffic in skeletal muscle. Traffic 5, 129-139, doi:10.1111/j.1600-0854.2003.00164.x (2004). 195 Gajic, O. et al. Ventilator-induced cell wounding and repair in the intact lung. Am J Respir Crit Care Med 167, 1057-1063, doi:10.1164/rccm.200208-889OC (2003). 196 Jaiswal, J. K. et al. S100A11 is required for efficient plasma membrane repair and survival of invasive cancer cells. Nat Commun 5, 3795, doi:10.1038/ncomms4795 (2014). 197 Doherty, K. R. & McNally, E. M. Repairing the tears: dysferlin in muscle membrane repair. Trends Mol Med 9, 327-330 (2003). 198 Bazan, N. G., Marcheselli, V. L. & Cole-Edwards, K. Brain response to injury and neurodegeneration: endogenous neuroprotective signaling. Ann N Y Acad Sci 1053, 137-147, doi:10.1196/annals.1344.011 (2005). 199 Mina, E. W., Lasagna-Reeves, C., Glabe, C. G. & Kayed, R. Poloxamer 188 copolymer membrane sealant rescues toxicity of amyloid oligomers in vitro. J Mol Biol 391, 577-585, doi:10.1016/j.jmb.2009.06.024 (2009). 127

200 Cooper, S. T. et al. Dystrophinopathy carrier determination and detection of protein deficiencies in muscular dystrophy using lentiviral MyoD-forced myogenesis. Neuromuscul Disord 17, 276-284, doi:10.1016/j.nmd.2006.12.010 (2007). 201 Jaiswal, J. K. et al. Patients with a non-dysferlin Miyoshi myopathy have a novel membrane repair defect. Traffic 8, 77-88, doi:10.1111/j.1600-0854.2006.00505.x (2007). 202 van der Kooi, A. J. et al. Limb-girdle muscular dystrophy in the Netherlands: gene defect identified in half the families. Neurology 68, 2125-2128, doi:10.1212/01.wnl.0000264853.40735.3b (2007). 203 Wenzel, K. et al. Dysfunction of dysferlin-deficient hearts. J Mol Med (Berl) 85, 1203-1214, doi:10.1007/s00109-007-0253-7 (2007). 204 Lemckert, F. A. et al. Lack of MG53 in human heart precludes utility as a biomarker of myocardial injury or endogenous cardioprotective factor. Cardiovasc Res 110, 178-187, doi:10.1093/cvr/cvw017 (2016). 205 Hendricks, B. K. & Shi, R. Mechanisms of neuronal membrane sealing following mechanical trauma. in Neurosci Bull Vol. 30 627-644 (2014). 206 Bittner, G. D., Spaeth, C. S., Poon, A. D., Burgess, Z. S. & McGill, C. H. Repair of traumatic plasmalemmal damage to neurons and other eukaryotic cells. Neural Regen Res 11, 1033-1042, doi:10.4103/1673-5374.187019 (2016). 207 Houang, E. M. et al. Membrane-stabilizing copolymers confer marked protection to dystrophic skeletal muscle in vivo. Mol Ther Methods Clin Dev 2, 15042, doi:10.1038/mtm.2015.42 (2015). 208 Moloughney, J. G. & Weisleder, N. Poloxamer 188 (p188) as a membrane resealing reagent in biomedical applications. Recent Pat Biotechnol 6, 200-211 (2012). 209 Lee, J. J. A., Maruyama, R., Sakurai, H. & Yokota, T. Cell Membrane Repair Assay Using a Two-photon Laser Microscope. J Vis Exp, doi:10.3791/56999 (2018). 210 Sreetama, S. C. et al. Membrane Stabilization by Modified Steroid Offers a Potential Therapy for Muscular Dystrophy Due to Dysferlin Deficit. Mol Ther 26, 2231-2242, doi:10.1016/j.ymthe.2018.07.021 (2018). 211 Quattrocelli, M. et al. Intermittent Glucocorticoid Dosing Improves Muscle Repair and Function in Mice with Limb-Girdle Muscular Dystrophy. Am J Pathol 187, 2520-2535, doi:10.1016/j.ajpath.2017.07.017 (2017). 212 Shin, J. E., Geisler, S. & DiAntonio, A. Dynamic regulation of SCG10 in regenerating axons after injury. Exp Neurol 252, 1-11, doi:10.1016/j.expneurol.2013.11.007 (2014). 213 Lerch, J. K. et al. Stress Increases Peripheral Axon Growth and Regeneration through Glucocorticoid Receptor-Dependent Transcriptional Programs. in eNeuro Vol. 4 (2017). 214 Abe, N., Borson, S. H., Gambello, M. J., Wang, F. & Cavalli, V. Mammalian target of rapamycin (mTOR) activation increases axonal growth capacity of

128

injured peripheral nerves. J Biol Chem 285, 28034-28043, doi:10.1074/jbc.M110.125336 (2010). 215 Cho, Y., Sloutsky, R., Naegle, K. M. & Cavalli, V. Injury-induced HDAC5 nuclear export is essential for axon regeneration. Cell 155, 894-908, doi:10.1016/j.cell.2013.10.004 (2013). 216 Krause, T. L., Fishman, H. M., Ballinger, M. L. & Bittner, G. D. Extent and mechanism of sealing in transected giant axons of squid and earthworms. J Neurosci 14, 6638-6651 (1994). 217 Li, H. et al. Modulation of wound healing and scar formation by MG53 protein- mediated cell membrane repair. J Biol Chem 290, 24592-24603, doi:10.1074/jbc.M115.680074 (2015). 218 Anwar, M. A., Al Shehabi, T. S. & Eid, A. H. Inflammogenesis of Secondary Spinal Cord Injury. Front Cell Neurosci 10, 98, doi:10.3389/fncel.2016.00098 (2016). 219 Shi, K., Zhang, J., Dong, J. F. & Shi, F. D. Dissemination of brain inflammation in traumatic brain injury. Cell Mol Immunol 16, 523-530, doi:10.1038/s41423- 019-0213-5 (2019). 220 Guan, F. et al. MG53 attenuates lipopolysaccharide-induced neurotoxicity and neuroinflammation via inhibiting TLR4/NF-κB pathway in vitro and in vivo. Prog Neuropsychopharmacol Biol Psychiatry 95, 109684, doi:10.1016/j.pnpbp.2019.109684 (2019). 221 Torii, T., Miyamoto, Y. & Yamauchi, J. Cellular Signal-Regulated Schwann Cell Myelination and Remyelination. Adv Exp Med Biol 1190, 3-22, doi:10.1007/978- 981-32-9636-7_1 (2019). 222 Wu, H. K. et al. Glucose-Sensitive Myokine/Cardiokine MG53 Regulates Systemic Insulin Response and Metabolic Homeostasis. Circulation 139, 901- 914, doi:10.1161/circulationaha.118.037216 (2019). 223 Balastik, M. et al. Deficiency in ubiquitin ligase TRIM2 causes accumulation of neurofilament light chain and neurodegeneration. Proc Natl Acad Sci U S A 105, 12016-12021, doi:10.1073/pnas.0802261105 (2008). 224 Khazaei, M. R. et al. The E3-ubiquitin ligase TRIM2 regulates neuronal polarization. J Neurochem 117, 29-37, doi:10.1111/j.1471-4159.2010.06971.x (2011). 225 Li, JJ. et al.A Recessive Trim2 Mutation Causes an Axonal Neuropathy in Mice. Neurobiology of disease 140, doi:10.1016/j.nbd.2020.104845 (2020). 226 Reymond, A. et al. The tripartite motif family identifies cell compartments. in EMBO J Vol. 20 2140-2151 (2001). 227 Witte, M. E. et al. Calcium Influx through Plasma-Membrane Nanoruptures Drives Axon Degeneration in a Model of Multiple Sclerosis. Neuron 101, 615- 624.e615, doi:10.1016/j.neuron.2018.12.023 (2019). 228 Julien, C. et al. In vivo induction of membrane damage by beta-amyloid peptide oligomers. Acta Neuropathol Commun 6, 131, doi:10.1186/s40478-018-0634-x (2018).

129

229 Guan, F. et al. The TRIM protein Mitsugumin 53 enhances survival and therapeutic efficacy of stem cells in murine traumatic brain injury. Stem Cell Res Ther 10, 352, doi:10.1186/s13287-019-1433-4 (2019). 230 Thompson, S. et al. Identification of a novel Bcl-2-interacting mediator of cell death (Bim) E3 ligase, tripartite motif-containing protein 2 (TRIM2), and its role in rapid ischemic tolerance-induced neuroprotection. J Biol Chem 286, 19331- 19339, doi:10.1074/jbc.M110.197707 (2011). 231 Fang, C. et al. MicroRNA-181c Ameliorates Cognitive Impairment Induced by Chronic Cerebral Hypoperfusion in Rats. Mol Neurobiol, doi:10.1007/s12035- 016-0268-6 (2016). 232 Ylikallio, E. et al. Deficiency of the E3 ubiquitin ligase TRIM2 in early-onset axonal neuropathy. Hum Mol Genet 22, 2975-2983, doi:10.1093/hmg/ddt149 (2013). 233 Zhu, Q., Couillard-Despres, S. & Julien, J. P. Delayed maturation of regenerating myelinated axons in mice lacking neurofilaments. Exp Neurol 148, 299-316, doi:10.1006/exnr.1997.6654 (1997). 234 Yadave, P. et al. Neurofilament Depletion Improves Microtubule Dynamics via Modulation of Stat3/stathmin Signaling. Acta neuropathologica 132, doi:10.1007/s00401-016-1564-y (2016). 235 McNeil, P. L. & Steinhardt, R. A. Plasma membrane disruption: repair, prevention, adaptation. Annu Rev Cell Dev Biol 19, 697-731, doi:10.1146/annurev.cellbio.19.111301.140101 (2003). 236 Cooper, ST. & McNeil, PL. Membrane Repair: Mechanisms and Pathophysiology. Physiological reviews 95, doi:10.1152/physrev.00037.2014 (2015). 237 Buddhe, S. et al. Cardiac Management of the Patient With Duchenne Muscular Dystrophy. Pediatrics 142, doi:10.1542/peds.2018-0333I (2018). 238 Gilroy, J. et al. Cardiac and Pulmonary Complications in Duchenne's Progressive Muscular Dystrophy. Circulation 27, doi:10.1161/01.cir.27.4.484 (1963). 239 Meyers, TA., & Townsend, D. Cardiac Pathophysiology and the Future of Cardiac Therapies in Duchenne Muscular Dystrophy. International journal of molecular sciences 20, doi:10.3390/ijms20174098 (2019). 240 Zhang, Y. et al. MG53 participates in ischaemic postconditioning through the RISK signalling pathway. in Cardiovasc Res Vol. 91 108-115 (2011). 241 Zhang, C. et al. MG53 is dispensable for T-tubule maturation but critical for maintaining T-tubule integrity following cardiac stress. J Mol Cell Cardiol 112, 123-130, doi:10.1016/j.yjmcc.2017.08.007 (2017). 242 Eddleman, CS. Bittner, Gd. & Fishman, HF. SEM comparison of severed ends of giant axons isolated from squid (Loligo pealeii) and crayfish (Procambarus clarkii). The Biological Bulletin 203, 219-220, doi:10.2307/1543406 (2020). 243 Pramono, ZAD. et al. Identification and Characterization of a Novel Human Dysferlin Transcript: dysferlin_v1. Human genetics 120, doi:10.1007/s00439- 006-0230-1 (2006).

130

244 Anderson, LV. et al. Dysferlin Is a Plasma Membrane Protein and Is Expressed Early in Human Development. Human molecular genetics 8, doi:10.1093/hmg/8.5.855 (1999). 245 Liu, J. et al. Dysferlin, a Novel Skeletal Muscle Gene, Is Mutated in Miyoshi Myopathy and Limb Girdle Muscular Dystrophy. Nature genetics 20, doi:10.1038/1682 (1998). 246 Galvin, JE. et al. The Muscle Protein Dysferlin Accumulates in the Alzheimer Brain. Acta neuropathologica 112, doi:10.1007/s00401-006-0147-8 (2006). 247 Berti, C., Messali, S., Ballabio, A., Reymond, A. & Meroni, G. TRIM9 is specifically expressed in the embryonic and adult nervous system. Mech Dev 113, 159-162, doi:10.1016/s0925-4773(02)00013-8 (2002). 248 Ohkawa, N. et al. Molecular cloning and characterization of neural activity- related RING finger protein (NARF): a new member of the RBCC family is a candidate for the partner of myosin V. J Neurochem 78, 75-87 (2001). 249 Tsujita, K. et al. Proteome of acidic phospholipid-binding proteins: spatial and temporal regulation of Coronin 1A by phosphoinositides. J Biol Chem 285, 6781- 6789, doi:10.1074/jbc.M109.057018 (2010). 250 Boye, T. L. et al. Annexin A4 and A6 induce membrane curvature and constriction during cell membrane repair. Nature Communications 8, 1-11, doi:doi:10.1038/s41467-017-01743-6 (2017). 251 Davis, D. B., Doherty, K. R., Delmonte, A. J. & McNally, E. M. Calcium- sensitive phospholipid binding properties of normal and mutant ferlin C2 domains. J Biol Chem 277, 22883-22888, doi:10.1074/jbc.M201858200 (2002). 252 Pehlivan, D. et al. Exome sequencing reveals homozygous TRIM2 mutation in a patient with early onset CMT and bilateral vocal cord paralysis. Hum Genet 134, 671-673, doi:10.1007/s00439-015-1548-3 (2015). 253 Cox, T. C. The microtubule-associated C-I subfamily of TRIM proteins and the regulation of polarized cell responses. Adv Exp Med Biol 770, 105-118, doi:10.1007/978-1-4614-5398-7_8 (2012).

131