Functional Characterization of Novel PFN1 Mutations Causative for Familial Amyotrophic Lateral Sclerosis: a Dissertation

University of Massachusetts Medical School eScholarship@UMMS

GSBS Dissertations and Theses Graduate School of Biomedical Sciences

2015-12-17

Functional Characterization of Novel PFN1 Mutations Causative for Familial Amyotrophic Lateral Sclerosis: A Dissertation

Chi-Hong Wu University of Massachusetts Medical School

Let us know how access to this document benefits ou.y Follow this and additional works at: https://escholarship.umassmed.edu/gsbs_diss

Part of the Genetics and Genomics Commons, Molecular and Cellular Neuroscience Commons, and the Nervous System Diseases Commons

Repository Citation Wu C. (2015). Functional Characterization of Novel PFN1 Mutations Causative for Familial Amyotrophic Lateral Sclerosis: A Dissertation. GSBS Dissertations and Theses. https://doi.org/10.13028/M2WS38. Retrieved from https://escholarship.umassmed.edu/gsbs_diss/815

This material is brought to you by eScholarship@UMMS. It has been accepted for inclusion in GSBS Dissertations and Theses by an authorized administrator of eScholarship@UMMS. For more information, please contact [email protected]. FUNCTIONAL CHARACTERIZATION OF NOVEL PFN1 MUTATIONS CAUSATIVE FOR FAMILIAL AMYOTROPHIC LATERAL SCLEROSIS

A Dissertation Presented

Chi-Hong Wu

Submitted to the faculty of the University of Massachusetts Graduate School of Biomedical Sciences, Worcester in partial fulfillment of the requirement for the degree of

DOCTOR OF PHILOSOPHY

December 17, 2015

INTERDISCIPLINARY GRADUATE PROGRAM

FUNCTIONAL CHARACTERIZATION OF NOVEL PFN1 MUTATIONS CAUSATIVE FOR FAMILIAL AMYOTROPHIC LATERAL SCLEROSIS

A Dissertation Presented By Chi-Hong Wu

The signatures of the Dissertation Defense Committee signify completion and approval as to style and content of the Dissertation

______John E. Landers, Ph.D., Thesis Advisor

______Fen-Biao Gao, Ph.D., Member of Committee

______Marc R. Freeman, Ph.D., Member of Committee

______Zuoshang Xu, MD, Ph.D., Member of Committee

______Udai B. Pandey, Ph.D., Member of Committee

The signature of the Chair of the Committee signifies that the written dissertation meets the requirements of the Dissertation Committee

______Lawrence J. Hayward, MD, Ph.D., Chair of Committee

The signature of the Dean of the Graduate School of Biomedical Sciences signifies that the student has met all graduation requirements of the School

______Anthony Carruthers, Ph.D. Dean of the Graduate School of Biomedical Sciences

Interdisciplinary Graduate Program December 17, 2015 ‘

ii ACKNOWLEDGEMENTS

There are many people whom I would like to thank for their support throughout the course of my graduate study. First and foremost, I would like to thank my thesis advisor, John Landers, for taking me as your first student and providing a dynamic and interactive research environment. I am also grateful that you always encourage me as a scientist, and are open-minded and supportive for my projects throughout the years. I have also learned greatly from our discussions, brainstorming, and your feedback for my proposals. Thank you very much for everything!

I have also enjoyed working with all the members of Landers lab. In particular, I would like to thank Claudia Fallini. You never hesitate to provide me with your help and critical comments. You also teach me to keep positive despite the downtime in research, which I really appreciate. Apart from the scientific stuffs, from you I have also learned what authentic pizza should be like. And many thanks to all post-docs in the lab, Eric Danielson, Kevin Kenna and

Anthony Giampretruzzi, for your help, encouragement and good laughs, especially when I was applying for post-doc positions. Also, I would like to thank all the current and former members of Bosco lab, especially Sivakumar

Boopathy, Catherine Ward, Melissa Rotunno, and Maeve Tischbein. You guys have encouraged me tremendously and are always generous enough to share with me your reagents and equipment.

iii I would also like to thank Helene Tran-Ladam from Gao lab. You have

been there for me since the day one I started the fly project. You not only support

me, encourage me how to move forward as a researcher, help me with all kinds

of experiments but also give me the most critical feedback on my proposals,

experiment designs and research rationales. Likewise, I would like to thank Tsai-

Yi Lu from Freeman lab. You have always made my life a lot easier but sharing your experience and humor. And many thanks to all the fly people in Gao lab for your support: I really appreciate that you guys are always willing to help me and answer my all kinds of questions about flies.

For their advice and guidance throughout the years I would also like to thank the members of my thesis advisory committee: Dr. Zuoshang Xu, Dr. Fen-

Biao Gao, Dr. Marc Freeman, Dr. Daryl Bosco and Dr. Larry Hayward. I would also like to thank Dr. Udai Pandey for being part of my dissertation examination committee. I really appreciate your time.

All of you have made my years in graduate school a wonderful and tremendous journey.

iv ABSTRACT

Amyotrophic lateral sclerosis (ALS) is a progressive adult neurodegenerative disease that causes death of both upper and lower motor neurons.

Approximately 90 percent of ALS cases are sporadic (SALS), and 10 percent are inherited (FALS). Mutations in the PFN1 gene have been identified as causative for one percent of FALS. PFN1 is a small actin-binding protein that promotes actin polymerization, but how ALS-linked PFN1 mutations affect its cognate functions or acquire gain-of-function toxicity remains largely unknown.

To elucidate the contribution of ALS-linked PFN1 mutations to neurodegeneration, we have characterized these mutants in both mammalian cultured cells and Drosophila models. In mammalian neuronal cells, we demonstrate that ALS-linked PFN1 mutants form ubiquitinated aggregates and alter neuronal morphology. We also show that ALS-linked PFN1 mutants have partial loss-of-function effects on actin polymerization in growth cones of mouse primary motor neurons and larval neuromuscular junctions (NMJ) in Drosophila.

In Drosophila, we also observe that PFN1 level influences integrity of adult motor neurons, as demonstrated by locomotion, lifespan, and leg NMJ morphology.

In sum, the work presented in this dissertation has shed light on PFN1- linked ALS pathogenesis by demonstrating a loss-of-function mechanism. We have also developed a Drosophila PFN1 model that will serve as a valuable tool to further uncover PFN1-associated cellular pathways that mediate motor neuron functions.

v TABLE OF CONTENTS

APPROVAL……………………………………………………………………………...ii ACKNOWLEDGEMENTS ...... iii ABSTRACT ...... v TABLE OF CONTENTS ...... vi LIST OF TABLES ...... viii LIST OF FIGURES ...... viii LIST OF THIRD PARTY COPYRIGHTED MATERIAL ...... x LIST OF ABBREVIATIONS ...... xi PREFACE ...... xiv CHAPTER I: INTRODUCTION ...... 1 1.1 What is Amyotrophic lateral sclerosis? ...... 1 1.2 What genetics factors cause ALS? ...... 3 1.3 ALS and frontotemporal lobar dementia ...... 4 1.4 What cellular and molecular mechanisms underlie ALS? ...... 5 1.5 How to identify gene mutations that cause FALS? ...... 14 1.6 Discovery of ALS-linked mutations in the PFN1 gene ...... 16 1.7 What is Profilin 1? ...... 22 Functions of profilin 1 ...... 23 1.8 What does profilin 1 do in neurons? ...... 28 Neuronal localization of profilin ...... 28 Role of profilin in axonal outgrowth ...... 28 Role of profilin in postsynaptic actin network ...... 29 Profilin 1 vs. profilin 2 ...... 30 1.9 Profilin and neurodegeneration ...... 31 Spinal muscular atrophy (SMA) ...... 31 Huntington’s disease ...... 32 ALS ...... 32 1.10 Summary ...... 33 PREFACE TO CHAPTER II: ...... 35 CHAPTER II: CHARACTERIZATION OF ALS-LINKED PFN1 MUTATIONS .... 36 IN VITRO ...... 36 2.1 Introduction ...... 36 2.2 Results ...... 38 Mutant PFN1 form insoluble aggregates...... 38 Mutant PFN1 form ubiquitinated aggregates in cultured cells...... 40 Proteasome inhibition enhances aggregate formation by mutant PFN1...... 40 PFN1 aggregates sequester TDP-43...... 45 PFN1 aggregates did not sequester wild-type PFN1...... 48

vi Mutant PFN1 have reduced affinity to actin...... 50 Mutant PFN1 lead to collapsed growth cones and reduced F-actin/G-actin ratio in primary motor neurons...... 51 Mutant PFN1 lead to shortened axon outgrowth in primary motor neurons...... 52 Mutant PFN1 do not increase cell death in HEK293 cells...... 52 2.3 Discussion ...... 57 2.4 Acknowledgements ...... 62 2.5 Material and methods ...... 62 PREFACE TO CHAPTER III: ...... 68 CHAPTER III: FUNCTIONAL DISSECTION OF DROSOPHILA MODELS EXPRESSING ALS-LINKED PFN1 MUTAIONS ...... 69 3.1 Introduction ...... 69 3.2 Results ...... 73 Human PFN1 rescues lethality of chickadee knockdown in motor neurons and glia...... 73 PFN1 overexpression increases protrusion and ghost bouton formation in larval neuromuscular junctions...... 77 M114T PFN1 has a loss-of-function effect on F-actin formation in larval NMJ...... 81 PFN1 overexpression reduces basal number of satellite boutons and depletes synaptic vesicles ...... 84 PFN1 overexpression leads to adult locomotion defects and reduces lifespan ...... 87 PFN1 overexpression results in adult leg NMJ degeneration ...... 93 Superoxide dismutase (SOD) genetically interacts with PFN1 in locomotion and NMJ morphology ...... 95 Mutant PFN1 do not form visible aggregates in this Drosophila model ...... 97 3.3 Discussion ...... 103 3.4 Acknowledgements ...... 111 3.5 Material and methods ...... 111 CHAPTER IV: GENERAL DISCUSSION AND FUTURE DIRECTION ...... 117 4.1 Propensity to aggregate of ALS-linked PFN1 mutants ...... 119 4.2 Loss of function of ALS-linked PFN1 mutations in actin polymerization .... 123 4.3 Gain of function of ALS-linked PFN1 mutations in neuronal integrity ...... 126 4.4 PFN 1 level regulates adult locomotion and lifespan ...... 127 4.5 Non-cell autonomous role of PFN1 in neuronal integrity ...... 129 4.6 Advantages of Drosophila models expressing human PFN1 ...... 130 Identification of genetic interaction between SOD1 and PFN1 ...... 130 To identify novel cellular pathways regulated by wild-type PFN1 ...... 133 To uncover loss-of-function or gain-of-function mechanisms of mutant PFN1 ..... 134 BIBLIOGRAPHY ...... 136

vii LIST OF TABLES

Chapter I Table I-1 – Genetic factors that cause FALS Table I-2 – FALS-linked genes are involved in various functional pathways Table I-3 – Genotype analysis of PFN1 mutants in ALS and control populations Table I-4 – PLP ligands interacting with PFN1

Chapter III Table III-1 – Summary of phenotypes of PFN1 overexpression

Chapter IV Table IV-1 – Comparison of phenotypes of ALS-linked PFN1 mutations in different models

LIST OF FIGURES

Chapter I Figure I-1 – Exome sequencing identifies PFN1 gene mutations in FALS families Figure I-2 – The evolutionary conservation of PFN1 mutations Figure I-3 – PFN1 promotes actin polymerization Figure I-4 – Functional motifs of PFN1

Chapter II Figure II-1 – Mutant PFN1 trans-localize in insoluble fractions Figure II-2 – Mutant PFN1 produce ubiquitinated aggregates Figure II-3 – Proteasome inhibition enhances mutant PFN1 aggregates Figure II-4 – PFN1 aggregates sequester TDP-43 Figure II-5 – TDP-43 aggregates do not co-localize with wild-type PFN1 Figure II-6 – Mutant PFN1 aggregates do not sequester wild-type PFN1

viii Figure II-7 – Mutant PFN1 reduce affinity to actin Figure II-8 – Mutant PFN1 reduce growth cone size and F-/G-actin expression Figure II-9 – Mutant PFN1 inhibit axon outgrowth

Chapter III Figure III-1 – Generation of Drosophila models expressing human PFN1 Figure III-2 – Human PFN1 rescue pupal death caused by RNAi knockdown of chickadee in motor neurons and glia Figure III-3 – PFN1 overexpression induces formation of protrusions and ghost boutons without altering major bouton morphology in larval NMJ Figure III-4 – PFN1 overexpression in motor neurons does not cause synaptic retraction or crawling defects in 3rd instar larval stage Figure III-5 – PFN1 M114T induces fewer F-actin rich filopodia than PFN1 WT in larval NMJ Figure III-6 – PFN1 overexpression increases F-actin patches within the whole synapse in larval NMJ Figure III-7 – PFN1 overexpression leads to reduction in synaptic vesicles and satellite bouton number in larval NMJ Figure III-8 – PFN1 overexpression in motor neurons leads to adult locomotion defects and shorter lifespan Figure III-9 – Chickadee overexpression leads to similar locomotion defects and shorter lifespan compared to PFN1 Figure III-10 – PFN1- induced toxicity is cell-type specific Figure III-11 – PFN1 overexpression leads to NMJ disorganization in adult legs. Figure III-12 – Amorphic SOD mutants enhance PFN1- induced toxicity Figure III-13 – PFN1 overexpression and SODn1 mutant combine to promote NMJ overgrowth Figure III-14 – ALS-linked PFN1 mutants do not form aggregates in Drosophila tissue

ix LIST OF THIRD PARTY COPYRIGHTED MATERIAL

The following material was reproduced from published work:

Publisher License number Table I-2 Nature Publishing Group 3757810686690 Table I-4 Elsevier 3757810109442

Figure I-3 Elsevier 3757810109442

x LIST OF ABBREVIATIONS

ALS – Amyotrophic lateral sclerosis

AMPA – α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid ANG – Angiogenin ATXN2 – Ataxin 2 BRP – Bruchpilot protein C9ORF72 – Chromosome 9 open reading frame 72 ChAT – Choline acetyltransferase CHMP2B – Charged multivesicular body protein 2B CREST – Calcium-responsive transactivator CRISPR – clustered regulatory interspaced short palindromic repeats CSP – Cysteine-string-protein Dlg – Disc-large DAO – D-amino-acid oxidase dbSNP132 – Single nucleotide polymorphism database, build 132 DCTN1 – Dynactin DGluR III – Drosophila glutamate receptor subunit III d.p.e – Days post-eclosion eIF2α – Eukaryotic translation initiation factor 2α GFP – Green fluorescent protein GMR – Glass multiple reporter EWSR1 – Ewing sarcoma RNA-binding protein 1 FALS – Familial amyotrophic lateral sclerosis FTLD – Frontotemporal lobar dementia FUS or FUS/TLS – Fused in sarcoma/Translocated in liposarcoma HEK293 – Human embryonic kidney cells HL 3.1 – Hemolymph-like solution hnRNP – Heterogeneous ribonuclear protein

xi HRP – Horseradish peroxidase iPSCs – Induced pluripotent stem cells

JNK – c-Jun N-terminal protein kinases kDa – Kilo dalton LDH – Lactate dehydrogenase MATR3 – Matrin 3 mDia1 – Mouse diaphanous-related formin 1 MEM – Minimum essential media Mena – Mammalian enabled

MAPK – Mitogen-activated protein kinases N2A – Neuroblastoma-2A cells NEFH – Neurofilament, heavy polypeptide

NMDA – N-methyl-D-aspartate NMJ – Neuromuscular junction NP-40 – Nonyl phenoxypolyethoxylethanol-40 NSC-34 – Neuroblastoma x spinal cord hybrid 34 cells OPTN – Optineurin PBS – Phosphate-buffered saline PC12 – Pheochromocytoma cells PFA – Paraformaldehyde PFN1 – Profilin 1 PFY – Yeast profilin

PI(4,5)P2 – Phosphatidylinositide-4, 5-bisphosphate PLP – Poly-L-proline PMN – Primary motor neurons PRPH – Peripherin PSD – Postsynaptic density ROCK – Rho-kinase ROS – Reactive oxygen species

xii SALS – Sporadic amyotrophic lateral sclerosis SDS – PAGE -– Sodium dodecyl sulfate – Polyacrylamide gel electrophoresis SETX – Senataxin SMA – Spinal muscular atrophy SMN – Survival of motoneuron SNP – Single nucleotide polymorphism SOD1 – Cu/Zn dependent superoxide dismutase SOD2 – Mg dependent Superoxide dismutase SPG11 – spastic paraplegia 11 SQSTM1 – Sequestosome-1 TAF15 – TATA box binding protein associated factor 68 kDa TDP-43 or TARDBP – Transactive response DNA binding protein 43 TUBA4A –Tubulin α-4A UAS – Upstream activation sequence UBA – Ubiquitin-associated domain UBIQLN2 – Ubiquilin 2 UPR – Unfolded protein response VAPB – Vesicle-associated membrane protein VASP – Vasodilator-stimulated phosphoprotein VCP – Valosin containing protein WASP – Wiskott-Aldrich syndrome protein WAVE – WASP-family verprolin-homologous protein Wnt/Wg – Wingless gene WIP – WASP-interacting protein

xiii PREFACE

Parts of this dissertation appeared in:

Wu CH, Fallini C, Ticozzi N, Keagle PJ, Sapp PC, Piotrowska K, Lowe P, Koppers M, McKenna-Yasek D, Baron DM, Kost JE, Gonzalez-Perez P, Fox AD, Adams J, Taroni F, Tiloca C, Leclerc AL, Chafe SC, Mangroo D, Moore MJ, Zitzewitz JA, Xu ZS, van den Berg LH, Glass JD, Siciliano G, Cirulli ET, Goldstein DB, Salachas F, Meininger V, Rossoll W, Ratti A, Gellera C, Bosco DA, Bassell GJ, Silani V, Drory VE, Brown RH Jr, Landers JE. Mutations in the profilin 1 gene cause familial amyotrophic lateral sclerosis. Nature. 2012 Aug 23; 488(7412): 499-503.

xiv CHAPTER I: INTRODUCTION

1.1 What is Amyotrophic lateral sclerosis?

Amyotrophic lateral sclerosis (ALS), also known as Lou Gehrig’s disease,

is a debilitating neurodegenerative disease that selectively causes motor neuron

death in both brain and spinal cord and leads to weakness of limb and bulbar

muscles (reviewed in Wijesejera and Leign, 2009). The paralysis often starts

focally and spreads throughout different regions, resulting in impaired critical functions such as swallowing, breathing, and speaking. The average age of onset is around 50 years, and juvenile ALS is uncommon. Patients usually die within 3 to 5 years after initial symptoms, although the clinical course and

disease severity can vary, with some patients surviving for decades.

Compared with other common neurodegenerative diseases, ALS is often

considered an “orphan disease”. About 2 per 100,000 new cases are diagnosed

each year, and its prevalence is about 5 per 100,000 total cases each year,

making it the most common neuromuscular disease worldwide (reviewed in

Peters et at, 2014). Currently there is no effective therapy for this disease, and

most treatments involve symptom relief and supportive care. The only FDA-

approved drug, riluzole, unsatisfactorily prolongs survival by only 3 to 5 months

(Bensimon, 1994; Lacomblez et al, 1996; Ludolph and Jesse, 2009). Therefore,

the progressive character of ALS and lack of effective treatment require the

1 awareness of the public, and make it an important disease for scientists to

investigate.

Histo-pathologically, protein aggregates serve as a hallmark in post- mortem brain tissues of ALS patients (Kanning et al 2010). These aggregates, which localize mostly in the cytoplasm of affected motor neurons, are ubiquitinated and formed by misfolded proteins. In many cases, they contain phosphorylated TAR DNA binding protein 43 (TDP-43) (Neumann et al, 2006).

Interestingly, mutations in TDP-43 have also been identified as causative for ALS

(Sreedharan et al, 2008). Apart from aggregates, loss of motor neurons in affected regions can also be detected, which is usually accompanied by gliosis

(Wijesejera and Leign, 2009).

Peripherally, neuromuscular junction denervation can also be found in biopsies of muscles from ALS patients (Fischer et al, 2004; Bruneteau et al,

2015). Correspondingly, electrophysiological changes in electromyographic measurements are early signs in the course of the disease, indicative of the development of early distal axonopathy (Dengler et al, 1990; Frey et al, 2000; de

Carvalho et al, 2008). In fact, a “dying-back” mechanism that views ALS as an axonopathy has been suggested in recent years (Fischer et al, 2004, reviewed in

Moloney et al, 2014). For instance, in ALS mouse models, it has been shown that axonal retraction occurs before loss of neuronal cell bodies in the spinal cord, suggesting degenerating programs may start distally in synaptic regions and progress toward cell bodies, eventually leading to cell death (Fischer et al, 2004).

2 Furthermore, in ALS mouse model, blockage of cell apoptosis pathway prevents

motor neuron death but not neuromuscular denervation, providing another line of

evidence that neuromuscular denervation results from local degenerating

molecular programs, rather than being a secondary effect of dying cell bodies

(Gould et al, 2006). On the other hand, considering most ALS-associated mutant

proteins have high propensity to aggregate and may impair neurons in multiple

aspects, it is also likely that neuromuscular denervation is a general sign of

neuronal dysfunctions (both in cell bodies and in axons). In all, while it is still not

fully understood whether neuromuscular denervation is a primary cause of ALS

or secondary to other pathogenic mechanisms, these studies highlight the

importance of NMJ integrity in ALS pathogenesis.

1.2 What genetics factors cause ALS?

Approximately 90% of ALS cases are sporadic (SALS), and 10% are

inherited (familial ALS, FALS) (reviewed in Renton et al, 2014). FALS is mostly autosomal dominant and caused by mutations in heterogeneous genes. About 20 years ago, by using linkage analysis, SOD1 was identified as the first gene that harbored missense mutations causing FALS (Rosen et al, 1993). SOD1 encodes superoxide dismutase 1, a protein involved in reducing reactive oxygen species, and up to date over a hundred mutations in SOD1 have been reported

(Andersen, 2006). Since the discovery of SOD1 mutations, more genes that are involved in different cellular pathways have also been identified that contain

3 FALS-linked mutations (reviewed in Renton et al, 2014), e.g. TARDBP

(Sreedharan et al, 2008), FUS (Kwiatkowski et al, 2009; Vance at al, 2009), VCP

(Johnson et al, 2010), and UBQLN2 (Deng et al, 2011). Most of these mutations reside within coding regions that cause missense or nonsense alterations of amino acid sequences. However, several groups identified a hexanucleotide

GGGGCC repeat expansion in the first intron of C9ORF72 (DeJesus-Hernandez et al, 2011; Renton, 2011; Glijselinck et al, 2012), which were estimated to account for about 40% of FALS cases. This finding thus stresses the relevance of alterations in noncoding genome to ALS genetics. Nonetheless, despite these progresses in gene discovery, it is still unknown what genetic factors contribute to approximately one third of FALS patients (Table I-1). Notably, de novo mutations in genes involved in FALS, such as SOD1, FUS, and C9ORF72, are also reported in SALS patients (reviewed in Renton, 2014). This highlights the importance of identification of novel gene mutations in FALS, which will help to shed light on common cellular pathways that underlie pathogenesis of both sporadic and familial ALS.

1.3 ALS and frontotemporal lobar dementia

Frontotemporal lobar dementia (FTLD) is the second most common type of dementia with onset of disease under 60 years of age. The clinical characteristics of FTLD include behavioral changes and language impairment, e.g. aphasia, which are accompanied by atrophy of frontal and temporal lobes.

4 Protein inclusions containing TDP-43 can often be found in cortical neurons in

these affected regions, reminiscent of those found in motor neurons in ALS

(reviewed in Rademakers et al, 2012). Interestingly, ALS sometimes develops

coincidentally with FTLD; about 15% of ALS patients developed FTLD (Ringholz

et al, 2005; Wheaton et al, 2007) and similarly, 15% of FTLD patients have

clinical features of ALS (Ringholz et al, 2005). Therefore, these two diseases are

often regarded as two ends of the spectrum of a single disorder. In support of this

view, FALS-associated genes have been shown to contribute to FTLD to varied

degrees. For instance, while mutations in TDP-43, FUS and C9orf72 are also

found in FTLD cases, SOD1 is strictly an ALS-associated gene (Table I-2; reviewed in Robberecht and Philips, 2013). This variability emphasizes the heterogeneity of ALS-FTLD pathogenesis.

1.4 What cellular and molecular mechanisms underlie ALS?

The heterogeneity of ALS-linked gene mutations also implicates the diversity of molecular mechanisms that contribute to neurodegeneration (Table I-

2, reviewed in Robberecht and Philips, 2013; Peters et al, 2015). Over the past decades, a variety of distinct, but not mutually exclusive, cellular pathways have been reported to be dysfunctional or impaired in ALS models. The accountable mechanisms have included protein toxicity, aberrant RNA processing, cytoskeletal disruption, altered oxidative stress and mitochondria damage,

5 excitotoxicity, and non-cell autonomous toxicity. Mechanisms related to this dissertation are discussed below in more detail:

Proteostasis and proteinopathy

ALS-associated mutant proteins are often misfolded and subject to different posttranslational modifications, e.g. ubiquitination (reviewed in

Robberecht and Philips, 2013; Peters et al, 2015). Their clearance relies on two

protein degradation pathways, proteasome and autophagy (reviewed in

Tyedmers et al, 2010). Misfolded mutant proteins first form oligomeric species and then accumulate into aggregates. They alter proteostasis possibly through overwhelming the protein degradation machineries or directly inhibiting functions

of components in the machineries. On the other hand, mutations have been reported in several components of ubiquitin-protesome system: VCP (Johnson et

al, 2010) UBQLN2 (Deng et al. 2011) SQSTM1 (Fecto et al, 2011) and

autophagy: CHMP2B (Parkinson et al, 2006), OPTN (Maruyama et al, 2010). For

instance, ubiquilin 2 belongs to the ubiquitin-like protein family, and transports ubiquitinated proteins to the proteasome for degradation. Several mutations are identified in the gene UBQLN2 in FALS patients, and they reside in the UBA domain where ubiquilin 2 binds ubiquitinated proteins (Deng et al. 2011).

Although it needs to be clarified how these mutations affects functions of

proteasome pathways, it indicates the importance of protein quality control in the pathogenesis of ALS.

Apart from compromising ubiquitin-proteasome system, ALS-linked mutations can cause protein toxicity through several other mechanisms. Firstly, the accumulation of misfolded proteins causes endoplasmic reticular (ER) stress,

9 triggers unfolded protein response (UPR), and eventually leads to cellular apoptosis (reviewed in Matus et al, 2013). Secondly, the presence of aggregates may pose a toxic effect by sequestering other cellular factors, thus impairing their functions. For instance, Bonetto and colleagues utilized proteomic approach to characterize detergent-insoluble proteins in spinal cord tissues from SOD1 G93A mouse models and ALS patients (Basso et al, 2009). They report that cellular factors important in different pathways are highly enriched in insoluble fractions from ALS models. This enrichment includes mitochondrial proteins, cytoskeletal elements, chaperones, and enzymes involved in energy metabolism, which are often abnormally tyrosine-nitrated. While it remains unknown as to the functional impacts on these sequestered proteins, their study lends support to the hypothesis that ALS-associated protein aggregates may cause protein toxicity through this gain-of-function mechanism.

RNA processing

Up to date, many RNA/DNA binding proteins have been found to associate with ALS (reviewed in Renton et al, 2014 and Peters at al, 2015). TDP-

43 and FUS are two most common ALS-linked RNA binding proteins, mutations of which account for about 4% of FALS cases. These RNA interacting proteins mainly localize in the nucleus, and are involved in almost every step of RNA processing, from transcription to mRNA splicing to micro-RNA processing.

Mutant forms of these proteins are frequently translocated into the cytoplasm

10 (Kwiatkowski et al, 2009; Mutilac et al, 2015), raising the possibility of a loss-of- function effect in the nucleus. However, it is also likely that these mislocalized proteins acquire toxic property in the cytoplasm, thereby causing a gain-of- function outcome (reviewed in Da Cruz and Cleveland, 2011; Lee et al, 2011).

How ALS-linked mutations affect functions of these RNA-interacting proteins is still inconclusive, but there is increasing evidence showing that both mechanisms can contribute to neurodegeneration. For instance, ALS-linked TDP-43 mutants have been shown to alter splicing of over 1,000 RNA transcripts in rodent models

(Arnold et al, 2013). On the cytoplasmic side, the role of stress granules in the pathogenesis of ALS has been increasingly investigated in recent years. Stress granule is a type of ribonucleoprotein granules that assemble under cellular

stress and sequester mRNA to maintain RNA homeostasis (Li et al, 2013). TDP-

43 has been detected in stress granules, and its ALS-linked mutants can alter stress dynamics (Liu-Yesucevitz et al, 2010; McDonald et al, 2011). Likewise, mutant forms of FUS have been shown to trans-locate to the cytoplasm, target to

stress granules and alter their dynamics (Bosco et al, 2010; Baron et al, 2013).

Taken together, these observations indicate that ALS-linked mutations can impair

RNA processing at different stages and disrupt RNA homeostasis, particularly during cellular stress.

11 Oxidative stress

Oxidative stress has long been considered a contributing factor to the pathogenesis of ALS (reviewed in Barber et al, 2006). Clinically, increased oxidative stress has also been detected in ALS patients (Nagase et al, 2015).

The first ALS-linked gene identified, SOD1, encodes superoxide dismutase 1, an enzyme accountable for reducing reactive oxygen species (ROS) (Fridovich,

1986). Although most SOD1 mutations do not alter its enzyme activity (reviewed in Sau et al, 2007), in line with the clinical data, evidence of increased oxidative stress has been observed in rodent models that express ALS-linked SOD1 mutations (Ferrante et al, 1997; Andrus et al, 1998; Liu et al, 1998). How these mutant proteins trigger oxidative stress is still unknown. One of the enticing explanations is that misfolded proteins can associate with mitochondria, the major site of ROS production, and cause damage to the organelle (reviewed in

Barber et al, 2006). Indeed, abnormal neuronal mitochondria morphologies have been reported in various ALS models including motor neuron-like NSC-43 cell lines expressing mutant SOD1 (Menzies et al, 2002). Increased oxidative stress may further initiate a vicious cycle by damaging membrane integrity of mitochondria and altering protein conformation (reviewed in Barber et al, 2006).

For instance, Bosco and colleagues have demonstrated oxidation of wild-type

SOD1 can trigger misfolding of the protein, resulting in an aberrant conformation that mimics mutant SOD1 (Bosco et al, 2010). Thus altered oxidative stress can

12 be associated with both the initiation and aggravation of neurodegeneration in

ALS.

Cytoskeletal dysfunction and axonal failure

Cytoskeleton serves as fundamental scaffold that supports axonal and

synaptic structures and functions. In particular, axonal transport is one of major

neuronal functions that heavily rely on cytoskeleton, and disrupted axon transport involving various cargos, such as mitochondria (De Vos et al, 2007), ChAT

(Tateno et al, 2009) and mRNA granules (Alami et al, 2014), has been reported in ALS models harboring SOD1 and TDP-43 mutations. Mutations of cytoskeletal

components have also been identified as causative for FALS. For instance,

dynactin is a protein that links cargos to the motor protein dynein, thereby

facilitating microtubule-dependent retrograde axonal transport. Mutations in

DCTN1, which encodes dynactin, have been reported to account for less than

1% of FALS (Puls et al, 2003). In rodent models, these mutant dynactin have led

to neuromuscular denervation and accumulation of vesicles in cell bodies of

motor neurons, indicative of disrupted axonal transport (Laird et al, 2008).

Mutations in genes encoding other cytoskeletal components including neurofilament heavy chain (Figlewicz et al, 1994; Al-chalabi et al, 1999) and peripherin (Corrado et al, 2011) have also been found in rare FALS cases. Apart from that, our genetic study discovered ALS-linked mutations in the PFN1 gene

(discussed in more detail in Section I.6). Profilin 1 is an actin-binding protein that

13 promotes actin polymerization, and throughout this dissertation, we will discuss the functional impacts of ALS-linked PFN1 mutations. In all, these findings provide strong evidence that cytoskeletal structures and functions are closely linked to ALS pathogenesis, and how their disruption leads to neurodegeneration have just started being investigated intently.

1.5 How to identify gene mutations that cause FALS?

As mentioned previously, discovery of novel gene mutations that cause

FALS will pave the road to decipher common pathogenic mechanisms that underlie this devastating disease. Traditionally, linkage analysis was a standard method to identify gene mutations in Mendelian neurological diseases, but it has several limitations (reviewed in Bras et al, 2012). Firstly, it requires large multi- generational pedigrees that contain available affected and unaffected family members. This usually presents difficulties in the case of rare late-onset neurodegenerative diseases such as ALS, for which the source of complete pedigrees may be limited. Secondly, complete penetrance of the disease is not always observed. Lastly, the genomic regions that linkage analysis identifies usually consist of hundreds of genes, which still require follow-up Sanger sequencing to precisely identify the actual mutation sites.

With the advent of next-generation sequencing, many alternative methods of gene discovery have been introduced in order to uncover rare disease-linked

mutations. In particular, exome sequencing has gained increasing popularity as a

14 cost-effective approach to identify gene mutations causative for rare neurological diseases including ALS (reviewed in Singleton, 2011; Bras et al, 2012). In principle, amino-acid coding sequences of genomic DNA are captured by arrays that contain probes complementary to approximately 90% of human exome. The captured DNAs then undergo next-generation sequencing and are compared among affected and unaffected samples. Because most ALS associated mutations reside within coding regions of the genome, this approach is less expensive than whole-genome sequencing and proves effective to identify rare disease-causing variants. However, exome sequencing does have its limitations.

Since it only targets coding regions, it will not detect potential disease-linked

mutations in non-coding regions that may be important in genetic and epigenetic

regulation. Similarly, not all genes are captured and analyzed properly due to

sensitivity of the arrays used and intrinsic property of the coding sequences, with

those of high GC content being more difficult to amplify. Lastly, most variants that

are identified in an individual sample are benign single nucleotide polymorphisms

(SNPs). Therefore, further filtering process is needed to confirm real disease-

linked mutations. Despite these limitations, in the past few years, several novel

ALS-causing gene mutations, such as VCP (Johnson et al, 2010) and Matrin 3

(Johnson et al, 2014), were discovered using this approach, supporting exome

sequencing as a useful method to investigate ALS genetics.

15 1.6 Discovery of ALS-linked mutations in the PFN1 gene

To identify causative genes for FALS, we performed exome capture followed by deep sequencing on two large ALS families of Caucasian (family 1) and Sephardic Jewish (family 2) origin (Figure I-1A, B). Both displayed a dominant inheritance mode and were negative for known ALS-causing mutations, including the hexanucleotide repeat expansion in C9orf72 gene. For each family, two affected members with maximum genetic distance were selected for exome sequencing. To identify candidate causative mutations, variants were identified and filtered, as in previous exome sequencing reports (Johnson et al, 2010), using several criteria: the variant is observed in both family members, alters the amino acid sequence and is not excluded by linkage analysis. The variant is

absent from current SNP database including dbSNP132, the 1000 Genomic

Project (May 2011 release) or the National Heart, Lung, and Blood Institute

(NHLBI) Exome Sequencing Project (ESP) Exome Variant Server (5,379

sequenced exomes; http://evs.gs.washington.edu/EVS/). Remaining variants

were confirmed by Sanger sequencing and tested for Mendelian segregation in

all affected family members. The resulting number of candidate causative

mutations identified was two within family 1 and three within family 2.

Interestingly, the two families contain different mutations (C71G and M114T)

within a single common gene: PFN1, located on chromosome 17p13.2. PFN1 is

a 140-amino-acid protein and major growth regulator of filamentous (F)-actin

through its binding of monomeric (G)-actin (discussed in later sections).

16 To determine whether PFN1 mutations cause familial ALS, the coding region was sequenced in a panel of 272 further FALS cases prescreened for common causative mutations. Five other familial ALS cases containing alterations in the PFN1 gene were identified. Interestingly, the C71G alteration originally identified in family 1 was discovered in two other families. For one of these families (family 3) (Figure I-1C), DNA was available for three other affected family members. Sequencing of these samples showed that the mutation co-segregated with ALS. A second M114T mutation was identified in an

ALS family of Italian origin (family 4). DNA available for one sibling was shown by sequencing to contain this mutation. PFN1 variants were observed in two other

FALS cases: a consecutive base-pair change (AA to GT) resulting in an E117G mutation, and a G-to-T transversion resulting in a G118V mutation. DNA was not available from other family members for these cases. Sequencing of the PFN1 coding region in 816 sporadic ALS (SALS) samples identified two samples containing the E117G mutation.

To confirm further that the newly identified variants (E117G and G118V) represent causal mutations, not benign polymorphisms, each was interrogated in the 1000 Genomes Project database and the NHLBI ESP Exome Variant Server.

The G118V mutation was not identified in either database. However, E117G variant was observed in 2 samples out of 5,379 at the NHLBI ESP Exome

Variant Server. We extended this analysis by genotyping all mutations in an independent set of 1,089 control samples. Three of the mutations (C71G, M114T

17 and G118V) were not observed in the 1,089 control samples; however, the

E117G variant was observed in a single control. By combining data from the

1000 Genome Project, the NHLBI ESP Exome Variant Server and independent

genotyping, three of the ALS-linked variants were not present in 7,560 control samples, whereas the fourth (E117G) was found in 3 out of 1,090 ALS cases and in 3 out of 7,560 control samples. We hypothesized that the E117G variant is

either non-pathogenic or, more likely, represents a less pathogenic mutation.

Extending our findings, Shaw and colleagues screened 383 more ALS patients

and concluded that E117G is a moderate risk factor for ALS (Fratta et al, 2014).

Furthermore, another group later reported 2 more novel mutations (A20T and

Q139L) by sequencing of a cohort of 485 ALS patients (Smith and Vance et al,

2015).

In total, we identified 4 mutations in 7 out of 274 FALS cases (Table I-3).

In each case, the altered amino acid was evolutionarily conserved down to the

level of zebrafish, supporting the possibility that these mutations are pathogenic

(Figure I-2). The age of onset for FALS cases with PFN1 mutations is 44.8 ± 7.4.

All PFN1 mutant cases displayed limb onset and no bulbar onset were observed,

suggesting a common clinical phenotype among patients with PFN1 mutations.

Taken together, our findings, along with others, suggest that mutations in

the PFN1 gene cause about 1 % of FALS cases.

1.7 What is Profilin 1?

Profilin 1 is a 140-amino-acid protein that belongs to the profilin family

(Kwiatkowski and Bruns, 1988). It is mostly localized in the cytoplasm, although

nuclear profilin 1 has been reported with elusive roles (reviewed in Witke, 2004).

PFN1 is an essential gene, and its knockout in mice leads to early embryonic

death (Witke, 2001). In mammals, there are four profilin isoforms that have been

identified (reviewed in Birbach, 2008). Profilin 1 is expressed ubiquitously except

in skeletal muscles, while the expression of profilin 2 is restricted to the central

nervous system. Profilin 3 and 4 have been detected only in testis. Although

profilin is conserved during evolution and their homologs and orthologs can be

found in different species, homology of their primary sequences is remarkably

low (Witke, 2004). Even different profilin isoforms within the same species are

not highly identical (e.g. human profilin 2 is approximately 65% identical to profilin

1). However, crystal structures of profilin resolved from different isoforms and species are largely superimposable (Metzler, 1995; Nodelman et al, 1999), indicating that conservation of profilin structures may be more critical in maintaining their functions than similarity of primary sequences. To further support this concept, Gitler and colleagues demonstrate that human profilin 1

rescues toxic phenotype of yeast PFY null mutants despite their low sequence

conservation (Figley et al, 2014).

22 Functions of profilin 1

Profilin 1 is an actin binding protein, and its major function is to promote actin polymerization by recruiting monomeric globular (G) -actin to growing ends of polymeric filamentous (F) –actin (Figure I-3) (Witke, 2004). The facilitation of

this process can be spontaneous or mediated by other actin-associated factors

such as formin (Romero et al, 2004). Profilin 1 is also thought to facilitate the

ADP-ATP exchange of G-actin, thus increasing G-actin pools that are readily

available for F-actin incorporation (Mockrin and Korin, 1980). On the other hand,

protein levels of profilin 1 seem to be critical in regulation of actin polymerization,

as in vitro studies show that at high concentration, profilin 1 can inhibit actin

polymerization by sequestering G-actin (Polland and Cooper, 1984).

Apart from actin binding, profilin 1 binds phosphatidylinositide-4, 5-

bisphosphate (PI(4,5)P2) (Lassing et al, 1988; Lu et al, 1996; Lambrechts, 1997), suggesting it may be involved in phospholipid-mediated signaling pathway.

Likewise, Profilin 1 also interacts with protein ligands that contain poly-L-proline

(PLP) rich stretches (Lindberg et al, 1988; Lambrechts, 1997; Mahoney et al,

1999). Structurally, the PLP binding motif of profilin 1 consists of its N- and C- terminal helices and is separated from its actin binding domain by a central β- sheet, while PI(4,5)P2 binding motif partially overlaps with the actin binding domain (Figure I-4). Up to date, more than 50 cellular ligands have been reported to contain PLP rich stretches and to interact with profilin 1 (Table I-4).

They are involved in various actin-dependent and actin-independent cellular

23 pathways. Most of these interactions were reported in vitro, and their

physiological roles have yet to be elucidated. Nevertheless, the diversity of this

interaction profile indicates that profilin 1 has multi-functional roles in regulating different cellular pathways.

1.8 What does profilin 1 do in neurons?

Neuronal localization of profilin

Actin cytoskeleton provides a crucial scaffolding network in both

presynaptic and postsynaptic neurons (reviewed in Cingolani and Goda, 2008).

For instance, it contributes to the maintenance of synaptic vesicle machinery

(presynaptically) and distribution of neurotransmitter receptors (post-

synaptically). Given its association with actin, the role of profilin in the regulation

of neuronal cytoskeleton has long been discussed. Studies have demonstrated

that profilin 1 is present in both presynaptic (Faivre-Sarrailh et al, 1993) and postsynaptic regions (Giesemann et al, 2003). It interacts with corresponding scaffolding proteins including presynaptic aczonin (Wang et al, 1999) and postsynaptic gephyrin and drebrin (Mammoto et al, 1998), thus implying its involvement in synaptic actin network.

Role of profilin in axonal outgrowth

Furthermore, in drosophila embryo, null mutants of profilin leads to arrested axonal outgrowth in motor neurons (Wills et al, 1999). Similarly, in mice,

Mammalian enabled (Mena), a PLP rich ligand, is shown to interact with profilin

1, and Mena deficiency leads to defects in axonal pathfinding at embryonic stage. This possibly acts through regulation of profilin 1 because Mena-deficient mice heterozygous for profilin 1 show severe defects in neurulation (Lanier et al,

28 1999). These data indicate that profilin 1 is important for correct axonal path finding during development. It has also been suggested that different binding motifs of profilin 1 affect neuritogenesis in specific ways. Ampe and colleagues proposed this notion by studying neuritogenesis in PC12 cells, a cell line derived from rate pheochromocytoma (Lambrechts et a, 2006). They found that while actin binding-defective profilin 1 mutant reduced neurite length and filopodia formation, PLP binding-defective mutants led to longer and more branched neurites. On the other hand, Dotti and colleagues showed that profilin 2 negatively regulated neurite outgrowth in rat hippocampal neurons, which was dependent on the Rho/ROCK pathway (Silva et al, 2003). Taken together, these studies emphasize the complexity in profilin-mediated neurite outgrowth.

Role of profilin in postsynaptic actin network

Several studies have also investigated postsynaptic roles of profilin.

Presynaptic glutamate and electrical stimulations induce targeting of profilin 2 to dendritic spines of hippocampal neurons (Ackermann et al, 2003). Blocking this targeting process destabilizes their structures. Likewise, potassium chloride treatment of primary hippocampal neurons leads to a similar effect by recruiting profilin 1 into dendritic spine regions (Neuhoff et al, 2005). These findings imply that profilin may regulate postsynaptic actin dynamics in an activity-controlled manner. Interestingly, it has also been shown that behavioral paradigms such as fear conditioning result in enrichment of profilin in dendritic spines in the lateral

29 amygdala, thereby altering their structures (Lamprecht et al, 2006). This study

further supports the hypothesis that distribution of profilin in dendritic spines has

regulatory effects in important neuronal functions such as learning and memory.

Profilin 1 vs. profilin 2

It is intriguing that there are two profilin isoforms, profilin 1 and 2,

coexisting in the central nervous system. Both isoforms are detected in

presynaptic, and more abundantly, postsynaptic regions (Murk et al, 2012).

However, they respond differently to activity manipulation. In cultured rat hippocampal neurons, pharmacological inhibition of NMDA receptors leads to reduction of synaptic profilin 2, while profilin 1 is spared. On the other hand, neuronal stimulation by brain derived neurotrophic factor (BDNF) increases profilin 1 level to a greater degree than profilin 2 (Murk et al, 2012).

Biochemically, mass spectrometric study of mouse brain extracts also shows that proteins interacting with profilin 1 and 2 do not fully overlap (Witke et al, 1998).

Likewise, proflin 1 is also shown to have higher affinity to actin than profilin 2

(Giesselmann et al, 1995). Functionally, in hippocampal neurons, profilin 1 cannot fully rescue dendritic abnormalities caused by profilin 2 knockdown

(Michaelsen et al, 2010). Taken together, these findings suggest that in the central nervous system, profilin 1 may have distinct functions that cannot be fully replaced by profilin 2, and vice versa.

30 1.9 Profilin and neurodegeneration

Spinal muscular atrophy (SMA)

SMA is a common autosomal recessive neurological disorder, characterized by loss of motor neurons in the spinal cord and brain stem. It is caused by deletion or missense mutations in the survival of motoneuron 1

(SMN1) gene that encodes SMN (Brzustowicz et al, 1990). SMN is present both in the nucleus and the cytoplasm. Nuclear SMN localizes within nuclear foci called “gems” (Liu et al, 1996); it is involved in multiple pathways including RNA processing and is required for normal motor neuron development. Notably, SMN protein contains a PLP rich motif (Giessemann et al, 1999) and has been shown to interact with profilin (Witke et al, 1998). In particular, profilin 2 has much higher affinity to SMN than profilin 1. They also demonstrate subcellular co-localization, and this interaction is abolished by SMA-linked missense mutations and deletions (Sharma et al, 2005). On a molecular level, knockdown of SMN leads

to hyperphosphorylation of profilin through action of Rho-kinase (ROCK), and

phospho-mimic mutant of profilin 2 reduces neurite outgrowth in PC12 cell

models (Bowerman et al, 2007). Notably, increased F-actin/G-actin ratio is also detected in primary motor neurons derived from SMA mouse models (Nöll et al,

2011). Taken together, these studies demonstrate that SMN deficiency can cause alteration of actin cytoskeleton and eventually abnormal neurite extension through dysregulation of profilin 2. It also highlights the contribution of profilin to motor neuron integrity.

31 Huntington’s disease

Huntington’s disease is another common neurodegenerative disorder

caused by polyglutamine (polyQ) expansion in the huntingtin protein (Walker,

2007). The expanded huntingtin tends to form aggregates and presents

cytotoxicity that leads to neurodegeneration (Rubinsztein, 2003). Interestingly,

huntingtin also has PLP binding motifs and is an interacting partner of profilin

(Witke et al, 1996). In cultured cells, overexpression of profilin 1 reduces

huntingtin aggregation (Bauer et al, 2009), and this effect requires its actin

binding activity (Shao et al, 2008). Hyperphosphorylation of profilin by ROCK negatively regulates both its ability to bind actin and to reduce huntingtin

aggregates (Shao et al, 2008). On the other hand, another study showed that

expanded polyQ tracts promote degradation of profilin through ubiquitin-

proteasome system (Burnett et al, 2008). In line with other reports, this study also

demonstrated that overexpression of profilin mitigates polyQ-induced cellular

toxicity in both cellular and Drosophila models (Burnett et al, 2008). These data imply the interaction between huntingtin and profilin has reciprocal modulatory effects. They also lend support to the hypothesis that profilin plays a regulatory role in aggregate formation in neurodegenerative diseases.

ALS

As stated in earlier sections, stress granules are important coping machinery during cellular stress (Li et al, 2013), and altered stress granule

32 dynamics have been associated with ALS (Baron et al, 2013). Notably, Gitler and

colleagues reported that profilin 1 is associated with stress granules in primary

cortical neurons and multiple cell lines, and ALS-linked profilin 1 mutants fail to

be targeted into oxidative stress-induced stress granules (Figley et al, 2014).

This observation indicates that profilin 1 may engage in maintenance of RNA

homeostasis through regulating stress granule dynamics.

1.10 Summary

ALS is a devastating neurodegenerative disease with varied gene

etiology. By using exome-sequencing methods, we identified mutations C71G,

M114T, E117G, and G118V in the PFN1 gene. How these mutations alter functions of profilin 1 and lead to neurodegeneration is still unknown. The work presented in this dissertation aims to characterize the neuronal impacts of ALS- linked PFN1 mutations and to elucidate underlying mechanisms using in vitro

cellular and in vivo Drosophila models. Chapter II investigates affinity to actin and

propensity to aggregate of ALS-linked profilin 1 mutants in mammalian cultured

cells. In this chapter, co-immunoprecipitation studies demonstrate that mutant

profilin 1 reduce affinity to actin and immunofluorescence in primary motor

neurons confirms reduced F-actin/G-actin ratio, indicative of a loss-of-function

mechanism. Molecular solubility analysis and immunofluorescence show mutant

profilin 1 form ubiquitinated aggregates, which is enhanced by proteasome

inhibition. Overexpression of mutant profilin 1 in primary motor neurons also

33 reduces axonal length and growth cone area, supporting their deleterious effects on neuronal morphologies. Chapter III introduces a novel Drosophila model overexpressing PFN1 to further characterize these mutations in vivo. In this

Chapter, immunohistochemistry of larval neuromuscular junctions (NMJs) demonstrates that profilin 1 overexpression promotes formation of filopodia without altering overall structure of NMJs. In this analysis, M114T has less filopodia formation compared to WT profilin 1. Behavioral analyses then reveal that profilin 1 overexpression leads to adult-onset locomotion defects.

Combination of immunohistochemical and behavioral approaches spanning different developmental stages corroborates the hypothesis that ALS-linked

PFN1 mutations have a loss-of-function effect on F-actin formation. Collectively, the work presented in this dissertation sheds light on largely elusive PFN1- associated ALS pathogenesis, by demonstrating aggregation of mutant profilin1 in vitro and, more importantly, a loss-of-function mechanism in actin polymerization in vitro and in vivo. The findings in this novel Drosophila model also lay the groundwork for future genetic screens to uncover other cellular pathways affected by mutant profilin 1.

34 PREFACE TO CHAPTER II:

Chi-Hong Wu performed the experiments presented in this Chapter with the following exceptions:

Claudia Fallini performed all experiments on primary motor neurons.

Immunohistochemistry of human tissues were performed at the Emory University

Neuropathology Core (with the assistance of Marla Gearing and Deborah

Cooper).

The work presented in this Chapter aimed to characterize novel PFN1 mutations causative for FALS. All data were published with the exception of Figure 6,

Figure 7C and Figure 9C.

35 CHAPTER II: CHARACTERIZATION OF ALS-LINKED PFN1 MUTATIONS

IN VITRO

2.1 Introduction

By using exome-sequencing, we identified C71G, M114T, E117G and

G118V mutations in the PFN1 gene that contribute to about 1% of FALS cases.

Profilin 1 is a small actin-binding protein of about 15 kDa. It localizes mostly in

the cytoplasm and is ubiquitously expressed except in skeletal muscles

(reviewed in Witke, 2004). The major function of profilin 1 is to promote actin polymerization by binding monomeric G-actin and delivering them to the growing ends of polymeric F-actin. In vitro studies show that proflin 1 also interacts with phosphatidylinositide-4, 5-bisphosphate (PI(4,5)P2) (Lassing et al, 1988; Lu et al,

1996; Lambrechts, 1997) and cellular ligands containing poly-L-proline (PLP) rich

stretches (Lindberg et al, 1988; Lambrechts, 1997; Mahoney et al, 1999; Witke,

2004), suggesting it functions in various cellular pathways. Profilin regulates the

neuronal cytoskeleton and is crucial in maintaining multiple aspects of neuronal

functions (reviewed in Birbach, 2008). For example, alteration of its expression

leads to arrested axonal outgrowth and incorrect axon pathfinding in embryonic

Drosophila and rodent models (Lanier et al, 1999; Wiils et al, 1999). Likewise,

profilin interacts with various scaffolding proteins in synaptic regions (Giesemann

et al, 2003), and is recruited to dendritic spines in an activity-regulated manner,

suggesting a role in the regulation of postsynaptic activities (Ackermann and

36 Matus, 2003). Witke and colleagues also demonstrated that profilin depletion

alters synaptic vesicle exocytosis, thereby affecting synaptic transmission (Pilo

Boyl et al, 2007). Nonetheless, how ALS-linked PFN1 mutations affect functions of profilin 1 on a molecular and cellular level is still unknown.

Ubiquitinated aggregates are histopathological hallmarks of ALS. These aggregates (Kanning et al, 2010), which localize in affected motor neurons, consist of misfolded proteins that harbor ALS-linked mutations. In many cases, they also contain phosphorylated TDP-43 (Neumann et al, 2006). It is unknown whether PFN1 mutations also increase its propensity to aggregate.

To investigate if mutant PFN1 form aggregates, we first transfected wild- type or mutant PFN1 in cultured mammalian cells including neuroblastoma-2A

(N2A) cell line and primary motor neurons. We then used this system to measure protein solubility by molecular solubility analysis and immunofluorescence. To elucidate if PFN1 mutations lead to a loss of function in actin polymerization, we first performed co-immunoprecipitation to test the interaction between actin and

PFN1. We then extended this experiment by transfecting primary motor neurons with wild-type or mutant PFN1 and measuring F-actin/G-actin ratio. Lastly, to test if mutant PFN1 impair neuronal morphologies, we again transfected primary motor neurons with wild-type or mutant PFN1 and measured axonal outgrowth and area of growth cones.

37 2.2 Results

Mutant PFN1 form insoluble aggregates.

To investigate whether the observed PFN1 mutants form insoluble aggregates, Western blot analysis of NP-40-soluble and insoluble fractions was performed on Neuro-2A (N2A) cells transfected with wild-type or one of the four

PFN1 mutants. PFN1 protein was present predominantly in the soluble fraction of cells transfected with the wild-type construct as compared with the insoluble fraction. Conversely, a considerable proportion of the C71G, M114T and G118V mutant proteins were detected in the insoluble fraction (Figure II-1A).

Furthermore, several higher molecular mass species were observed, indicative of

SDS-resistant PFN1 oligomers. However, the E117G mutant displayed a pattern more similar to wild-type PFN1 with most of the expressed protein in the soluble fraction. Differential expression of PFN1 constructs was ruled out by Western blot analysis of whole cell lysates (Figure II-1B). Analysis of lymphoblast cell lines derived from affected and unaffected members of the ALS family harboring C71G mutation did not display any difference in PFN1 protein solubility (Figure II-1C).

Autopsy material was not available for any affected individual.

Mutant PFN1 form ubiquitinated aggregates in cultured cells.

We extended these observations by staining the PFN1 protein in transfected cells. Wild-type PFN1 exhibited a diffuse cytoplasmic expression pattern in transfected N2A cells, as previously reported (Suetsugu et al, 1998).

By contrast, ALS-linked PFN1 mutants often assembled into cytoplasmic aggregates (Figure II-2A). In this experiment, cytoplasmic V5-staining positive granules with concentrated fluorescence intensity that constained with ubiquitin were defined as aggregates. Image analysis determined that 15-61% of mutant- expressing cells contain cytoplasmic aggregates (Figure II-2B), including the

E117G mutant, which showed minimal insoluble PFN1 protein by Western blot analysis. No aggregates were observed for cells expressing wild-type PFN1. Co- staining revealed that these aggregates were also ubiquitinated. Primary motor neurons (PMNs) expressing the C71G, M114T and G118V mutants similarly demonstrated ubiquitinated aggregates, albeit at a lower percentage; aggregates were not observed in cells expressing the E117G mutant or wild-type PFN1

(Figure II-2C). Immunoprecipitation of the PFN1 protein followed by Western blot analysis confirmed the insoluble mutant PFN1 protein is poly-ubiquitinated

(Figure II-2D).

Proteasome inhibition enhances aggregate formation by mutant PFN1.

To determine whether ubiquitin-proteasome system impairment causes accumulation of mutant PFN1 aggregates, transfected N2A cells were exposed to the proteasome inhibitor MG132. ALS-linked PFN1 mutants, including E117G,

40 showed increased insoluble protein levels and increased levels of higher

molecular mass species by Western blot analysis (Figure II-3A). Minimal insoluble protein was observed for the wild-type PFN1. PFN1 staining in N2A

cells confirmed these results (Figure II-3B). Cells expressing C71G, M114T and

G118V mutant PFN1 had numerous, large aggregates after MG132 treatment.

E117G mutants showed a moderate aggregate level, and the wild-type PFN1

displayed minimal levels.

Figure II-2. Mutant PFN1 produces ubiquitinated insoluble aggregates.

N2A cells (A) and PMNs (C) were transfected with constructs expressing either wild-type or mutant V5-PFN1. The PMNs were additionally co-transfected with a HA-ubiquitin expressing construct. The cells were then fixed and stained with antibodies directed against V5, HA (PMNs) and ubiquitin (N2A). Examples of aggregates, indicated by arrows, are enlarged in the inset in (C). Scale bar: 5 µm (A), 10 µm (C). (B) Transfected N2A cells displaying insoluble aggregates were counted using the software package MetaMorph. Comparisons were made using one-way ANOVA testing with Dunnett’s multiple test comparison. *P<0.05, ***P<0.001, n.s. P>0.05. Error bars indicate SEM. (D) Insoluble mutant PFN1 is polyubiquitinated. N2A cells were transfected with control vector (lanes 3, 4, 9,10), V5-PFN1-C71G (lanes 5, 6, 11, 12) and a HA-tagged Ubiquitin expressing vector (lanes 3-6, 9-12). The cells were treated with (lanes 2, 4, 6, 8, 10, 12) and without (lanes 1, 3, 5, 7, 9, 11) the proteasome inhibitor MG132. The cells were then lysed and NP-40 soluble (lanes 1-6) and insoluble (lanes 7-12) fractions were isolated and immunoprecipitated with a V5 antibody. The immunoprecipitates were separated by SDS-PAGE and then immunoblotted with antibodies for either V5 or HA, as indicated.

PFN1 aggregates sequester TDP-43.

Given the propensity of mutant PFN1 to form aggregates, we investigated whether other ALS-related proteins may be present within these aggregates.

Thus, we transfected PMNs with mutant PFN1 and tested for co-aggregation of the ALS-related proteins FUS and TDP-43. Furthermore, we also tested for aggregation of the SMA-related protein SMN owing to its ability to bind PFN1

(Giesemann et al, 1999). About 30-40% of cells contained cytoplasmic PFN1 aggregates co-stained with TDP-43 (Figure II-4A). However, no co-aggregation of either FUS or SMN with mutant PFN1 was observed (Figure II-4B,C). These results suggest that mutant PFN1 may contribute to ALS pathogenesis by inducing aggregation of TDP-43. On the basis of these observations, we investigated whether aggregates of TDP-43 contain PFN1 by staining spinal cord tissues from 18 SALS cases displaying TDP-43 pathology and 6 non-ALS controls without TDP-43 pathology. Abnormal PFN1 pathology was not discovered in SALS cases, suggesting that TDP-43 aggregation does not induce

PFN1 aggregation (Figure II-5A). Expression of carboxyl-terminal fragment of

TDP-43, which produces insoluble aggregates, in PMNs also failed to co- aggregate with wild-type PFN1, thus supporting this observation (Figure II-5B).

PFN1 aggregates did not sequester wild-type PFN1.

The propensity of mutant PFN1 to aggregate also raises the possibility that it may pose a gain-of-function mechanism by sequestering wild-type PFN1 into aggregates, thereby interfering with their cellular functions. To test this hypothesis, a V5-tagged wild-type PFN1 construct and a HA-tagged C71G PFN1 construct were co-transfected in N2A cells. Cellular staining showed that wild- type PFN1 remained diffuse in the cytoplasm and depleted from the aggregates formed by C71G PFN1 (Figure II-6A). To exclude the effect of different tags on aggregation, a V5-tagged C71G PFN1 construct was co-transfected with HA- tagged C71G construct. The staining result showed that C71G PFN1 tagged with

V5 and HA epitopes co-aggregated (Figure II-6B). These observations indicate that aggregates formed by mutant PFN1 do not sequester wild-type PFN1.

Mutant PFN1 have reduced affinity to actin.

Evaluation of the profilin 1-actin complex crystal structure showed that all

ALS-linked mutations lie in close proximity to the actin-binding residues of profilin

1 (Schutt et al, 1993; Figure II-7A). Therefore, we investigated whether the ALS- linked mutations display a decreased level of bound actin. Towards this end, we performed immunoprecipitation and Western blot analysis of cells transfected with wild-type and mutant PFN1. As a control, we also transfected cells with a construct expressing a synthetic profilin 1 (H120E) mutant protein. This alteration is located at a crucial residue previously shown to abolish PFN1 binding to actin

(Suetsugu et al, 1998). We observed that C71G, M114T, G118V and H120E mutants, but not the E117G, had reduced levels of bound actin relative to wild- type PFN1 (Figure II-7B). Apart from actin, we also wondered if ALS-linked

PFN1 mutations disrupted interaction with other cellular factors such as PLP ligands. We chose to test mDia1, a PLP ligand that coordinates with PFN1 to facilitate actin polymerization (Watanabe et al, 1997). Co-immunoprecipitation showed that PFN1 C71G had reduced interaction with mDia1 (Figure II-7C). This interaction did not seem to depend on affinity to actin, because PFN1 H120E, the actin binding-defective mutant, still maintained its interaction with mDia1. This result suggests that PFN1 C71G may affect actin polymerization through interfering with multiple cellular factors.

50 Mutant PFN1 lead to collapsed growth cones and reduced F-actin/G-actin ratio in

primary motor neurons.

The regulation of actin dynamics in the growth cone is necessary for axon

outgrowth. To determine whether ALS-linked PFN1 mutants have a similar phenotype, PMNs transfected with wild-type and two mutant PFN1 constructs

(C71G and G118V) were stained to detect F- and G-actin localization patterns in the highly dynamic and actin-rich growth cone (Figure II-8A). PFN1 mutant expression in PMNs led to a significantly reduced growth cone size (~43-52%) relative to wild-type PFN1 (Figure II-8B). Also, mutant PFN1 expression significantly altered growth cone morphology. In wild-type PFN1-expressing cells, growth cones had elaborate structures with several F-actin rich filopodia, whereas virtually no filopodia were visible in the mutant PFN1 growth cones.

Similar results were observed for the synthetic H120E mutant defective in actin binding. The growth cones in mutant-expressing PMNs also had a lower ratio of

F/G-actin relative to wild-type-expressing PMNs. In particular, the C71G mutant- expressing PMNs displayed an F/G-actin ratio 24.4% of the wild-type expressing

PMNs (Figure II-8C). These results suggest that mutant PFN1 are less efficient

in the conversion of G-actin to F-actin within the growth cone region, thus affecting its morphology.

51 Mutant PFN1 lead to shortened axon outgrowth in primary motor neurons.

Previous reports have shown that PFN1 alterations inhibit neurite outgrowth (Suetsugu et al, 1998; Wills et al, 1999). We thus investigated whether

ALS-linked PFN1 mutants inhibit neurite outgrowth by measuring axonal length in

PMNs transfected with wild-type or mutant PFN1. As a positive control, we also

transfected PMNs with the H120E-expressing construct. In addition to lacking

actin-binding ability, H120E protein inhibits neurite outgrowth (Suetsugu et al). As

expected, the H120E-expressing cells displayed a pronounced decrease in axon

length relative to the wild-type construct. Three ALS-linked PFN1 mutations

(C71G, M114T and G118V) also showed a significant decrease in axon

outgrowth. In particular, the G118V-associated reduction is similar to that

observed with H120E construct. Axon outgrowth inhibition was observed with the

E117G mutant but did not reach statistical significance (Figure II-9A,B). There

results suggest that mutations in PFN1 may contribute to ALS pathogenicity in

part by inhibiting axon dynamics.

Mutant PFN1 do not increase cell death in HEK293 cells.

To determine if mutant PFN1 affect cell viability, HEK293 cells (derived

from human embryonic kidney cells), were transfected with WT and mutant

PFN1, and subject to LDH membrane integrity analysis 24 hours after

transfection. Lactate dehydrogenase (LDH) is a cytosolic enzyme that is released

when integrity of cellular membrane is compromised. It can therefore be used as

52 an indicator for increased cell death. When compared with control cells, no

increase in LDH level in the medium was detected from cells transfected with

either wild-type or C71G PFN1 (Figure II-9C). This result suggests mutant PFN1 does not lead to immediate cell death under these conditions, and its lethal effects may require longer time to manifest.

2.3 Discussion

In this Chapter, we characterized functional impacts of novel ALS-linked

mutations in the PFN1 gene. Molecular solubility analysis showed that mutations

C71G, M114T and G118V led to the presence of mutant PFN1 in NP-40 resistant

fractions, while wild-type PFN1 remained in NP-40 soluble fractions.

Immunofluorescence confirmed that mutant PFN1 formed ubiquitinated

aggregates in N2A cells and primary motor neurons. Pharmacological inhibition

of proteasome further enhanced aggregate formation both in solubility test and

cellular staining. These results indicate mutant PFN1 has higher propensity to

aggregate, similar to other ALS-linked mutant proteins. Bosco and colleagues

later conducted in vitro chemical and thermal denaturation analyses of

recombinant mutant PFN1 and confirmed that ALS-linked mutations destabilize

PFN1 (Boopathy et al, 2014). X-ray crystallography of M114T PFN1 further

revealed that the destabilization is induced by a moderate structural change

(Boopathy et al, 2014). This observation thus provides further evidence that ALS- linked mutations induces PFN1 aggregation.

Contrary to other mutants, our solubility analysis and cell staining showed that E117G variant acts similar to wild-type PFN1, with minimal NP-40 insoluble

fractions and aggregate formation. Nonetheless, when proteasome was inhibited,

E117G variant resulted in more prominent aggregate formation than wild-type

PFN1. This was consistent with Bosco and colleagues’ study where E117G only

has moderate destabilizing effects compared with other ALS-linked PFN1

57 mutants (Boopathy et al, 2014). It is noteworthy that this variant was also

detected in SALS patients and controls. Therefore, these data support the notion

that E117G is an ALS-associated risk factor (Fratta et al, 2014) and manifestation of its toxicity may require secondary impairment of other cellular

factors, such as proteasome inhibition. The observation that E117G maintains

affinity to actin also raises the possibility that it may affect other actin-

independent functions.

We then determined if PFN1 aggregates sequestered other cellular factors, thereby resulting in a gain-of-function effect. We did not detect sequestration of SMN, FUS and wild-type PFN1. However, ~30-40% of cells

containing cytoplasmic PFN1 aggregates co-stained with TDP-43. We also stained spinal cord tissues from SALS patients displaying TDP-43 pathology and expressed carboxyl-terminal fragment of TDP-43, which produces insoluble aggregates, in PMNs. Both of these approaches failed to detect co-staining of

PFN1 in TDP-43 aggregates. These results suggest that TDP-43 aggregation induced by mutant PFN1 is specific and PFN1 may contribute to ALS pathogenesis via a gain-of–function mechanism by inducing aggregation of TDP-

43.

We also wondered whether ubiquitinated aggregates could be detected in

ALS patients harboring PFN1 mutations. Solubility analysis of lymphoblast cells derived from ALS patients harboring C71G mutations showed no increase in insoluble PFN1 fractions compared to controls. We argue that the inherent

58 cellular environments for proteostasis may differ in central nervous system and immune system, and a “second-hit” factor, such as oxidative stress and dysfunctional protein clearance machinery, may be required for protein aggregates to form. With the advent of inducible pluripotent stem cells and

CRSIPR genomic editing, it will be possible in the near future to confirm this cell- type specific vulnerability. We also attempted to determine if mutant PFN1 increased cell death by measuring LDH release. In this assay, no increase in

LDH levels was detected in cells transfected with mutant PFN1. Because in this experiment we used HEK293 cells, a non-neuronal cell line, to achieve higher transfection efficiency, the absence of increased cell death may still be due to differential cell-type vulnerability. Because ALS is an adult-onset progressive neurodegenerative disease, it is also plausible that the expression of mutant

PFN1 may not cause immediate cell death, but accumulate its toxicity over time.

The major function of PFN1is to promote actin polymerization and crystal structure of profilin 1-actin complex reveals that ALS-linked residues are in close proximity to actin binding sites. Therefore, we tested if ALS-linked mutations would affect actin polymerization. Co-immunoprecipitation in high-salt buffers showed that all ALS-linked variants except for E117G had reduced affinity to actin. When expressed in primary motor neurons, mutant PFN1 also led to reduced F-actin/G-actin ratio in grown cones, indicating impaired actin polymerization. Taken together, our results suggest a loss of function effect on actin polymerization particularly in axonal termini. On the other hand, Bosco and

59 colleagues’ in vitro actin affinity analysis showed no difference in actin binding

among wild-type and mutant PFN1 (Boopathy et al, 2014). Their analysis was

based on the observation that at high concentration, PFN1 can sequester G-actin

and inhibit spontaneous actin nucleation (Polland and Cooper, 1984). PFN1

concentration-dependent suppression of actin polymerization was thus measured

to indirectly indicate its affinity to G-actin. Nonetheless, in cellular environments, many factors, such as formin, can interact and coordinate with PFN1 to facilitate actin polymerization (Romero et al, 2004). Therefore this discrepancy raises the possibility that mutant PFN1 may impair actin polymerization and dynamics

through aberrant interactions with other protein components of actin

cytoskeleton. To support this hypothesis, our co-immunoprecipitation showed

that at least PFN1 C71G reduces its interaction with mDia1, a formin isoform that

mediates actin polymerization. Because this interaction does not depend on

affinity to actin, it remains unclear whether this disruption is due to loss of

interaction with PLP stretches or due to conformational changes that alter affinity

to specific proteins. Nonetheless, our data indicate ALS-linked PFN1 mutants

can affect actin polymerization through multiple mechanisms.

The actin cytoskeleton plays an important role in multiple aspects of

neuronal functions including axon outgrowth and synaptic integrity (Luo, 2002).

Thus we asked if mutant PFN1 would impair neuronal morphologies.

Overexpression of ALS-linked PFN1 variants in primary motor neurons resulted

in shorter axons and collapsed growth cones. This is likely to be caused by

60 altered actin dynamics in the grown cone through a loss-of-function mechanism.

However, we cannot exclude the contribution of gain-of-function toxic effects of

PFN1 aggregates, possibly by sequestering other cellular factors or overwhelming the ubiquitin-proteasome system. Since PFN1 has a variety of interacting factors other than G-actin (reviewed in Witke, 2004), it is also possible

that ALS-linked mutations may affect other actin-independent pathways that lead

to neuronal defects. While further work is required to clarify the contribution of

each pathway, our findings are significant in that they demonstrate the relevance

of ALS-linked PFN1 mutations to neurodegeneration. Although neuritogenesis is

an early developmental event, axon outgrowth also occurs during re-innervation

(reviewed in Harel and Strittmatter, 2006; Bradke et al, 2012). Alteration of this

process may promote neurodegeneration secondary to axonal injury or other

degenerating stimuli. Furthermore, the “dying-back” hypothesis suggests

degenerating mechanisms may be initiated focally in the axonal and synaptic

regions (reviewed in Moloney et al, 2014). Maintenance of structures and

functions in these regions rely heavily on intact actin dynamics. Our observation

that mutant PFN1 reduce F-actin/G-actin ratio in axonal termini supports the

possibility that ALS-linked mutations may disrupt axonal and synaptic functions

by altering local actin dynamics.

61 2.4 Acknowledgements

We thank the laboratory of Dr. Stephen Doxsey, the UMass Medical

School Digital Light Microscopy Core, the UMass Medical School Deep

Sequencing Core, the Emory University Neuropathology Core, Dr. Marla Gearing and Deborah Cooper for their assistance.

2.5 Material and methods

Cell culture and transfections of N2A cells

N2A cells were maintained in MEM with 10% fetal bovine serum, 2 mM L- glutamine and 100 U penicillin per 100 µg streptomycin at 5% CO2. For cell lysis and western blot analyses, transfections were performed in 6-well plates with

4 µg of plasmid DNA using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol. For immunofluorescence, cells were plated onto 12-mm round coverslips in 24-well plates and transfected with 0.8 µg of DNA.

Immunofluorescence of N2A cells

Immunofluorescence was performed as previously described (Bosco et al, 2010)

In brief, at 48 h after transfection, cells were fixed with 4% paraformaldehyde at room temperature for 15 min, permeabilized with 1% Triton X-100 at room temperature for 10 min, and then blocked with blocking buffer (50 mM NH4Cl,

10 mg ml−1 BSA, 2% natural goat serum, 0.1% Triton X-100 in Dulbecco’s PBS)

at 37 °C for 1 h. Appropriate fluorophore-conjugated antibody was diluted in

62 dilution buffer (0.1% Triton X-100, 0.15% goat serum in Dulbecco’s PBS) and

added at 4 °C overnight. After three washes with PBST (0.1% Tween-20), cells

were mounted with Vectashield Hard Set Mounting Medium containing DAPI

(Vector Laboratories). Confocal images were obtained with a ×100 Plan Apo oil

objective lens on a Nikon TE-2000E2 inverted microscope (Nikon Instruments)

with a Yokogawa CSU10 Spinning Disk Confocal Scan Head (Solamere

Technology Group) and processed with ImageJ software. For confocal imaging, mouse Dylight 549-V5 antibody (1:100, AbDserotec) and rabbit Dylight 488- ubiquitin antibody (1:100, Enzo Life Sciences) were used. For aggregate counting, mouse Alexa Fluor 488-V5 antibody (1:100, AbDserotec) was used. To count the number of cells containing insoluble aggregates, images were obtained with a Zeiss Axioskop 2 microscope with a ×100 objective. The images were processed with MetaMorph (v.7.5, Molecular Devices) image analysis software.

Fractionation of insoluble/soluble PFN1 and western blotting

Transfected cells were washed with cold PBS and then scraped directly in NP-40 lysis buffer (1% NP-40, 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 10% glycerol, 1 mM dithiothreitol (DTT), 10 mM sodium fluoride, 1 mM sodium orthovanadate and 5 mM sodium pyrophosphate) with EDTA-free protease inhibitors (Roche). The lysates were rotated for 30 min at 4 °C followed by centrifugation at 15,800g for 20 min. The supernatant was removed and used as the soluble fraction. To remove carryovers, the pellet was washed with lysis buffer, and then resuspended in urea-SDS buffer (NP-40 lysis buffer with 8 M

63 urea, 3% SDS) followed by sonication. The lysate was then spun again for 20 min

at 4 °C and the supernatant was removed (insoluble fraction). Protein

concentrations were determined by the BCA assay. Western blot detection was

performed on Odyssey Infrared Imager (Li-Cor Biosciences). Antibodies for

Western blotting are as follows: mouse monoclonal anti-GAPDH (1:1,000,

Sigma-Aldrich); mouse monoclonal anti-V5 (1:2,000, Invitrogen); polyclonal

IRDye 800CW goat anti-mouse IgG (1:8,000, LI-COR); polyclonal IRDye 680 goat anti-rabbit IgG (1:8,000, LI-COR); and polyclonal IRDye 800CW goat anti- rabbit IgG (1:8,000, LI-COR). To inhibit proteasome activity in N2A cells, cells were incubated with 10 µM MG132 (Sigma-Aldrich) after transfection for 16 h before collection.

Immunoprecipitations

V5–PFN1-transfected HEK293 cells were lysed at 24 h with RIPA buffer (150 mM

NaCl, 50 mM Tris-HCl, pH 7.5, 1% NP40, 0.1% SDS, 0.5% sodium deoxycholate,

5 mM EDTA, 10 mM sodium fluoride, 1 mM sodium orthovanadate and 1× protease inhibitor cocktail). The lysates were precleared with Dynabeads Protein

G (Invitrogen) followed by immunoprecipitation with 1 µg anti-V5 antibody at 4 °C overnight followed by incubation with Dynabeads Protein G for 1 h. The protein– bead complexes were washed four times with RIPA buffer, eluted by boiling at

95 °C for 5 min and then subjected to western blot analysis to detect V5–PFN1 and actin. Antibodies: mouse anti-V5 (1:5,000, Invitrogen); mouse anti-β-actin

(1:1,000, Sigma); and goat anti-mouse 800CW (1:10,000, LICOR). To

64 demonstrate conjugation of mutant PFN1 by ubiquitin, HA–ubiquitin was co- transfected into N2A with either V5–PFN1(C71G) or V5 vector at a ratio of 1:1. At

48 h after transfection, cells were lysed and soluble fractions were prepared as described above. NP-40-resistant pellets were further dissolved in RIPA buffer and immunoprecipitated with 1 µg anti-V5 antibody and washed with lysis buffer three times. Western blot analysis was performed with an anti-HA-biotin antibody

(1:2,000, sigma) and IRDye 800CW streptavidin (1:2,000, LI-COR). After stripping the membrane, it was reprobed with mouse anti-V5 antibody (1:5,000,

Invitrogen).

Primary motor neuron culture, transfection and immunofluorescence

Primary motor neurons from embryonic day (E)13.5 mouse embryos were isolated, cultured and transfected as previously described (Fallini et al, 2010).

Ubiquitin protein fused to an HA tag was cloned into the pcDNA plasmid by PCR amplification. To inhibit proteasome activity, cells were treated with 10 µM

MG132 for 12 h. Motor neurons were fixed for 15 min with 4% paraformaldehyde in PBS two or three days after transfection, as indicated. Anti-V5 (1:1,000;

Invitrogen), HA (1:1,000; Cell Signaling Technology), TDP-43 (1:500,

ProteinTech Group), FUS (1:500, Bethyl labs), SMN (1:500, Santa Cruz) and

PFN1 (1:500, Sigma) antibodies were incubated overnight at 4 °C. Dylight488-,

Cy3-, Cy2- or Cy5-conjugated secondary antibodies (Jackson Immunoresearch) were incubated for 1 h at room temperature. F-actin and G-actin were labelled with rhodamine-conjugated phalloidin (1:1,000, Invitrogen) and Alexa488-DNase

65 I (1:250, Invitrogen), respectively. Z-series (5–10 sections, 0.2 µm thickness)

were acquired with an epifluorescence microscope (Ti, Nikon) equipped with a

cooled CCD camera (HQ2, Photometrics). Images were deconvolved

(Autoquant, MediaCybernetics) and analysed. The growth cone area and the

fluorescence intensity for F-actin and G-actin staining was measured using

ImageJ software. The ratio between the two values was averaged and compared

between different conditions. Statistical significance was assessed using

Kruskas–Wallis one-way ANOVA test and Dunn’s post-hoc test. For the analysis

of axon length, cells were fixed and stained to detect V5–PFN1 at 3 days after

transfection. Low magnification images (×10) were acquired and, if necessary,

reassembled in Adobe Photoshop. The axon was identified morphologically as

the longest neurite. The length of the longest axon branch was measured using

ImageJ plug-in NeuronJ (Meijering et al, 2004). Statistical significance was assessed with the Kolmogorov–Smirnov test.

Immunohistochemistry

Paraffin-embedded sections from post-mortem human spinal cord (8 µm thick) were deparaffinized and endogenous peroxidase activity was blocked with 3% hydrogen peroxide at 40 °C. Sections were then incubated with normal horse serum for 15 min at 40 °C, followed by anti-PFN1 primary antibody (1:2,000,

Sigma rabbit polyclonal) or anti-pTDP-43 (1:8,000, Cosmo Bio) diluted in 1%

BSA overnight at 4 °C. The next day, sections were incubated with biotinylated secondary antibody for 30 min at 37 °C followed by avidin–biotin peroxidase

66 complex (Vector Laboratories) for 60 min at 37 °C. DAB (3,3′-diaminobenzidine) was used as the chromogen (for colour development); tissues were then counterstained with haematoxylin.

LDH cell viability assay

HEK293 cells were plated into 96-well plates at a density of 10,000 cells, and transfected with PFN1 constructs using Lipofectamine 2000 (Invitrogen). Cells with transfection cargo were used as controls. At 24 hours after transfection, cells were subject to cell viability analysis using CytoTox-ONE Homogeneous

Membrane Integrity Assay (Promega) according to manufacturer’s protocol.

67 PREFACE TO CHAPTER III:

Chi-Hong Wu performed and analyzed all the experiments presented in this

Chapter.

The work presented in this Chapter aimed to characterize ALS-linked PFN1 mutations in a novel Drosophila model. All data were included and being prepared for publication with the exception of Figure 2A-C, Figure 10A, Figure

12, Figure 13 and Figure 14D.

68 CHAPTER III: FUNCTIONAL DISSECTION OF DROSOPHILA MODELS

EXPRESSING ALS-LINKED PFN1 MUTAIONS

3.1 Introduction

Drosophila melanogaster has been widely used as an in vivo organism to model neurodegenerative diseases including ALS (Lanson et al., 2011; Li et al.,

2010; Watson et al., 2008). It presents many advantages over other common model organisms. It has relatively shorter lifespan, which allows us to monitor adult phenotypes caused by ALS mutations in months. It is also advantageous to use GAL4-UAS binary expression system to dissect functions of gene of interest in a tissue specific manner (Brand and Perrimon, 1993). Furthermore, Drosophila

has small and non-redundant genome, and more than 60 percent of fly genes

have homologs in the human genome (Wangler et al, 2015); therefore it can be

used to conduct genetic screens in order to elucidate relevant cellular pathways.

Overexpression of ALS-linked proteins in Drosophila have led to many

cell-type specific phenotypes such as rough eyes associated with retinal

degeneration (Hanson et al, 2010; Li et al, 2010; Ritson et al, 2010; Lanson et al,

2011) and wing defects (Xia et al, 2012; Yang et al, 2015). Importantly, when

expressed in motor neurons, these proteins reduce lifespan and impair locomotion (Watson et al, 2008; Estes et al, 2011), reflecting motor dysfunctions observed in ALS patients.

On a cellular level, neuromuscular junction denervation associated with synaptic loss is frequently observed in ALS patients and rodent models (Fischer

69 et al, 2004). Thus it is plausible ALS-linked protein mutants focally impair

synaptic functions and eventually lead to axon retraction and neurodegeneration.

Fly larval NMJ is a well-established system for studying synaptic function and

morphology (Keshishan et al, 1996; Menon et al, 2013). Termini of larval

segmental nerves form a series of varicosities where they contact and innervate

the muscles. These “boutons” contain active zones where neurotransmitters are

released, and have a stereotyped beads-on-a-string pattern, allowing us to

quantify and compare bouton morphology among different genotypes. At 3rd-intar larval stage, ALS-linked proteins have led to abnormal morphologies of neuromuscular junctions including altered bouton numbers and sizes (Estes et al,

2011; Xia et al, 2012) and increased satellite bouton numbers (Estes et al, 2011), indicating the importance of ALS-linked proteins in maintaining the integrity of

NMJs. Notably, overexpression of wild-type ALS-linked proteins can also cause these phenotypes in different degrees of severity (Hanson et al, 2010; Li et al,

2010; Estes et al, 2011; Xia et al, 2012).

The identification of these phenotypes is invaluable because they can serve as readout for forward or reverse genetic screens in order to uncover novel cellular pathways altered by ALS-linked proteins. In 2015, Taylor, Gao and colleagues introduced a new C9orf72 model and by conducting genetic screens using deficiency lines, they discovered that in the presence of expanded G4C2 repeats, nucleocytoplasmic transport was impaired (Freibaum and Lu et al,

2015). Their finding is pivotal in that it sheds light on largely elusive C9orf72

70 pathogenesis. Likewise, Freeman and colleagues first reported that TDP-43

induced adult-onset NMJ degeneration in fly legs. They further utilized this novel

adult leg NMJ assay to uncover genetic interactions between TDP-43 and

several proteins involved in chromatic modeling and RNA export (Sreedharan et

al, 2015). On the other hand, from a reverse genetic approach, Bonini and

colleagues confirmed TDP-43 toxicity was modified by phosphorylated eIF2α, a

component in stress granules (Kim et al, 2014). These important findings not only

contribute to understanding of ALS pathogenesis, they also emphasize the

strength of Drosophila as a tool to decode cellular mechanisms that underlie neurodegeneration.

Chickadee is the Drosophila PFN1 ortholog (Cooley et al, 1992), and the homology between chickadee and PFN1 is approximately 30 percent (Figure III-

1C). Although the amino acid residues where ALS mutations locate are not conserved, previous literature suggests that structural similarity of profilin from different species may be more crucial than sequence identity in conserving its functions (Witke, 2004). In support of this notion, Gitler and colleagues reported that in yeast, human PFN1 functionally rescued PFY mutants despite their low sequence conservation (Figley et al, 2014).

To further characterize how ALS-linked PFN1 mutations lead to neurodegeneration in vivo, we generated a novel Drosophila model expressing human wild-type and mutant PFN1. We first determined if human PFN1 was functionally conserved in the Drosophila system. To investigate if mutant PFN1

71 impair neuromuscular junctions, we expressed PFN1 in motor neurons and stained larval NMJs with various synaptic markers. We conducted image analysis focusing on structural changes and alteration of F-actin content. To clarify if mutant PFN1 cause progressive locomotion defects in adults, we expressed

PFN1 in motor neuron and measured locomotion at different ages. We also monitored their lifespan to determine if PFN1 overexpression affects longevity.

Finally, to test if mutant PFN1 aggregates in vivo, we performed immunohistochemistry and molecular solubility analysis to detect the presence of

PFN1 aggregates.

72 3.2 Results

Human PFN1 rescues lethality of chickadee knockdown in motor neurons and glia.

To characterize the contribution of PFN1 mutations to neurodegeneration in vivo, we generated transgenic Drosophila models expressing human PFN1 with an N-terminal V5 tag. Because these human PFN1 constructs were randomly inserted into the Drosophila genome with P-element, we selected

clones with similar expression levels among genotypes for further studies

(Figure III-1A,B). Since human PFN1 and Drosophila chickadee are not

conserved (Figure III-1C), we first asked whether human PFN1 could functionally

substitute Drosophila chickadee. In larval NMJs, we first determined chickadee

was enriched in peripheral glia enwrapping segmental neurons and expressed

weakly in axons (Figure III-2A-C). Chickadee is an essential gene (Verheyen et

al, 1994), and its knockdown by RNAi in these tissues led to pupal death (Figure

III-2D-E). When we expressed human PFN1 under chickadee-deficient

background, it fully rescued pupal death in both motor neurons and glia. The expression of UAS-mCD8GFP in the same RNAi background failed to rescue the lethality, indicating that this effect was not due to dilution of UAS lines. This result suggested that human PFN1 could functionally replace Drosophila chickadee in the central nervous system. Surprisingly, the expression of PFN1 mutants also fully rescued pupal death, suggesting that these mutants are functional at early developmental stage.

A. Repo-Gal4> mcherry-NLS mcherry chic HRP

B. Repo-Gal4> + C. Repo-Gal4> chic RNAi HRP chic view orthogonal

D. E.

75 Figure III-2. Human PFN1 rescue pupal death caused by RNAi knockdown of chickadee in motor neurons and glia.

(A) Endogenous chickadee is enriched in peripheral glia. UAS-mcherry-NLS was driven by Repo-Gal4 and used as a glial nucleus marker. (B,C) Confirmation of presence of chickadee in peripheral glia. Chic knowndown by HMS00550 depleted chickadee in glia. Noted that chickadee is expressed weakly in segmental axons. Scale bar = 10 um. (D) Human PFN1 rescue of pupal death caused by chickadee knockdown in motor neurons. Genotypes= OK371- Gal4>UAS-PFN1 WT, UAS-chic RNAi (HMS00550); OK371-Gal4>UAS-PFN1 C71G, UAS-chic RNAi; OK371-Gal4>UAS-PFN1 M114T, UAS-chic RNAi; OK371-Gal4>UAS-mCD8-GFP, UAS-chic RNAi. Rescue efficiency was normalized to PFN1 WT. (E) Human PFN1 rescue of pupal death caused by chickadee knockdown in glia. Genotypes= repo-Gal4>UAS-PFN1 WT, UAS-chic RNAi; repo-Gal4>UAS-PFN1 C71G, UAS-chic RNAi; repo-Gal4>UAS-PFN1 M114T, UAS-chic RNAi; repo-Gal4>UAS-mCD8-GFP, UAS-chic RNAi. Error bars represent S.E.M. throughout the study. n.s. not significant, *P< 0.05, **P< 0.005, ***P< 0.0005, ****P<0.0001, one-way ANOVA and Tukey post hoc test throughout the study unless stated otherwise. chic = chickadee.

76 PFN1 overexpression increases protrusion and ghost bouton formation in larval

neuromuscular junctions.

To further investigate if mutant PFN1 have a loss-of-function effect on F- actin formation at the NMJ, we looked at 3rd instar larval neuromuscular junctions (NMJ). We chose NMJs on muscle 4 because they are generally flatter and provide a better platform to monitor actin cytoskeleton (Pawson et al, 2008).

Overexpression of profilin1, either wild type or mutant, did not alter the number and size of major boutons (Figure III-3A-C). However, PFN1 overexpression led to the formation of filamentous protrusions extending from boutons and less frequently, from the axonal shafts (Figure III-3D). We also observed an increase in ghost boutons, which are a type of immature small boutons that lack the postsynaptic marker Disc-large (Figure III-3E,F). When comparing wild-type and mutant PFN1, we found that M114T PFN1 had significantly fewer protrusions and ghost boutons while C71G profilin 1 yielded more variable results (Figure III-

3D,F). Despite these morphological changes, we did not detect any synaptic retraction (Figure III-4A-D). Larval crawling assay also showed that 3rd instar

PFN1-expressing larvae expressing either wild-type or mutant PFN1 demonstrated normal crawling ability compared to controls (Figure III-4E). Taken together, our data suggest that although PFN1 overexpression did not lead to locomotion dysfunction or synapse retraction at 3rd instar larval stage, PFN1

M114T has a partial loss-of-function effect on promoting protrusion formation in larval NMJs.

77 A. OK371-Gal4 OK371-Gal4> PFN1 C71G

HRP Dlg OK371-Gal4> PFN1 WT OK371-Gal4> PFN1 M114T

B. C. D.

E. F.

OK371-Gal4> PFN1 WT HRP Dlg

78 Figure III-3. PFN1 overexpression induces formation of protrusions and ghost boutons without altering major bouton morphology in larval NMJ.

(A) Visualization of 3rd instar larval NMJs in muscle 4 at segment A3. PFN1 overexpression did not alter general synaptic morphology including number and size of major boutons but induced formation of protrusions and ghost boutons at axon termini. Horseradish peroxidase(HRP) labels axonal membrane. Disc large (Dlg) is homolog of PSD-95 and used as a post-synaptic marker. Scale bar = 10 µm. (B) Quantification of major bouton number (animals analyzed: Driver only =10, WT =10, C71G =13, M114T =14) (C) Quantification of major bouton size (animals analyzed: Driver only =6, WT =9, C71G =13, M114T =8). (D) Quantification of protrusion number (animals analyzed: Driver only =10, WT =10, C71G =13, M114T =14) (E) Visualization of ghost boutons in NMJs overexpressing PFN1 WT. Ghost bouton (indicated by arrows) was defined as a small immature bouton extending from the axonal shaft or major boutons without apposition to Dlg. Scale bar = 10 µm. (F) Quantification of ghost bouton number. PFN1 M114T induced significantly less protrusion and ghost bouton formation than PFN1 WT. (animals analyzed: Driver only =10, WT =10, C71G =13, M114T =14).

M114T PFN1 has a loss-of-function effect on F-actin formation in larval NMJ.

To further investigate whether the formation of these protrusions depended on actin polymerization, we utilized flies transgenic for UAS-lifeact-

GFP. Lifeact is a 17-amino-acid peptide that binds to F-actin and has been used widely to monitor F-actin dynamics in vivo (Riedl et al, 2008). Lifeact-GFP coexpression with profilin1 confirmed that the protrusions induced by PFN1 overexpression were rich in F-actin (Figure III-5A-D). We also confirmed that

PFN1 M114T formed much fewer filopodia than wild type profilin1 (Figure III-5E).

However, the average length of filopodia was similar among wild type and mutant

PFN1 (Figure III-5F). We also quantified F-actin levels within the whole synaptic regions to measure the effects of mutant PFN1 on overall actin polymerization. In the absence of PFN1 overexpression, F-actin formed discrete GFP positive patches within boutons and along axonal shafts (Figure III-6A). When PFN1 was overexpressed, the number and the average size of F-actin patches drastically increased (Figure III-6B-E). We also observed a trend toward a mild reduction in the number of F-actin patches when M114T mutant was expressed compared to

WT overexpression (Figure III-6F), supporting the hypothesis that PFN1 M114T has a partial loss-of-function effect on actin polymerization, particularly in the presynaptic regions of NMJs.

PFN1 overexpression reduces basal number of satellite boutons and depletes

synaptic vesicles

Satellite boutons are another type of small boutons that sprout from major

boutons. Unlike ghost boutons, satellite boutons contain active zones and are

apposed by postsynaptic markers (Torroja et al, 1999), suggesting a functional difference between these two types of boutons. In our experiments, satellite boutons were occasionally found in control NMJs fro, animals reared at 29°C

(Figure III-7A). We found that overexpression of both wild type and mutant PFN1 reduced basal number of satellite boutons (Figure III-7B). Because alterations in the number of satellite boutons have been observed in mutants that affect endocytosis and actin cytoskeleton (Marie et al, 2004; Dickman et al, 2006;

Rodal et al, 2008), we then asked whether PFN1 overexpression would change synaptic content. Using cysteine-string-protein (CSP) to label synaptic vesicles, we saw a reduction in synaptic vesicles in NMJs of PFN1-overexpressing flies

(Figure III-7C). By comparison, mutant PFN1 reduced the synaptic vesicles to a similar extent to wild type PFN1 (Figure III-7D), implying the mechanisms underlying protrusion formation and satellite bouton reduction may be different.

84 A. B.

OK371-Gal4

Dlg HRP

C. OK371-Gal4 CSP HRP D.

OK371-Gal4> PFN1 WT

OK371-Gal4> PFN1 C71G

OK371-Gal4> PFN1 M114T

85 Figure III-7. PFN1 overexpression leads to reduction in synaptic vesicles and satellite bouton number in larval NMJ.

(A) Representative image of satellite boutons in control NMJs. Arrows indicates satellite boutons. Scale bar = 10 µm. (B) Quantification of satellite bouton number. PFN1 overexpression reduces number of basal satellite boutons at 29°C. (animal analyzed: Driver only =10, WT =10, C71G =13, M114T =14) (C) NMJs stained with CSP and HRP. Cysteine string protein (CSP) was used as a presynaptic marker to label synaptic vesicles. Scale bar = 10 µm. (D) Quantification of CSP intensity per synaptic region. Synaptic area was delineated by HRP. Profilin 1 overexpression resulted in a reduction in CSP intensity in the whole synapse. (animal analyzed: Driver only =8, WT =6, C71G =7, M114T =8)

86 PFN1 overexpression leads to adult locomotion defects and reduces lifespan

We then asked whether PFN1 overexpression would cause locomotion

defects at adult age, which reflects adult-onset motor dysfunction in ALS.

Negative geotaxis is a behavior commonly used to measure age-associated

locomotion decline (Gargano et al, 2005) and the effects of ALS-associated

genes on adult locomotion in flies (Watson et al, 2008; Estes et al, 2011). We

thus used OK371-Gal4 to drive expression of PFN1 in glutamatergic neurons,

which included motor neurons (Mahr et al, 2006), and measured negative

geotaxis (climbing assay) at different ages. At five days post-eclosion (d.p.e.),

flies expressing wild-type and M114T PFN1 had climbing ability similar to driver-

only controls, while there was a mild reduction in flies expressing C71G PFN1. At

10 d.p.e., all PFN1-expressing flies showed significant climbing defects, and by

day 20, almost none could climb even without external agitation (Figure III-8A).

These defects were reproduced using multiple clones of PFN1 expressing flies to exclude possible position effects (Figure III-8B). Interestingly, we observed a trend whereby wild-type PFN1 overexpression led to more severe defects than

mutant PFN1 from day 10 onwards, although it did not reach statistical significance. PFN1 overexpression in motor neurons also drastically reduced lifespan and in this analysis, wild-type PFN1 was significantly more severe than mutant PFN1 (Figure III-8C). Bosco and colleagues reported that ALS-linked mutations destabilize PFN1 in vitro and increase protein turnover in mammalian cultured cells (Boopathy et al, 2014). To test if the difference of toxicity was due

87 to change in expression level secondary to protein degradation, we checked

PFN1 protein level from flies at 10 d.p.e., when the climbing defects started to manifest. We did not detect reduced protein levels of mutant PFN1 (Figure III-

8D,E), implying that the difference in toxicity did not result from protein

degradation. Interestingly, Western blot analysis of PFN1 protein level at 10 d.p.e. showed that PFN1 M114T had higher level than WT and C71G. It has yet to be clarified whether this difference is due to the hypothesis that PFN1 WT is more toxic, thereby resulting in increased motor neuron death. We then repeated

climbing assays and lifespan analysis with chickadee-expressing flies and found similar defects compared to wild-type human PFN1 (Figure III-9A-C). This suggested that the toxicity was not an artificial outcome of overexpression of exogenous human PFN1, but more likely caused by dysregulation of cellular pathways in which native PFN1 is involved. To test whether PFN1

overexpression had a similar impact in different cell types, we expressed PFN1 in glia and measured lifespan. Interestingly, as in motor neurons, wild-type PFN1 overexpression drastically reduced lifespan, while flies expressing mutant PFN1 had lifespan similar to controls (Figure III-10A). On the contrary, PFN1 overexpression in the eye tissue including photoreceptor cells (driven by GMR-

Gal4) did not cause any rough eye phenotypes (Figure III-10B). Taken together, our data show PFN1 overexpression results in adult progressive locomotion defects, on which mutant PFN1 has a partial loss-of-function effect. It also indicates the cellular sensitivity to increased PFN1 level is cell-type specific.

88 A.

B. C.

D. E.

DriverPFN1 only PFN1WT PFN1C71G M114TDriverPFN1 only PFN1WT PFN1C71G M114T V5

tubulin

3 d.p.e. 3 10 d.p.e. 10 OK371-Gal4

89 Figure III-8. PFN1 overexpression in motor neurons leads to adult locomotion defects and shorter lifespan.

(A) Climbing assay of adult flies expressing PFN1. PFN1 overexpression in motor neurons (driven by OK371-Gal4) led to severe climbing defects from 10 d.p.e. (B) Confirmation of climbing defects. Overexpression of multiple clones of wild-type PFN1 all led to severe toxicity. (C) Lifespan of adult flies expressing PFN1. PFN1 overexpression in motor neurons resulted in shorter lifespan compared to driver only control. PFN1 WT was significantly more toxic than C71G and M114T. Log-rank test: WT vs. control, P < 0.0001; WT vs. C71G, P < 0.0001; WT vs. M114T, P < 0.0001; C71G vs. control, P < 0.0001; C71G vs. M114T, P < 0.0001; M114T vs. control, P < 0.0001. (D) Western blot analysis of PFN1 expression in motor neurons at different ages. There was no decrease in protein level in C71G and M114T both at 3 and 10 d.p.e. (E) Quantification of profilin 1 expression level at 10 d.p.e. d.p.e = days post-eclosion.

PFN1 overexpression results in adult leg NMJ degeneration

Because PFN1 overexpression in motor neurons resulted in adult locomotion defects and shorter lifespan, we wondered if there were signs of axon degeneration at adult stage. mCD8-GFP expressed in motor neurons (driven by

OK371-Gal4) has been reported to outline structures of adult NMJ and label active zones (Sreedharan et al, 2015, Figure III-11A). Utilizing this assay, we co- expressed PFN1 and 2 copies of mCD8-GFP in motor neurons and monitored

NMJ integrity at 10 d.p.e., when climbing defects manifested. While the NMJ remained intact in control flies, PFN1 overexpression resulted in enlargement and disorganization of the NMJ, accompanied by loss of GFP puncta (Figure III-

11B). This result , together with the finding in 3rd instar larval stage, supports the hypothesis that PFN1 overexpression causes adult-onset neuromuscular degeneration.

Superoxide dismutase (SOD) genetically interacts with PFN1 in locomotion and

NMJ morphology

Given the observed morphological and behavioral phenotypes, we

decided to investigate possible molecular mechanisms that account for PFN1-

induced toxicity. Thus, we conducted epistatic studies to test if other ALS-

associated genes interact with PFN1 in common cellular pathways. Interestingly, we found that heterozygous amorphic SOD mutant SODn1 enhanced PFN1-

induced climbing defects, while loss-of-function ataxin-2 and TDP-43 mutants

failed to modify this phenotype (Figure III-12A). SOD1 is a cytoplasmic

superoxide dismutase responsible for scavenging cellular free superoxide

radicals (Fridovich, 1986), and it is one of the most common ALS-linked genes

(reviewed in Renton et al, 2014). To be noted, SODn1 is a loss-of-function allele

of Drosophila SOD1 homolog that contains a point mutation that disrupts the

hydrogen bond on the SOD dimer surface. It also showed that SODn1 mutant

further reduced lifespan of PFN1-over expressing flies, although the

heterozygous allele alone also led to a mild reduction in longevity (Figure III-

12B). Western blot analysis of 10-day-old flies demonstrated that SODn1 did not

alter PFN1 level. (Figure III-12C,D). Heterozygous SODn1 mutant alone did not

impair climbing, suggesting this enhancement was not additive. (Figure III-12E).

Importantly, SODn1 mutant failed to enhance climbing phenotypes caused by

PFN1 C71G and M114T (Figure III-12E). We attempted to confirm this genetic

interaction using another amorphic SOD mutatnt, SODn64, which affected

95 enzymatic activity of SOD more drastically than SODn1 (Phillips et al, 1995).

Surprisingly, SODn64 mutant enhanced climbing defects in both wild-type and mutant PFN1 (Figure III-12F). This differential modification hinted at a partial loss-of-function effect of ALS-linked PFN1 mutations on their genetic interaction

with SOD. SOD is a cytoplasmic superoxide dismutase involved in free radical reduction. To investigate if PFN1-induced phenotypes could also be modified by other oxidative stress-associated components, we tested SOD2, a mitochondria- specific superoxide dismutase. Unlike SOD, loss-of-function SOD2Δ02 mutant did not modify PFN1-induced climbing defects (Figure III-12G). We then wondered if

SODn1 mutant could modify larval NMJ changes caused by PFN1 overexpression. Sweeney and colleagues have used homozygous amorphic

SOD mutants and pharmacological stressors to demonstrate that oxidative stress results in NMJ overgrowth (Milton et al, 2011). Our data supported their finding that homozygous SODn1 mutant increased bouton numbers (Figure III-13A-D).

Interestingly, while neither heterozygous SODn1 mutant alone nor PFN1 overexpression led to NMJ overgrowth, the combination of these two conditions significantly increased bouton numbers, indicative of an epistatic relationship

(Figure III-13E). On the contrary, heterozygous SODn1 mutant did not alter

PFN1-induced filopodia formation (Figure III-13F). In all, this result corroboratea

our finding in the locomotion assay that PFN1 genetically interacts with SOD.

Importantly, ALS-linked PFN1 mutants may have a partial loss-of-function effect

on this interaction.

96 Mutant PFN1 do not form visible aggregates in this Drosophila model

Since mutant PFN1 forms cytoplasmic aggregates when overexpressed in mammalian cultured cells (Chapter II), we wondered whether these mutants also

led to protein aggregation in vivo. To test this, we expressed either wild type or

mutant PFN1 in motor neurons and stained larval ventral nerve cords for

presence of aggregates. However, we did not detect visible aggregate formation

of mutant PFN1 (Figure III-14A,B). We also confirmed this result by performing

solubility analysis of aged fly tissues, in which both wild type and mutant PFN1

remained in the detergent-soluble protein fractions (Figure III-14C). Proteasome inhibition has been reported to enhance aggresome formation (shown in Chapter

II). Therefore, we co-expressed human PFN1 and Prosbeta21, a temperature- sensitive dominant-negative mutant of proteasome subunits, which created a proteasome-defective cellular environment (Smyth et al, 1999).

Immunohistochemistry analysis failed to detect insoluble aggregates (Figure III-

14D).

Figure III-12. Amorphic SOD mutants enhance PFN1- induced toxicity.

(A) Modification of PFN1-induced climbing toxicity by other ALS-associated proteins. Amorphic SOD mutant SODn1 enhanced climbing defects of PFN1 overexpression (driven by OK371-Gal4). (B) Lifespan analysis. SODn1 mutant reduced longevity of both driver only and PFN1-overexpressing flies. Log-rank test: OK371/+ vs. OK371/+; SODn1/+, P < 0.0001; OK371/+; PFN1 WT/+ vs. OK371/+; PFN1/SODn1, P < 0.0001. (C,D) Western blot analysis and quantification of PFN1 level co-expressed with SODn1 mutant. SODn1 mutant did not alter protein level of PFN1. (E) SODn1 mutant enhanced climbing phenotype of wild-type PFN1 overexpression but not C71G and M114T. (F) SODn64 mutant enhanced both wild-type and mutant PFN1 toxicity. (G) Loss-of- function SOD2 mutant did not modify PFN1 toxicity. All climbing assays were preformed at 10 d.p.e. SOD = superoxide dismutase, Atx2= ataxin 2, TBPH= TAR DNA-binding protein-43 homolog, SOD2 =superoxide dismutase 2.

A. PFN1 WT; mcherry-NLS B. PFN1 C71G; mcherry-NLS

OK371-Gal4 OK371-Gal4

V5 mCherry-NLS

C. D. PFN1 WT; Prosbeta21 OK371-Gal4 Driver onlyPFN1 WTPFN1 C71G S I S I S I PFN1 C71G; Prosbeta21 V5

tubulin

tub-Gal4 OK371-Gal4

101 Figure III-14. ALS-linked PFN1 mutants do not form aggregates in Drosophila tissues.

(A,B) Immunofluorescence staining of profilin 1 in ventral nerve cords from 3rd instar larvae. Either PFN1 WT or PFN1 C71G were driven by OK371-Gal4, expressed in motor neurons, and stained with V5 antibody. Both proteins were mostly localized in the cytoplasm without forming visible aggregates. UAS- mcherry-NLS was used to label nuclei of motor neurons. Scar bar = 30 µm. (C) Western blot analysis of fly head tissues subject to Triton-X 100 soluble (S) and insoluble (I) fractionation. Fly heads were collected at 24 days post-eclosion. (D) Staining of PFN1 in ventral nerve cords from 3rd instar larvae under proteasome- defective background. UAS-Prosβ21 was co-expressed with PFN1 to inhibit proteasome. V5 antibody labeled profilin 1, and both WT and C71G remained mostly in the cytoplasm without aggregate formation.

102 3.3 Discussion

The cytoskeleton is crucial in maintaining axonal function and structure,

and gene mutations of several cytoskeletal components have been reported to

be associated with ALS, such as PFN1, TUBA4A (Smith et al, 2014), DCTN1

(Puls et al, 2003), NEFH (Figlewicz et al, 1994; Al-chalabi et al, 1999), and

PRPH (Gros-Louis et al, 2004). However, it is still not fully understood how mutations of these cytoskeletal proteins contribute to neurodegeneration. Here we established a novel Drosophila model expressing either wild-type or mutant

human PFN1 in order to elucidate mechanisms underlying neurodegeneration

caused by mutant PFN1. Although human PFN1 and its Drosophila ortholog

chickadee are lowly conserved, we showed that human PFN1 rescues pupal

death caused by chickadee knockdown in motor neurons and glia, indicative of

its conserved functions.

Profilin is involved in various aspects of neuronal functions including

axonal outgrowth of motor neurons in Drosophila embryos (Wills et al, 1999; Kim

et al, 2001). It has been shown to be present both in presynaptic (Faivre-Sarrailh

et al, 1993) and postsynaptic (Giesemann, et al, 2003) sites. Likewise, it is also

known that synaptic morphology and functions rely on intact F-actin network, and

mutations in components of the actin cytoskeleton have been shown to cause

abnormal larval NMJ morphologies (Zhao et al, 2013, Piccioli et al, 2014). To

clarify how PFN1 regulates integrity of neuromuscular junctions, we expressed

wild-type or mutant human PFN1 in motor neurons and measured morphological

103 changes in larval NMJs. Although PFN1 overexpression does not alter size and

number of major boutons, we found that it leads to the formation of protrusions

and ghost boutons. Davis and colleagues reported similar protrusion formation

when hts/adducin, an actin capping protein, was depleted in larval NMJ (Pielage

et al, 2011). This implies that the protrusion formation is likely a result of F-actin

upregulation. Enrichment of Lifeact-GFP in these protrusions further confirmed

they are F-actin rich filopodia. This result supports a previous study showing that

in mouse embryonic fibroblasts, PFN1 overexpression preferentially induces

filopodia formation (Rotty et al, 2015).

Interestingly, the same phenotypes have been reported in wild type NMJs

when they undergo spaced high potassium stimulation (Ataman et al, 2008). In

the study, Budnik and colleagues found that activity-induced synaptopods were

detected in live NMJs but hardly preserved after tissue fixation. However, in our

case, PFN1-induced protrusions are well detected in post-fixed NMJs. It is therefore plausible that PFN1 plays an activity-regulated role in the presynaptic region. Importantly, we found PFN1 M114T produces fewer filopodia than wild- type PFN1. While the number of filopodia is reduced, the average length is unchanged, suggesting that M114T may affect the initiation, rather than elongation, of unbranched F-actin during filopodia formation. It will thus be worthwhile further clarifying which components in actin polymerization are affected by mutant PFN1. We aslo observed a mild but non-significant decrease

104 in F-actin patches in whole synaptic regions in PFN1 M114T mutant, supporting that mutant PFN1 has partial loss-of-function effects on actin polymerization.

The F-actin network serves as scaffold for regulating the pools and

recycling of synaptic vesicles (reviewed in Cingolani and Goda, 2008). For

instance, WAVE complex regulates F-actin assembly in presynaptic NMJ

terminals, and mutants of Cyfip, a component of WAVE complex, showed an

increase in satellite boutons and led to endocytic defects (Zhao et al, 2013). We

found that PFN1 overexpression reduces basal satellite bouton numbers and

CSP content in whole synaptic areas. We speculate that PFN1 overexpression may either promote synaptic vesicle release or regulate its recycling through actin-dependent or –independent pathways. However, unlike filopodia formation, mutant PFN1 lead to similar reduction in satellite number and CSP signal compared to wild-type PFN1. This implies the pathways through which PFN1 regulate synaptic vesicles and filopodia formation may not overlap.

In Davis and colleagues’ study, loss-of-function hts/adducin mutant resulted in synaptic retractions in larval NMJs (Pielage et al, 2011). These retracting synapses maintained residual HRP and postsynaptic DGluR III staining but lost presynaptic active zone marker BRP. In our experiments, PFN1 overexpression does not cause synaptic retraction or alter larval crawling ability despite the morphological changes in larval NMJs. Furthermore, although we observed partial loss of function in filopodia formation in M114T PFN1, mutant

PFN1 are also able to rescue pupal death caused by chickadee knockdown and

105 do not affect larval locomotion. These results suggest that mutant PFN1 do not have a deleterious effect (i.e. life or death) at early developmental stage.

Nonetheless, in adult flies our data showed that PFN1 overexpression in

motor neurons leads to progressive climbing defects and reduced lifespan. By

using an adult leg NMJ assay, we also found that PFN1 overexpression leads to

the disorganization of NMJs. We argue that PFN1 overexpression may up-

regulate its cellular functions to the extent that it leads to neural toxicity.

Considering that ALS is an adult-onset progressive neurodegenerative disease,

our findings are important in that they demonstrate an age-dependent motor dysfunction due to PFN1 overexpression. Because mutant PFN1 is significantly less toxic than wild-type PFN1 despite similar expression level, we also speculate that PFN1 mutants have a partial loss of function, thereby resulting in milder toxicity. We cannot exclude the possibility that there are unknown pathways affected by mutant PFN1 through gain-of-function mechanisms, and that these phenotypes are masked by toxicities of actin dysregulation. Therefore, genetic screening on PFN1 mutants will be needed to explore this possibility.

On a molecular level, our epistatic study also revealed a genetic interaction between PFN1 and SOD, another ALS-linked protein. In this study, we tested two different amorphic SOD mutants. SODn1 allele contains a point mutation that disrupts the hydrogen bond on the SOD dimer surface, and

SODn1/SODwt heterozygotes maintain 37% of SODwt/SODwt activity. On the other hand, SODn64 allele harbors a point mutation that affects binding of SOD1 to it

106 cofactors Cu and Zn ions. SODn64/SODwt heterozygotes result in 12% of

SODwt/SODwt activity (Phillips et al, 1995). These two SOD1 mutants

consistently exacerbated the climbing defects caused by wild-type PFN1

overexpression. We further confirmed this genetic interaction by showing that

heterozygous SODn1 allele and PFN1 overexpression have a concerted effect on promoting NMJ overgrowth. Notably, only the more severe form SODn64 allele modifies the toxicity of ALS-linked PFN1 mutants. This differential modification hints that mutant PFN1 may have a partial loss-of-function effect on the genetic interaction with SOD. It has yet to be clarified whether this partial loss of function is actin-dependent (which is in line with our other results) or through other unknown actin-independent pathways.

Contrary to SOD1, loss-of-function SOD2 mutant alleles fail to alter the climbing phenotype. This result implies that this interaction is specific to SOD1

and not shared by other cellular components associated with ROS reduction.

Nonetheless, the SOD alleles we tested possess dominant negative

characteristics, while SOD2 mutant is a simply loss-of-function amorph. Thus their heterozygotes may increase cellular vulnerability to oxidative stress to different levels. Likewise, while heterozygous SODn1 mutant does not lead to

NMJ overgrowth, which is induced by oxidative stress, it moderately reduced longevity of adult flies, suggesting that heterozygous SODn1 mutant may still induce oxidative stress in adult flies. Therefore, while it is possible that PFN1 overexpression somehow regulates the activity of SOD1, it is more likely that

107 SOD1 amorph-dependent induction of oxidative stress modulates the functions of

PFN1, thereby enhancing its toxicity. Future investigation on the molecular

mechanisms underlying this genetic interaction will shed light on how PFN1

regulate functions of adult motor neurons during oxidative stress.

Lastly, in this Drosophila model, we did not detect visible aggregates in

motor neurons by cell staining or insoluble mutant PFN1 by solubility analysis.

Proteasome inhibition by co-expression of a dominant-negative proteasome subunit mutant Prosbeta21 also failed to promote aggregate formation. It is not

known if mutant PFN1 possesses the same protein stability or propensity to

aggregate in different models (e.g. mammalian cell cultures and Drosophila). It is

also possible that mutant PFN1 gradually develop aggregates during aging,

which cannot be detected because of early-onset lethality resulting from PFN1

overexpression. The contribution of aggregates to pathogenesis of ALS has been

discussed heatedly over the past decades. While some argue that aggregation

leads to toxic gain-of-function effects that compromise cellular viability (reviewed

in Peters et al, 2015), it is increasingly recognized that in many aggregate-

forming neurodegenerative diseases, soluble misfolded proteins are the actual

toxic species, and aggregate formation serves as a protective mechanism by

sequestering these toxic proteins (Arrasate, et al, 2004; Tanaka et al, 2004;

Cowan et al, 2013; Cragnaz et al, 2014). Although it is still inconclusive which

mechanism plays a major role, absence of aggregates in our Drosophila system

provides a good platform for us to dissect loss-of-function and gain-of-function

108 mechanisms of soluble ALS-linked mutant PFN1 without being confounded by the effects contributed by aggregates.

Taken together, in this Chapter, our data support the findings in Chapter II that ALS-linked PFN1 mutants have a loss-of-function effect on actin polymerization, particularly in presynaptic NMJ terminals. We also demonstrate that PFN1 overexpression results in progressive locomotion defects, shorter lifespan, and neuromuscular degeneration in adult flies. Summary of these phenotypes is listed at the end of this chapter (Table III-1). These phenotypes not only reflect symptoms in ALS patients, but will also be valuable as readout for future genetic screens to better understand PFN1-associated neurodegeneration.

109

3.4 Acknowledgements

We thank members in laboratories of Dr. Fen-Biao Gao and Dr. Marc

Freeman for discussions, technical assistance and sharing of equipment and reagents.

3.5 Material and methods

Fly stocks and transgenic flies

All flies were maintained on standard fly food (molasses, cornmeal, yeast, and agar) at 25°C. To increase expression of transgenes, crosses were set at 25°C for two days and then shifted to 29°C afterwards in all experiments unless stated otherwise. W1118 strain was used as control flies. To generate UAS-V5-hPFN1 transgenic flies, cDNA encoding wild-type, C71G and M114T PFN1 were amplified from constructs previously reported (Wu et al, 2012). They were then cloned into pUAST vector to express profilin 1 protein tagged with an N-terminal

V5 epitope, under control of UAS sequence in the promoter region. The constructs were sent for fly injection (BestGene Inc.) The following stocks were obtained from Bloomington Stock Center: OK371-Gal4, repo-Gal4, gmr-Gal4, act5c-Gal4, tubulin-Gal4, UAS-mCD8::GFP, UAS-mcherry-NLS, UAS-

Prosbeta21, UAS-lifeact-GFP, SODn1, SODn64, and SOD2Δ02. RNAi line against chickadee (HMS00550) was obtained from TRiP RNAi Screening Center at

Harvard Medical School. UAS-chicE36.5, which was used in initial control

111 experiments, was kindly provided by Dr. Lynn Cooley. TBPHΔ142/Cyo was kindly provided by Dr. Fen-Biao Gao.

Western blot and solubility analysis

Heads of 10 one-to-three-day-old male flies were collected and homogenized in

SDS-containing loading buffers, and equal amount were loaded onto 4-20%

SDS-PAGE (Bio-rad), transferred to nitrocellulose membranes through Turboblot

(Bio-rad), and blotted with antibodies accordingly. For solubility test, heads of male flies were first homogenized in lysis buffer containing 0.1% Triton X100,

150mM NaCl, and 0.1M Tris, pH 8.0 supplemented with protease inhibitor cocktail (Roche), and centrifuged at 13000 rpm for 15 minutes. The supernatants were collected as soluble fractions. The pellets were then washed with the same lysis buffer at least twice, lysed in 8M urea buffers and centrifuged for another 20 minutes. The final supernatants were collected as insoluble fractions. Antibodies: mouse anti-V5 (1:5000, Novus Biologicals), rabbit anti-V5 (1:5000, Novus biological), mouse anti-tubulin (E7, 1:1000, DSHB).

Climbing assay and lifespan

One to three day-old male flies from corresponding crosses were collected, and transferred to fresh food vials with density of 20 flies per vial. Flies were allowed to recover from anesthesia for at least 2 days before any behavior analysis, and food were changed every 2 days to prevent flies from getting trapped in the food.

112 Climbing assays were conducted in the early evenings on day 5, day 10, day 15,

and day 20. To induce negative geotaxis, all the flies were gently tapped down to

the bottom of the vials, and their ability to climb up was recorded using a

standard video camera. Percentage of flies that climbed above 4 cm lines at 10

seconds was calculated as climbing index. This measurement point was

empirically determined using control flies. For lifespan analysis, 1 to 3 days old

male flies were collected into fresh food vials with density of 10 flies per vial.

Flies were transferred to fresh food vials every 2 days, and surviving flies were

counted at the time.

Larval crawling assay

Larval crawling assay was performed as previously described (Nichols et al,

2012). In brief, 3rd instar wandering larvae were collected and transferred onto a

1% agarose plate. The crawling distance within one minute was then measured.

For each genotype, at least 20 animals from two crosses were tested.

Immunohistochemistry

Ventral nerve cord (VNC) dissection and staining were performed as previously

described (Wu and Luo, 2006). In brief, VNCs of 3rd instar larvae were dissected

in cold PBS and fixed in 4% PFA for 20 minutes. Fixed samples were blocked

with blocking solution containing 5% normal goat serum and incubated with

primary antibodies at 4C overnight. They were then washed for three times with

113 PBT (0.3% Triton-X 100), incubated with secondary antibodies at room

temperature for three hours and mounted in Vectashield medium. At lease 3

animals per genotype were analyzed.

Larval NMJ preparation

Third instar wandering larvae were dissected in HL3.1 buffer and prepared as

previously described (Brent et al, 2009a, 2009b). In brief, after dissection, larval

fillets were fixed in 4% paraformaldehyde for 15 minutes, blocked for 30 minutes

with PBTN buffer containing 0.3% Triton X100, 5% normal goat serum, and

200mg/ml BSA, and incubated with primary antibodies at 4C overnight. The fillets

were then washed for several times, incubated with secondary antibodies at

room temperature for 1.5 hours, and mounted in Vectashield medium. For lifeact-

GFP analysis, ant-GFP antibody was used to amplify the signals. Antibodies:

mouse anti-Dlg (1:100, DHSB), mouse anti-CSP (1:100, DHSB), rabbit anti-GFP

(1:1000, Abcam), Cy3-HRP or Cy5-HRP (1:200, Jackson ImmunoResearch).

Larval NMJ analysis

NMJs on muscle 4 of segment A3 were chosen for all analyses. HRP were used

to delineate axons and Dlg were used as a post-synaptic marker. Bead-like varicosities that were apposed to Dlg were defined as major boutons. Ghost boutons referred to small immature boutons lacking Dlg apposition. Satellite boutons were designated if small boutons growing from major boutons or axonal

114 shafts overlapped with Dlg. For filopodia analysis, thin filaments protruding from

boutons and axonal shafts were labeled with either HRP or lifeact-GFP and

counted manually. For F-actin patch analysis, Lifeact-GFP positive patches

larger than 0.1 um2 were counted using particle analysis on ImageJ. The same setting and thresholds were applied to all genotypes during image acquisition and analysis. For all images, Z-stack images were acquired under 60X objective lens using an epifluorescence microscope (Ti, Nikon) equipped with a cooled

CCD camera. Images were then deconvolved (Autoquant, MediaCybernetics).

Maximum projections of processed images were then analyzed using ImageJ software.

Adult leg NMJ analysis

Leg NMJ dissection and analysis were performed as previously described

(Sreedharan et al, 2015). In short, first legs of male flies at 10 d.p.e were dissected and mounted directly in halocarbon oil 27. Images were acquired under

60X objective lens using a spinning-disc confocal microscope (Nikon D-Eclipse

C1). Maximum projection of NMJs at mid-femur were used for analysis.

115 Statistics analysis

One-way ANOVA followed by Tukey post-hoc test was used for all experiments except for lifespan analysis. For lifespan, Log-rank test was applied. P <0.05 was considered statistically significant.

116 CHAPTER IV: GENERAL DISCUSSION AND FUTURE DIRECTION

ALS is a progressive neurodegenerative disease that causes both upper

and lower motor neuron death. Approximately 90 percent of ALS cases are

sporadic (SALS), and 10 percent are inherited (FALS) (reviewed in Renton et al,

2014). Through genetic studies over the past decades, many gene mutations

have been identified as causative or associated with FALS, which account for

about 60% of cases. Nonetheless, it remains unresolved what genetic factors

contribute to the remaining cases. Adding hues of complexity to the picture,

these identified genes are involved in various cellular pathways and many gain-

of-function and loss-of-function mechanisms have been proposed for ALS-linked

gene mutations (reviewed in Robberecht and Philips, 2013). Although these

efforts contribute to our understanding of ALS pathogenesis, they highlight the molecular heterogeneity that underlies this disease.

Among these identified cellular pathways is the cytoskeletal network.

Cytoskeleton consists of two major networks, microtubule and actin, each of which consists of numerous associated proteins that fine-tune their functions.

They provide major support for neuronal structures, but also maintain crucial functions such as axonal transport and synaptic transmission (Cingolani and

Goda, 2008). They are particularly important in the case of ALS in that axons of motor neurons travel long distance from the spinal cord to distal neuromuscular junctions, making them highly susceptible to disruption of cytoskeletal functions.

ALS-associated gene mutations of several cytoskeletal components have also

117 been reported, such as PFN1, TUBA4A (Smith and Fallini et al, 2014), DCTN1

(Puls et al, 2003), and NEFH (Figlewicz et al, 1994; Al-chalabi et al, 1999).

Studies on these mutations have shed light on how cytoskeletal disruption can

contribute to neurodegeneration. For instance, Landers and colleagues reported

ALS-linked mutations in the TUBA4A gene, one of the major subunits of microtubules, and these mutants alter incorporation of TUBA4A into microtubules and disrupt microtubule dynamics through a dominant-negative mechanism

(Smith and Fallini et al, 2014). These findings thus lend strong support to the

relevance of cytoskeletal alteration to ALS pathogenesis.

Using exome sequencing, our genetic study identified several mutations in

the PFN1 gene, C71G, M114T, E117G and G118V. Profilin 1 is a mostly

cytoplasmic protein of about 15 kDa (reviewed in Witke, 2004). It has high affinity

to G-actin, and promotes actin polymerization under optimal concentration. Apart

from actin, in vitro studies demonstrate that profilin 1 also interacts with

phospholipids (PI(4,5)P2) and ligands containing PLP stretches, indicating it engages in various signaling pathways. Nonetheless, it is still unknown whether

ALS-linked PFN1 mutations disrupt these functions or even pose gain-of-function

mechanisms that lead to neurodegeneration.

Towards this end, in this dissertation we have characterized ALS-linked

PFN1 mutations in both mammalian cultured cells and Drosophila models. In

cultured cells, we first demonstrate that ALS-linked PFN1 mutants form

ubiquitinated aggregates. This finding is important in that it recaptures common

118 histopathological hallmarks for ALS (Kanning et al, 2010). We also show that

ALS-linked PFN1 mutations reduce affinity to actin and alter F-actin/G-actin ratio

in growth cones of motor neurons, indicating a loss-of-function mechanism.

These mutants also have a deleterious impact on neuronal morphology (Chapter

II). In order to better understand how ALS-linked PFN1 mutants lead to

neurodegeneration in vivo, we generate a novel Drosophila model expressing human PFN1 (Chapter III). In larval NMJs, ALS-linked PFN1 mutants have a

partial loss-of-function effect on filopodia formation. In adult flies, overexpression

of wild-type PFN1 alone causes locomotion defects, shorter lifespan and NMJ

disorganization. These defects indicate that PFN1 level influences function of

adult motor neurons. Interestingly, ALS-linked PFN1 mutants result in milder

toxicity compared to wild-type PFN1. This difference further supports ALS-linked

PFN1 mutants possess partial loss-of-function effects in regulating its cognate

cellular pathways. Taken together, our work has contributed to understanding of

PFN1-associated ALS pathogenesis. Because PFN1 overexpression results in

locomotion defects and adult NMJ disorganization, our Drosophila model will also

serve as a valuable tool to dissect molecular mechanisms of PFN1-induced

neurodegeneration.

4.1 Propensity to aggregate of ALS-linked PFN1 mutants

In Chapter II, we demonstrate ALS-linked profilin 1 mutants possess the

propensity to aggregate by showing they form ubiquitinated aggregates in

119 primary motor neurons and immortalized cell lines. Molecular analyses also

confirm these mutants are re-localized in detergent- resistant fractions and conjugated with poly-ubiquitin. Proteasome inhibition by MG-132 further enhances aggresome formation, indicating ubiquitin-proteasome system is major machinery for clearance of misfolded ALS-linked mutant proteins (Tyedmers et al, 2010). We also attempt to confirm whether we can detect aggregates in tissues from ALS patients harboring PFN1 mutations. Surprisingly, western blot analysis shows there is no increased insoluble PFN1 fraction in patient-derived lymphoblast cell lines harboring C71G mutation. We speculate that cellular environments may be different in immune and central nervous system, thereby rendering different cellular susceptibility to aggregation. Future studies on neural tissues or motor neurons derived from iPSCs of ALS patients will help to confirm this speculation.

It is also interesting that different ALS-linked PFN1 mutations have varied propensity to aggregate. In Chapter II, using cellular staining and solubility analysis, we demonstrate that C71G mutation confers the highest propensity to aggregate. This result is also supported by a later study showing recombinant profilin 1 C71G mutant misfolds most drastically under in vitro denaturing condition (Boopathy et al, 2014). In comparison, another mutation E117G, which was observed in non-ALS controls and SALS patients in our genetic study, does not form aggregates in cultured cells and remains soluble in solubility analysis.

Because this mutant also maintains similar affinity to actin compared to wild-type

120 profilin 1, we argue that this mutation may be non-pathogenic or pathogenic but

less penetrant. Our data suggest the latter because when the proteasome is

inhibited, profilin 1 E117G starts forming aggregates while wild-type profilin 1

remains mostly diffuse. In support of this finding, another genetic study reports

that E117G is a moderate risk factor for FALS (Fratta et al, 2014), indicative of its

inherent pathogenicity.

The presence of profilin 1 aggregates also raises the possibility that they

may pose a gain-of-function mechanism by sequestering other cellular factors.

Indeed, in primary motor neurons transfected with mutant profilin 1, we find that

approximately 30% of profilin 1 aggregates co-stain TDP-43, another ALS-linked

protein. This finding is important for two reasons. Firstly, TDP-43 can often be

detected in aggregates in ALS patients harboring different gene mutations

(Neumann et al, 2006). The presence of TDP-43 in mutant profilin 1 aggregates

further supports that ALS-linked PFN1 mutations recapture common ALS-

relevant histopathological characteristics. Secondly, TDP-43 is an RNA-binding

protein that localizes mostly in the nucleus and regulates various aspects of RNA

processing. Thus, translocalization of TDP-43 into the cytoplasm may affect its

nuclear functions and result in dys-regulated RNA processing and gene expression. This sequestration is specific to TDP-43, because we do not detect

other nuclear proteins, such as SMN and FUS, in PFN1 aggregates. It remains

unknown whether profilin 1 aggregates also sequester cellular factors other than

TDP-43, thereby altering different cellular pathways.

121 In Chapter III, we attempt to test if ALS-linked profilin 1 mutants also form

aggregates in the Drosophila model. In larval ventral nerve cord, where motor

neurons reside, both wild-type and C71G profilin 1 remain diffuse in the

cytoplasm without formation of visible aggregates. Because in Chapter II, we

demonstrate that proteasome inhibition enhances profilin 1 aggregation in

mammalian cultured cells, we wonder if it also promotes aggregate formation in

fly tissues. Prosbeta21 is a temperature-sensitive dominant-negative mutant of proteasome subunit, which would create a proteasome-defective cellular environment (Smyth and Belote 1999). Nonetheless, co-expression of

Prosbeta21 and profilin 1 C71G in motor neurons does not promote aggregate formation, as demonstrated in larval ventral nerve cord. Mild disorganization of cellular morphology is found, although it is more likely due to the effect of proteasome inhibition. There are several possible explanations for the discrepancy between results from mammalian cultured cells and Drosophila models. Firstly, it is not known if mutant profilin 1 possess the same protein stability or propensity to aggregate in Drosophila, which may have different cellular factors involved in protein folding and thus have different thresholds for protein aggregation. Secondly, it is also possible that mutant profilin 1 gradually develop aggregates during aging, which cannot be detected because of early lethal toxicity resulting from profilin 1 overexpression. Lastly, formation of aggregates may require a second-hit triggering factor, such as oxidative stress and axonal injury. That said, despite the absence of aggregates in our Drosophila

122 system, this model provides a good platform for us to dissect loss-of-function and

gain-of-function effects of soluble mutant profilin 1 without being confounded by

effects contributed by aggregates, as discussed in Chapter III.

4.2 Loss of function of ALS-linked PFN1 mutations in actin polymerization

Structural studies of profilin 1 have shown that ALS-linked residues are located in close proximity to actin-binding residues (Schutt et al, 1993), raising the possibility that ALS-linked PFN1 mutations can impair its actin binding, thereby altering actin polymerization. To address this question, we utilized both in vitro mammalian cell cultures and in vivo Drosophila models. As demonstrated in

Chapter II, when transiently transfected in mouse primary motor neurons, ALS- linked PFN1 mutations reduce F-actin/G-actin ratio in growth cones compared to wild-type profilin 1. This result indicates F-actin formation is less efficient in axonal termini in the presence of ALS-linked profilin 1. Co-immunoprecipitation experiments support this finding by showing that ALS-linked profilin1 mutants reduce affinity to actin under stringent buffer condition. In Drosophila, profilin is crucial for motor axon outgrowth in the Drosophila embryo because depletion of profilin leads to growth cone arrest (Wills et al, 1999). As shown in Chapter III, overexpression of either wild-type or mutant profilin 1 in motor neurons does not alter synapse formation at early developmental stages, as demonstrated by normal bouton number and size in larval NMJs. Nonetheless, it results in filopodia formation and increase in F-actin patches in the synapse. This result supports a recent study in which profilin 1 overexpression favors formation of

123 filopodia over lamellopodia through promoting linear Ena/VASP-dependent actin

assembly (Rotty et al, 2015). This study also stresses the importance of profilin

1 level in balancing F-actin substructures through coordinating with different

actin-interacting components. Interestingly, filopodia formation is also observed

in wild-type larval NMJs when they undergo spaced high-potassium stimulation.

This filopodia formation is accompanied by up-regulated trans-synaptic Wnt/Wg

signaling pathway (Ataman et al, 2008). Thus it will be worthwhile investigating

whether profilin 1 regulates (or is regulated by) activity-dependent synaptic

function through these pathways. When comparing with wild-type profilin 1, we

find M114T mutant forms fewer filopodia, and leads to mild reduction in F-actin content in the synapse. This result is important in that it corroborates our finding in mammalian neuronal cultures that ALS-linked profilin 1 mutants have partial loss-of-function effects on actin polymerization, particularly in presynaptic regions. Although the number of filopodia is reduced, their average length is unaltered by M114T mutant. This raises the possibility that profilin 1 M114T may affect initiation, rather than elongation, of unbranched F-actin filaments. Future experiments on which actin-interacting components are affected by mutant profilin 1 will help to decipher this loss-of-function mechanism. On the other hand, while profilin 1 C71G mutant affects actin polymerization in vitro, in

Drosophila larval NMJs, this mutant leads to a more variable result in filopodia formation, difference of which is insignificant compared to wild-type profilin 1.

One possible explanation is that considering low conservation of sequences

124 between profilin 1 and chickadee, profilin 1 C71G may not reduce its affinity to

actin or other actin-associated cellular components in Drosophila system as much as in mammalian cells. Comparison of structures of profilin 1 C71G and

chickadee and investigation of their interaction profiles will thus help to confirm this speculation. The other possibility is that in mammalian cells, profilin 1 C71G possesses highest propensity to aggregate, and therefore may affect actin polymerization through sequestering functional actin-associated proteins into aggregates. Because we do not detect visible mutant profilin 1 aggregates in this

Drosophila model (discussed below), we cannot confirm this gain-of-function mechanism in vivo.

In this dissertation, we focus on effects of ALS-linked profilin 1 mutants on axonal termini, including growth cones in mammalian motor neurons and presynaptic NMJs in flies. Nonetheless, F-actin network also serves as important scaffold in postsynaptic regions (reviewed in Cingolani and Goda, 2008). Apart from structural support, actin dynamics contributes to anchoring and trafficking of neurotransmitter receptors, such as AMPA and NMDA receptors. Profilin has been shown to associate with this postsynaptic scaffolding network (Giesemann et al, 2003). For instance, profilin interacts with gephyrin, a major component of postsynaptic densities (PSDs) in inhibitory synapses. Gephyrin competes with G- actin and phospholipids for binding to profilin (Giesemann et al, 2003), implying the regulatory role of profilin in postsynaptic actin dynamics. Likewise, profilin is shown to target to dendritic spines upon electrical and chemical stimulation, and

125 blockage of this targeting alters structure of dendritic spines (Ackermann and

Matus, 2003). It is thus worthwhile investigating if ALS-linked profilin 1 mutants

also have loss-of-function effects on post-synaptic F-actin network and result in

functional consequences. A recent study supports the importance of this direction by demonstrating that profilin 1 C71G leads to postsynaptic morphological changes including increased dendritic arborization and spine formation in hippocampal neurons (Brettle et al, 2015). While the underlying molecular mechanisms have yet to be invesitigated, it stresses the potential impacts of

ALS-linked profilin 1 mutants on both presynaptic and postsynaptic sites of neurons.

4.3 Gain of function of ALS-linked PFN1 mutations in neuronal integrity

In section 4.2, we have discussed loss of function of ALS-linked PFN1 mutations in actin polymerization. Nontheless, since these mutants have high propensity to aggregate, it is also likely that the observation of altered actin dynamics and neuronal morphology in mouse primary motor neurons is in part, if not all, due to gain of function of these profilin 1 mutants. In Chapter II, our coimmunoprecipitation result suggests that ALS-linked PFN1 mutations reduce its affinity to actin; however, this interaction was measured indirectly using the whole cellular lysates. Therefore, it cannot be excluded that this alteration may result from toxic effects of profilin 1 aggregates, rather than reduction in actin binding. Future experiments utilizing recombinant mutant profilin 1 and actin are required to clarify whether their direct interaction is inherently affected. Profilin 1

126 aggregates may impair actin polymerization through several mechanisms. Firstly,

the formation of profilin 1 aggregates may reduce free profilin 1 available for actin

binding and thus actin polymerization. Secondly, profilin 1 aggregates may sequester other celluar components involved in actin polymerization, thereby altering actin dynamics in the growth cones of primary motor neruons. This alteration further impairs neuronal morphologies including axonal length and growth cone area. Likewise, the presence of profilin aggregates can also cause abnormal neuronal phenotypes through other actin-independent pathways, such as sequestering other cellular factors important in neuronal integrity and disrupting protein quality control. Thus it will be worthwhile invesitating the contributions of these hypotheses to altered neuronal morophologies. On the other hand, in Chapter III, studies on larval NMJs suggest that mutant profilin 1 have less filopodia formation despite the lack of aggregates . This result hints that ALS-linked PFN1 mutations can still affect actin polymerization independent of aggregate formation. Therefore, it is likely that both gain of function and loss of function of ALS-linked PFN1 mutations play a role in neurodegeneration.

4.4 PFN 1 level regulates adult locomotion and lifespan

The disease onset of ALS is around 50 years of age, accompanied by rapid loss of motor functions (reviewed in Wijesekera and Leigh, 2009). To test if

ALS-linked profilin 1 mutants cause adult-onset locomotion defects, we measured climbing ability and lifespan in adult profilin 1-overexpressing flies at

127 different ages. Interestingly, profilin overexpression in glutamatergic neurons,

which include motor neurons, results in progressive climbing defects and shorter

lifespan. The behavioral defects are accompanied by signs of neuromuscular

degeneration in adult legs. These phenotypes are important in that they reflect

ALS-associated pathogenesis in several aspects. Firstly, profilin 1-induced

toxicity is specific to ALS-relevant cell types. While profilin 1 overexpression in

glia results in similarly shorter lifespan, its overexpression in the eye tissue,

which include photoreceptor cells, does not cause rough eye phenotypes.

Secondly, profilin 1-induced locomotion defects manifests only in aging flies. At

3rd instar larval stage, despite morphological changes discussed above, profilin 1 overexpression does not impair larval crawling ability. Likewise, at adult stage, the locomotion defect is not observed in young flies and shows progressive pattern with age. Collectively, our data indicate the importance of profilin 1 level in maintenance of locomotion function and integrity of neuromuscular junctions in adult flies. It will thus be interesting to see if profilin 1 expression level is altered in tissues from ALS patients harboring different gene mutations. Intriguingly, when comparing the toxicity caused by wild-type and mutant profilin 1, we find that toxicity of wild-type profilin 1 is significantly more severe than ALS-linked profilin 1 mutants, particularly in lifespan analysis. This difference does not seem due to increased protein turnover of mutant profilin 1, because western blot analysis shows no reduced expression level. Therefore we argue that mutant profilin 1 have partial loss-of-function effects in up-regulating their cognate

128 cellular pathways to the same toxic level as wild-type profilin 1. It has yet to be

clarified as to what molecular mechanisms, actin-dependent or –independent, underlie this differential toxicity.

4.5 Non-cell autonomous role of PFN1 in neuronal integrity

Astrocytes are one of the integral components of tripartite synapses that

contribute to maintenance of synaptic structure and functions (reviewed in Eroglu

and Barres, 2010). Apart from axons, profilin 1 has been shown to modulate

astrocytic processes in an actin-dependent manner (Molotkov et al, 2013;

Schweinhuber et al, 2015). Peripherally, another study demonstrates that proflin

1 serves as an effector of Rho/ROCK signaling pathway and promotes

myelination of Schwann cells (Montani et al, 2014). These studies highlight that

profilin 1 is a multi-functional player in different glia subtypes. In Chapter III, at

3rd instar larval stage, we first demonstrate that endogenous chickadee is enriched in peripheral glia that enwrap segmental nerves. But in order to clarify the contribution of glial profilin 1 to neuronal integrity, it will require further confirmation as to whether glial profilin 1 depletion causes wrapping defects and affects axonal functions. Likewise, at adult stage, our observation that profilin 1 overexpression in glia results in toxicity further confirms the importance of profilin

1 level in different ALS-relevant cell types. Over the last few years, non-cell autonomous toxicity has been brought into the picture of ALS pathogenesis

129 (reviewed in llieva et al, 2009 and Valori et al, 2014). For instance, in both ALS

patients (Rothstein et al, 1995) and ALS rodent models (Howland et al, 2002),

loss of glial glutamate transporters has been reported, which may cause

excitotoxicity through disruption of glutamate buffering by glia. By investigating if neuronal activity or synaptic transmission is altered with increased profilin 1 level in glia, it will elucidate non-cell-autonomous contribution of profilin 1 to

neurodegeneration. Lastly, similar to motor neurons, wild-type profilin 1

overexpression in glia results in significantly shorter lifespan than mutant profilin

1. This result again suggest ALS-linked profilin 1 mutants may have a loss-of-

function effect in glia as well as in motor neurons.

4.6 Advantages of Drosophila models expressing human PFN1

Throughout this dissertation, we have compared several differences

between mammalian cultured cells and Drosophila models expressing ALS-

linked PFN1 mutants (summarized in Table IV-1). Despite the lack of

aggregates, our Drosophila models provide a good platform for following

applications:

Identification of genetic interaction between SOD1 and PFN1

In Chapter III, we discussed the advantage of ALS Drosophila models as a

genetic tool to identify relevant cellular pathways. In our Drosophila model,

130 profilin 1 overexpression in motor neurons leads to adult phenotypes including

locomotion defects, shorter lifespan, and leg NMJ degeneration. These

phenotypes not only stress the importance of profilin 1 level in adult NMJs, but

can also be used to dissect motor neuron-specific cellular pathways associated

with profilin 1. Towards this end, we first wonder if profilin 1 and other ALS-linked

proteins are involved in common cellular pathways. We observe that amorphic

SOD1 mutants, but not TDP-43 and ataxin-2, modify profilin 1-induced locomotion defects, indicative of their genetic interaction. SOD1 is a cytoplasmic superoxide dismutase responsible for scavenging cellular free superoxide radicals (Fridovich, 1986). We also test SOD2, a mitochondria-specific isozyme, to clarify if this modifying effect is shared by other factors involved in reduction of oxidative stress. Loss-of-function SOD2 mutants fail to enhance profilin 1- induced climbing defects, implying the genetic interaction is specific to SOD1.

We further confirm this genetic interaction by measuring NMJ morphology at 3rd

instar larval stage. Sweeney and colleagues have used chemical stressors and

homozygous amorphic SOD mutants to demonstrate that oxidative stress

induces larval NMJ overgrowth (Milton et al, 2011). In line with their finding, we

show that homozygous SODn1 mutant results in an increase in bouton number.

Interestingly, while neither heterozygous SODn1 allele nor profilin 1

overexpression alone alters bouton number, these two genetic conditions concert

to induce NMJ overgrowth. Because NMJ overgrowth has been associated with

oxidative stress, it will be worthwhile to investigate if profilin 1 overexpression

131 contributes to formation of oxidative stress in heterozygous SODn1 background,

thereby resulting in increased bouton numbers. It is plausible that profilin 1 level

regulates SOD1 activity, dimerization, or maturation through an unknown

pathway. Likewise, although SOD1 localizes in the cytoplasm, it is also reported

to be present in the intermembrane space of mitochondria with elusive roles.

While some reported that SOD1 in the mitochondrial intermembrane space increased toxic ROS (Goldsteins et al, 2008), others showed that targeting of

SOD1 into the same space rescued axonopathy in Sod1-/- mouse models

(Fischer et al, 2011). Furthermore, quality control of mitochondria involves their fission and fusion, which are particularly important during cellular stress

(reviewed in Youle and van der Bliek, 2012), and these processes are regulated

by cytoskeletal dynamics (reviewed in van der Bliek et al, 2013). It is thus

possible that profilin 1 overexpression can alter actin network that maintains

mitochondria dynamics, thus affecting the distribution of intermembrane SOD1.

On the other hand, considering that heterozygous SODn1 mutant

moderately affects longevity of adult flies, it may still induce oxidative stress, despite to a milder level than the homozygous background. Therefore, it is more likely that it is the formation of oxidative stress that modulates functions of profilin

1, thereby enhancing its toxicity. it will thus be helpful to test if small moleculte- induced oxidative stress can also modify phenotypes of profilin 1 overexpression.

It is interesting that in larval NMJ, SODn1 allele does not alter profilin 1-induced filopodia formation but contributes to overall NMJ overgrowth. This differential

132 modification hints that in response to oxidative stress, profilin 1 may regulate

specific F-acitn substructures in presynaptic sites of NMJs. It will be worthwhile

investigating the functional outcome of this synaptic alteration, e.g. whether synaptic transmission is affected. In fact, the relevance of profilin 1 level to cellular stress is supported by a study showing that in Drosophila S2 cells, osmotic stress upregulates gene expression of chickadee through MAPK/JNK pathway (Suganuma et al, 2010). While it still remains unknown as to the functional link between profilin and cellular stress, our finding does identify a possible role of profilin 1 in regulation of motor neuron function during oxidative stress.

To identify novel cellular pathways regulated by wild-type PFN1

As discussed in Chapter I, profilin 1 interacts with different cellular factors, which include G-actin, PLP ligands, and PI(4,5)P2, and is involved in various cellular pathways. Profilin 1 has also been implicated in multiple neuronal functions, in particular during early development (reviewed in Birbach, 2008).

However, how it contributes to integrity of adult motor neurons and neuromuscular junctions still remains elusive. In the earlier section, we have discussed the advantage of this Drosophila model by demonstrating the genetic interaction between SOD1 and profilin 1. In order to uncover other novel profilin

1-associated cellular pathways, the adult NMJ phenotype of profilin 1 overexpression will serve as valuable readout for genetic screens. Considering

133 ALS is an adult-onset motor neuron disease, this phenotype is age- and cell

type- relevant. Identification of molecular mechanisms that maintain adult NMJs

will help us better understand functional roles of profilin 1 in aging motor neurons

and how its dysregulation may associate with neurodegeneration.

To uncover loss-of-function or gain-of-function mechanisms of mutant PFN1

Throughout this dissertation, we have demonstrated that ALS-linked PFN1

mutations lead to partial loss of function in actin polymerization. In particular, in

our Drosophila model, this observation is mostly inferred from the difference in toxicity compared to wild-type profilin 1. Since profilin 1 overexpression alone can

dys-regulate actin cytoskeleton and cause strong locomotion defects, this toxicity

may mask any adverse effects that are specific to ALS-linked profilin 1 mutants.

Thus, it is possible that these mutations have loss-of-function effects in other actin-independent pathways or even acquire gain-of-function mechanisms that contribute to neurodegeneration. Future genetic screens on NMJ phenotypes in

flies expressing ALS-linked profilin 1 mutants will thus reveal cellular pathways

that are altered by mutant prfoilin 1 and shed light on profilin 1-associated ALS

pathogenesis.

134

BIBLIOGRAPHY

Ackermann, M., and Matus, A. (2003). Activity-induced targeting of profilin and stabilization of dendritic spine morphology. Nat Neurosci 6, 1194–1200.

Alami, N.H., Smith, R.B., Carrasco, M.A., Williams, L.A., Winborn, C.S., Han, S.S.W., Kiskinis, E., Winborn, B., Freibaum, B.D., … Taylor, J.P. (2014). Axonal transport of TDP-43 mRNA granules is impaired by ALS-causing mutations. Neuron 81, 536–543.

Al-Chalabi, A., Andersen, P.M., Nilsson, P., Chioza, B., Andersson, J.L., Russ, C., Shaw, C.E., Powell, J.F., and Leigh, P.N. (1999). Deletions of the Heavy Neurofilament Subunit Tail in Amyotrophic Lateral Sclerosis. Hum. Mol. Genet. 8, 157–164.

Andersen, P.M. (2006). Amyotrophic lateral sclerosis associated with mutations in the CuZn superoxide dismutase gene. Curr Neurol Neurosci Rep 6, 37–46.

Andersen, P.M., and Al-Chalabi, A. (2011). Clinical genetics of amyotrophic lateral sclerosis: what do we really know? Nat Rev Neurol 7, 603–615.

Andrus, P.K., Fleck, T.J., Gurney, M.E., and Hall, E.D. (1998). Protein oxidative damage in a transgenic mouse model of familial amyotrophic lateral sclerosis. J. Neurochem. 71, 2041–2048.

Arnold, E.S., Ling, S.-C., Huelga, S.C., Lagier-Tourenne, C., Polymenidou, M., Ditsworth, D., Kordasiewicz, H.B., McAlonis-Downes, M., Platoshyn, O., … Cleveland, D.W. (2013). ALS-linked TDP-43 mutations produce aberrant RNA splicing and adult-onset motor neuron disease without aggregation or loss of nuclear TDP-43. PNAS 110, E736–E745.

Arrasate, M., Mitra, S., Schweitzer, E.S., Segal, M.R., and Finkbeiner, S. (2004). Inclusion body formation reduces levels of mutant huntingtin and the risk of neuronal death. Nature 431, 805–810.

Ataman, B., Ashley, J., Gorczyca, M., Ramachandran, P., Fouquet, W., Sigrist, S.J., and Budnik, V. (2008). Rapid Activity-Dependent Modifications in Synaptic Structure and Function Require Bidirectional Wnt Signaling. Neuron 57, 705–718.

Barber, S.C., Mead, R.J., and Shaw, P.J. (2006). Oxidative stress in ALS: A mechanism of neurodegeneration and a therapeutic target. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease 1762, 1051–1067.

136 Baron, D.M., Kaushansky, L.J., Ward, C.L., Sama, R.R.K., Chian, R.-J., Boggio, K.J., Quaresma, A.J.C., Nickerson, J.A., and Bosco, D.A. (2013). Amyotrophic lateral sclerosis-linked FUS/TLS alters stress granule assembly and dynamics. Mol Neurodegener 8, 30.

Basso, M., Samengo, G., Nardo, G., Massignan, T., D’Alessandro, G., Tartari, S., Cantoni, L., Marino, M., Cheroni, C., … Bonetto, V. (2009). Characterization of Detergent-Insoluble Proteins in ALS Indicates a Causal Link between Nitrative Stress and Aggregation in Pathogenesis. PLoS ONE 4, e8130.

Bauer, P.O., Wong, H.K., Oyama, F., Goswami, A., Okuno, M., Kino, Y., Miyazaki, H., and Nukina, N. (2009). Inhibition of Rho Kinases Enhances the Degradation of Mutant Huntingtin. J. Biol. Chem. 284, 13153–13164.

Bensimon, G., Lacomblez, L., and Meininger, V. (1994). A Controlled Trial of Riluzole in Amyotrophic Lateral Sclerosis. New England Journal of Medicine 330, 585–591.

Birbach, A. (2008). Profilin, a multi-modal regulator of neuronal plasticity. Bioessays 30, 994–1002.

Bliek, A.M. van der, Shen, Q., and Kawajiri, S. (2013). Mechanisms of Mitochondrial Fission and Fusion. Cold Spring Harb Perspect Biol 5, a011072.

Boopathy, S., Silvas, T.V., Tischbein, M., Jansen, S., Shandilya, S.M., Zitzewitz, J.A., Landers, J.E., Goode, B.L., Schiffer, C.A., and Bosco, D.A. (2015). Structural basis for mutation-induced destabilization of profilin 1 in ALS. PNAS 112, 7984–7989.

Bosco, D.A., Lemay, N., Ko, H.K., Zhou, H., Burke, C., Kwiatkowski, T.J., Sapp, P., McKenna-Yasek, D., Brown, R.H., and Hayward, L.J. (2010a). Mutant FUS Proteins that Cause Amyotrophic Lateral Sclerosis Incorporate into Stress Granules. Hum. Mol. Genet. ddq335.

Bosco, D.A., Morfini, G., Karabacak, N.M., Song, Y., Gros-Louis, F., Pasinelli, P., Goolsby, H., Fontaine, B.A., Lemay, N., … Jr, R.H.B. (2010b). Wild-type and mutant SOD1 share an aberrant conformation and a common pathogenic pathway in ALS. Nat Neurosci 13, 1396–1403.

Bowerman, M., Shafey, D., and Kothary, R. (2007). Smn depletion alters profilin II expression and leads to upregulation of the RhoA/ROCK pathway and defects in neuronal integrity. J. Mol. Neurosci. 32, 120–131.

Bowerman, M., Anderson, C.L., Beauvais, A., Boyl, P.P., Witke, W., and Kothary,

137 R. (2009). SMN, profilin IIa and plastin 3: A link between the deregulation of actin dynamics and SMA pathogenesis. Molecular and Cellular Neuroscience 42, 66– 74.

Bradke, F., Fawcett, J.W., and Spira, M.E. (2012). Assembly of a new growth cone after axotomy: the precursor to axon regeneration. Nat. Rev. Neurosci. 13, 183–193.

Brand, A.H., and Perrimon, N. (1993). Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118, 401– 415.

Bras, J., Guerreiro, R., and Hardy, J. (2012). Use of next-generation sequencing and other whole-genome strategies to dissect neurological disease. Nat Rev Neurosci 13, 453–464.

Brent, J., Werner, K., and McCabe, B.D. (2009a). Drosophila Larval NMJ Immunohistochemistry. Journal of Visualized Experiments.

Brent, J.R., Werner, K.M., and McCabe, B.D. (2009b). Drosophila Larval NMJ Dissection. Journal of Visualized Experiments.

Brettle, M., Suchowerska, A.K., Chua, S.W., Ittner, L.M., and Fath, T. (2015). Amyotrophic lateral sclerosis-associated mutant profilin 1 increases dendritic arborisation and spine formation in primary hippocampal neurons. Neuroscience Letters 609, 223–228.

Bruneteau, G., Bauché, S., Gonzalez de Aguilar, J.L., Brochier, G., Mandjee, N., Tanguy, M.-L., Hussain, G., Behin, A., Khiami, F., … Hantaï, D. (2015). Endplate denervation correlates with Nogo-A muscle expression in amyotrophic lateral sclerosis patients. Ann Clin Transl Neurol 2, 362–372.

Brzustowicz, L.M., Lehner, T., Castilla, L.H., Penchaszadeh, G.K., Wilhelmsen, K.C., Daniels, R., Davies, K.E., Leppert, M., Ziter, F., and Wood, D. (1990). Genetic mapping of chronic childhood-onset spinal muscular atrophy to chromosome 5q11.2-13.3. Nature 344, 540–541.

Burnett, B.G., Andrews, J., Ranganathan, S., Fischbeck, K.H., and Di Prospero, N.A. (2008). Expression of expanded polyglutamine targets profilin for degradation and alters actin dynamics. Neurobiology of Disease 30, 365–374.

Carrì, M.T., Valle, C., Bozzo, F., and Cozzolino, M. (2015). Oxidative stress and mitochondrial damage: importance in non-SOD1 ALS. Front. Cell. Neurosci. 9, 41.

138 de Carvalho, M., Dengler, R., Eisen, A., England, J.D., Kaji, R., Kimura, J., Mills, K., Mitsumoto, H., Nodera, H., … Swash, M. (2008). Electrodiagnostic criteria for diagnosis of ALS. Clinical Neurophysiology 119, 497–503.

Cingolani, L.A., and Goda, Y. (2008). Actin in action: the interplay between the actin cytoskeleton and synaptic efficacy. Nat Rev Neurosci 9, 344–356.

Cooley, L., Verheyen, E., and Ayers, K. (1992). chickadee encodes a profilin required for intercellular cytoplasm transport during Drosophila oogenesis. Cell 69, 173–184.

Corrado, L., Carlomagno, Y., Falasco, L., Mellone, S., Godi, M., Cova, E., Cereda, C., Testa, L., Mazzini, L., and D’Alfonso, S. (2011). A novel peripherin gene (PRPH) mutation identified in one sporadic amyotrophic lateral sclerosis patient. Neurobiology of Aging 32, 552.e1–e552.e6.

Cowan, C.M., and Mudher, A. (2013). Are tau aggregates toxic or protective in tauopathies? Front. Neurol. 4, 114.

Cragnaz, L., Klima, R., Skoko, N., Budini, M., Feiguin, F., and Baralle, F.E. (2014). Aggregate formation prevents dTDP-43 neurotoxicity in the Drosophila melanogaster eye. Neurobiology of Disease 71, 74–80.

Da Cruz, S., and Cleveland, D.W. (2011). Understanding the role of TDP-43 and FUS/TLS in ALS and beyond. Curr Opin Neurobiol 21, 904–919.

DeJesus-Hernandez, M., Mackenzie, I.R., Boeve, B.F., Boxer, A.L., Baker, M., Rutherford, N.J., Nicholson, A.M., Finch, N.A., Flynn, H., … Rademakers, R. (2011). Expanded GGGGCC Hexanucleotide Repeat in Noncoding Region of C9ORF72 Causes Chromosome 9p-Linked FTD and ALS. Neuron 72, 245–256.

Dengler, R., Konstanzer, A., Küther, G., Hesse, S., Wolf, W., and Strupplerdr, A. (1990). Amyotrophic lateral sclerosis: Macro–EMG and twitch forces of single motor units. Muscle Nerve 13, 545–550.

Deng, H.-X., Chen, W., Hong, S.-T., Boycott, K.M., Gorrie, G.H., Siddique, N., Yang, Y., Fecto, F., Shi, Y., … Siddique, T. (2011). Mutations in UBQLN2 cause dominant X-linked juvenile and adult-onset ALS and ALS/dementia. Nature 477, 211–215.

De Vos, K.J., Chapman, A.L., Tennant, M.E., Manser, C., Tudor, E.L., Lau, K.-F., Brownlees, J., Ackerley, S., Shaw, P.J., … Grierson, A.J. (2007). Familial amyotrophic lateral sclerosis-linked SOD1 mutants perturb fast axonal transport to reduce axonal mitochondria content. Hum. Mol. Genet. 16, 2720–2728.

139 Dickman, D.K., Lu, Z., Meinertzhagen, I.A., and Schwarz, T.L. (2006). Altered Synaptic Development and Active Zone Spacing in Endocytosis Mutants. Current Biology 16, 591–598.

Eroglu, C., and Barres, B.A. (2010). Regulation of synaptic connectivity by glia. Nature 468, 223–231.

Estes, P.S., Boehringer, A., Zwick, R., Tang, J.E., Grigsby, B., and Zarnescu, D.C. (2011). Wild-type and A315T mutant TDP-43 exert differential neurotoxicity in a Drosophila model of ALS. Hum. Mol. Genet. 20, 2308–2321.

Faivre-Sarrailh, C., Lena, J.Y., Had, L., Vignes, M., and Lindberg, U. (1993). Location of profilin at presynaptic sites in the cerebellar cortex; implication for the regulation of the actin-polymerization state during axonal elongation and synaptogenesis. J. Neurocytol. 22, 1060–1072.

Fallini, C., Bassell, G.J., and Rossoll, W. (2010). High-efficiency transfection of cultured primary motor neurons to study protein localization, trafficking, and function. Molecular Neurodegeneration 5, 17.

Fecto, F., Yan, J., Vemula, S.P., Liu, E., Yang, Y., Chen, W., Zheng, J.G., Shi, Y., Siddique, N., … Siddique, T. (2011). SQSTM1 mutations in familial and sporadic amyotrophic lateral sclerosis. Arch. Neurol. 68, 1440–1446.

Ferrante, R.J., Shinobu, L.A., Schulz, J.B., Matthews, R.T., Thomas, C.E., Kowall, N.W., Gurney, M.E., and Beal, M.F. (1997). Increased 3-nitrotyrosine and oxidative damage in mice with a human copper/zinc superoxide dismutase mutation. Ann. Neurol. 42, 326–334.

Figlewicz, D.A., Krizus, A., Martinoli, M.G., Meininger, V., Dib, M., Rouleau, G.A., and Julien, J.P. (1994). Variants of the heavy neurofilament subunit are associated with the development of amyotrophic lateral sclerosis. Hum. Mol. Genet. 3, 1757–1761.

Figley, M.D., Bieri, G., Kolaitis, R.-M., Taylor, J.P., and Gitler, A.D. (2014). Profilin 1 Associates with Stress Granules and ALS-Linked Mutations Alter Stress Granule Dynamics. J. Neurosci. 34, 8083–8097.

Fischer, L.R., Culver, D.G., Tennant, P., Davis, A.A., Wang, M., Castellano- Sanchez, A., Khan, J., Polak, M.A., and Glass, J.D. (2004). Amyotrophic lateral sclerosis is a distal axonopathy: evidence in mice and man. Experimental Neurology 185, 232–240.

Fischer, L.R., Igoudjil, A., Magrané, J., Li, Y., Hansen, J.M., Manfredi, G., and

140 Glass, J.D. (2011). SOD1 targeted to the mitochondrial intermembrane space prevents motor neuropathy in the Sod1 knockout mouse. Brain 134, 196–209.

Fratta, P., Charnock, J., Collins, T., Devoy, A., Howard, R., Malaspina, A., Orrell, R., Sidle, K., Clarke, J., … Fisher, E.M.C. (2014). Profilin1 E117G is a moderate risk factor for amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry 85, 506–508.

Freibaum, B.D., Lu, Y., Lopez-Gonzalez, R., Kim, N.C., Almeida, S., Lee, K.-H., Badders, N., Valentine, M., Miller, B.L., … Taylor, J.P. (2015). GGGGCC repeat expansion in C9orf72 compromises nucleocytoplasmic transport. Nature 525, 129–133.

Frey, D., Schneider, C., Xu, L., Borg, J., Spooren, W., and Caroni, P. (2000). Early and Selective Loss of Neuromuscular Synapse Subtypes with Low Sprouting Competence in Motoneuron Diseases. J. Neurosci. 20, 2534–2542.

Fridovich, I. (1986). Superoxide dismutases. Adv. Enzymol. Relat. Areas Mol. Biol. 58, 61–97.

Gargano, J.W., Martin, I., Bhandari, P., and Grotewiel, M.S. (2005). Rapid iterative negative geotaxis (RING): a new method for assessing age-related locomotor decline in Drosophila. Experimental Gerontology 40, 386–395.

Gieselmann, R., Kwiatkowski, D.J., Janmey, P.A., and Witke, W. (1995). Distinct biochemical characteristics of the two human profilin isoforms. Eur. J. Biochem. 229, 621–628.

Giesemann, T., Schwarz, G., Nawrotzki, R., Berhörster, K., Rothkegel, M., Schlüter, K., Schrader, N., Schindelin, H., Mendel, R.R., … Jockusch, B.M. (2003). Complex Formation between the Postsynaptic Scaffolding Protein Gephyrin, Profilin, and Mena: A Possible Link to the Microfilament System. J. Neurosci. 23, 8330–8339.

Giesemann, T., Rathke-Hartlieb, S., Rothkegel, M., Bartsch, J.W., Buchmeier, S., Jockusch, B.M., and Jockusch, H. (1999). A Role for Polyproline Motifs in the Spinal Muscular Atrophy Protein SMN PROFILINS BIND TO AND COLOCALIZE WITH SMN IN NUCLEAR GEMS. J. Biol. Chem. 274, 37908–37914.

Gijselinck, I., Van Langenhove, T., van der Zee, J., Sleegers, K., Philtjens, S., Kleinberger, G., Janssens, J., Bettens, K., Van Cauwenberghe, C., … Van Broeckhoven, C. (2012). A C9orf72 promoter repeat expansion in a Flanders- Belgian cohort with disorders of the frontotemporal lobar degeneration-

141 amyotrophic lateral sclerosis spectrum: a gene identification study. The Lancet Neurology 11, 54–65.

Goldsteins, G., Keksa-Goldsteine, V., Ahtoniemi, T., Jaronen, M., Arens, E., Åkerman, K., Chan, P.H., and Koistinaho, J. (2008). Deleterious Role of Superoxide Dismutase in the Mitochondrial Intermembrane Space. J. Biol. Chem. 283, 8446–8452.

Gould, T.W., Buss, R.R., Vinsant, S., Prevette, D., Sun, W., Knudson, C.M., Milligan, C.E., and Oppenheim, R.W. (2006). Complete Dissociation of Motor Neuron Death from Motor Dysfunction by Bax Deletion in a Mouse Model of ALS. J. Neurosci. 26, 8774–8786.

Gros-Louis, F., Larivière, R., Gowing, G., Laurent, S., Camu, W., Bouchard, J.-P., Meininger, V., Rouleau, G.A., and Julien, J.-P. (2004). A frameshift deletion in peripherin gene associated with amyotrophic lateral sclerosis. J. Biol. Chem. 279, 45951–45956.

Hanson, K.A., Kim, S.H., Wassarman, D.A., and Tibbetts, R.S. (2010). Ubiquilin Modifies TDP-43 Toxicity in a Drosophila Model of Amyotrophic Lateral Sclerosis (ALS). J. Biol. Chem. 285, 11068–11072.

Harel, N.Y., and Strittmatter, S.M. (2006). Can regenerating axons recapitulate developmental guidance during recovery from spinal cord injury? Nat Rev Neurosci 7, 603–616.

H Keshishian, K Broadie, A Chiba, and Bate, M. (1996). The Drosophila Neuromuscular Junction: A Model System for Studying Synaptic Development and Function. Annual Review of Neuroscience 19, 545–575.

Howland, D.S., Liu, J., She, Y., Goad, B., Maragakis, N.J., Kim, B., Erickson, J., Kulik, J., DeVito, L., … Rothstein, J.D. (2002). Focal loss of the glutamate transporter EAAT2 in a transgenic rat model of SOD1 mutant-mediated amyotrophic lateral sclerosis (ALS). Proc. Natl. Acad. Sci. U.S.A. 99, 1604–1609.

Ilieva, H., Polymenidou, M., and Cleveland, D.W. (2009). Non–cell autonomous toxicity in neurodegenerative disorders: ALS and beyond. J Cell Biol 187, 761– 772.

Johnson, J.O., Mandrioli, J., Benatar, M., Abramzon, Y., Van Deerlin, V.M., Trojanowski, J.Q., Gibbs, J.R., Brunetti, M., Gronka, S., … Traynor, B.J. (2010). Exome Sequencing Reveals VCP Mutations as a Cause of Familial ALS. Neuron 68, 857–864.

142 Johnson, J.O., Pioro, E.P., Boehringer, A., Chia, R., Feit, H., Renton, A.E., Pliner, H.A., Abramzon, Y., Marangi, G., … Traynor, B.J. (2014). Mutations in the Matrin 3 gene cause familial amyotrophic lateral sclerosis. Nat Neurosci 17, 664–666.

Kanning, K.C., Kaplan, A., and Henderson, C.E. (2010). Motor Neuron Diversity in Development and Disease. Annual Review of Neuroscience 33, 409–440.

Kim, H.-J., Raphael, A.R., LaDow, E.S., McGurk, L., Weber, R.A., Trojanowski, J.Q., Lee, V.M.-Y., Finkbeiner, S., Gitler, A.D., and Bonini, N.M. (2014). Therapeutic modulation of eIF2α phosphorylation rescues TDP-43 toxicity in amyotrophic lateral sclerosis disease models. Nat Genet 46, 152–160.

Kim, Y.-S., Furman, S., Sink, H., and VanBerkum, M.F.A. (2001). Calmodulin and profilin coregulate axon outgrowth in Drosophila. J. Neurobiol. 47, 26–38.

Kwiatkowski, T.J., Bosco, D.A., LeClerc, A.L., Tamrazian, E., Vanderburg, C.R., Russ, C., Davis, A., Gilchrist, J., Kasarskis, E.J., … Brown, R.H. (2009). Mutations in the FUS/TLS Gene on Chromosome 16 Cause Familial Amyotrophic Lateral Sclerosis. Science 323, 1205–1208.

Kwiatkowski, D.J., and Bruns, G.A. (1988). Human profilin. Molecular cloning, sequence comparison, and chromosomal analysis. J. Biol. Chem. 263, 5910– 5915.

Lacomblez, L., Bensimon, G., Leigh, P.N., Guillet, P., and Meininger, V. (1996). Dose-ranging study of riluzole in amyotrophic lateral sclerosis. Amyotrophic Lateral Sclerosis/Riluzole Study Group II. Lancet 347, 1425–1431.

Laird, F.M., Farah, M.H., Ackerley, S., Hoke, A., Maragakis, N., Rothstein, J.D., Griffin, J., Price, D.L., Martin, L.J., and Wong, P.C. (2008). Motor neuron disease occurring in a mutant dynactin mouse model is characterized by defects in vesicular trafficking. J. Neurosci. 28, 1997–2005.

Lambrechts, A. (1997). The mammalian profilin isoforms display complementary affinities for PIP2 and proline-rich sequences. The EMBO Journal 16, 484–494.

Lambrechts, A., Jonckheere, V., Peleman, C., Polet, D., Vos, W.D., Vandekerckhove, J., and Ampe, C. (2006). Profilin-I-ligand interactions influence various aspects of neuronal differentiation. Journal of Cell Science 119, 1570– 1578.

Lamprecht, R., Farb, C.R., Rodrigues, S.M., and LeDoux, J.E. (2006). Fear conditioning drives profilin into amygdala dendritic spines. Nat Neurosci 9, 481– 483.

143 Lanier, L.M., Gates, M.A., Witke, W., Menzies, A.S., Wehman, A.M., Macklis, J.D., Kwiatkowski, D., Soriano, P., and Gertler, F.B. (1999). Mena Is Required for Neurulation and Commissure Formation. Neuron 22, 313–325.

Lanson, N.A., Maltare, A., King, H., Smith, R., Kim, J.H., Taylor, J.P., Lloyd, T.E., and Pandey, U.B. (2011). A Drosophila model of FUS-related neurodegeneration reveals genetic interaction between FUS and TDP-43. Hum. Mol. Genet. 20, 2510–2523.

Lassing, I., and Lindberg, U. (1988). Specificity of the interaction between phosphatidylinositol 4,5-bisphosphate and the profilin:actin complex. J. Cell. Biochem. 37, 255–267.

Lee, E.B., Lee, V.M.-Y., and Trojanowski, J.Q. (2011). Gains or losses: molecular mechanisms of TDP43-mediated neurodegeneration. Nat Rev Neurosci 13, 38– 50.

Li, Y.R., King, O.D., Shorter, J., and Gitler, A.D. (2013). Stress granules as crucibles of ALS pathogenesis. J Cell Biol 201, 361–372.

Lindberg, U., Schutt, C.E., Hellsten, E., Tjäder, A.C., and Hult, T. (1988). The use of poly(L-proline)-Sepharose in the isolation of profilin and profilactin complexes. Biochim. Biophys. Acta 967, 391–400.

Liu, Q., and Dreyfuss, G. (1996). A novel nuclear structure containing the survival of motor neurons protein. EMBO J 15, 3555–3565.

Liu, R., Althaus, J.S., Ellerbrock, B.R., Becker, D.A., and Gurney, M.E. (1998). Enhanced oxygen radical production in a transgenic mouse model of familial amyotrophic lateral sclerosis. Ann. Neurol. 44, 763–770.

Liu-Yesucevitz, L., Bilgutay, A., Zhang, Y.-J., Vanderwyde, T., Citro, A., Mehta, T., Zaarur, N., McKee, A., Bowser, R., … Wolozin, B. (2010). Tar DNA Binding Protein-43 (TDP-43) Associates with Stress Granules: Analysis of Cultured Cells and Pathological Brain Tissue. PLoS ONE 5, e13250.

Li, Y., Ray, P., Rao, E.J., Shi, C., Guo, W., Chen, X., Woodruff, E.A., Fushimi, K., and Wu, J.Y. (2010). A Drosophila model for TDP-43 proteinopathy. PNAS 107, 3169–3174.

Lu, P.-J., Shieh, W.-R., Rhee, S.G., Yin, H.L., and Chen, C.-S. (1996). Lipid Products of Phosphoinositide 3-Kinase Bind Human Profilin with High Affinity. Biochemistry 35, 14027–14034.

144 Ludolph, A.C., and Jesse, S. (2009). Evidence-Based Drug Treatment in Amyotrophic Lateral Sclerosis and Upcoming Clinical Trials. Ther Adv Neurol Disord 2, 319–326.

Luo, L. (2002). Actin cytoskeleton regulation in neuronal morphogenesis and structural plasticity. Annu. Rev. Cell Dev. Biol. 18, 601–635.

Mahoney, N.M., Rozwarski, D.A., Fedorov, E., Fedorov, A.A., and Almo, S.C. (1999). Profilin binds proline-rich ligands in two distinct amide backbone orientations. Nat Struct Mol Biol 6, 666–671.

Mahr, A., and Aberle, H. (2006). The expression pattern of the Drosophila vesicular glutamate transporter: A marker protein for motoneurons and glutamatergic centers in the brain. Gene Expression Patterns 6, 299–309.

Mammoto, A., Sasaki, T., Asakura, T., Hotta, I., Imamura, H., Takahashi, K., Matsuura, Y., Shirao, T., and Takai, Y. (1998). Interactions of drebrin and gephyrin with profilin. Biochem. Biophys. Res. Commun. 243, 86–89.

Marie, B., Sweeney, S.T., Poskanzer, K.E., Roos, J., Kelly, R.B., and Davis, G.W. (2004). Dap160/Intersectin Scaffolds the Periactive Zone to Achieve High- Fidelity Endocytosis and Normal Synaptic Growth. Neuron 43, 207–219.

Maruyama, H., Morino, H., Ito, H., Izumi, Y., Kato, H., Watanabe, Y., Kinoshita, Y., Kamada, M., Nodera, H., … Kawakami, H. (2010). Mutations of optineurin in amyotrophic lateral sclerosis. Nature 465, 223–226.

Matus, S., Valenzuela, V., Medinas, D.B., Hetz, C., Matus, S., Valenzuela, V., Medinas, D.B., and Hetz, C. (2013). ER Dysfunction and Protein Folding Stress in ALS, ER Dysfunction and Protein Folding Stress in ALS. International Journal of Cell Biology, International Journal of Cell Biology 2013, 2013, e674751.

McDonald, K.K., Aulas, A., Destroismaisons, L., Pickles, S., Beleac, E., Camu, W., Rouleau, G.A., and Vande Velde, C. (2011). TAR DNA-binding protein 43 (TDP-43) regulates stress granule dynamics via differential regulation of G3BP and TIA-1. Hum. Mol. Genet. 20, 1400–1410.

Meijering, E., Jacob, M., Sarria, J.-C.F., Steiner, P., Hirling, H., and Unser, M. (2004). Design and validation of a tool for neurite tracing and analysis in fluorescence microscopy images. Cytometry A 58, 167–176.

Menon, K.P., Carrillo, R.A., and Zinn, K. (2013). Development and plasticity of the Drosophila larval neuromuscular junction. WIREs Dev Biol 2, 647–670.

145 Menzies, F.M., Cookson, M.R., Taylor, R.W., Turnbull, D.M., Chrzanowska‐ Lightowlers, Z.M.A., Dong, L., Figlewicz, D.A., and Shaw, P.J. (2002). Mitochondrial dysfunction in a cell culture model of familial amyotrophic lateral sclerosis. Brain 125, 1522–1533.

Metzler, W.J., Farmer, B.T., Constantine, K.L., Friedrichs, M.S., Mueller, L., and Lavoie, T. (1995). Refined solution structure of human profilin I. Protein Science 4, 450–459.

Michaelsen, K., Murk, K., Zagrebelsky, M., Dreznjak, A., Jockusch, B.M., Rothkegel, M., and Korte, M. (2010). Fine-tuning of neuronal architecture requires two profilin isoforms. PNAS 107, 15780–15785.

Milton, V.J., Jarrett, H.E., Gowers, K., Chalak, S., Briggs, L., Robinson, I.M., and Sweeney, S.T. (2011). Oxidative stress induces overgrowth of the Drosophila neuromuscular junction. PNAS 108, 17521–17526.

Mockrin, S.C., and Korn, E.D. (1980). Acanthamoeba profilin interacts with G- actin to increase the rate of exchange of actin-bound adenosine 5’-triphosphate. Biochemistry 19, 5359–5362.

Moloney, E.B., de Winter, F., and Verhaagen, J. (2014). ALS as a distal axonopathy: molecular mechanisms affecting neuromuscular junction stability in the presymptomatic stages of the disease. Front. Neurosci 8, 252.

Molotkov, D., Zobova, S., Arcas, J.M., and Khiroug, L. (2013). Calcium-induced outgrowth of astrocytic peripheral processes requires actin binding by Profilin-1. Cell Calcium 53, 338–348.

Montani, L., Buerki-Thurnherr, T., Faria, J.P. de, Pereira, J.A., Dias, N.G., Fernandes, R., Gonçalves, A.F., Braun, A., Benninger, Y., … Relvas, J.B. (2014). Profilin 1 is required for peripheral nervous system myelination. Development 141, 1553–1561.

Murk, K., Wittenmayer, N., Michaelsen-Preusse, K., Dresbach, T., Schoenenberger, C.-A., Korte, M., Jockusch, B.M., and Rothkegel, M. (2012). Neuronal Profilin Isoforms Are Addressed by Different Signalling Pathways. PLoS ONE 7, e34167.

Mutihac, R., Alegre-Abarrategui, J., Gordon, D., Farrimond, L., Yamasaki-Mann, M., Talbot, K., and Wade-Martins, R. (2015). TARDBP pathogenic mutations increase cytoplasmic translocation of TDP-43 and cause reduction of endoplasmic reticulum Ca2 + signaling in motor neurons. Neurobiology of

146 Disease 75, 64–77.

Nagase, M., Yamamoto, Y., Miyazaki, Y., and Yoshino, H. (2015). Increased oxidative stress in patients with amyotrophic lateral sclerosis and the effect of edaravone administration. Redox Rep.

Neuhoff, H., Sassoè-Pognetto, M., Panzanelli, P., Maas, C., Witke, W., and Kneussel, M. (2005). The actin-binding protein profilin I is localized at synaptic sites in an activity-regulated manner. European Journal of Neuroscience 21, 15– 25.

Neumann, M., Sampathu, D.M., Kwong, L.K., Truax, A.C., Micsenyi, M.C., Chou, T.T., Bruce, J., Schuck, T., Grossman, M., … Lee, V.M.-Y. (2006). Ubiquitinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclerosis. Science 314, 130–133.

Nichols, C.D., Becnel, J., and Pandey, U.B. (2012). Methods to Assay Drosophila Behavior. Journal of Visualized Experiments.

Nodelman, I.M., Bowman, G.D., Lindberg, U., and Schutt, C.E. (1999). X-ray structure determination of human profilin II: a comparative structural analysis of human profilins1. Journal of Molecular Biology 294, 1271–1285.

Nölle, A., Zeug, A., Bergeijk, J. van, Tönges, L., Gerhard, R., Brinkmann, H., Rayes, S.A., Hensel, N., Schill, Y., … Claus, P. (2011). The spinal muscular atrophy disease protein SMN is linked to the rho-kinase pathway via profilin. Hum. Mol. Genet. 20, 4865–4878.

Parkinson, N., Ince, P.G., Smith, M.O., Highley, R., Skibinski, G., Andersen, P.M., Morrison, K.E., Pall, H.S., Hardiman, O., … FReJA Consortium (2006). ALS phenotypes with mutations in CHMP2B (charged multivesicular body protein 2B). Neurology 67, 1074–1077.

Pawson, C., Eaton, B.A., and Davis, G.W. (2008). Formin-Dependent Synaptic Growth; Evidence that Dlar Signals via Diaphanous to Modulate Synaptic Actin and Dynamic Pioneer Microtubules. J Neurosci 28, 11111–11123.

Peters, O.M., Ghasemi, M., and Brown, R.H. (2015). Emerging mechanisms of molecular pathology in ALS. Journal of Clinical Investigation 125, 1767–1779.

Phillips, J.P., Tainer, J.A., Getzoff, E.D., Boulianne, G.L., Kirby, K., and Hilliker, A.J. (1995). Subunit-destabilizing mutations in Drosophila copper/zinc superoxide dismutase: neuropathology and a model of dimer dysequilibrium. Proc Natl Acad Sci U S A 92, 8574–8578.

147 Pielage, J., Bulat, V., Zuchero, J.B., Fetter, R.D., and Davis, G.W. (2011). Hts/Adducin Controls Synaptic Elaboration and Elimination. Neuron 69, 1114– 1131.

Pilo Boyl, P., Di Nardo, A., Mulle, C., Sassoè-Pognetto, M., Panzanelli, P., Mele, A., Kneussel, M., Costantini, V., Perlas, E., … Witke, W. (2007). Profilin2 contributes to synaptic vesicle exocytosis, neuronal excitability, and novelty- seeking behavior. The EMBO Journal 26, 2991–3002.

Pollard, T.D., and Cooper, J.A. (1984). Quantitative analysis of the effect of Acanthamoeba profilin on actin filament nucleation and elongation. Biochemistry 23, 6631–6641.

Puls, I., Jonnakuty, C., LaMonte, B.H., Holzbaur, E.L.F., Tokito, M., Mann, E., Floeter, M.K., Bidus, K., Drayna, D., … Fischbeck, K.H. (2003). Mutant dynactin in motor neuron disease. Nat Genet 33, 455–456.

Rademakers, R., Neumann, M., and Mackenzie, I.R. (2012). Advances in understanding the molecular basis of frontotemporal dementia. Nat Rev Neurol 8, 423–434.

Renton, A.E., Chiò, A., and Traynor, B.J. (2014). State of play in amyotrophic lateral sclerosis genetics. Nat Neurosci 17, 17–23.

Renton, A.E., Majounie, E., Waite, A., Simón-Sánchez, J., Rollinson, S., Gibbs, J.R., Schymick, J.C., Laaksovirta, H., van Swieten, J.C., … Traynor, B.J. (2011). A Hexanucleotide Repeat Expansion in C9ORF72 Is the Cause of Chromosome 9p21-Linked ALS-FTD. Neuron 72, 257–268.

Riedl, J., Crevenna, A.H., Kessenbrock, K., Yu, J.H., Neukirchen, D., Bista, M., Bradke, F., Jenne, D., Holak, T.A., … Wedlich-Soldner, R. (2008). Lifeact: a versatile marker to visualize F-actin. Nat Meth 5, 605–607.

Ringholz, G.M., Appel, S.H., Bradshaw, M., Cooke, N.A., Mosnik, D.M., and Schulz, P.E. (2005). Prevalence and patterns of cognitive impairment in sporadic ALS. Neurology 65, 586–590.

Ritson, G.P., Custer, S.K., Freibaum, B.D., Guinto, J.B., Geffel, D., Moore, J., Tang, W., Winton, M.J., Neumann, M., … Taylor, J.P. (2010). TDP-43 Mediates Degeneration in a Novel Drosophila Model of Disease Caused by Mutations in VCP/p97. J. Neurosci. 30, 7729–7739.

Robberecht, W., and Philips, T. (2013). The changing scene of amyotrophic lateral sclerosis. Nat Rev Neurosci 14, 248–264.

148 Rodal, A.A., Motola-Barnes, R.N., and Littleton, J.T. (2008). Nervous Wreck and Cdc42 Cooperate to Regulate Endocytic Actin Assembly during Synaptic Growth. J. Neurosci. 28, 8316–8325.

Romero, S., Le Clainche, C., Didry, D., Egile, C., Pantaloni, D., and Carlier, M.-F. (2004). Formin Is a Processive Motor that Requires Profilin to Accelerate Actin Assembly and Associated ATP Hydrolysis. Cell 119, 419–429.

Rosen, D.R., Siddique, T., Patterson, D., Figlewicz, D.A., Sapp, P., Hentati, A., Donaldson, D., Goto, J., O’Regan, J.P., … Brown, R.H. (1993). Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 362, 59–62.

Rothstein, J.D., Van Kammen, M., Levey, A.I., Martin, L.J., and Kuncl, R.W. (1995). Selective loss of glial glutamate transporter GLT-1 in amyotrophic lateral sclerosis. Ann. Neurol. 38, 73–84.

Rotty, J.D., Wu, C., Haynes, E.M., Suarez, C., Winkelman, J.D., Johnson, H.E., Haugh, J.M., Kovar, D.R., and Bear, J.E. (2015). Profilin-1 Serves as a Gatekeeper for Actin Assembly by Arp2/3-Dependent and -Independent Pathways. Developmental Cell 32, 54–67.

Rubinsztein, D.C., and Carmichael, J. (2003). Huntington’s disease: molecular basis of neurodegeneration. Expert Rev Mol Med 5, 1–21.

Sau, D., Biasi, S.D., Vitellaro-Zuccarello, L., Riso, P., Guarnieri, S., Porrini, M., Simeoni, S., Crippa, V., Onesto, E., … Poletti, A. (2007). Mutation of SOD1 in ALS: a gain of a loss of function. Hum. Mol. Genet. 16, 1604–1618.

Schutt, C.E., Myslik, J.C., Rozycki, M.D., Goonesekere, N.C.W., and Lindberg, U. (1993). The structure of crystalline profilin β-actin. Nature 365, 810–816.

Schweinhuber, S.K., Meßerschmidt, T., Hänsch, R., Korte, M., and Rothkegel, M. (2015). Profilin Isoforms Modulate Astrocytic Morphology and the Motility of Astrocytic Processes. PLoS ONE 10, e0117244.

Shao, J., Welch, W.J., DiProspero, N.A., and Diamond, M.I. (2008). Phosphorylation of Profilin by ROCK1 Regulates Polyglutamine Aggregation. Mol. Cell. Biol. 28, 5196–5208.

Sharma, A., Lambrechts, A., Hao, L. thi, Le, T.T., Sewry, C.A., Ampe, C., Burghes, A.H.M., and Morris, G.E. (2005). A role for complexes of survival of motor neurons (SMN) protein with gemins and profilin in neurite-like cytoplasmic

149 extensions of cultured nerve cells. Experimental Cell Research 309, 185–197.

Silva, J.S.D., Medina, M., Zuliani, C., Nardo, A.D., Witke, W., and Dotti, C.G. (2003). RhoA/ROCK regulation of neuritogenesis via profilin IIa–mediated control of actin stability. J Cell Biol 162, 1267–1279.

Singleton, A.B. (2011). Exome sequencing: a transformative technology. The Lancet Neurology 10, 942–946.

Smith, B.N., Ticozzi, N., Fallini, C., Gkazi, A.S., Topp, S., Kenna, K.P., Scotter, E.L., Kost, J., Keagle, P., … Landers, J.E. (2014). Exome-wide Rare Variant Analysis Identifies TUBA4A Mutations Associated with Familial ALS. Neuron 84, 324–331.

Smith, B.N., Vance, C., Scotter, E.L., Troakes, C., Wong, C.H., Topp, S., Maekawa, S., King, A., Mitchell, J.C., … Shaw, C.E. (2015). Novel mutations support a role for Profilin 1 in the pathogenesis of ALS. Neurobiology of Aging 36, 1602.e17–e1602.e27.

Smyth, K.A., and Belote, J.M. (1999). The Dominant Temperature-Sensitive Lethal DTS7 of Drosophila melanogaster Encodes an Altered 20S Proteasome β- Type Subunit. Genetics 151, 211–220.

Sreedharan, J., Blair, I.P., Tripathi, V.B., Hu, X., Vance, C., Rogelj, B., Ackerley, S., Durnall, J.C., Williams, K.L., … Shaw, C.E. (2008). TDP-43 mutations in familial and sporadic amyotrophic lateral sclerosis. Science 319, 1668–1672.

Sreedharan, J., and Brown, R.H. (2013). Amyotrophic lateral sclerosis: Problems and prospects. Ann. Neurol. 74, 309–316.

Sreedharan, J., Neukomm, L.J., Brown Jr., R.H., and Freeman, M.R. (2015). Age-Dependent TDP-43-Mediated Motor Neuron Degeneration Requires GSK3, hat-trick, and xmas-2. Current Biology 25, 2130–2136.

Suetsugu, S. (1998). The essential role of profilin in the assembly of actin for microspike formation. The EMBO Journal 17, 6516–6526.

Suganuma, T., Mushegian, A., Swanson, S.K., Abmayr, S.M., Florens, L., Washburn, M.P., and Workman, J.L. (2010). The ATAC Acetyltransferase Complex Coordinates MAP Kinases to Regulate JNK Target Genes. Cell 142, 726–736.

Takenawa, T., and Suetsugu, S. (2007). The WASP–WAVE protein network: connecting the membrane to the cytoskeleton. Nat Rev Mol Cell Biol 8, 37–48.

150 Tanaka, M., Kim, Y.M., Lee, G., Junn, E., Iwatsubo, T., and Mouradian, M.M. (2004). Aggresomes Formed by α-Synuclein and Synphilin-1 Are Cytoprotective. J. Biol. Chem. 279, 4625–4631.

Tateno, M., Kato, S., Sakurai, T., Nukina, N., Takahashi, R., and Araki, T. (2009). Mutant SOD1 impairs axonal transport of choline acetyltransferase and acetylcholine release by sequestering KAP3. Hum. Mol. Genet. 18, 942–955.

Torroja, L., Packard, M., Gorczyca, M., White, K., and Budnik, V. (1999). The Drosophila β-Amyloid Precursor Protein Homolog Promotes Synapse Differentiation at the Neuromuscular Junction. J. Neurosci. 19, 7793–7803.

Tyedmers, J., Mogk, A., and Bukau, B. (2010). Cellular strategies for controlling protein aggregation. Nat. Rev. Mol. Cell Biol. 11, 777–788.

Valori, C.F., Brambilla, L., Martorana, F., and Rossi, D. (2013). The multifaceted role of glial cells in amyotrophic lateral sclerosis. Cell. Mol. Life Sci. 71, 287–297.

Vance, C., Rogelj, B., Hortobágyi, T., De Vos, K.J., Nishimura, A.L., Sreedharan, J., Hu, X., Smith, B., Ruddy, D., … Shaw, C.E. (2009). Mutations in FUS, an RNA Processing Protein, Cause Familial Amyotrophic Lateral Sclerosis Type 6. Science 323, 1208–1211.

Walker, F.O. (2007). Huntington’s disease. Lancet 369, 218–228.

Wang, X., Kibschull, M., Laue, M.M., Lichte, B., Petrasch-parwez, E., and Kilimann, M.W. (1999). Aczonin, a 550-kD putative scaffolding protein of presynaptic active zones, shares homology regions with Rim and Bassoon and binds profilin. J Cell Biol 151–162.

Wangler, M.F., Yamamoto, S., and Bellen, H.J. (2015). Fruit Flies in Biomedical Research. Genetics 199, 639–653.

Watanabe, N., Madaule, P., Reid, T., Ishizaki, T., Watanabe, G., Kakizuka, A., Saito, Y., Nakao, K., Jockusch, B.M., and Narumiya, S. (1997). p140mDia, a mammalian homolog of Drosophila diaphanous, is a target protein for Rho small GTPase and is a ligand for profilin. EMBO J 16, 3044–3056.

Watson, M.R., Lagow, R.D., Xu, K., Zhang, B., and Bonini, N.M. (2008). A Drosophila Model for Amyotrophic Lateral Sclerosis Reveals Motor Neuron Damage by Human SOD1. J. Biol. Chem. 283, 24972–24981.

Wheaton, M.W., Salamone, A.R., Mosnik, D.M., McDonald, R.O., Appel, S.H., Schmolck, H.I., Ringholz, G.M., and Schulz, P.E. (2007). Cognitive impairment in

151 familial ALS. Neurology 69, 1411–1417.

Wijesekera, L.C., and Leigh, P.N. (2009). Amyotrophic lateral sclerosis. Orphanet Journal of Rare Diseases 4, 3.

Wills, Z., Marr, L., Zinn, K., Goodman, C.S., and Van Vactor, D. (1999). Profilin and the Abl Tyrosine Kinase Are Required for Motor Axon Outgrowth in the Drosophila Embryo. Neuron 22, 291–299.

Witke, W. (1998). In mouse brain profilin I and profilin II associate with regulators of the endocytic pathway and actin assembly. The EMBO Journal 17, 967–976.

Witke, W. (2004). The role of profilin complexes in cell motility and other cellular processes. Trends in Cell Biology 14, 461–469.

Witke, W., Sutherland, J.D., Sharpe, A., Arai, M., and Kwiatkowski, D.J. (2001). Profilin I is essential for cell survival and cell division in early mouse development. Proc Natl Acad Sci U S A 98, 3832–3836.

Wu, J.S., and Luo, L. (2006). A protocol for dissecting Drosophila melanogaster brains for live imaging or immunostaining. Nat. Protocols 1, 2110–2115.

Xia, R., Liu, Y., Yang, L., Gal, J., Zhu, H., and Jia, J. (2012). Motor neuron apoptosis and neuromuscular junction perturbation are prominent features in a Drosophila model of Fus-mediated ALS. Molecular Neurodegeneration 7, 10.

Yang, D., Abdallah, A., Li, Z., Lu, Y., Almeida, S., and Gao, F.-B. (2015). FTD/ALS-associated poly(GR) protein impairs the Notch pathway and is recruited by poly(GA) into cytoplasmic inclusions. Acta Neuropathol 130, 525–535.

Youle, R.J., and Bliek, A.M. van der (2012). Mitochondrial Fission, Fusion, and Stress. Science 337, 1062–1065.

Zhao, L., Wang, D., Wang, Q., Rodal, A.A., and Zhang, Y.Q. (2013). Drosophila cyfip Regulates Synaptic Development and Endocytosis by Suppressing Filamentous Actin Assembly. PLoS Genet 9, e1003450.

152