Modeling and Analyzing Cellular Defects Associated With

MODELING AND ANALYZING CELLULAR DEFECTS ASSOCIATED WITH

MOVEMENT DISORDERS IN CAENORHABDITIS ELEGANS

CHUAN XU

GUY A. CALDWELL, COMMITTEE CHAIR KIM A. CALDWELL, CO-CHAIR JANIS O’DONNELL CAROL DUFFY PATRICK A. FRANTOM

A DISSERTATION

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Biological Sciences in the Graduate School of The University of Alabama

TUSCALOOSA, ALABAMA

2018

ABSTRACT

As a powerful modern organism, Caenorhabditis elegans has been widely used to study pathologies behind movement disorders including Parkinson’s disease (PD) and dystonia. PD is the second most common neurodegenerative disease, in which more than 90% of cases are idiopathic. The etiology of PD has long been thought to involve both genetic and environmental factors. The classical pathological hallmarks of PD are the progressive loss of dopaminergic neurons within the substantia nigra, accompanied by the accumulation of - synuclein (-syn) in the form of Lewy bodies. Here, we identified four compounds

(cyclosporin A, meclofenoxate hydrochloride, sulfaphenazole, and choline) that can rescue mitochondrial phospholipid depletion induced neurodegeneration in C. elegans with -syn expression in dopaminergic neurons. To examine putative epigenetically-regulated modifiers of -syn induced dopaminergic neurodegeneration in C. elegans, we demonstrated a specific microRNA, mir-239, when mutated, showed a robust resistance to neurotoxicity resulted from -syn. By functionally investigating a suite of expression-validated targets of mir-239 regulation via conditional knockdown using a dopaminergic neuron-specific RNAi-sensitive

-syn strain, we discerned a subset of downstream targets contributing to neuroprotection afforded by mir-239. These findings support the predictive nature of C. elegans in validating potential modifiers of -syn neurotoxicity and discovering potential neuroprotective chemicals associated with PD. Human torsinA, encoded by the DYT1 gene, is an ER resident chaperone protein that has been identified to be responsible for a human movement disorder

ii called early-onset torsion dystonia. Here we revealed that tor-2, the C. elegans homologue of human torsinA, was indicated as a possible regulator in the trafficking process of an AMPA receptor subunit, GLR-1; implying a possible connection between the glutamatergic transmission and the etiology of dystonia. The combined outcomes of these studies of movement disorders strongly support C. elegans is a very desirable model organism for the analysis of cell biology and genetic features associated with PD and dystonia.

iii

DEDICATION

I would like to dedicate this dissertation to everyone who helped me and supported me over the past years to achieve this masterpiece. In particular, this manuscript is dedicated to my parents and my best friend Zhiwei for providing me unwavering support and strength in all my endeavors.

LIST OF ABBREVIATIONS AND SYMBOLS

6-OHDA 6-Hydroxydopamine

 Alpha

AD Alzheimer disease

AD Autosomal dominant

ADE neuron Anterior deirid neuron

AMPA Amino-3-hydroxy-5-methylisoxazoleproprionic acid

AR Autosomal recessive

ATP Adenosine triphosphate

β Beta

Botox Botulinum toxin bp base pair

C Celsius cDNA Complementary DNA

CDP Cytidine diphosphate

CEP neuron Cephalic neuron

C. elegans Caenorhabditis elegans

CGC Caenorhabditis genetics center

Cho Choline

CL Phospholipid cardiolipin

CNS Central nervous system

CNIHs Cornichon homologs

CsA Cyclosporine A

D2 Dopamine 2

DA Dopamine

DG Diacylglycerol

DMAE Dimethylethanolamine

DMSO Dimethyl sulfoxide

DNA Deoxyribonucleic acid dsRNA Double-stranded RNA

ε Epsilon

E. coli Escherichia coli

EOTD Early-onset torsion dystonia

ER Endoplasmic reticulum

ETA Ethanolamine

EV Empty vector

FDA Food and drug administration

GTP Guanosine triphosphate

GWAS Genome-wide association study iGluRs ionotropic glutamate receptors

IOD Integrated optical density

IPTG Isopropyl β-D-1-thiogalactopyranoside

vi kDa Kilodalton

L4 Larval stage 4

L-DOPA L-3,4-dihydroxyphenylalanine

LINC Linker of nucleoskeleton and cytoskeleton

LSD Lysosomal storage disorder

MAMs Mitochondrial associated membranes

MFX Meclofenoxate hydrochloride

MGs Monoglycerides miRNAs MicroRNAs

MMP Mitochondrial membrane potential

MPP+ 1-Methyl-4-phenylpyridinium mRNA Messenger RNA mPTP Mitochondrial permeability transition pore

l Microliter mg Milligram ml Milliliter mM  Millimolar n Number

N/A Not applicable

NBIA Neurodegeneration with brain iron accumulation ncRNA Non-coding ribonucleic acid

NE Nuclear envelope

vii

NGM Nematode growth medium

NMDA N-methyl-d-aspartate

Ω Omega

PARK Parkinson disease gene

PBMCs Peripheral blood mononuclear cells

PC Phosphatidyl choline p-Cho Phosphorylated-choline

PCR Polymerase chain reaction

PD Parkinson’s disease

PE Phosphatidyl ethanolamine p-ETA Phosphorylated ETA

PM Plasma membrane

Pre-miRNA Precursor miRNA

Pri-mRNA Primary miRNA

PS Phosphatidylserine qRT-PCR Quantitative reverse transcription-PCR

Ran Ras-related nuclear protein

REP1 Repeat sequence

RISC RNA-induced silencing complex

RNA Ribonucleic acid

RNAi RNA interference

RNase Ribonuclease

viii

Roc Ras of complex

ROS Reactive oxidative species rRNA Ribosomal ribonucleic acid

SCI Spinal cord injury

SD Standard deviation

SNPs Single nucleotide polymorphisms

SUL Sulfaphenazole

TARPs Transmembrane AMPAR regulatory proteins

TBI Traumatic brain injury tRNA Transfer ribonucleic acid

UTR Untranslated region

VGLUT Vesicular glutamate transporter

VNC Ventral nerve cord

WT Wildtype

Proteins/Genes

Aβ Amyloid-beta

AMPARs AMPA receptors

ANO3 Anoctamin3

ATP1A3 ATPase Na+/K+ transporting subunit alpha 3

ATP13A2 Lysosomal type 5P-type ATPase

syn Alpha synuclein

CACNA1B Calcium voltage-gated channel subunit alpha 1B

CED-10 Cell-corpse engulfment protein

CoA Coenzyme A

COL6A3 Collagen Type VI alpha3 chain

CRLS-1 Cardiolipin synthase

CYP2C9 Cytochrome P450 family 2 subfamily C polypeptide 9

DAT-1 Dopamine transporter

DGCR8 DiGeorge syndrome critical region 8

DJ-1 Oncogene DJ-1

DNAJC6 DnaJ heat shock protein family (Hsp40) member C6

DNAJC13 DnaJ heat shock protein family (Hsp40) member C13

DPCK Dephospho-CoA kinase

EIF4G1 Eukaryotic translation initiation factor 4 gamma 1

FBX07 F-Box protein 7

GBA Glucocerebrosidase

GCH1 GTP cyclohydrolase 1

GFP Green fluorescent protein

GIGYF2 GRB10 interacting GYF protein 2

GNAL G protein subunit alpha L

GTPase Guanosine triphosphatase

HPCA Hippocalcin

HSL Hormone-sensitive lipase

HTRA2 HtrA serine peptidase 2

KCTD17 Potassium channel tetramerization domain containing 17

KMT2B Lysine methyltransferase 2B

LIPE Hormone-sensitive lipase

LRRK2 Leucine-rich repeat kinase 2

MAGUKs Membrane-associated guanylate kinases

MAPKKK Mitogen-activated protein kinase kinase kinase

MECR Mitochondrial trans-2-enoyl-CoA reductase

NCEH-1 Neutral cholesterol ester hydrolase 1

PANK Pantothenate kinase

PDR-1 Parkinson's disease related

PINK-1 PTEN induced putative kinase 1

PLA2G6 Phospholipase A2 group VI

PNKD2 Paroxysmal nonkinesigenic dyskinesia 2

PPAT Phosphopantetheine adenylyltransferase

Ppcdc Phosphopantothenoylcysteine decarboxylase

PRRT2 Proline rich transmembrane protein 2

PRKN Parkin

PRKRA Protein activator of interferon induced protein kinase

PSD-1 Phosphatidylserine decarboxylase

RAB-8 Ras-related protein 8

ROL-6 Roller 6 (collagen)

SGCE Sarcoglycan epsilon

SLC2A1 Solute carrier family 2 member 1

SLC17A5 Solute carrier family 17 member 5 Soluble N-ethylmaleimide-sensitive factor attachment SNARE protein receptor

SNB-1 Synaptobrevin-1

SNCA Synuclein, alpha

LRK-1 Leucine-rich repeat serine/threonine-protein kinase 1

SPR Sepiapterin reductase

SYNJ1 Synaptojanin 1

TAF1 TATA-Box binding protein associated factor 1

TH Tyrosine hydroxylase

THAP1 THAP domain containing 1

TOR-2 Torsin2 (torsinA)

TOR1A Torsin family1 member A

TUBB4A Tubulin Beta4 A class Iva

UCHL-1 Ubiquitin C-terminal hydrolase L1

VGLUT Vesicular glutamate transporter

VPS35 Vacuolar protein sorting 35

VPS41 Vacuolar protein sorting-associated protein 41

xii

ACKNOWLEDGMENTS

First of all, I would like to thank my academic advisors Prof. Guy A. Caldwell and Prof.

Kim A. Caldwell for offering me the opportunity to perform scientific research in their wonderful lab at this institution. I am also grateful for their mentorship and endless support to not only my study, but also my life. I appreciate them very much from the bottom of my heart. Secondly, I want to thank my colleagues in Caldwell Lab. Your assistance, the happy memories, and funny moments we shared during my Ph.D. study will always bear in my mind. Additionally, I am thankful for the help of former and current undergraduate students,

Michael Teal, Samantha Glukhova and especially Blake Parker who helped me with my projects in the past several years. I also want to take this opportunity to thank our wonderful lab manager Dr. Laura Berkowitz and Post Doctor Xiaohui Yan who provides me great help in my research. I thank The University of Alabama, the Department of Biological Sciences and Chinese Scholarship Council (CSC) for the funding support. And my graduate committee members, Dr. Janis O’Donnell, Dr. Carol Duffy, Dr. Patrick A. Frantom, Dr. Stevan Marcus, thank you for guiding me throughout my entire graduate study. In the end, I want to thank you for the support and encouragement from my family members and friends which enabled me to persist, focus on and achieve the goals of my Ph.D. study.

xiii

CONTENTS

ABSTRACT ...... ii

DEDICATION ...... iv

LIST OF ABBREVIATIONS AND SYMBOLS ...... v

ACKNOWLEDGMENTS ...... xiii

LIST OF TABLES ...... xvii

LIST OF FIGURES ...... xviii

CHAPTER ONE: INTRODUCTION ...... 1 a. Parkinson’s disease ...... 1 b. MicroRNA and PD ...... 3 c. Dystonia ...... 8 d. Early onset torsion dystonia (EOTD) ...... 9 e. C. elegans models of movement disorders ...... 10 f. Focus of dissertation ...... 19 g. References ...... 20 CHAPTER TWO: CHEMICAL COMPENSATION OF MITOCHONDRIAL PHOSPHOLIPID DEPLETION IN YEAST AND ANIMALS MODELS OF PARKINSON’S DISEASE ...... 26 a. Abstract ...... 27 b. Introduction ...... 28 c. Results ...... 30

xiv

d. Discussion ...... 43

e. Materials and methods ...... 47

f. References ...... 51 CHAPTER THREE: MICRORNA-MEDIATED PROTECTION AGAINST -SYNUCLEIN-INDUCED DOPAMINERGIC NEURODEGENERATION IN C. ELEGANS ...... 56 a. Abstract ...... 57

b. Introduction ...... 57

c. Results ...... 59

d. Discussion ...... 71

e. Materials and methods...... 75

f. References ...... 79 CHAPTER FOUR: INVESTIGATING A ROLE OF TOR-2 IN CONTROLLING ER EXPORT OF AMPA RECEPTORS IN THE REGULATION OF EXCITABILITY IN C. ELEGANS ...... 83 a. Abstract ...... 84

b. Introduction ...... 84

c. Results ...... 88

d. Discussion ...... 96

e. Materials and methods ...... 98

f. References ...... 101

CHAPTER FIVE: CONCLUSIONS AND FUTURE DIRECTIONS...... 105

a. Introduction ...... 105 b. Conclusion ...... 106

c. Discussion and future direction ...... 110

xv d. References ...... 116

xvi

LIST OF TABLES 1.1. Summary of mutations in PD genes linked to PD pathogenesis and their corresponding C. elegans orthologs ...... 11

1.2. Current DYT loci with brief descriptions of associated phenotypes, mode of inheritance and corresponding C. elegans orthologs ...... 15

2.1. Phenotypes of yeast and worms expressing -syn after treatment with candidate drugs ...... 32

3.1. Predicted function of mir-239b gene targets and corresponding human orthologs...... 64

xvii

LIST OF FIGURES 1.1. Schematic representation of the biogenesis pathway of microRNA processing ...... 5

1.2. Schematic representation of coding and non-coding DNA in human genome ...... 6 2.1. PE and CL synthesis in mitochondria and ER ...... 30

2.2. Structures of positive candidates identified from screen...... 33 2.3. -syn-induced dopaminergic neurodegeneration in C. elegans is rescued by MFX ...... 36

2.4. -syn-induced dopaminergic neurodegeneration in C. elegans is rescued by choline chloride ...... 38

2.5. -syn-induced dopaminergic neurodegeneration in C. elegans is rescued by CsA...... 40

2.6. Depletion of psd-1 or crls-1 in -syn-expressing dopaminergic neurons enhance neurodegeneration ...... 42

2.7. Choline, MFX, CsA, and SUL exposure do not affect -syn expression...... 43

3.1. mir-239 mutation enhances resistance to -syn induced neurotoxicity in C. elegans...... 61

3.2. mir-239 mutation extends lifespan and rescues -syn-induced neurodegeneration ...... 63

3.3. Examination of 26 validated target genes of mir-239b via RNAi knockdown in the dopaminergic-neuron sensitive RNAi -syn strain with and without mir-239 mutation ...... 68

3.4. mir-239 mutation partly rescues aging-associated locomotive deficits in -syn expressing worms ...... 71

4.1. Systematic tor-2 RNAi increases ER stress in worm strain expressing Phsp-4::GFP ...... 89

xviii

4.2. A schematic of locomotion circuitry that controls the initiation of backward movement in C. elegans...... 90

4.3. Treatment of pan-neuronal RNAi worm strain TU3401 with dsRNA tor-2 causes significant defects on backwards movements and nose touch response...... 91

4.4. Loss of function of tor-2 mutation causes significant defects on backwards movement and nose touch response in C. elegans ...... 92

4.5. Loss of function of tor-2 mutation and pan-neuronal tor-2 RNAi decreases omega turn frequency in C. elegans ...... 93

4.6. Synaptic levels of Pglr-1::GLR-1::GFP are decreased in pan neuronal TU3401 worm strain followed by tor-2 RNAi ...... 96

xix

CHAPTER ONE

INTRODUCTION

Parkinson’s disease

Parkinson’s disease was first named “shaking palsy” by a British physician, James

Parkinson in 1817. For the first time, he described the clinical manifestation of six patients suffering from a slowly progressing disease characterized by "involuntary tremulous motion.” Now we called it Parkinson’s disease (PD) and it is the second most common neurodegenerative movement disorder that affects about 2% of the population above 65 years old of age in the world 1. The symptoms and signs of PD are variable in different individuals.

Briefly, the main characteristics of this disease include tremor, which usually begins in a limb, slowed movement, rigid muscle which may occur in any part of the body, impaired posture and balance, loss of automatic movement such as blinking, smiling, speech and writing problems 2. Since it develops gradually, early signs may be unnoticeable. For example, the patient’s face may show little or no expression, and the arms may not swing when walking 2.

Classically, PD is associated with a loss of striatal dopaminergic neurons (DA) in the substantia nigra, a mesencephalic structure of the basal ganglia motor circuit 3. The reduced dopamine levels cause abnormal brain activity, leading to signs of PD. The classical neuropathological hallmark of PD is the intracellular accumulation of Lewy bodies, which

1 are mainly composed of -synuclein aggregates, a presynaptic protein thought to be involved in neuronal plasticity. SNCA, coding for -synuclein, was the first locus identified that linked genetics with PD 4. A mutation of the -synuclein gene was identified in an Italian family and three unrelated families of Greek origin and this mutation resulted in early-onset autosomal dominant PD.

Thus far, at least 20 genes have been associated with PD as genetic risk factors such as the vesicular trafficking protein and retromer sorting protein VPS35, the leucine-rich repeat kinase 2 LRRK2, and the mitochondrial stress response proteins PINK1, Parkin and DJ-1 5.

Most of the identified PD genes are highly conserved across species such as mice, fruit fly or

C. elegans.

Environment factors also contribute to the onset of PD. For instance, rural residence, well- water drinking and exposure to agricultural chemicals have been reported as risk factors associated with PD 6. In fact, interactions between genetic susceptibility factors and environmental exposures are thought to be the contributors of PD. The identified PD genes can interact with environmental risk factors to cause mitochondrial dysfunction, oxidative stress, inflammation, autophagy, and apoptosis. These cellular dysfunctions can initiate a cascade that can result in neurodegeneration 7,8. For instance, the increased length of a dinucleotide repeat sequence (REP1) within the SNCA promoter in -synuclein was found to increase PD risk, but the shorter REP1 has been indicated as a protective factor. However, the protective effect of shorter REP1 appears to be lost upon pesticide paraquat exposure, which indicated an interaction between them 9.

The mechanism behind neurodegeneration of PD is very complicated. It associates with multiple pathways and mechanisms, such as -synuclein proteostasis, mitochondrial function, oxidative stress, calcium homeostasis, axonal transport, neuroinflammation or dysfunction of ubiquitin-proteasome pathway. For example, the first identified PD gene, -

2 synuclein (-syn), was indicated to be present in the mitochondrial and mitochondrial associated membranes (MAMs) in neurons 10,11. Evidence shows that -synuclein plays a role in maintaining mitochondrial homeostasis through different processes. For instance, in yeast, phosphatidylserine decarboxylase (Psd-1), which is embedded in the inner mitochondrial membrane, is responsible for converting phosphatidylserine (PS) to phosphatidylethanolamine (PE) in mitochondrial. Low levels of PE in the phosphatidylserine decarboxylase deletion mutant (psd1Δ) with -syn expression triggered mitochondrial defects and disrupted the homeostasis of -syn which led to the accumulation of cytoplasmic

-syn foci. Furthermore, these results increased ER stress and inhibited cell growth in yeast.

In C. elegans, knocking down psd-1 which encodes Psd, enhanced the toxicity of -syn and increased DA neuron neurodegeneration. This result further confirmed that the homeostasis of -syn was disrupted by low PE in mitochondrial 12. Moreover, overexpression wildtype

(WT) -syn or the mutated variants A30P or A53T induced significant mitochondrial transports defects and fragmentation in human-derived neuron 13. Studies in Drosophila revealed that overexpression human -syn showed strong neurodegeneration and further evidence addressed that -syn induced mitochondrial dysfunction through spectrin and the actin cytoskeleton 7. Although some progress has been made to help us understand pathological mechanisms, the full etiology of this disorder remains elusive, due in part to deficiencies in our understanding of molecular and cellular sensitizing risk factors.

MicroRNA and PD

MicroRNAs (miRNAs) represent a large group of small, 18-25 nucleotides long, non- coding RNA molecules 14. They are short inverted repeats with a double-stranded RNA

(dsRNA) stem loop structure and are found in both introns and intergenic clusters in the genome. There are multiple steps and specific cellular machinery involved in the biogenesis 3 process of miRNAs 15. miRNAs are firstly transcribed as primary miRNAs (pri-mRNAs) transcripts in the nucleus and then processed by two nuclear RNase-III enzymes termed

Drosha and DGCR8. Next, pri-mRNAs are processed to form the precursor miRNAs (pre- miRNAs) whereby the stem loop structure is removed before it is transported into the cytoplasm. Finally, they are cleaved by the Dicer RNAase III endonuclease to produce mature miRNA nucleotides of ~20bp 16-18. miRNAs have been implicated as one of the key regulators of gene expression at the post-transcriptional level. The main mechanism of repressing gene expression by miRNAs is to interact with the RNA-induced silencing complex (RISC); following this interaction, there is binding to the 3’ UTR of the targeted gene mRNA, resulting in subsequent degradation of mRNA and inhibition of protein translation 19. (Fig. 1.1)

Figure 1.1. Schematic representation of the biogenesis pathway of microRNA processing. The miRNA processing pathway has long been viewed as linear and universal to all mammalian miRNAs. This canonical maturation includes the production of the primary miRNA transcript (pri-miRNA, ~70bp) by RNA polymerase II or III and cleavage of the pri- miRNA by the microprocessor complex Drosha–DGCR8 (Pasha) in the nucleus. The resulting precursor hairpin, the pre-miRNA, is exported from the nucleus by Exportin-5–Ran- GTP. In the cytoplasm, the RNase Dicer in complex cleaves the pre-miRNA hairpin to its mature length (~20bp). The functional strand of the mature miRNA is loaded into the RNA- induced silencing complex (RISC), where it guides RISC to silence target mRNAs through mRNA cleavage, translational repression.

The first human genome sequence was published in 2001 20. However, less than 25,000 protein coding genes were discovered, which is 3-4 times lower than expected in 1980’s. This discovery resulted in a shift of research directions from mRNAs to non-coding RNA

(ncRNA), which is now known to be an important regulator of gene expression. Considering

5 that less than 2% of the human genome consists of protein coding genes while for invertebrates this number is 10-30% 8,21 (Fig. 1.2).

Figure 1.2. Schematic representation of coding and non-coding DNA in human genome. Only around 2% of the entire human genome is protein-coding gene. The left 98% is transcribed as non-coding RNA (ncRNA) which consist of rRNA, tRNA, introns, 5’ and 3’ untranslated regions, transposable elements, intergenic regions, and microRNAs. Modified from Source: Ardekani et al., 2010 Avicenna J Med Biotechnol

As the human genome was unveiled, non-coding RNA function was also gradually deciphered. miRNAs are one of the members of ncRNAs that also includes rRNA, tRNA, introns, 5’ and 3’ untranslated regions, transposable elements, and intergenic regions. miRNAs have been found in all eukaryotic cells 22. Approximately 2000 miRNA genes have been reported in the human genome (www.miRNA.org) and they are estimated to regulate

~30% of all protein-coding gene activities. Each miRNA is predicted to regulate about 200 target genes 23. miRNAs have been suggested to be essential and evolutionary ancient component of gene regulatory networks 24 because they are involved in the regulation of

6 almost every cellular process examined, such as development, differentiation, growth and metabolism 25.

Since miRNA plays a crucial role in cellular function, it is not surprising that mutation, deficiency or excesses of miRNA is associated with a large number of human diseases such as type 2 diabetes, cancer, cardiovascular disease, inflammatory disease, and neurodevelopmental disease 26. This small molecule was indicated as a regulator to restore imbalanced or dysregulated pathophysiological pathways in PD. A recent study found that there are significant differences of cortical expression patterns of miRNA between PD patient’s brains and non-diseased people’s brains. Many PD-associated miRNAs have been found, these miRNAs are associated with the regulation of apoptosis, autophagy, inflammation, mitochondrial dysfunction and oxidative stress genes 8. Moreover, a large number of altered miRNAs level were identified in human PD brain tissue samples including striatum, midbrain, prefrontal cortex, frontal cortex, amygdala, putamen, cingulate gyri. For example, miR-198, -485-5p, -339-5p, -208b, -135b, -299-5p, -330-5p, -542-3p, -379, -337-

5p, -34b, -34c were identified downregulated in substantia nigra; miR-200b*, -200a*, -195*,

424* were identified upregulated in frontal cortex 27.

miRNAs can be detected in circulatory biofluid (blood or serum/plasma) as either bound with a protein or encapsulated with microvesicle particles, called exosomes. Since their discovery in blood plasma in 2008 28, it has been shown that miRNAs are quite stable in the circulatory system 29,30. Therefore, miRNA has also been examined as a biomarker for aging and age-related diseases such as PD and Alzheimer’s disease. In this regard, the expression levels of miR-335, miR-374a/b, miR-199, miR-126, miR-151-5p, miR-29b/c, miR-147, miR-

28-5p, miR-30b/c, miR-301a, and miR-26a in peripheral blood mononuclear cells (PBMCs) from PD patients are lower than those from healthy people 31. The clinical diagnosis of PD often relies on late onset of motor impairments. Unfortunately, most of the DA neuron at this

7 stage may have been lost in patients and there will only be limited clinical benefit of current therapeutics. The detection of miRNAs in early symptomatic PD patients may provide a benefit as a compensated way for this drawback 32. Or, perhaps, in the future, there could be preventative screening in all individuals if the blood biomarkers become highly sensitive and predictive. Thus, miRNA has been proposed as putative biomarker to diagnose the PD at an early stage.

Dystonia

The term “Dystonia” was coined by a German neurologist named Hermann Oppenheim in

1911. During that century, many dystonia examples have been described in medical literatures 33,34. Dystonia is a movement disorder characterized by involuntary, sustained muscle contractions that are often painful. It is estimated that approximately 15 to 30 out of

100,000 people have dystonia, which makes it the third most common neurological disorder after PD and Alzheimer's disease. To manage muscle contractions, treatment options often include pharmacological therapy such as botulinum toxin (Botox) injected into specific muscles to reduce muscle contraction and improve abnormal postures, physical therapy or surgical approaches, such as deep brain stimulation or selective denervation surgery, in severe situations. However, the diagnosis and treatment of dystonia remain challenging 35.

Dystonia can be classified into three ways: age of onset, etiology, and distribution of body parts 36. Based on the age of onset, dystonia can be divided into early (younger than 26 years) and late-onset (older than 26 years) forms. Based on the distribution of affected body parts, dystonia can occur in a single body region, such as focal, legs, arms or the whole trunk, however, it can also spread to contiguous or non-contiguous regions with time. For example, onset in the legs is most frequent during childhood, and with increased age, dystonia can spread to the arms and hands, neck, and then cranial muscle. Based on the etiology, dystonia

8 can be mainly classified into primary dystonia and secondary dystonia. Primary dystonia usually has a hereditary component. At the molecular level, almost 30 genes related to dystonia have been discovered. Most of them are inherited in an autosomal dominant manner with reduced penetrance such as torsinA (DYT1), GTP-cyclohydrolase 1 (DYT5a) and ε- sarcoglycan (DYT11) 37-39.

Conversely, secondary dystonia is often caused by an external stressor, such as taking certain medications (neuroleptics and calcium channel blockers), exposure to toxins (carbon monoxide and wasp sting) or even a forceful blow (car accident) or an epidural during childbirth. A common cause are antipsychotic drugs used as a treatment of schizophrenia and other psychotic disorders that can induce tardive dystonia 40.

Early onset torsion dystonia (EOTD)

Early-onset torsion dystonia (EOTD) is one of the most severe types of dystonia, which usually typically presents in childhood or adolescence and only on occasion in adulthood in human 41. It is mainly caused by an autosomal dominantly inherited in frame deletion of 3 bp

(GAG) in the DYT1 gene on chromosome 9q34 which most commonly leads to a deletion of a C-terminal glutamate residue in position 302/303 in the protein torsinA 42.

TorsinA belongs to the AAA+ ATPase family that includes many functionally diverse proteins that all use energy released through ATP hydrolysis to remodel target substrates.

Among the family members, there are other torsin related gene products, torsinB, TOR2A, and TOR3A in mammals. TorsinA is the most well-characterized of the torsin molecules and it is localized at the endoplasmic reticulum (ER) lumen where it is functions as a molecular chaperone 43-45. Substrate-specific roles associated with torsinA include associations with the secretory pathway, protein folding and degradation 46-48. Like many chaperones, torsinA has been shown to interact with a variety of substrates, including the dopamine transporter,

9 cytoskeleton, components of the secretory pathway, and synaptic vesicle machinery 48,49. The majority of torsinA resides within the ER. C. elegans study demonstrated that the ability of torsinA to help protein folding and maintain a homeostatic threshold against ER stress is dependent upon its proper localization within the ER 50. When mutated, torsinA associates more closely with the nuclear envelope 51,52. Additionally, mutant torsinA interacts with different substrates such as SUN-domain and Nesprin proteins that assemble into the LINC complex in the nuclear envelope (NE) 53. Moreover, overexpression mutant torsinA in C. elegans significant increased ER stress indicating a disrupted proteostasis 44.

C. elegans models of movement disorders

Caenorhabditis elegans was first utilized as a model organism to study animal development and behavior by Sydney Brenner in 1965 54. The genome sequence for C. elegans was completed in 1988 and it was the first animal to have its entire genome revealed

55. The C. elegans hermaphrodite nervous system consists of 302 neurons with all major classes of neurotransmitters represented within defined neuronal subtypes. In recent years, the neuron circuit diagram has provided the foundation to explain phenotypes of behavioral and locomotory mutants 56.

The nematode Caenorhabditis elegans possesses a lot of advantages such as short generation time (3 days from egg to adult), short lifespan and is easy for laboratory cultivation using Escherichia coli as the food source. This free-living small animal has been proven to be a useful model system to explore genomics, cell biology, cell death, epigenetics and aging. A widely useful technique, RNA interference (RNAi), was applied to knockdown selective target genes by feeding targeted dsRNA in C. elegans. Andrew Fire initially described this method in 1998 and won the 2006 Nobel Prize in Physiology or Medicine 57.

Many human diseases have been modeled in C. elegans, including PD and dystonia, as the

cellular features of these diseases are conserved in the nematode. Considering the short

lifespan of the nematode, there are advantages to performing genetic studies in this organism.

Mammalian PD models remain invaluable in studying PD, especially considering the wide

range clinical manifestation associated with PD such as motor dysfunction, anxiety, and

gastrointestinal dysfunction 58. However, mammalian models can be limited, considering that

neurodegeneration is often not observed and these models are very expensive.

Human genetics has shown that PD is associated with a series of genetic risk factors. So

far, more than 20 genes termed PARK have been identified. Most of these genes are well

conserved with homologues in C. elegans 5. (Table 1.1) Thus, it provides a great research

advantage for RNAi knockdown and genetics studies in C. elegans within the DA neurons to

build C. elegans PD models of these specific heritable genetic risk factors. Additionally,

overexpression models can also be constructed via the creation of transgenic C. elegans. In

total, these genetic tools can be used to identify potential neuroprotective targets, study

genetic interactions of PD-associated genes and examine PD environmental factors.

Table 1.1 Summary of mutations in PD genes linked to PD pathogenesis and their corresponding C. elegans orthologs

PD Locus PD gene Function Inheritance C. elegans Ortholog/ E value PARK1 SNCA/- Unknown AD N/A synuclein PARK2 Parkin E3 ligase; mitophagy AR pdr-1; 6e-25 PARK3 Possibly Biopterin biosynthesis in AD dhs-21;6e-09 sepiapterin DA metabolism reductase(SPR) PARK4 Regulatory Unknown N/A N/A elements of SNCA PARK5 UCHL-1 Hydrolysis of a peptide AD ubh-1; 2e-29 bond at the C-terminal glycine of ubiquitin; bind 11

to free monoubiquitin to prevent its degradation in lysosomes PARK6 PINK-1 Serine threonine kinase; AR pink-1; 2e-55 protects against mitochondrial dysfunction during cellular stress; Involved in mitophagy PARK7 DJ-1 Cell protection against AR djr1.1; 2e-46 oxidative stress; act as djr1.2; 1e-34 oxidative stress sensor and redox-sensitive chaperone and protease in cell death PARK-8 LRRK2 Leucine-rich repeat AD lrk-1; 2e-45 kinase; autophagy; synaptic vesicle trafficking; intracellular signaling but largely unknown PARK9 ATP13A2 ATPase; intracellular AR catp-7; e-167 cation homeostasis and catp-6; e-118 the maintenance of catp-5; e-166 neuronal integrity; transporter of glucose and other sugars, bile salts and organic acids, metal ions and amine compounds; Lysosomal and mitochondrial maintenance PARK10 Unknown; N/A N/A N/A likely not one gene PARK11 Unknown; PloyQ protein associated AR uncharacterized possibly with receptor tyrosine protein GIGYF2 kinase C18H9.3; 4e-04 PARK12 Unknown N/A N/A N/A PARK13 Potentially Serine protease; apoptosis N/A N/A HTRA2 PARK14 PLA2G6 Phospholipase; normal AR ipla-3; 6e-77 phospholipid remodeling ipla-2; 4e-75 PARK15 Possibly F-box protein; AR N/A FBX07 phosphorylation dependent ubiquitination PARK16 Unknown N/A N/A N/A PARK17 VPS35 vascular trafficking and AD vps-35; 0 retromer sorting 12

PARK18 EIF4G1 Translation initiation AD ifg-1; 2e-24 factor PARK19 DNAJC6 Heat shock protein; AR dnj-25; 7e-31 molecular chaperone PARK20 SYNJ1 Phosphoinositide AR unc-26; 0 phosphatase; synaptic transmission and membrane trafficking PARK21 DNAJC13 Heat shock protein; AD rme-8; 0 molecular chaperone N/A: Not applicable; AD= Autosomal dominant; AR= Autosomal recessive

Various cellular and transgenic models have been generated dopaminergic

neurodegeneration, as a pathological hallmark of PD in C. elegans. A classical hallmark of

PD occurs in Lewy body inclusions, where the main component is the protein -synuclein. It

is thought that it disrupts normal synaptic cellular function and subsequently induces

neurodegeneration. Mutations in SNCA cause dominant early-onset PD 59. Our lab created a

worm neurodegenerative model whereby wild-type -syn was overexpressed in DA neurons

and identified age- and dose-dependent neurodegeneration 60. Since C. elegans do not have

an endogenous -synuclein gene, this expression was essentially performed in a “null”

background without concern for the endogenous -syn background. This model has been

used in numerous studies to examine neuroprotective gene targets 61,62. Some of these targets

were predictive of results later identified to be neuroprotective in mammalian systems 63-66.

LRRK2, a signaling component, encodes a large, 2527 amino-acid protein with leucine-

rich repeat, Roc GTPase, COR, MAPKKK, and WD40 domains. The LRRK2 (G2019S)

mutation (in the MAPKKK domain) has been frequently detected in autosomal dominant and

idiopathic PD patients 67,68. C. elegans LRRK2 PD models were generated by overexpressing

either wild-type or G2019S LRRK2 under control of a pan-neuronal promoter (snb-1).

Results showed that enhanced DA neurodegeneration was accompanied by reduction of

dopamine levels 69. C. elegans also was used to explore how LRRK2 alters cellular

13 vulnerability to various forms of stress inducers. It was reported that depletion of lrk-1 (worm homologue of human LRRK2) and mutant analysis increased toxicity in the nematodes to rotenone treatment 69. However, overexpression of wild-type LRRK2 significantly increased resistance against rotenone 69.

Drug screening in C. elegans neurodegeneration models has also yielded promising results. For example, a C. elegans MPP+ model of neurodegeneration has been developed to study high-throughput drug screening together with verification by exposing worms to pharmacological compounds already used for the treatment of PD patients 70. In this study, the researchers identified several anti-PD drugs ameliorated MPP+ induced mobility defects such as L-DOPA, a precursor to dopamine, dopamine and dopamine receptor agonists including lisuride, apomorphine.

Another widely used treatment to model the neurodegeneration associated with PD in worms is 6-OHDA. This chemical can induce DA neurotoxicity by autooxidation-derived

ROS and the inhibition of the mitochondrial respiratory chain (mitochondrial complex I and

IV) 71.Based on this, researchers identified two D2 receptor agonists, bromocriptine and quinpirole, that protected against cell death induced by 6-OHDA toxicity in a dose-dependent manner 72.

Although mammalian models often provide many clinical features associated with dystonia such as tremors, muscle contraction, it is often expensive. In this regard, C. elegans has been utilized for genetic studies to identify genes and drugs that might be therapeutically relevant. This is possible because homologues of several primary dystonia genes have been identified in C. elegans. (Table 1.2) Many cellular, genetic and chemical studies have been performed using the C. elegans homologs. For example, a drug screen in a C. elegans model of the chaperone defects associated with early-onset torsion dystonia, identified two classes of antibiotics, quinolones and aminopenicillins as chemical effectors which enhanced WT

14 torsinA activity. Notably, one of these drug classes, ampicillin, also rescued beam-walking behavior defect in heterozygous Dyt1 ΔE knock-in mouse model and also improved torsinA- dependent secretory function in DYT1 patient fibroblasts 73.

Table 1.2 Current DYT loci with brief descriptions of associated phenotypes, mode of inheritance and corresponding C. elegans orthologs.

Dystonia Phenotype Dystonia Function Inheritance C. elegans Gene Ortholog/ E value DYT1: Early-onset TOR1A ATPase; molecular AD tor-2; 2e-22 primary torsion chaperone ooc-5; 5e-20 dystonia DYT2: Early-onset HPCA Calcium-binding AR ncs-1; 3e-63 primary dystonia protein; possibly ncs-2; 3e-50 with prominent regulate voltage- ncs-3; 8e-57 cranio-cervical dependent calcium involvement channels DYT3: Adult onset TAF1 The largest X-linked taf-1;0 dystonia- component of recessive parkinsonism, transcription factor prevalent in the TFIID; contain novel Philippines. N- and C-terminal Ser/Thr kinase domains; Acetyltransferase activity DYT4: Whispering TUBB4A A member of the beta AD tbb-1; 0 dystonia (adult onset tubulin family; cell tbb-2; 0 spasmodic cycle tbb-4; 0 dysphonia) with mec-7;0 generalization and ben-1; 0 “hobby horse” gait DYT5a: Progressive GCH1 The cofactor for AD cat-4; 1e-75 DOPA-responsive tyrosine hydroxylase; dystonia or complex the rate-limiting encephalopathy enzyme for dopamine synthesis DYT5b: Akinetic TH Tyrosine Hydroxylase; AR cat-2; e-106 rigid syndrome with involve in the DOPA-responsive conversion of tyrosine dystonia with diurnal to dopamine; the rate- variation limiting enzyme in the synthesis of catecholamines

DYT6: Adult-onset THAP1 DNA-binding AD N/A torsion dystonia with transcription prominent cranio- regulator; pro- cervical and apoptotic activity laryngeal involvement DYT7: Adult-onset chromosome N/A AD N/A primary cervical 18p dystonia DYT8: Paroxysmal MR1 Antigen-presenting AD N/A non-kinesigenic molecule specialized dyskinesia in presenting microbial vitamin B metabolites DYT10: Paroxysmal PRRT2 Proline rich AD N/A kinesigenic transmembrane dyskinesia protein 2; component of the outer core of AMPAR complex DYT11: Myoclonic SGCE Component of the AD sgca-1; 9e-06 dystonia (often with sarcoglycan alcohol complex; a linker responsiveness) between the F-actin cytoskeleton and the extracellular matrix DYT12: Rapid onset ATP1A3 Na+/K+ -ATPases; AD eat-6;0 dystonia catalyze the catp-4; 0 parkinsonism and hydrolysis of ATP alternating coupled with the hemiplegia of exchange of sodium childhood and potassium ions across the plasma membrane DYT13: Early-onset Chromosome N/A AD N/A torsinA dystonia in 1p36.32- one Italian family p36.13 DYT15: Myoclonic Chromosome N/A AD N/A dystonia with 18p11 alcohol responsiveness in one Canadian kindred DYT16: Early-onset PRKRA Protein activator of AR N/A dystonia- interferon induced parkinsonism protein kinase EIF2AK2; Apoptosis; RNA silencing

DYT17: Primary Chromosome N/A AR N/A focal dystonia with 20p11.2- progression in one q13.12 Lebanese family DYT18: Paroxysmal SLC2A1 Facilitative glucose AD fgt-1; 6e-84 exercise-induced transporter dyskinesia+ epilepsy DYT19: Episodic Chromosome N/A AD N/A kinesigenic 16q13-q22.1 dyskinesia 2 DYT20: Paroxysmal Chromosome N/A AD Y17G7B.3; non-kinesiogenic 2q31; 2e-35 dyskinesia 2, in one candidate large Canadian locus: family PNKD2 DYT21: Adult-onset Chromosome N/A AD N/A mixed dystonia with 2q14.3-q21.3 generalization in one Swedish family DYT22: Reserved, N/A N/A N/A N/A but not published DYT23: Autosomal CACNA1B Calcium Voltage- AD unc-2; 0 dominant, often Gated Channel egl-19; 0 tremulous cranio- Subunit Alpha1 B; cervical dystonia+ calcium-dependent upper limb tremor intracellular processes DYT24: Adult-onset ANO3 Intracellular calcium AD anoh-1; 6e-49 cranio-cervical activated chloride primary torsion channel activity and dystonia phospholipid scramblase activity DYT25: Adult onset GNAL Guanine nucleotide- AD gsa-1; e-142 of focal dystonia binding proteins (G proteins); modulators or transducers in various transmembrane signaling systems DYT26: Myoclonic KCTD17 Potassium channel AD inso-1; 2e-51 dystonia Tetramerization Domain Containing 17; A positive regulator of ciliogenesis; endoplasmic reticulum calcium ion homeostasis

DYT27: Onset of COL6A3 One member of type AR N/A segmental isolated VI collagen family; dystonia serine-type endopeptidase inhibitor activity DYT28: childhood- KMT2B Histone AD set-2; 3e-40 onset dystonia methyltransferase DYT29: childhood- MECR Mitochondrial Trans- AR mecr-1; onset dystonia with 2-Enoyl-CoA 8e-77 optic atrophy and Reductase; catalyze Y48A6B.9; basal ganglia the last step in 2e-64 abnormalities mitochondrial fatty (DYTOABG) acid synthesis N/A: Not applicable; AD= Autosomal dominant; AR= Autosomal recessive; DYT9, DYT14 are not included in the table as they are now known to be synonymous with DYT18, DYT5a, respectively.

Focus of dissertation

My research has been based on learning more about movement disorders using the animal model organism, C. elegans. In Chapter 2, I describe collaborative research where the

Prestwick library of 1121 Food and Drug Administration-approved drugs were screened in yeast expressing α-syn and four compounds were identified that rescued a mitochondrial phenotype. To determine if these compounds could rescue dopaminergic neurodegeneration, these four compounds were examined in C. elegans, where we demonstrated that a subset of these compounds were indeed neuroprotective. We discuss how these drugs might block - syn pathology in DA neurons 74.

In Chapter 3, I continue my investigations of PD pathogenesis in C. elegans by investigating the role of a specific miRNA, mir-239, and its targets in dopaminergic neurodegeneration in a worm strain that expresses -synuclein. Through these studies we identified that depletion of 9 gene products significantly reduce neuron protection afforded by the mir-239 mutant, indicating these genes act as downstream targets.

In Chapter 4, I turned my attention to dystonia, and investigated a role for tor-2, the C. elegans homolog of human torsinA, in controlling ER export of glutamate AMPA receptors.

These studies primarily involved genetic analyses where I connected tor-2 with glr-1, an

AMPA receptor subunit, using behavioral and cellular readouts.

Chapter 5 describes specific outcomes and future directions of these three research projects. Collectively, these works provide more proof that C. elegans is a powerful model system to evaluate genetic factors underlying the neuronal dysfunction of PD and dystonia. In total, the genes, proteins, biological pathways and compounds characterized within this dissertation expand our understanding of mechanisms underlying these movement disorders and builds a foundation for further validation in mammalian models.

References

1. Poewe, W. et al. Parkinson disease. Nat Rev Dis Primers 3, 17013, doi:10.1038/nrdp.2017.13 (2017).

2. Massano, J. & Bhatia, K. P. Clinical Approach to Parkinson's Disease: Features, Diagnosis, and Principles of Management. Cold Spring Harbor Perspectives in Medicine 2, a008870, doi:10.1101/cshperspect.a008870 (2012).

3. Wood-Kaczmar, A., Gandhi, S. & Wood, N. W. Understanding the molecular causes of Parkinson's disease. Trends Mol Med 12, 521-528, doi:10.1016/j.molmed.2006.09.007 (2006).

4. Spillantini, M. G. et al. Alpha-synuclein in Lewy bodies. Nature 388, 839-840, doi:10.1038/42166 (1997).

5. Martinez, B. A., Caldwell, K. A. & Caldwell, G. A. C. elegans as a model system to accelerate discovery for Parkinson disease. Curr Opin Genet Dev 44, 102-109, doi:10.1016/j.gde.2017.02.011 (2017).

6. Di Monte, D. A., Lavasani, M. & Manning-Bog, A. B. Environmental factors in Parkinson's disease. Neurotoxicology 23, 487-502 (2002).

7. Surmeier, D. J., Obeso, J. A. & Halliday, G. M. Selective neuronal vulnerability in Parkinson disease. Nat Rev Neurosci 18, 101-113, doi:10.1038/nrn.2016.178 (2017).

8. Saghazadeh, A. & Rezaei, N. MicroRNA machinery in Parkinson's disease: a platform for neurodegenerative diseases. Expert Rev Neurother, 1-27, doi:10.1586/14737175.2015.1114886 (2015).

9. Gatto, N. M. et al. Alpha-Synuclein gene may interact with environmental factors in increasing risk of Parkinson's disease. Neuroepidemiology 35, 191-195, doi:10.1159/000315157 (2010).

10. Guardia-Laguarta, C. et al. Alpha-Synuclein is localized to mitochondria-associated ER membranes. The Journal of neuroscience : the official journal of the Society for Neuroscience 34, 249-259, doi:10.1523/jneurosci.2507-13.2014 (2014).

11. Parihar, M. S., Parihar, A., Fujita, M., Hashimoto, M. & Ghafourifar, P. Alpha- synuclein overexpression and aggregation exacerbates impairment of mitochondrial functions by augmenting oxidative stress in human neuroblastoma cells. The international journal of biochemistry & cell biology 41, 2015-2024, doi:10.1016/j.biocel.2009.05.008 (2009).

12. Wang, S. et al. Phosphatidylethanolamine deficiency disrupts α-synuclein homeostasis in yeast and worm models of Parkinson disease. Proceedings of the

National Academy of Sciences 111, E3976-E3985, doi:10.1073/pnas.1411694111 (2014).

13. Pozo Devoto, V. M. et al. Alpha-Synuclein control of mitochondrial homeostasis in human-derived neurons is disrupted by mutations associated with Parkinson's disease. Sci Rep 7, 5042, doi:10.1038/s41598-017-05334-9 (2017).

14. Garcia-Gimenez, J. L. et al. Epigenetic biomarkers: A new perspective in laboratory diagnostics. Clin Chim Acta 413, 1576-1582, doi:10.1016/j.cca.2012.05.021 (2012).

15. Hastings, M. L. & Krainer, A. R. Pre-mRNA splicing in the new millennium. Curr Opin Cell Biol 13, 302-309 (2001).

16. Morey, C. & Avner, P. Employment opportunities for non-coding RNAs. FEBS Lett 567, 27-34, doi:10.1016/j.febslet.2004.03.117 (2004).

17. Bilen, J., Liu, N. & Bonini, N. M. A new role for microRNA pathways: modulation of degeneration induced by pathogenic human disease proteins. Cell Cycle 5, 2835- 2838, doi:10.4161/cc.5.24.3579 (2006).

18. Winter, J., Jung, S., Keller, S., Gregory, R. I. & Diederichs, S. Many roads to maturity: microRNA biogenesis pathways and their regulation. Nat Cell Biol 11, 228- 234, doi:10.1038/ncb0309-228 (2009).

19. Shivdasani, R. A. MicroRNAs: regulators of gene expression and cell differentiation. Blood 108, 3646-3653, doi:10.1182/blood-2006-01-030015 (2006).

20. Venter, J. C. et al. The Sequence of the Human Genome. Science 291, 1304-1351, doi:10.1126/science.1058040 (2001).

21. Ardekani, A. M. & Naeini, M. M. The Role of MicroRNAs in Human Diseases. Avicenna Journal of Medical Biotechnology 2, 161-179 (2010).

22. Kung, J. T. Y., Colognori, D. & Lee, J. T. Long Noncoding RNAs: Past, Present, and Future. Genetics 193, 651-669, doi:10.1534/genetics.112.146704 (2013).

23. Urbich, C., Kuehbacher, A. & Dimmeler, S. Role of microRNAs in vascular diseases, inflammation, and angiogenesis. Cardiovasc Res 79, 581-588, doi:10.1093/cvr/cvn156 (2008).

24. Tufekci, K. U., Oner, M. G., Meuwissen, R. L. & Genc, S. The role of microRNAs in human diseases. Methods Mol Biol 1107, 33-50, doi:10.1007/978-1-62703-748-8_3 (2014).

25. Filipowicz, W., Bhattacharyya, S. N. & Sonenberg, N. Mechanisms of post- transcriptional regulation by microRNAs: are the answers in sight? Nat Rev Genet 9, 102-114, doi:10.1038/nrg2290 (2008)

26. Ha, T.-Y. MicroRNAs in Human Diseases: From Cancer to Cardiovascular Disease. Immune Network 11, 135-154, doi:10.4110/in.2011.11.3.135 (2011).

27. Martinez, B. & Peplow, P. V. MicroRNAs in Parkinson's disease and emerging therapeutic targets. Neural Regeneration Research 12, 1945-1959, doi:10.4103/1673- 5374.221147 (2017).

28. Cortez, M. A. et al. MicroRNAs in body fluids--the mix of hormones and biomarkers. Nat Rev Clin Oncol 8, 467-477, doi:10.1038/nrclinonc.2011.76 (2011).

29. Nalls, M. A. et al. Large-scale meta-analysis of genome-wide association data identifies six new risk loci for Parkinson's disease. Nat Genet 46, 989-993, doi:10.1038/ng.3043 (2014).

30. Burgos, K. et al. Profiles of extracellular miRNA in cerebrospinal fluid and serum from patients with Alzheimer's and Parkinson's diseases correlate with disease status and features of pathology. PloS one 9, e94839, doi:10.1371/journal.pone.0094839 (2014).

31. Zhang, X. et al. Reduced Circulating Levels of miR-433 and miR-133b Are Potential Biomarkers for Parkinson's Disease. Frontiers in cellular neuroscience 11, 170, doi:10.3389/fncel.2017.00170 (2017).

32. Arshad, A. R. et al. MicroRNAs and Target Genes As Biomarkers for the Diagnosis of Early Onset of Parkinson Disease. Frontiers in molecular neuroscience 10, 352, doi:10.3389/fnmol.2017.00352 (2017).

33. Fahn, S. Classification of movement disorders. Movement disorders : official journal of the Movement Disorder Society 26, 947-957, doi:10.1002/mds.23759 (2011).

34. Goetz, C. G. & Horn, S. S. Treatment of tremor and dystonia. Neurologic Clinics 19, 129-144, doi:10.1016/S0733-8619(05)70009-8.

35. Albanese, A. et al. Phenomenology and classification of dystonia: a consensus update. Movement disorders : official journal of the Movement Disorder Society 28, 863-873, doi:10.1002/mds.25475 (2013).

36. Phukan, J., Albanese, A., Gasser, T. & Warner, T. Primary dystonia and dystonia-plus syndromes: clinical characteristics, diagnosis, and pathogenesis. The Lancet. Neurology 10, 1074-1085, doi:10.1016/s1474-4422(11)70232-0 (2011).

37. LeDoux, M. S. The Genetics of Dystonias. Advances in genetics 79, 35-85, doi:10.1016/B978-0-12-394395-8.00002-5 (2012).

38. Caldwell, K. A., Shu, Y., Roberts, N. B., Caldwell, G. A. & O'Donnell, J. M. Invertebrate models of dystonia. Current neuropharmacology 11, 16-29, doi:10.2174/157015913804999504 (2013).

39. Charlesworth, G., Bhatia, K. P. & Wood, N. W. The genetics of dystonia: new twists in an old tale. Brain 136, 2017-2037, doi:10.1093/brain/awt138 (2013).

40. Mathews, M. et al. Antipsychotic-induced movement disorders: evaluation and treatment. Psychiatry (Edgmont (Pa. : Township)) 2, 36-41 (2005).

41. O'Riordan, S. et al. Age at onset as a factor in determining the phenotype of primary torsion dystonia. Neurology 63, 1423-1426 (2004).

42. Ozelius, L. J. et al. The TOR1A (DYT1) gene family and its role in early onset torsion dystonia. Genomics 62, 377-384, doi:10.1006/geno.1999.6039 (1999).

43. Caldwell, G. A. et al. Suppression of polyglutamine-induced protein aggregation in Caenorhabditis elegans by torsin proteins. Hum Mol Genet 12, 307-319 (2003).

44. Chen, P. et al. The early-onset torsion dystonia-associated protein, torsinA, is a homeostatic regulator of endoplasmic reticulum stress response. Hum Mol Genet 19, 3502-3515, doi:10.1093/hmg/ddq266 (2010).

45. Burdette, A. J., Churchill, P. F., Caldwell, G. A. & Caldwell, K. A. The early-onset torsion dystonia-associated protein, torsinA, displays molecular chaperone activity in vitro. Cell Stress Chaperones 15, 605-617, doi:10.1007/s12192-010-0173-2 (2010).

46. Zacchi, L. F. et al. The BiP Molecular Chaperone Plays Multiple Roles during the Biogenesis of TorsinA, an AAA(+) ATPase Associated with the Neurological Disease Early-onset Torsion Dystonia. The Journal of Biological Chemistry 289, 12727- 12747, doi:10.1074/jbc.M113.529123 (2014).

47. Zhao, C., Brown, R. S. H., Tang, C.-H. A., Hu, C.-C. A. & Schlieker, C. Site-specific Proteolysis Mobilizes TorsinA from the Membrane of the Endoplasmic Reticulum (ER) in Response to ER Stress and B Cell Stimulation. The Journal of biological chemistry 291, 9469-9481, doi:10.1074/jbc.m115.709337 (2016).

48. Zacchi, L. F. et al. Early-onset torsion dystonia: a novel high-throughput yeast genetic screen for factors modifying protein levels of torsinADeltaE. Dis Model Mech 10, 1129-1140, doi:10.1242/dmm.029926 (2017).

49. Nery, F. C. et al. TorsinA participates in endoplasmic reticulum-associated degradation. Nature communications 2, 393-393, doi:10.1038/ncomms1383 (2011).

50. Chen, P. et al. The early-onset torsion dystonia-associated protein, torsinA, is a homeostatic regulator of endoplasmic reticulum stress response. Human Molecular Genetics 19, 3502-3515, doi:10.1093/hmg/ddq266 (2010).

51. Goodchild, R. E., Kim, C. E. & Dauer, W. T. Loss of the Dystonia-Associated Protein TorsinA Selectively Disrupts the Neuronal Nuclear Envelope. Neuron 48, 923-932, doi:10.1016/j.neuron.2005.11.010.

52. Naismith, T. V., Heuser, J. E., Breakefield, X. O. & Hanson, P. I. TorsinA in the nuclear envelope. Proc Natl Acad Sci U S A 101, 7612-7617, doi:10.1073/pnas.0308760101 (2004).

53. Jungwirth, M. T., Kumar, D., Jeong, D. Y. & Goodchild, R. E. The nuclear envelope localization of DYT1 dystonia torsinA-ΔE requires the SUN1 LINC complex component. BMC Cell Biology 12, 24-24, doi:10.1186/1471-2121-12-24 (2011).

54. Brenner, S. The genetics of Caenorhabditis elegans. Genetics 77, 71-94 (1974).

55. Consortium, C. e. S. (1998).

56. Riddle, D. L., Blumenthal, T., Meyer, B. J. & Priess, J. R. in C. elegans II (Cold Spring Harbor Laboratory Press, 1997).

57. Fire, A. et al. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 806-811, doi:10.1038/35888 (1998).

58. Lees, A. J., Hardy, J. & Revesz, T. Parkinson's disease. Lancet (London, England) 373, 2055-2066, doi:10.1016/s0140-6736(09)60492-x (2009).

59. Corti, O., Lesage, S. & Brice, A. What Genetics Tells us About the Causes and Mechanisms of Parkinson's Disease. Physiological Reviews 91, 1161-1218, doi:10.1152/physrev.00022.2010 (2011).

60. Cao, S., Gelwix, C. C., Caldwell, K. A. & Caldwell, G. A. Torsin-mediated protection from cellular stress in the dopaminergic neurons of Caenorhabditis elegans. The Journal of neuroscience : the official journal of the Society for Neuroscience 25, 3801-3812, doi:10.1523/jneurosci.5157-04.2005 (2005).

61. Harrington, A. J., Yacoubian, T. A., Slone, S. R., Caldwell, K. A. & Caldwell, G. A. Functional analysis of VPS41-mediated neuroprotection in Caenorhabditis elegans and mammalian models of Parkinson's disease. The Journal of neuroscience : the official journal of the Society for Neuroscience 32, 2142-2153, doi:10.1523/jneurosci.2606-11.2012 (2012).

62. Ray, A., Zhang, S., Rentas, C., Caldwell, K. A. & Caldwell, G. A. RTCB-1 mediates neuroprotection via XBP-1 mRNA splicing in the unfolded protein response pathway. The Journal of neuroscience : the official journal of the Society for Neuroscience 34, 16076-16085, doi:10.1523/jneurosci.1945-14.2014 (2014).

63. Cooper, A. A. et al. -Synuclein Blocks ER-Golgi Traffic and Rab1 Rescues Neuron Loss in Parkinson’s Models. Science (New York, N.Y.) 313, 324-328, doi:10.1126/science.1129462 (2006).

64. Ruan, Q., Harrington, A. J., Caldwell, K. A., Caldwell, G. A. & Standaert, D. G. VPS41, a protein involved in lysosomal trafficking, is protective in Caenorhabditis elegans and mammalian cellular models of Parkinson's disease. Neurobiology of disease 37, 330-338, doi:10.1016/j.nbd.2009.10.011 (2010).

65. Tardiff, D. F., Tucci, M. L., Caldwell, K. A., Caldwell, G. A. & Lindquist, S. Different 8-hydroxyquinolines protect models of TDP-43 protein, alpha-synuclein, and polyglutamine proteotoxicity through distinct mechanisms. J Biol Chem 287, 4107- 4120, doi:10.1074/jbc.M111.308668 (2012).

66. Knight, A. L. et al. The glycolytic enzyme, GPI, is a functionally conserved modifier of dopaminergic neurodegeneration in Parkinson's models. Cell Metab 20, 145-157, doi:10.1016/j.cmet.2014.04.017 (2014).

67. Di Fonzo, A. et al. A frequent LRRK2 gene mutation associated with autosomal dominant Parkinson's disease. Lancet (London, England) 365, 412-415, doi:10.1016/s0140-6736(05)17829-5 (2005).

68. Gilks, W. P. et al. A common LRRK2 mutation in idiopathic Parkinson's disease. Lancet (London, England) 365, 415-416, doi:10.1016/s0140-6736(05)17830-1 (2005).

69. Saha, S. et al. LRRK2 modulates vulnerability to mitochondrial dysfunction in Caenorhabditis elegans. The Journal of neuroscience : the official journal of the Society for Neuroscience 29, 9210-9218, doi:10.1523/jneurosci.2281-09.2009 (2009).

70. Braungart, E., Gerlach, M., Riederer, P., Baumeister, R. & Hoener, M. C. Caenorhabditis elegans MPP+ model of Parkinson's disease for high-throughput drug screenings. Neuro-degenerative diseases 1, 175-183, doi:10.1159/000080983 (2004).

71. Glinka, Y. Y. & Youdim, M. B. H. Inhibition of mitochondrial complexes I and IV by 6-hydroxydopamine. European Journal of Pharmacology: Environmental Toxicology and Pharmacology 292, 329-332, doi:10.1016/0926-6917(95)90040-3 (1995).

72. Marvanova, M. & Nichols, C. D. Identification of neuroprotective compounds of caenorhabditis elegans dopaminergic neurons against 6-OHDA. Journal of molecular neuroscience : MN 31, 127-137 (2007).

73. Cao, S. et al. Chemical enhancement of torsinA function in cell and animal models of torsion dystonia. Disease Models & Mechanisms 3, 386-396, doi:10.1242/dmm.003715 (2010).

74. Wang, S. et al. Chemical Compensation of Mitochondrial Phospholipid Depletion in Yeast and Animal Models of Parkinson's Disease. PloS one 11, e0164465, doi:10.1371/journal.pone.0164465 (2016).

CHAPTER TWO

CHEMICAL COMPENSATION OF MITOCHONDRIAL PHOSPHOLIPID DEPLETION IN YEAST AND ANIMAL MODELS OF PARKINSON’S DISEASE

This work was published in PLOS ONE, October 13, 2016 under the following citation:

Chuan Xu ³, Shaoxiao Wang, Siyuan Zhang, Addie Barron, Floyd Galiano, Dhaval Patel,

Yong Joo Lee, Guy A. Caldwell, Kim A. Caldwell, Stephan N. Witt (2016) Chemical

Compensation of Mitochondrial Phospholipid Depletion in Yeast and Animal Models of

Parkinson's Disease. PLoS ONE 11(10): e0164465. doi:10.1371/journal. pone.0164465.

Chuan Xu, Shaoxiao Wang, Siyuan Zhang, Addie Barron, Floyd Galiano, Dhaval Patel,

Yong Joo Lee collected all the data. Chuan Xu, Shaoxiao Wang, Siyuan Zhang, Dr. Guy A.

Caldwell, Dr. Kim A. Caldwell and Dr. Stephan N. Witt co-wrote the manuscript.

Abstract

We have been investigating the role that phosphatidylethanolamine (PE) and phosphatidylcholine (PC) content plays in modulating the solubility of the Parkinson's disease protein alpha-synuclein (-syn) using Saccharomyces cerevisiae and Caenorhabditis elegans. One enzyme that synthesizes PE is the conserved enzyme phosphatidylserine decarboxylase (Psd1/yeast; PSD-1/worms), which is lodged in the inner mitochondrial membrane. We previously found that decreasing the level of PE due to knockdown of

Psd1/psd-1 affects the homeostasis of -syn in vivo. In S. cerevisiae, the co-occurrence of low PE and -syn in psd1Δ cells triggers mitochondrial defects, stress in the endoplasmic reticulum, misprocessing of glycosylphosphatidylinositol-anchored proteins, and a 3-fold increase in the level of -syn. The goal of this study was to identify drugs that rescue this phenotype. We screened the Prestwick library of 1121 Food and Drug Administration- approved drugs using psd1Δ + -syn cells and identified cyclosporin A, meclofenoxate hydrochloride, and sulfaphenazole as putative protective compounds. The protective activity of these drugs was corroborated using C. elegans in which -syn is expressed specifically in the dopaminergic neurons, with psd-1 depleted by RNAi. Worm populations were examined for dopaminergic neuron survival following psd-1 knockdown. Exposure to cyclosporine, meclofenoxate, and sulfaphenazole significantly enhanced survival at day 7 in -syn- expressing worm populations whereby 50±55% of the populations displayed normal neurons, compared to only 10±15% of untreated animals. We also found that all three drugs rescued worms expressing -syn in dopaminergic neurons that were deficient in the phospholipid cardiolipin following cardiolipin synthase (crls-1) depletion by RNAi. We discuss how these drugs might block -syn pathology in dopaminergic neurons.

Introduction

Parkinson’s disease (PD) affects 1–2% of the population over 65 years of age and is the most common movement disorder. The disease is a consequence of the selective degeneration of dopaminergic neurons in a region of the mid-brain called the substantia nigra 1. Loss of these neurons results in slowness of movement, rigidity, and postural instability. The affected neurons often display cytoplasmic inclusions called Lewy bodies, whose main component is the protein -syn 2. Post-translationally modified forms of -syn or an accumulation of -syn due to age-related declines in the protein degradation pathways likely cause sporadic cases of PD. Missense mutations in -syn 3 or duplications/triplications

4 of the locus result in early-onset PD. In some individuals, if -syn slowly accumulates over time, eventually toxic, oligomeric conformations may form and disrupt cell function, leading to cell death. The toxic conformations kill the host neurons and spread to healthy neighboring neurons.

Highly expressed in the brain, -syn is also present in red blood cells, intestinal cells, liver cells, and melanocytes. -syn, which has sequence similarity to lipid binding proteins 5, binds to membranes, vesicles, and even sequesters large numbers of lipid molecules to form nanoparticles 6,7, consistent with it being a lipid carrier. -syn has also been proposed to act in concert with soluble N-ethylmaleimide-sensitive factor attachment protein receptor

(SNARE) proteins to facilitate synaptic vesicle fusion with the presynaptic membrane 8. A wealth of evidence is consistent with -syn changing its structure in a context-dependent manner. That is, -syn is intrinsically disordered in solution 9 but upon binding to membranes it adopts a α-helical conformation 10. If -syn builds up in cells, then it self- associates into a myriad array of soluble protofibrils, some of which may be toxic 11. -syn can also form amyloid fibers. Preformed fibers of -syn, when injected into healthy mice,

28 cause a rapid neurodegenerative disease consistent with PD 12. The molecular details as to how-syn changes conformations, kills, and spreads are the subjects of intense investigations.

PE and its metabolites can decline in the brain with age 13-17. -syn is thought to slowly aggregate and form inclusions in neurons with age. In light of these phenomena, we hypothesized that decreasing the level of PE in cells would affect -syn homeostasis, possibly leading to inclusion/ foci formation. To this end, we used S. cerevisiae and C. elegans models of PD. The various pathways for the formation of PE and the enzymes that synthesize PE are conserved in yeast, worms, flies and mammals 18,19 (Fig. 2.1). First, lodged in the inner membrane, the enzyme Psd1 converts phosphatidylserine (PS) to PE 20. PE synthesized in the inner mitochondrial membrane can spread via mitochondrial-associated membranes to other cellular compartments 21,22. Second, the cytidine diphosphate (CDP)- ethanolamine (Kennedy) pathway consists of three enzymes that convert the metabolite ethanolamine into PE 23; the last enzyme in this pathway is embedded in the membranes of the endoplasmic reticulum (ER). In some cells, Psd1 may synthesize most of the PE whereas in other cells the Kennedy pathway may synthesize most of the PE. Using yeast and worms, we showed that decreasing the level of PE by knocking down the gene coding for phosphatidylserine decarboxylase triggers mitochondrial defects, stress in the ER, misprocessing of glycosylphosphatidylinositol-anchored-anchored proteins, and a 3-fold increase in the level of -syn 24. Supplementation of yeast or worms with ethanolamine, which converts to PE via the CDP-ethanolamine pathway, abolished the extramitochondrial defects due to the co-occurrence of low PE (psd1Δ) and -syn. We were curious whether any

Food and Drug Administration (FDA)-approved drugs would rescue cells with low PE and - syn. High throughput screening identified three drugs—meclofenoxate hydrochloride (MFX), cyclosporineA (CsA), and sulfaphenazole (SUL)—that rescued the slow growth phenotype of

29 psd1Δ cells expressing -syn. The drugs were then further evaluated in a C. elegans model of

-syn-induced dopaminergic neurodegeneration.

Figure 2.1. PE and CL synthesis in mitochondria and ER 19. CDP, cytidine diphosphate; Cho, choline; DG, diacylglycerol; ER, endoplasmic reticulum; ETA, ethanolamine; p-ETA/p-Cho, phosphorylated ETA/choline; PM, plasma membrane; PS, phosphatidylserine. Mitochondrial PE deficiency causes mitochondrial defects, ER and cell wall stress, misprocessing of glycosylphosphatidylinositol-anchored proteins, accumulation of -syn. Cardiolipin deficiency causes defects in mitochondrial bioenergetics.

Results

High throughput screen of prestwick library

A high throughput screen of the Prestwick library of 1121 FDA-approved drugs was conducted to identify drugs that rescue the slow growth phenotype of psd1Δ yeast cells expressing human wild-type -syn under the control of the Gal1 promoter. Ethanolamine was a positive control. Two hits—MFX and SUL—were identified that rescued the growth defect almost as well as ethanolamine, whereas CsA was one of the weaker hits. These three drugs

(Fig. 2.1) of the highest purity were purchased from Sigma and re-tested in yeast. To estimate

30 dose response relationships, each drug was retested in a growth assay in a liquid medium over a range of concentrations in psd1Δ/-syn cells. The dose-response curves showed half- maximal responses of 10 μM for MFX and SUL and 100 μM for CsA (data not shown). We have previously shown that psd1Δ cells expressing -syn display stress in the ER and cell wall. Therefore, we tested each of the three drug candidates for their ability to inhibit ER stress using yeast carrying the reporter plasmid pMCZ-Y 25. After culturing psd1Δ/-syn cells for 8 h in inducing medium with the indicated drug cells were lysed and the LacZ activity, which is proportional to ER stress, was measured in the clarified lysate. SUL decreased ER stress by 88% compared to the control (DMSO) (p <0.0001) (data not shown), whereas CsA decreased ER stress by 32% (p=0.0001) and MFX slightly increased ER stress

(p=0.02).The three drug candidates were also tested for their ability to inhibit cell wall stress using a reporter plasmid in which the bacterial lacZ gene is controlled by the Rlm1-regulated promoter of PRM5 (PPRM5::lacZ) 24,25. Prm5 is induced in response to activation of the cell wall integrity pathway. After culturing psd1Δ/-syn cells for 8 h in inducing medium with the various drugs cells were lysed and the LacZ activity, which is proportional to cell wall stress, was measured. SUL and MFX each decreased cell wall stress by ~50% compared to control (DMSO) (0.001 ≤ p ≤ 0.002) (data not shown). CsA increased cell wall stress compared to the control (p<0.0001). The three drugs appear to affect different pathways in psd1Δ cells (Table 2.1).

Table 2.1. Phenotypes of yeast and worms expressing -syn after treatment with candidate drugs

Yeast  -syn /low PE (psd1) Worm  -syn /low PE or CL Drug Growth ER stress Cell wall DA neuron stress loss DMSO negligible yes yes yes MFX strong rescue slightly rescue rescue enhance SUL strong rescue strong rescue rescue rescue CsA weak rescue weak rescue enhance rescue

Corroboration using a C. elegans neurodegeneration model

To further investigate the findings from yeast, we tested the four drugs including meclofenoxate, cyclosporine A, sufaphenazole and choline chloride (Fig. 2.2.-Fig. 2.6.) in a

C. elegans neurodegeneration model where expression of wild-type (non-mutated) human - syn cDNA under control of a dopamine transporter-specific promoter [Pdat-1::-syn + Pdat-

26,27 1::GFP] results in progressive, dose-dependent neurodegeneration . To enable depletion of neuronal genes using RNAi (RNA interference)-mediated silencing, the dopaminergic neuron-sensitive RNAi strain of C. elegans (UA196 [sid-1(pk3321); Pdat-1::-syn, Pdat-

28 1::GFP; Pdat-1::sid-1, Pmyo-2::mCherry])was used .We have shown that RNAi silencing of psd-1 expression enhances the toxicity of -syn in the dopaminergic neurons 24. Drugs were tested in UA196 worms with EV (empty vector) RNAi or psd-1 knocked down by RNAi.

Figure 2.2. Structures of positive candidates identified from screen. (A) Sulfaphenazole (SUL), (B) meclofenoxate hydrochloride (MFX) and (C) cyclosporine A (CsA).

To expand our knowledge of the effects of lipid deficiencies in neurons, we asked how

RNAi-depletion of cardiolipin synthase (crls-1) affects the survival of nematode

33 dopaminergic neurons expressing -syn. The phospholipid cardiolipin (CL) is critical for proper functioning of the mitochondrial respiratory complexes and it predominantly localizes to the inner mitochondrial membrane. PE and CL are cone-shaped phospholipids that tend to impart negative curvature to bilayers 29. Each of these phospholipids forms hexagonal phases and are considered “non-bilayer forming lipids”. Because PE and CL have similar biophysical properties and overlapping functions 30, we hypothesized that drugs that rescue cells with -syn and low PE would also rescue those with -syn and low or no CL. Drugs were also tested in UA196 worms with EV RNAi or crls-1 RNAi.

Using the dopaminergic-specific -syn RNAi strain, we depleted psd-1 or crls-1 by RNAi and analyzed dopaminergic neurodegeneration in worm populations compared to EV

(negative control) RNAi. The solvent control treatments (DMSO or water) did not significantly interfere with the RNAi conditions. As described previously 24, at day 7 there was a significant difference between -syn -expressing worms treated with EV + DMSO, where about 30% of the population displayed a full complement of dopaminergic neurons and -syn expressing worms treated with psd-1 RNAi + DMSO, where about ~10%–15% of the population retained a full complement of neurons (Fig. 2.3A–Fig. 2.6 A; p < 0.05; two- way ANOVA). Depleting crls-1 in a DMSO background also results in ~10–15% of the population preserving their dopaminergic neurons (Fig.2.3B- Fig.2.6B; p < 0.05; two-way

ANOVA).

MFX, CsA, SUL, and Choline protect worm dopaminergic neurons that express -syn and with psd-1 or crls-1 depleted by RNAi from degeneration

MFX. Based on the results from yeast, we initially tested MFX at four doses ranging from

3.75 to 30 μM). MFX at 3.75 μM protected against dopaminergic cell loss in psd-1 dsRNA and EVRNAi worms (Fig. 2.3 A). The rescue by MFX on psd-1 depleted -syn transgenic

34 worms went from 9% at baseline treatment to ~53% at all concentrations tested (p < 0.0001, two-wayANOVA). Similarly, MFX increased the population of EV RNAi worms that displayed the full complement of neurons from 30% to 50% at all concentrations tested (p <

0.05). MFX also increased the percentage of -syn expressing crls-1 worms with a full complement of neurons in a dose-dependent fashion from 20 to 50% (p < 0.05, two-way

ANOVA), with a half-maximal response at 8μM MFX (Fig. 2.3 B). Given that the activity of

MFX was detected at the lowest dose tested (3.75 μM), we proceeded to examine additional decreasing concentrations (Fig. 2.3 C) in both psd-1 and crls-1 knockdown worms. These latter studies revealed that concentrations 1μM and above will rescue DA neurodegeneration in psd-1 and crls-1 depleted animals.

Figure 2.3. -syn -induced dopaminergic neurodegeneration in C. elegans is rescued by MFX. Treatment of -syn -expressing dopaminergic neurons with psd-1 or crls-1 RNAi, which causes enhanced neurodegeneration compared to -syn alone, is also rescued by MFX (A-C). Graphical representation of C. elegans strain UA196[sid-1(pk3321); Pdat-1::-syn, Pdat- 1::GFP; Pdat-1::sid-1, Pmyo-2::mCherry] following psd-1 (A, C) or crls-1 (B, C)knockdown. For RNAi experimental conditions, synchronized C. elegans were analyzed at day 7 post- hatching. RNAi bacteria, which do not express an RNAi clone (EV), were used as a negative control. A worm was scored as normal when it had a full complement of six anterior dopaminergic neurons. Data are reported as the mean ± SD, n = 90 worms. *p < 0.05, two- way ANOVA. (A) C. elegans fed with EV or psd-1 dsRNA were treated with MFX (0, 3.75, 7.5, 15, 30 μM dissolved in 0.1% v/v DMSO). (B) The same MFX concentrations were analyzed following dopaminergic neuron-specific EV or crls-1 knockdown. (C) C. elegans fed with EV, psd-1, or crls-1 dsRNA were treated with MFX (0, 0.5, 1, 2, 3.75 μM dissolved in 0.1% v/v DMSO).

MFX is an ester of 4-chlorophenoxyacetic acid and dimethylethanolamine (DMAE). MFX is thought to hydrolyze into these two components inside cells 31. Because DMAE (HO-CH2-

N(CH3)2) and choline (HO-CH2-N+(CH3)3), which is an essential nutrient, differ by only a single methyl group, we also tested choline.

Choline. Choline at 1.25 μM protected against-syn -induced dopaminergic cell loss inworms treated with psd-1 dsRNA, crls-1 dsRNA, and EV RNAi (Fig. 2.4 A and 2.4 B). For

36 example,after 7 days, there was a significant difference between -syn -expressing worms with either psd-1 or crls-1 depleted, where only 10% of the population displayed a full complement of neurons,but in -syn–expressing worms with psd-1 or crls-1–depleted treated with 1.25 μM choline, 50% or 40% of the respective populations displayed a full complement of dopaminergicneurons (p < 0.05, two-way ANOVA). Notably, after 7 days the population of -syn expressing EV worms that retained a full complement of neurons was 25–30%, whereas the population of the same worms treated with choline (1.25 μM) increased to 50%

(p< 0.05) (Fig. 2.4 A and 2.4 B, left panels). For the various strains, increasing the dose above 1.25 μM protected no further, suggesting that the response is saturated at this concentration and that the dose for the half maximal response is less than 1 μM.

Figure 2.4. -syn -induced dopaminergic neurodegeneration in C. elegans is rescued by choline chloride. Treatment of -syn expressing dopaminergic neurons with psd-1 or crls-1 RNAi, which causes enhanced neurodegeneration compared to -syn alone, is also rescued by choline chloride (A, B). Graphical representation of C. elegans strain UA196 [sid- 1(pk3321); Pdat-1::-syn, Pdat-1::GFP;Pdat-1::sid-1, Pmyo-2::mCherry] following psd-1 (A) or crls-1 (B) knockdown. For both RNAi experimental conditions, synchronized C. elegans were analyzed at day 7 post-hatching. RNAi bacteria, which do not express an RNAi clone (EV), were used as a negative control. A worm was scored as normal when it had a full complement of six anterior dopaminergic neurons. Data are reported as the mean ± SD, n = 90 worms. *p < 0.05, two-way ANOVA. (A) C. elegans fed with EV or psd-1 dsRNA were treated with choline chloride (0, 1.25, 2.5, 5, 10 μM dissolved in water). (B) The same choline concentrations were analyzed following dopaminergic neuron-specific EV or crls-1 knockdown.

CsA. CsA protected against dopaminergic cell loss in psd-1, crls-1, and EV RNAi worms

(Fig. 2.5 A and 2.5 B). CsA treatment of -syn expressing worms increased the number of worms with a full complement of dopaminergic neurons from 17% to 45% and 50% for worms with crls-1 and psd-1 knocked down, respectively. Full protection occurred in each case at the initial lowest dose of CsA tested, which was 3.75 μM (Fig. 2.5 A and 2.5 B), suggesting that the half-maximal concentration of CsA is less than 3.75 μM. CsA treatment of

-syn expressing EV control worms exhibited a dose-response curve with a half-maximal

CsA concentration of 8 μM. Therefore, we proceeded to examine additional lower concentrations (0.5–2 μM) (Fig. 2.5 C).These studies revealed that that concentrations 1 μM and above will rescue DA neurodegenerationin psd-1 and crls-1 depleted animals.

Figure 2.5. -syn -induced dopaminergic neurodegeneration in C. elegans is rescued by CsA. Treatment of -syn -expressing dopaminergic neurons with psd-1 or crls-1 RNAi, which causes enhanced neurodegeneration compared to -syn alone, is also rescued by CsA. (A-C). Graphical representation of C. elegans strain UA196[sid-1(pk3321); Pdat-1::-syn, Pdat-1::GFP; Pdat-1::sid-1, Pmyo-2::mCherry] following psd-1 (A, C) or crls-1 (B, C) knockdown. For both RNAi experimental conditions, synchronized C. elegans were analyzed at day 7 posthatching. RNAi bacteria, which do not express an RNAi clone (EV), were used as a negative control. A worm was scored as normal when it had a full complement of six anterior dopaminergic neurons. Data are reported as the mean ± SD, n = 90 worms. *p < 0.05, two-way ANOVA. (A) C. elegans fed with EV or psd-1 dsRNA treated with CsA (0, 3.75, 7.5, 15, 30 μM dissolved in 0.1% v/v DMSO). (B) The same CsA concentrations were analyzed following dopaminergic neuron-specific EV or crls-1 knockdown. (C) C. elegans EV, psd-1, or crls-1dsRNA were treated with CsA (0, 0.5, 1, 2, 3.75 μM dissolved in 0.1 v/v DMS0).

SUL. -syn-expressing worms were less sensitive to SUL than the other drugs. SUL rescued the enhanced neurodegeneration in psd-1 and crls-1 depleted worms only at the highest initial concentration tested (30 μM). For example, SUL at 30 μM boosted the population of wormswith a full complement of neurons from 11%/ 10% to 53%/ 40% in the psd-1/crls-1 -depleted,-syn-expressing worms (Fig. 2.6 A and 2.6 B) (p < 0.0005, two-way

ANOVA). Unlike the other drugs, SUL failed to protect the EV control worms from progressive -syn-dependent neuron loss. To potentially discern a more detailed response profile for SUL, we conducted another set of trials using a broader range of concentrations

(Fig. 2.6 C). Nevertheless, this subsequent analysis revealed that 30 μM is the optimal dosage yielding neuroprotection when psd-1 or crls-1 are knocked down by RNAi.

Figure 2.6. Depletion of psd-1 or crls-1 in -syn -expressing dopaminergic neurons enhances neurodegeneration; this neurodegeneration is rescued by 30 μM SUL. Additionally, when dopaminergic neurons are depleted for psd-1 and treated with 30 μM SUL they display enhanced protection beyond wild type levels (A-C). Graphical representation of C. elegans strain UA196 [sid-1(pk3321); Pdat-1::-syn, Pdat-1:: GFP; Pdat-1::sid-1, Pmyo-2::mCherry] following psd-1 (A, C) or crls-1 (B, C) knockdown. For both RNAi experimental conditions, synchronized C. elegans were analyzed at day 7 post-hatching. RNAi bacteria, which do not express an RNAi clone (EV), were used as a negative control. A worm was scored as normal when it had a full complement of six anterior dopaminergic neurons. Data are reported as the mean ± SD, n = 90 worms. *p < 0.0005, two-way ANOVA. (A) C. elegans fed with EV or psd-1 dsRNA were treated with SUL was treated (0, 3.75, 7.5, 15, 30 μM dissolved in 0.1% v/v DMSO). (B) The same SUL concentration range was tested following dopaminergic neuron-specific EV or crls-1 knockdown. (C) C. elegans fed EV, psd-1, or crls-1 dsRNA were treated with SUL (0, 10, 20, 30, 40 μM dissolved in 0.1% DMS0).

-syn expression is not altered following exposure to MFX, CsA, SUL or Choline in C. elegans

We were curious to know if the neuroprotection from -syn in C. elegans was due to lowered -syn gene expression.We treated -syn -expressing worms with the most effective rescuing concentration of each drug and then worms were harvested at day 7 post-hatching for -syn mRNA analysis. This was also the day used for DA neurodegeneration analysis.We determined that -syn expression is unchanged among various drug treatments and solvent

42 controls (Fig. 2.7 ). Thus, the DA neuroprotection observed is not altering -syn expression levels.

Figure 2.7. Choline, MFX, CsA, and SUL exposure do not affect -syn expression. Stained agarose gel image depicting the products of semi-quantitative RT-PCR reactions following C. elegans strain UA196 [sid-1(pk3321); Pdat-1::-syn, Pdat-1::GFP; Pdat-1::sid-1, Pmyo-2:: mCherry] exposure to chemicals or solvents. Worms were exposed to the following treatments: 10 μM choline chloride (in ddH2O solvent), 30 μM MFX, 30 μM CsA, 30 μM SUL (all in 0.1% DMSO solvent). The products of RT-PCR are shown for -syn primers (top panel) and ama-1 loading control primers (bottom panel). Equal amounts of PCR product were loaded in each lane. The normalized intensity values for the various drug treatments and solvent controls are shown below the blot images.

Discussion

We identified three drugs that compensate for mitochondrial phospholipid depletion in animal models of Parkinson’s disease. CsA and MFX have been previously reported to protect against neurodegenerative phenotypes in various models 32-35, whereas this is the first

43 report that SUL protects against -syn -induced DA neuron loss. None of the drugs affected the expression of -syn (Fig. 2.7 ). Guided by reports on MFX, CsA, and SUL, we discuss below how these drugs might function to compensate for mitochondrial phospholipid depletion in our PD models.

MFX. MFX protects nematode DA neurons with or without lipid depletion from age- and

-syn-associated cell death (Fig. 2.3 ). MFX (a.k.a. centrophenoxine) is a nootropic drug that is marketed as a memory enhancer. This drug,which readily crosses the blood-brain barrier, has been reported to inhibit enzymes involved in PC biosynthesis 36,37, increase acetylcholine, scavenges radicals 38, and ameliorate rotenone-induced motor dysfunction in rodents34,35.

MFX rapidly hydrolyzes into 4-chlorophenoxyacetic acid and dimethylethanolamine

(DMAE) at neutral pH, and DMAE is considered to be the active product because of its ability to scavenge hydroxyl radicals 39. DMAE can also couple with diacylglycerol to yield phosphatidyl-DMAE, which incorporates into membranes and scavenges radicals 31.

Phosphatidyl-DMAE can also be methylated to PC. We suggest that MFX protects nematode

DA neurons with or without lipid depletion from age- and -syn-associated cell death by scavenging radicals or increasing the level of PC.

Choline. Choline partially ameliorates -syn-induced neuron loss in worms (EV) without depletion of mitochondrial lipids (Fig. 2.4 ). Feeding choline to worms should increase the level of PC by stimulating the CDP-phosphatidylcholine pathway (Fig. 2.1). PC synthesized in the Kennedy pathway helps maintain -syn in a soluble non-toxic state 24, and as shown here it rescues mitochondrial PE or PC depletion. Given that the CL and PC synthetic pathways do not intersect, how does choline rescue mitochondrial CL deficiency? First, although PC comprises 50% of the mitochondrial membranes, little is known about its role in mitochondrial protein biogenesis or stability. Second, uncharacterized homeostatic

44 mechanisms exist that maintain the proper ratio of bilayer-forming lipids like PC to nonbilayer forming lipids like CL and PE. This is evidenced by two recent discoveries using the psd1Δ yeast cells 22. We suggest that supplemental choline increases PC, and that PC compensates for low CL by the same or similar homeostatic mechanism by which supplementalethanolamine increase CL.

CsA. CsA protected nematode dopamine neurons with or without lipid depletion from progressive -syn-associated cell death (Fig. 2.5). CsA is a powerful immunosuppressant that prevents T-cell activation by inhibiting Ca++-signaling 40. CsA has a secondary activity in many types of cells, i.e., CsA-cyclophilin D complexes inhibit the mitochondrial permeability transitionpore (mPTP) 41. The mPTP is a non-specific pore that forms in the innermitochondrial membranes in response to high matrix Ca++, arachidonic acid 42, ceramide43, inorganic phosphate, and many other factors. Pore opening is reversible, although prolonged opening of the mPTP results in the collapse of the proton motive force, release of matrix NADH, and leakage of cytochrome c into the cytosol, which in turn triggers cell death. The mPTP, which also exists in yeast, is relevant to cardiac injuries 44 and neurodegeneration 45,46.

CsA likely protects worm neurons by inhibiting the mPTP. In worm dopaminergic neurons that express -syn, we propose that there are two inciters of mPTP, and one of these is -syn.

For example, recombinant -syn binds to purified rat brain mitochondria, which depolarizes the membranes, activates the mPTP, and releases of cytochrome c 47-49. CsA (1 μM) blocks mPTP activation in human SH-SY5Y cells under conditions in which -syn accumulates in cells due to inhibition of the proteasome 49. A similar concentration (~3 μM) of CsA partially blocks neurodegeneration in C. elegans (Fig. 2.5C). The second mPTP inciter, we propose, is low PE or low CL, from aging 50 or modeled herein by RNAi depletion. An issue to address

45 is why the dose-response curves for CsA treated -syn/psd-1 depleted worms and -syn/crls-

1 depleted worms are so similar (Fig. 2.5). Our view is that mitochondrial depletion of CL or

PE in dopaminergic neurons yields identical output: the mPTP is activated. Consequently,

CsA inhibits the mPTP with a dose response curve that is independent of whether the instigatingevent was mitochondrial CL or PE depletion.

SUL. SUL is a sulfonamide antibiotic that targets bacterial replication by inhibiting folate biosynthesis. SUL is also a selective inhibitor of the mammalian cytochrome P450 isozyme

CYP2C9 51, which oxidizes as much as ~15% of drugs undergoing phase I clearance in theliver. CYP2C9 is also expressed in the heart and the brain 52. SUL was recently identified in a screen of the Prestwick drug library that sought to identify drugs that block light- induced, degenerative loss of photoreceptors that occurs in inherited and age-related retinal degenerative diseases 53. SUL inhibits light-induced necrosis and apoptosis of mouse-derived photoreceptor 661W cells. It was concluded that SUL blocks this cell death pathway by inhibiting CYP2C9. Since there are potentially at least two targets for SUL, perhaps this explains the bellshaped dose-response curves (Fig. 2.6 C), where at low doses (< 30 μM)

SUL partially rescues whereas at high doses (> 30 μM) it is toxic.

A BLAST search query using the human CYP2C9 protein (cytochrome P450 family 2 subfamily C polypeptide 9) sequence against the C. elegans protein database yielded the

CytochromeP450 cyp-33c11 (6e-74). cyp-33c11 is a homolog of human gene CYP2J2.

CYP2J2 is conserved in 86 organisms, including chimpanzee, rhesus monkey, dog, cow, mouse, rat, chicken, zebrafish, fruit fly, mosquito, and frog. In humans, the CYP2J2 and

CYP2C9 isozymes carry out the epoxidation of endogenous arachidonic acid in the heart and brain, respectively. If SUL inhibits worm cyp-33c11 the conversion of arachidonic acid from membrane phospholipids to vasoactive epoxyeicosatrienoic acids should be blocked, which

46 should increase the level of arachidonic acid in cells. Arachidonic acid upregulates syntaxin 1 and promotes its interaction with the SNARE complex 54. Strikingly, -syn binds to arachidonic acid, which blocks arachidonic acid-induced SNARE interactions, both in vitro and in vivo 54. One possibility is that in -syn /EV/SUL worms, soluble, monomeric-syn binds and sequesters most of the arachidonic acid molecules, thereby blocking SNARE- mediated exocytosis. In contrast, in -syn/psd-1/SULworms, aggregated -syn fails to bind to arachidonic acid; consequently, DA neuron loss is partially rescued compared to -syn

/EV/SUL worms (Fig. 2.6 C).

In summary, treating -syn expressing worms with MFX, CsA, or SUL caused a 200–

400% increase in the number of animals with the normal complement of dopaminergic neurons after 7 days of molecule exposure.We hypothesize that neuroprotection respectively observed for each drug comes from MFX scavenging radicals or converting to PC, CsA inhibiting the mPTP, or SUL inhibiting CYP2J2 and sub-family members.Whether these drug candidates, singly or in combination, protect against-syn-associated pathology in PD patient-derived iPSCs will be the subject of future investigations.

Materials and methods

High-throughput screen

The Prestwick library, containing 1121 number chemically and therapeutically diverse drugs, was screened for compounds that enhanced the growth of psd1Δ yeast cells with or without - syn. These cells were pre-grown in SC-Sucrose-URA (2% w/v) dropout media for approximately 24 h at 30°C with shaking. Cells were pelleted (4000 x g), washed, and resuspended in SC-URA galactose media to yield OD 600 nm of 0.4. 135 L aliquots of yeast cells were pipetted robotically into each well of the 96 well plates. Library compounds diluted in SCURA galactose (15 L) were added to each well robotically to yield a final drug

47 concentration of 5 M. Plates were incubated with gentle shaking for 20-24 h at 30°C in a humidified incubator. Quantitative changes in growth were assayed via optical density changes, whereby psd1Δ cells +EV with DMSO (1% v/v), a negative control, would typically have an OD600 nm reading of approximately 0.45, and psd1Δ cells + -syn treated with ethanolamine (5 mM), the positive control, would have an OD600 nm reading of 1.3. The two strongest hit compounds, SUL and MFX, rescued the growth with average z-scores of

17.8 (84% of positive control) and 11.8 (49% of positive control). CsA, a weaker hit, rescued growth with an average z-score of 2.6 (15% of positive control). Compounds were tested in duplicates. The z-factor 55 for the screen was 0.896.

RNA interference (RNAi)

psd-1 and crls-1 RNAi feeding clones were purchased from Geneservice. Bacteria containing these plasmids were isolated and grown overnight in LB media with 100μg/ml ampicillin. Nematode growth media plates containing 1μM IPTG were seeded with RNAi feeding clones and allowed to dry. L4 staged hermaphrodites were transferred to corresponding RNAi plates and allowed to lay eggs overnight to synchronize the F1 progeny.

The dopaminergic neurons in the F1 progeny of the RNAi-treated worms were analyzed for neurodegeneration at day 7 following incubation at 20°C. C. elegans strain UA196 [sid-

1(pk3321); Pdat-1::-syn, Pdat-1::GFP;Pdat-1::sid-1, Pmyo-2::mCherry] that expresses -syn,

GFP, and SID-1 in the dopaminergic neurons and is susceptible to RNAi selectively in dopaminergic neurons was used in this study 28.

Dopaminergic neurodegeneration analyses in C. elegans

C. elegans dopaminergic neurons were analyzed for degeneration as previously described

56. Strain UA196 was treated with psd-1, crls-1 or EV dsRNA. Nematodes were

48 synchronized, grown at 20°C, and analyzed at day 7 of development for -syn-induced dopaminergic neurodegeneration. On the day of analysis, the six anterior dopaminergic neurons were examined in 30 adult hermaphrodite worms, in triplicate. These worms were immobilized on glass coverslips using 3 mM levamisole and transferred onto 2% agarose pads on microscope slides. The analysis was carried out using a Nikon E800 with an Endow

GFP filter cube (Chroma). Worms were considered normal when all six anterior dopaminergic neurons were present without any signs of degeneration, as previously reported

56. In total, at least 90 adult worms were analyzed for each RNAi treatment (30 worms/trial; a total of 3 trials). Statistical analyses were performed using two-way ANOVA and a Tukey's or Sikdak's post hoc analysis and are means ± standard deviation (p < 0.05) using GraphPad

Prism (version 6).

Pharmacological treatment of C. elegans

SUL (Sigma-Aldrich), choline chloride (Avantor), MFX (Sigma-Aldrich) and CsA

(Sigma-Aldrich) were dissolved in corresponding solutions as DMSO or water and then added to preautoclaved media, with the volume of compound solution taken into account.

SUL, MFX and CsA were tested in C. elegans over a range of concentrations as described in the Results and Figure Legends (0, 1, 2, 3.75, 7.5, 15, 30, 40 μM with 0.1% DMSO in the media). C. elegans were exposed to a lower concentration range of choline chloride (0, 1.25,

2.5, 5, 10 μM) for consistency with the yeast experiments. All 35 mm worm plates were seeded with 300μL concentrated HT115 E. coli. Unless mentioned in the results section, C. elegans were exposed to drugs from hatching through day 7 of adulthood and analyzed for dopaminergic neurodegeneration.

C. elegans semi-quantitative RT-PCR analysis

Sixty worms of the strain UA196 [sid-1(pk3321); Pdat-1::-syn, Pdat-1::GFP; Pdat-1::sid-1,

Pmyo-2::mCherry]were grown to day 7 at 20°C. They were harvested from drug treatments or solvent control plates (30μM MFX/0.1% DMSO; 30μM CsA/0.1% DMSO; 30μM SUL/0.1%

DMSO;10μM CC/ddH2O). Worms were transferred to fresh drug/RNAi plates as needed to avoid starvation. Total RNAs were extracted from control and drug-treated worms as described previously 57. Total RNA was quantitated using a Nanodrop and 1μg of total RNA of each sample was used to synthesize first-strand cDNA using MMLV-RnaseH (-) transcriptase (Promega). Using the cDNA as templates, PCR reactions were conducted with

GoTaq polymerase (Promega) at 59°C annealing temperature. Primer sequences are as follows:

-syn forward primer: GGATGTATTCATGAAGGACTTTCAAAG

-syn reverse primer: GGCTTCAGGTTCGTAGTCTTG ama-1 forward primer: CGAGTCCAACGTACTCTCC ama-1 reverse primer: GATGTTGGAGAGTACTGAGC

PCR products were loaded onto a Gel Red (Sigma) stained 0.8% agarose gel. An image was captured by FujiFilm LAS 4000. Band intensities were compared by digital imaging using MetaMorph software. Fluorescent band intensities were normalized using the following equation: -syn/ama-1.

References

1. Moore, D. J., West, A. B., Dawson, V. L. & Dawson, T. M. Molecular pathophysiology of Parkinson's disease. Annual review of neuroscience 28, 57-87, doi:10.1146/annurev.neuro.28.061604.135718 (2005).

2. Spillantini, M. G. et al. Alpha-synuclein in Lewy bodies. Nature 388, 839-840, doi:10.1038/42166 (1997).

3. Kara, E. et al. Alpha-Synuclein mutations cluster around a putative protein loop. Neuroscience letters 546, 67-70, doi:10.1016/j.neulet.2013.04.058 (2013).

4. Singleton, A. B. et al. Alpha-Synuclein locus triplication causes Parkinson's disease. Science 302, 841, doi:10.1126/science.1090278 (2003).

5. Sharon, R. et al. Alpha-Synuclein occurs in lipid-rich high molecular weight complexes, binds fatty acids, and shows homology to the fatty acid-binding proteins. Proc Natl Acad Sci U S A 98, 9110-9115, doi:10.1073/pnas.171300598 (2001).

6. Varkey, J. et al. Alpha-Synuclein oligomers with broken helical conformation form lipoprotein nanoparticles. J Biol Chem 288, 17620-17630, doi:10.1074/jbc.M113.476697 (2013).

7. Eichmann, C. et al. Preparation and Characterization of Stable alpha-Synuclein Lipoprotein Particles. J Biol Chem 291, 8516-8527, doi:10.1074/jbc.M115.707968 (2016).

8. Burre, J., Sharma, M. & Sudhof, T. C. Alpha-Synuclein assembles into higher-order multimers upon membrane binding to promote SNARE complex formation. Proc Natl Acad Sci U S A 111, E4274-4283, doi:10.1073/pnas.1416598111 (2014).

9. Weinreb, P. H., Zhen, W., Poon, A. W., Conway, K. A. & Lansbury, P. T., Jr. NACP, a protein implicated in Alzheimer's disease and learning, is natively unfolded. Biochemistry 35, 13709-13715, doi:10.1021/bi961799n (1996).

10. Chandra, S., Chen, X., Rizo, J., Jahn, R. & Sudhof, T. C. A broken alpha -helix in folded alpha -Synuclein. J Biol Chem 278, 15313-15318, doi:10.1074/jbc.M213128200 (2003).

11. Volles, M. J. et al. Vesicle permeabilization by protofibrillar alpha-synuclein: implications for the pathogenesis and treatment of Parkinson's disease. Biochemistry 40, 7812-7819 (2001).

12. Luk, K. C. et al. Pathological alpha-synuclein transmission initiates Parkinson-like neurodegeneration in nontransgenic mice. Science 338, 949-953, doi:10.1126/science.1227157 (2012).

13. Riekkinen, P., Rinne, U. K., Pelliniemi, T. T. & Sonninen, V. Interaction between dopamine and phospholipids. Studies of the substantia nigra in Parkinson disease patients. Archives of neurology 32, 25-27 (1975).

14. Manyam, B. V., Ferraro, T. N. & Hare, T. A. Cerebrospinal fluid amino compounds in Parkinson's disease. Alterations due to carbidopa/levodopa. Archives of neurology 45, 48-50 (1988).

15. Ross, B. M., Mamalias, N., Moszczynska, A., Rajput, A. H. & Kish, S. J. Elevated activity of phospholipid biosynthetic enzymes in substantia nigra of patients with Parkinson's disease. Neuroscience 102, 899-904 (2001).

16. Hattingen, E. et al. Phosphorus and proton magnetic resonance spectroscopy demonstrates mitochondrial dysfunction in early and advanced Parkinson's disease. Brain 132, 3285-3297, doi:10.1093/brain/awp293 (2009).

17. Rappley, I. et al. Lipidomic profiling in mouse brain reveals differences between ages and genders, with smaller changes associated with alpha-synuclein genotype. J Neurochem 111, 15-25, doi:10.1111/j.1471-4159.2009.06290.x (2009).

18. Vance, J. E. & Tasseva, G. Formation and function of phosphatidylserine and phosphatidylethanolamine in mammalian cells. Biochim Biophys Acta 1831, 543-554, doi:10.1016/j.bbalip.2012.08.016 (2013).

19. Patil, V. A., Fox, J. L., Gohil, V. M., Winge, D. R. & Greenberg, M. L. Loss of cardiolipin leads to perturbation of mitochondrial and cellular iron homeostasis. J Biol Chem 288, 1696-1705, doi:10.1074/jbc.M112.428938 (2013).

20. Schuiki, I. & Daum, G. Phosphatidylserine decarboxylases, key enzymes of lipid metabolism. IUBMB life 61, 151-162, doi:10.1002/iub.159 (2009).

21. Kainu, V., Hermansson, M., Hanninen, S., Hokynar, K. & Somerharju, P. Import of phosphatidylserine to and export of phosphatidylethanolamine molecular species from mitochondria. Biochim Biophys Acta 1831, 429-437, doi:10.1016/j.bbalip.2012.11.003 (2013).

22. Baker, C. D., Basu Ball, W., Pryce, E. N. & Gohil, V. M. Specific requirements of nonbilayer phospholipids in mitochondrial respiratory chain function and formation. Molecular biology of the cell 27, 2161-2171, doi:10.1091/mbc.E15-12-0865 (2016).

23. Gibellini, F. & Smith, T. K. The Kennedy pathway--De novo synthesis of phosphatidylethanolamine and phosphatidylcholine. IUBMB life 62, 414-428, doi:10.1002/iub.337 (2010).

24. Wang, S. et al. Phosphatidylethanolamine deficiency disrupts alpha-synuclein homeostasis in yeast and worm models of Parkinson disease. Proc Natl Acad Sci U S A 111, E3976-3985, doi:10.1073/pnas.1411694111 (2014).

25. Wang, S. et al. Phosphatidylethanolamine deficiency disrupts α-synuclein homeostasis in yeast and worm models of Parkinson disease. Proceedings of the National Academy of Sciences 111, E3976-E3985, doi:10.1073/pnas.1411694111 (2014).

26. Cao, S., Gelwix, C. C., Caldwell, K. A. & Caldwell, G. A. Torsin-mediated protection from cellular stress in the dopaminergic neurons of Caenorhabditis elegans. The Journal of neuroscience : the official journal of the Society for Neuroscience 25, 3801-3812, doi:10.1523/jneurosci.5157-04.2005 (2005).

27. Hamamichi, S. et al. Hypothesis-based RNAi screening identifies neuroprotective genes in a Parkinson's disease model. Proc Natl Acad Sci U S A 105, 728-733, doi:10.1073/pnas.0711018105 (2008).

28. Harrington, A. J., Yacoubian, T. A., Slone, S. R., Caldwell, K. A. & Caldwell, G. A. Functional analysis of VPS41-mediated neuroprotection in Caenorhabditis elegans and mammalian models of Parkinson's disease. The Journal of neuroscience : the official journal of the Society for Neuroscience 32, 2142-2153, doi:10.1523/jneurosci.2606-11.2012 (2012).

29. Holthuis, J. C. & Menon, A. K. Lipid landscapes and pipelines in membrane homeostasis. Nature 510, 48-57, doi:10.1038/nature13474 (2014).

30. Gohil, V. M., Thompson, M. N. & Greenberg, M. L. Synthetic lethal interaction of the mitochondrial phosphatidylethanolamine and cardiolipin biosynthetic pathways in Saccharomyces cerevisiae. J Biol Chem 280, 35410-35416, doi:10.1074/jbc.M505478200 (2005).

31. Zs-Nagy, I. A survey of the available data on a new nootropic drug, BCE-001. Annals of the New York Academy of Sciences 717, 102-114 (1994).

32. Hansson, M. J. et al. Powerful cyclosporin inhibition of calcium-induced permeability transition in brain mitochondria. Brain research 960, 99-111 (2003).

33. Hansson, M. J. et al. Brain-derived respiring mitochondria exhibit homogeneous, complete and cyclosporin-sensitive permeability transition. J Neurochem 89, 715-729, doi:10.1111/j.1471-4159.2004.02400.x (2004).

34. Nehru, B., Verma, R., Khanna, P. & Sharma, S. K. Behavioral alterations in rotenone model of Parkinson's disease: attenuation by co-treatment of centrophenoxine. Brain research 1201, 122-127, doi:10.1016/j.brainres.2008.01.074 (2008).

35. Verma, R. & Nehru, B. Effect of centrophenoxine against rotenone-induced oxidative stress in an animal model of Parkinson's disease. Neurochem Int 55, 369-375, doi:10.1016/j.neuint.2009.04.001 (2009).

36. Parthasarathy, S., Cady, R. K., Kraushaar, D. S., Sladek, N. E. & Baumann, W. J. Inhibition of diacylglycerol:CDPcholine cholinephosphotransferase activity by dimethylaminoethyl p-chlorophenoxyacetate. Lipids 13, 161-164 (1978).

37. Parthasarathy, S., El-Rahman, A. & Baumann, W. J. Modulation of phosphatidylcholine synthesis in vitro. Inhibition of diacylglycerol cholinephosphotransferase and lysophosphatidylcholine acyltransferase by centrophenoxine and neophenoxine. Biochim Biophys Acta 665, 345-351 (1981).

38. Semsei, I. & Zs-Nagy, I. Superoxide radical scavenging ability of centrophenoxine and its salt dependence in vitro. Journal of free radicals in biology & medicine 1, 403-408 (1985).

39. Malanga, G., Aguiar, M. B., Martinez, H. D. & Puntarulo, S. New insights on dimethylaminoethanol (DMAE) features as a free radical scavenger. Drug metabolism letters 6, 54-59 (2012).

40. Liu, J. et al. Calcineurin is a common target of cyclophilin-cyclosporin A and FKBP- FK506 complexes. Cell 66, 807-815 (1991).

41. Halestrap, A. P. & Davidson, A. M. Inhibition of Ca2(+)-induced large-amplitude swelling of liver and heart mitochondria by cyclosporin is probably caused by the inhibitor binding to mitochondrial-matrix peptidyl-prolyl cis-trans isomerase and preventing it interacting with the adenine nucleotide translocase. The Biochemical journal 268, 153-160 (1990).

42. Penzo, D. et al. Arachidonic acid released by phospholipase A(2) activation triggers Ca(2+)-dependent apoptosis through the mitochondrial pathway. J Biol Chem 279, 25219-25225, doi:10.1074/jbc.M310381200 (2004).

43. Arora, A. S., Jones, B. J., Patel, T. C., Bronk, S. F. & Gores, G. J. Ceramide induces hepatocyte cell death through disruption of mitochondrial function in the rat. Hepatology (Baltimore, Md.) 25, 958-963, doi:10.1002/hep.510250428 (1997).

44. Halestrap, A. P. & Richardson, A. P. The mitochondrial permeability transition: a current perspective on its identity and role in ischaemia/reperfusion injury. Journal of molecular and cellular cardiology 78, 129-141, doi:10.1016/j.yjmcc.2014.08.018 (2015).

45. Martin, L. J., Semenkow, S., Hanaford, A. & Wong, M. Mitochondrial permeability transition pore regulates Parkinson's disease development in mutant alpha-synuclein transgenic mice. Neurobiology of aging 35, 1132-1152, doi:10.1016/j.neurobiolaging.2013.11.008 (2014).

46. Warne, J. et al. Selective Inhibition of the Mitochondrial Permeability Transition Pore Protects against Neurodegeneration in Experimental Multiple Sclerosis. J Biol Chem 291, 4356-4373, doi:10.1074/jbc.M115.700385 (2016).

47. Flower, T. R., Chesnokova, L. S., Froelich, C. A., Dixon, C. & Witt, S. N. Heat shock prevents alpha-synuclein-induced apoptosis in a yeast model of Parkinson's disease. Journal of molecular biology 351, 1081-1100, doi:10.1016/j.jmb.2005.06.060 (2005).

48. Smith, W. W. et al. Endoplasmic reticulum stress and mitochondrial cell death pathways mediate A53T mutant alpha-synuclein-induced toxicity. Hum Mol Genet 14, 3801-3811, doi:10.1093/hmg/ddi396 (2005).

49. Bir, A. et al. Alpha-Synuclein-induced mitochondrial dysfunction in isolated preparation and intact cells: implications in the pathogenesis of Parkinson's disease. J Neurochem 131, 868-877, doi:10.1111/jnc.12966 (2014).

50. Petrosillo, G., Ruggiero, F. M., Pistolese, M. & Paradies, G. Ca2+-induced reactive oxygen species production promotes cytochrome c release from rat liver mitochondria via mitochondrial permeability transition (MPT)-dependent and MPT-independent mechanisms: role of cardiolipin. J Biol Chem 279, 53103-53108, doi:10.1074/jbc.M407500200 (2004).

51. Granville, D. J. et al. Reduction of ischemia and reperfusion-induced myocardial damage by cytochrome P450 inhibitors. Proc Natl Acad Sci U S A 101, 1321-1326, doi:10.1073/pnas.0308185100 (2004).

52. Booth Depaz, I. M., Toselli, F., Wilce, P. A. & Gillam, E. M. Differential expression of cytochrome P450 enzymes from the CYP2C subfamily in the human brain. Drug metabolism and disposition: the biological fate of chemicals 43, 353-357, doi:10.1124/dmd.114.061242 (2015).

53. Chang, Q. et al. Cytochrome P450 2C epoxygenases mediate photochemical stress- induced death of photoreceptors. J Biol Chem 289, 8337-8352, doi:10.1074/jbc.M113.507152 (2014).

54. Darios, F. et al. Alpha-synuclein sequesters arachidonic acid to modulate SNARE- mediated exocytosis. EMBO Rep 11, 528-533, doi:10.1038/embor.2010.66 (2010).

55. Zhang, J. H., Chung, T. D. & Oldenburg, K. R. A Simple Statistical Parameter for Use in Evaluation and Validation of High Throughput Screening Assays. Journal of biomolecular screening 4, 67-73, doi:10.1177/108705719900400206 (1999).

56. Berkowitz, L. A. et al. Application of a C. elegans dopamine neuron degeneration assay for the validation of potential Parkinson's disease genes. Journal of visualized experiments : JoVE, doi:10.3791/835 (2008).

57. Ray, A., Zhang, S., Rentas, C., Caldwell, K. A. & Caldwell, G. A. RTCB-1 mediates neuroprotection via XBP-1 mRNA splicing in the unfolded protein response pathway. The Journal of neuroscience : the official journal of the Society for Neuroscience 34, 16076-16085, doi:10.1523/jneurosci.1945-14.2014 (2014).

CHAPTER THREE

MICRORNA-MEDIATED PROTECTION AGAINST -SYNUCLEIN-INDUCED DOPAMINERGIC NEURODEGENERATION IN C. ELEGANS

This work was performed by Chuan Xu, Siyuan Zhang, G. Blake Parker, Samantha A.

Glukhova, Paige M. Dexter, Susan M. DeLeon, Shusei Hamamichi, Laura A. Berkowitz,

Kim A. Caldwell and Guy A. Caldwell. Chuan Xu, Siyuan Zhang, G. Blake Parker, Samantha

A. Glukhova, Paige M. Dexter, Susan M. DeLeon, Shusei Hamamichi collected all the data.

Guy A. Caldwell and Kim A. Caldwell contributed to the editing.

Abstract

Current strategies to manage the devastating consequences of neurodegenerative diseases, like Parkinson’s disease (PD), are largely limited to symptomatic control and successful identification of more mechanistic interventions remains constrained by the range of potential targets available for therapeutic development. Epigenetic regulation of gene expression via the activity of microRNAs presents an unexplored strategy to more rapidly identify collections of putative effectors of neurodegeneration in vivo. Here we demonstrate, for the first time, that mutation of mir-239 induced a robust resistance to α-synuclein-induced neurodegeneration in transgenic Caenorhabditis elegans. We functionally investigated a suite of expression-validated targets of mir-239 regulation via conditional knockdown using a dopaminergic neuron-specific RNAi-sensitive strain of C. elegans. Depletion of a subset of these gene products reduced the neuroprotection imparted by the mir-239 mutant, thereby indicating these genes might act as downstream targets contributing to neuroprotection.

Furthermore, assaying for oxidative stress resistance revealed that mir-239 mutant animals exhibit enhanced resistance to stress. We further show that mutation of mir-239 promotes longevity, independent of the presence of -synuclein. The intersection of epigenetics, longevity and neurodegeneration exemplified by mir-239 highlights a strategy whereby progress toward therapeutic intervention is accelerated through functional investigation of miRNA-regulated genes to identify novel neuroprotective modulators.

Introduction

Parkinson’s disease (PD) is the second most common neurodegenerative disease with an extensive societal and economic burden. Idiopathic PD accounts for >90% of cases. Here, it is tempting to speculate that some of these cases could result from an interaction of environmental and/or complex, multigenic inherited factors. microRNAs (miRNAs), which

57 are short, noncoding RNA molecules, often negatively regulate gene expression by binding to the 3’UTR of targeted genes 1. This action represses protein translation, resulting in the coordinated control of a broad range of physiological processes. Progress toward halting neurodegeneration in PD is limited by the availability of functionally characterized gene products that represent new, potentially disease-modifying, effectors. The suites of genes epigenetically co-regulated by select miRNAs represent a largely uncharacterized source of potential therapeutic targets whereby phenotypic analysis to discern modifiers of dopamine

(DA) neuron survival can expedite the discovery process.

C. elegans is especially well suited for the identification of neuroprotective targets derived through miRNA target analysis. The ~3-week lifespan of this nematode accelerates studies of time-dependent neurodegeneration. In conjunction with conserved pathways for aging and neurodegenerative mechanisms, worm models have yielded both chemical and genetic modifiers that exhibit neuroprotective activities which translate to both mammalian and human datasets including rat primary neuron cultures, induced-pluripotent stem cells and

GWAS results2, 3,4,. Most usefully, an extensive collection of miRNA knockout strains, the vast majority of which are non-essential for organismal survival, has been generated to facilitate rapid genetic analysis of individual miRNAs 5. In this regard, based on published datasets where previously characterized targets of miRNAs were demonstrated to function as modulators of cytotoxicity in PD models, we selected four miRNAs that we hypothesized would have a potential modifying effect, when mutated, on -synuclein-induced dopaminergic neurodegeneration in C. elegans. Indeed, deletion of mir-239 significantly enhanced dopaminergic neuron survival, suggesting it normally suppresses the expression of one or more neuroprotective genes. Notably, this miRNA was originally prioritized for our analysis because it has a temporal association with aging whereby C. elegans lifespan was

58 promoted when mir-239 was mutant, but was shortened in transgenic worms with mir-239 overexpression 6.

In this study, we explore the functional impact of distinct epigenetically-regulated targets of mir-239 with an eye on their selective activity in modulating dopaminergic neurodegeneration in response to the proteostatic burden imposed by -syn overabundance in vivo. We focused our analysis of targets to those that were experimentally confirmed to be regulated by a distinct isoform of mir-239. Therefore, we systematically examined 26 expression-validated mir-239b targets within the context of -syn-induced dopaminergic neurodegeneration and determined that nine specific gene targets coordinated through mir-

239b function modulated -syn neurotoxicity.

Results mir-239 mutation promotes resistance to α-Syn-neurotoxicity

To examine putative epigenetically-regulated modifiers of -syn-induced dopaminergic neurodegeneration in C. elegans, we selected four miRNAs (mir-239, mir-251, mir-360, mir-

797) for which among their predicted targets were included genes that we previously reported to be neuroprotective in C. elegans and, in many cases, mammalian models. As examples, these gene targets include vps-41, a lysosomal trafficking component which contains predicted binding sites for both mir-239 and mir-251 7,8,9, rab-8, a Rab protein in the Ras

GTPase superfamily with a binding site for mir-360 9, and ced-10, a GTPase that is orthologous to RAC1, that contains a mir-797 binding site 10.

To determine if these four miRNAs would modulate -syn-induced dopaminergic neurotoxicity, transgenic C. elegans expressing -syn specifically in the DA neurons under a

DA transporter (dat-1) promoter, with GFP co-expressed separately from this promoter, were

11 used in this analysis [strain UA44(baIn11[Pdat-1::α-syn, Pdat-1::GFP])] . These worms were genetically crossed into the miRNA mutants to create homozygous -syn; mir-239, -syn; mir-251, -syn; mir-360, and -syn; mir-797 lines. Using established criteria 11,12, these resulting animals were scored for DA neurodegeneration at day 7 of lifespan. In one experiment, -syn; mir-251, -syn; mir-360, and -syn; mir-797 were analyzed (Fig. 3.1 A) while in a separate experiment (Fig. 3.1 B), -syn; mir-239 animals were analyzed; these analyses are displayed independently for this reason. As shown in Fig. 3.1 A, the mir-360 mutant did not alter DA neurodegeneration compared to the -syn control while there was significantly enhanced neurodegeneration with mir-251 (P <0.05) and neuroprotection from mir-797 (P <0.05). Strikingly, however, the mir-239 mutation exerted an unexpectedly strong rescue against -syn neurotoxicity, where ~98% of mir-239 mutant worms displayed intact

DA neurons compared to ~23% in -syn control worms (P <0.05) (Fig. 3.1 B).

Representative images display the neuroprotection conferred in a mir-239 mutant background

(Fig. 3.1 D).

Figure 3.1. mir-239 mutation enhances resistance to -syn induced neurotoxicity in C. elegans. (A, B) Bar graph depicts the percentage of worms with normal DA neurons at day 7 post-hatching in-syn worms (Strain UA44) and -syn worms in three different miRNAs mutant genetic background including -syn; mir-251, -syn; mir-360 and -syn; mir-797. (B) Bar graph depicts the percentage of worms with normal DA neurons at day 7 post- hatching in-syn worms and-syn worms with mir-239 mutation. (C, D) The images represent degenerative phenotypes of six DA neurons in the anterior region of -syn worms (Strain UA44) (C) and of wild type DA neurons in -syn; mir-239 mutant worms (D). Worms shown in the representative images are day 7 after hatching. GFP is driven under the dat-1 promoter to illuminate six intact DA neurons in the head. The six head DA neurons include two pairs of cephalic neurons (CEPs, the four anterior neurons) and one pair of anterior deirid neurons (ADEs, posterior to CEPs). Arrowheads indicate intact head DA neurons and arrows point to a missing cell body and/or process (C, D). The UA44 worm strain expresses Pdat-1::-syn; Pdat-1:: GFP. Data are represented as mean ± SD (N=3), 30 worms in triplicate (A, B). Data were analyzed using one-way ANOVA with a Dunnett post hoc test; *P <0.05 (A) or Student’s t-test; ****P < 0.0001 (B).

mir-239 mutant lifespan extension

It has been previously reported that mir-239 has a temporal association with aging in C. elegans; lifespan was extended when mir-239 was mutant and shortened in transgenic worms

61 with mir-239 overexpression 6. To confirm the observation of mir-239 regulating lifespan, a survival curve was generated for mir-239 mutant animals to measure viability over the course of aging. The result showed a consistent rightward shift of the survival curve in mir-239 mutant worms compared to WT (strain N2) controls (Fig. 3.2 A). While the difference in median lifespan between these two strains was about one day (N2 worms, 14 days; mir-239 worms,15 days), it was significant (P <0.005). We also evaluated the mir-239 background in the transgenic worms expressing human -syn in the DA neurons. Survival was recorded in both -syn control and -syn; mir-239 mutant worms. Here, we found that -syn control worms had a median lifespan of 13 days, whereas this was 14 days in -syn; mir-239 animals

(P <0.05) (Fig. 3.2 B). Considering the slightly delayed rate of aging caused by mir-239 mutant, we questioned whether mir-239-mediated neuroprotection at day 7 (chronological aging) might be a result of C. elegans developing somewhat slower than control animals.

Since they appeared to experience a one-day developmental lag, we further examined these animals at day 8 and found that mir-239 mutants also exhibited robust protection at this equivalent stage of biological aging. Thus, the extent of neurodegeneration in -syn worms harboring the mir-239 mutation was consistently attenuated (Fig. 3.2 C). These findings demonstrated that loss of mir-239 modified neuroprotection with respect to both chronological and biological aging.

Figure 3.2. mir-239 mutation extends lifespan and rescues -syn-induced neurodegeneration. (A) Survival curves for N2 worms and mir-239 mutant worms. (B) Survival curves for -syn worms (Strain UA44) and -syn worms with mir-239 mutation. (A, B) Survival was evaluated using a Mantel-Cox test. **P <0.005 (A); *P <0.05 (B). (C) Bar graph depicts the percentage of worms with normal DA neurons at day 7 or 8 post-hatching in -syn worms (Strain UA44) or -syn worms with mir-239 mutation. The UA44 worm strain expresses Pdat- 1::-syn; Pdat-1::GFP. Data are represented as mean ± SD (N=3), 30 worms in triplicate. Data were analyzed using two-way ANOVA with a Sidak post hoc test; ****P <0.0001. (D) Survival curves for -syn worms (Strain UA196) and -syn worms with mir-239 mutation. UA196 represents worm strain expressing [sid-1(pk3321); Pdat-1::sid-1, Pmyo-2::mCherry; Pdat- 1::GFP, Pdat-1::-syn]. Survival was evaluated using a Mantel-Cox test; *P <0.05.

Identification of neuroprotective mir-239b gene targets

To further investigate the neuroprotective nature of mir-239 deletion, we were curious to know if there were individual mir-239 targets that mediated a significant neuromodulating

63 effect in the -syn background. Out of the 624 predicted targets of mir-239b (listed as miR-

239b-5p; microRNA.org), 26 have been validated through manually curated experiments such as reporter genes, qPCR, Western blotting, microarrays/RNA-seq, biotin pull-down

[TarBase (v7.0)]13. Thus, we decided to examine these 26 validated targets and discern which were associated with the robust neuroprotective activity observed with the mir-239 mutation expressing -syn in DA neurons. The predicted function and human orthologs for these target genes are summarized in Table 3.1, where they are categorized into six functional groups.

Table 3.1 Predicted function of mir-239b gene targets and corresponding human orthologs.

C. elegans Gene product name, proposed function, Human E value gene ID or characteristics Ortholog Lipid storage-related genes F59E11.5 lipid storage; affected by rotenone and D- N/A N/A glucose hosl-1 hormone-sensitive lipase LIPE 7e-65 Reproduction and/or embryonic genes F25H9.6 enzyme in pathway that converts vitamin B5 Ppcdc 2e-40 to coenzyme A gtbp-1 G3BP stress granule assembly factor G3BP 1/2 2e-05 oaz-1 ornithine decarboxylase antizyme 1 OAZ1 0.001 ZK688.5 affected by rotenone, D-glucose, and N/A N/A resveratrol pkg-2 protein kinase, cGMP-dependent, type II PRKG1/2 e-128 Mediated by steroid hormones T27D12.1 solute carrier family 17 member 5 SLC17A5 8e-32 F57G12.1 regulated by 1-methylnicotinamide, N/A N/A progesterone, & testosterone F46G10.2 affected by rotenone and progesterone N/A N/A F21F3.6 regulated by progesterone, resveratrol, N/A N/A atrazine, and allantoin T24B8.3 regulated by 1-methylnicotinamide, N/A N/A progesterone, and quercetin Ion binding or transport genes B0393.5 latent transforming growth factor beta LTBP-1 4e-29 binding protein 1 hrg-3 exhibits heme binding activity N/A N/A

64 vdac-1 mitochondrial voltage dependent anion VDAC1 2e-57 channel 1 nra-3 affected by paraquat, quercetin, and N/A N/A progesterone Vulva development or muscle function genes R09A8.5 uncharacterized; expressed in muscle cells N/A N/A F45D3.2 uncharacterized; expressed in muscle cells N/A N/A R02F2.1 cyclin I CCNI 7e-11 B0336.3 RNA binding motif protein 27 RBM27 4e-18 C14C10.5 proteasome activator subunit 4 PSME4 0.0 Developmental genes T10G3.1 affected by levamisole and allantoin N/A N/A F27D4.2(lsy-22) exhibits repressing transcription factor N/A N/A binding activity K08F8.1(mak-1) mitogen-activated protein kinase-activated MAPKAPK2 2e-83 protein kinase 2/3 /3 H18N23.2 protein phosphatase 1 regulatory subunit 3B PPP1R3B/3C 2e-28 K08E3.5 UDP-glucose pyrophosphorylase 2 UGP2 e-179 N/A: Not applicable.

We utilized an -syn-expressing nematode strain (UA196) that exhibits sensitivity for

RNAi silencing selectively in the DA neurons (sid-1(pk3321); baIn33 [Pdat-1::sid-1, Pmyo-

8 2::mCherry]; baIn11[Pdat-1::-syn, Pdat-1::GFP]) . We genetically crossed the mir-239 mutation into this RNAi strain and analyzed -syn-induced neurodegeneration at day 7. In our experience, the baseline neurodegeneration observed in the -syn RNAi-sensitive strain is consistently lower than in the N2 -syn-expressing transgenic line. As displayed in Fig. 3.3

A, the control RNAi -syn transgenic worms displayed 30% normal DA neurons at day 7

(compared to the N2 -syn transgenic worms which had 23% normal neurons at day 7; Fig.

3.1 B). Once the mir-239 mutation was crossed into the -syn RNAi strain, neuroprotection was increased, with 60% of the population with normal neurons (Fig. 3.3 A); however, this mutation was not as protective as it was in the N2 wild-type background where 98% of the population displayed wildtype neurons (Fig. 3.3 A vs. 3.1 B).

We proceeded to examine the 26 mir-239b targets by RNAi knockdown of each of the

65 candidate genes, in triplicate, using the C. elegans DA neuron-specific -syn RNAi strain with the mir-239 mutation in the genetic background. The candidates were categorized to six broad functional groups (Table 3.1) that allowed us to knockdown a manageable number during each round of analysis, each with their own empty vector (EV) control, for experimental consistency and statistical considerations. We discerned that RNAi depletion of nine of these targeted genes in DA neurons significantly decreased the neuroprotective capacity of the mir-239 mutation (Fig. 3.3 B-3.3 G). To clarify whether these genes required mir-239b for resistance against -syn neurotoxicity, we exposed -syn control worms, that did not harbor the mir-239 mutation, to dsRNA specifically targeting each of the nine candidate genes. Notably, in the -syn control RNAi worms, there was no significant difference between knockdown of any gene or the EV control (Fig. 3.3 H). Thus, there is dependence on mir-239b for these nine genes in terms of their neuroprotective role against - syn-induced neurodegeneration. Furthermore, we examined the lifespan in both -syn RNAi sensitive worms and -syn; mir-239 mutant worms. We found the median lifespan of mir-

239 mutant worms were 16 days, while -syn control worms were 13 days. (Fig. 3.2 D) The result indicated a significant lifespan extension (P < 0.05), further confirmed that mir-239 mutation extends lifespan in -syn worms.

Figure 3.3.Examination of 26 validated target genes of mir-239b via RNAi knockdown in the dopaminergic-neuron sensitive RNAi -syn strain with and without mir-239 mutation. (A) Bar graph depicts the percentage of worms with normal DA neurons at day 7 after hatching in -syn worms and -syn worms with mir-239 mutation. (B-G) Bar graph depicts the percentage of worms with normal DA neurons at day 7 after hatching in -syn worms with mir-239 mutation under corresponding dsRNAs treatments of lipid storage-related genes (B), reproduction and/or embryonic genes (C), mediated by steroid hormones (D), ion binding or transport genes (E), vulva development or muscle function genes (F) and developmental genes (G). (H) Depletion of the nine targeted genes determined to confer neuroprotection by mir-239 mutation in using dopaminergic neuron-sensitive RNAi worm strain (UA196) with WT mir-239 retained. Bar graph depicts the percentage of worms with normal DA neurons at day 7 after hatching. UA196 represents worm strain expressing [sid- 1(pk3321); Pdat-1::sid-1, Pmyo-2::mCherry; Pdat-1::GFP, Pdat-1::-syn]. RNAi bacteria which do not express any knockdown-targeting clone are denoted “EV” for empty vector. Data are mean ± SD, (N=3), 30 worms in triplicate (A-H). Data were analyzed using Student’s t-test, **P <0.005 (A) or one-way ANOVA with a Dunnett post hoc test; *P <0.05 (B-H).

Aging-associated locomotive deficits in -syn-expressing worms were reduced by mir- 239 mutation

Aging is often associated with locomotor deficits 14-17. We revealed that mir-239 mutant worms with -syn expression have significant extended longevity compared with -syn control worms. (Fig. 3.2 B) Thus, we would like to know whether mir-239 mutation would help to rescue locomotive behavioral defects in -syn transgenic animals observed during aging process. Here we examined thrashing behavior by measuring body bends/min, which were defined as a change in direction of bending at the middle of worm body, between -syn

68 worms with or without mir-239 mutation every 5 days after day 4. We found that worms with or without mir-239 mutation all showed a decrease in thrashing behavior with increasing age, indicating the aging process was accompanied by locomotor defects (Fig. 3.4 A). However, when comparing with -syn control worms (strain UA196) to mir-239 mutant worms [strain

UA196; mir-239 (nDf62)], the mutant animals exhibit significantly increased body bends at days 4, 9 and 14, but not at day 19, thereby indicating that mir-239 mutation partly rescues thrashing behavior defects in -syn worms in an age-dependent manner. For example, at day

14, mir-239 mutant animals have ~80 body bends/min, whereas -syn control worms only have ~70 body bends/min. (Fig. 3.4 A)

mir-239 mutation enhances resistance to oxidative stress

Oxidative stress resistance is an established factor associated with extended longevity.

Several research studies have illustrated this in different model organisms, including

Drosophila melanogaster and C. elegans. For example, in D. melanogaster, a P-element insertion mutation conferring extended longevity was found to resist oxidative stress 18.

When mutated, daf-2, the C. elegans homolog of mammalian insulin growth factor receptor extended lifespan and resisted the oxidative stress 19. As our data revealed that mir-239 mutant worms expressing -syn extended lifespan, we asked that if mir-239 mutants have the ability of better resisting oxidative stress compared with -syn control worms. The thrashing data uncovered that mir-239 mutant worms rescued behavioral defects compared with -syn control worms. Thus, we wanted to discern whether or not mir-239 mutant animals would still exhibit a protective role in modulating thrashing behavior in the presence of additional oxidative stress. To examine this, we treated -syn worms, with or without mir-239 mutation, every five days after day 4 with 12 hours exposure of 100  sodium azide, and quantified body bends. We found that, during the aging process, mir-239 mutant worms and -syn 69 control worms all gradually decreased body bends with sodium azide treatment. However, mir-239 mutation worms [strain UA196; mir-239 (nDf62)] showed significantly increased body bends at days 4, 9, 14 and 19, indicating that mutation in this miRNA confers resistance to the oxidative stress and rescued the thrashing behavior defects compared with control worms. For example, at day 19, mir-239 worms have ~70 body bends/min, whereas control worms only have ~50 body bends/min (Fig. 3.4 B).

To further confirm the effect of mir-239 deletion on resistance to oxidative stress, we used paraquat to treat worms and monitored the occurrence of dead animals hourly for 6, as previously published 6,20. After treating wild-type (N2) control worms expressing -syn

(strain UA44) and worms expressing -syn with the mir-239 mutation [strain UA44; mir-

239(nDf62)], we found that mir-239 mutant worms survived longer than controls at day 7 post-hatching with 6 hours paraquat exposure. The median survival for -syn control worms was 3 hours, whereas mir-239 mutant worms exhibited a corresponding median survival of 4 hours, a statistically significant extension of survival hours (Fig. 3.4 C). We further examined this in the -syn-expressing dopaminergic neuron-specific RNAi strain (strain UA196) both with or without mir-239 mutation. Consistent with the wild-type N2 genetic background, mir-

239 mutant worms in the conditional RNAi-sensitive background also exhibited significantly longer survival hours (Fig. 3.4 D). Thus, mir-239 mutation results in an increased resistance to oxidative stress caused by paraquat in animals expressing -syn.

Figure 3.4. mir-239 mutation partly rescues aging-associated locomotive deficits in -syn expressing worms. (A) mir-239 mutation partly rescued thrashing behavior defects in -syn worms (Strain UA196) in an age-dependent manner. (B) mir-239 mutation rescued thrashing behavior defects in -syn worms (Strain UA196) with sodium azide treatment. Treat -syn worms and -syn worms with mir-239 mutation with 12 hours exposure of 100 μM sodium azide every five days after day 4 post-hatching. (A, B) Thrashing frequency of the animals were quantified by counting body bends/min in both -syn worms and -syn; mir-239 mutant worms at day 4, 9,14, and 19. Data are mean ± SD, (N=3), 10 worms in triplicate. Data were analyzed using two-way ANOVA with a Sidak-post hoc test; *P value < 0.05. (C, D) mir-239 mutation increased resistance to oxidative stress caused by paraquat in animals expressing -syn. Treat -syn worms (Strain UA44 (C) or UA196 (D)) and -syn worms with mir-239 mutation with 6 hours exposure of 200 mM paraquat at day 7 post-hatching. The mir-239 mutation significantly increased survival hours in -syn worms. (C, D) Statistics were analyzed using Coherent survival analysis followed by Mantel-Cox test; **P value < 0.005; UA44 represents worm strain expressing Pdat-1::GFP, Pdat-1::-syn in the N2 background; UA196 represents worm strain [sid-1(pk3321); Pdat-1::sid-1, Pmyo-2::mCherry; Pdat-1::GFP, Pdat-1::-syn].

Discussion

Neurodegenerative diseases present challenges both in common with each other and to other diseases, as well as unique pathological distinctions. One commonality that is most 71 poorly understood is the role environmental factors play in modulating susceptibility to neurodegeneration. In this context, epigenetic modifiers of gene expression comprise a primary regulatory nexus of cellular responses to a variety of potential environmental, as well as genetic, influences. It is therefore unsurprising that specific human miRNAs have been associated with PD that correlate with regulation of neuronal development, mitochondrial activity, and oxidative stress response 21. Furthermore, specific down-regulated miRNAs have been identified from human dopaminergic brain samples where well-studied familial

PD genes, including SNCA, PARK2, LRRK2, and GBA, are among the targets of some of these miRNAs 22,23. Moreover, in comparison to healthy individuals, differential expression levels of various miRNAs have been reported in both peripheral blood mononuclear cells and serum 24,25 as well as cerebral spinal fluid and exosomes 26,27. Overexpression of -syn in a mouse model identified miR-155 expression to be coincident with neuroinflammatory responses 28. Importantly, while these studies provide potential leads for diagnostic purposes, they remain largely theoretical in terms of identifying new therapeutic targets.

This study was designed to evaluate the hypothesis that select targets of miRNA regulation could be revealed as contributing to neuroprotection in vivo, and potentially represent previously unidentified candidates for therapeutic development. Here we report that mir-239 deletion in C. elegans displays significant and strong neuroprotection against -syn neurotoxicity in vivo (Fig. 3.1 B). While the impressive neuroprotective effect of mir-239 deletion on -syn-induced dopaminergic neurotoxicity in C. elegans suggests this miRNA, itself, could be developed into a human PD therapeutic target, it is unfortunately only conserved among invertebrates 6. The mir-239 mutation was previously determined to have an uncharacterized role in the aging process 6. We confirmed the promotion of lifespan in a

C. elegans mir-239 mutant, yet it is modest (one day). Furthermore, we parsed this

72 observation further to discern that in the presence of dopaminergic -syn expression, neuroprotection against proteotoxicity was not simply dependent on chronological time, as strong neuroprotection was also observed during biological aging (Fig. 3.1 C). Although most miRNAs do not share substantial sequence homology to a human counterpart, the network of co-regulated genes uncovered through this study implicates these epigenetically- regulated modifiers of neurodegeneration as putative therapeutic targets for further investigation.

In this regard, the three mir-239b gene targets identified through our screening efforts that are conserved in humans (Fig. 3.3 B-3.3 D) represent previously unreported modifiers of dopaminergic neuron survival that might be suitable for therapeutic development. For example, C. elegans F25H9.6, when knocked down, resulted in a severe neurodegenerative phenotype (Fig. 3.3 C). This gene is an ortholog of human Ppcdc, which encodes phosphopantothenoylcysteine decarboxylase, which is one of the last enzymes in the pathway that converts pantothenic acid (vitamin B5) to coenzyme A (CoA). Considering that some forms of neurodegeneration are associated with iron-accumulation caused by mutations in the

CoA biosynthesis pathway 29-31, further analysis of this gene candidate and pathway will be performed in future studies. Others and we have previously reported the neuroprotective impact across species of similar types of key metabolic enzymes, including those in glycolysis, pointing to evolutionarily-conserved pathways that could be potentially co-opted to modulate neuronal survival 4,32. Interestingly, a glucose-dependent effect on neuroprotection from the common PD-modeling toxin, 1-methyl-4-phenylpyridinium

(MPP+), that inhibits complex I of electron transport chain, was uncovered for human miR-7 in cultured SH-SY5Y neurons, wherein a functional glycolytic pathway was required for its protective effect 33.

The robust neuroprotection conferred by mir-239 knockdown in -syn-expressing C. elegans neurons (Fig. 3.1 B) suggests that either individual gene candidates, or perhaps combinations thereof, derived from the mir-239 positive list might represent novel and potent therapeutic leads. Going forward, the neuroprotective capacity of the hundreds of predicted targets of mir-239 that have not yet been validated should be considered. Nevertheless, caution in studying predicted mir-239 targets is warranted since most prospective regulated genes are not yet verified and may not be specific to this miRNA. In a more realistic scenario, as additional targets become expression verified and evaluated for neuroprotective capacity, bioinformatics may render it feasible to identify a human functional miRNA correlate to mir-239b based on the networked activity of these co-regulated targets in C. elegans that would putatively represent an effective target for intervention.

During the aging process, biological function progressively deteriorates with time and the role of miRNA expression is likely prominent in this decline. Moreover, aging often results in decreased resistance to multiple forms of stress and other related intracellular challenges resulting in increased disease susceptibility 34. While mechanisms underlying aging remain unresolved, the most prominent theory involves ROS accumulation leading to functional alterations, pathological conditions and finally cell death. Oxidative stress is defined as a result of excessive bioavailability of ROS whereby an imbalance exists between production and destruction of ROS, the latter relying on the tightly-regulated expression of antioxidant enzymes and scavengers 35,36. Locomotion, an essential feature of most animals, also characteristically slows with aging 37. The behavioral analyses conducted in this study indicated that mir-239 mutation, which partially resisted the effects of oxidative stress induced, as evidenced by a rescue of the diminished body bends observed in control worms.

This suggests that either the delayed aging and/or the activity of unrepressed target gene

74 expression in mir-239 mutant animals led to a physical improvement in locomotion. Taken together with the impact on neurodegeneration, these combined observations coalesce as a functional signature through which dopaminergic neuronal healthspan and lifespan can be assessed in C. elegans.

Finally, it should be noted that when the mir-239 mutation was crossed into the sid-1

RNAi strain along with -syn, it was not as neuroprotective as it was in the N2 wild-type background, as the percentage of worm populations with the normal complement of DA neurons decreased from 98% in the -syn N2 background to 60% in the -syn sid-1 background (Fig. 3.1 B vs. Fig. 3.3 A). We hypothesize that the sid-1 mutant background, which regulates the systemic efficiency of RNAi in C. elegans and is required for the tissue- specific RNAi in our model 38,12, also modulates the import of small RNAs into neurons that modulate dopaminergic neuron survival (Caldwell and Nourse, unpublished). Therefore, non- cell autonomous miRNAs that signal organismal responses to stressors and environmental cues may interface with the dopaminergic machinery to coordinately regulate neuronal functional and stave off degenerative effects over the course of aging.

Methods and materials

C. elegans strains

Standard laboratory procedures were followed for nematodes maintenance 39. The nematode strains MT15312 (mir-239 (nDf62) X) and sid-1(pk3321) were provided by the

Caenorhabditis Genetics Center (CGC). The following strains, UA44 (baIn11 [Pdat-1::-syn,

Pdat-1::GFP]) and UA196 (sid-1(pk3321);baIn33 [Pdat-1::sid-1, Pmyo-2::mCherry]; baIn11[Pdat-

8,40 1::-syn, Pdat-1::GFP]), were generated previously . MT15312 was out-crossed with wild

75 type N2 worms for three times and then crossed with UA44 to generate strain UA320 [mir-

239 (nDf62); baIn11[Pdat-1::-syn, Pdat-1::GFP]. MT15312 was out-crossed with sid-

1(pk3321) worms and then crossed with UA196 to generate strain UA321 [mir-239(nDf62) sid-1(pk3321); baIn33 [Pdat-1::sid-1, Pmyo-2::mCherry]; baIn11[Pdat-1::-syn, Pdat-1::GFP]).

The following primers were used to detect the mir-239(nDf62) mutation in worms: nDf62 forward: GTGACGTTCTCATTGAGATAAAA nDf62 reverse: CAGCGACAGATGCAATTTTTG

Lifespan assay

Lifespan was measured using the Cohort survival assay 41. 150 synchronized worms were equally distributed on 10 NGM plates and grown at 20oC. From day 4 after hatching on, the occurrence of death or censored events were monitored every 24 hours until all worms were dead. The dead worms were scored as “1” and censored worms were marked as “0” where phenotypes such as worm bagging, vulva protrusion or uncoordinated movement displayed.

Statistics were analyzed using a Coherent survival analysis followed by Mantel-Cox test

(GraphPad Prism software).

RNAi treatments

The E. coil bacteria for RNAi were obtained from the Ahringer C. elegans library 42. Each bacterial strain was cultured overnight at 37oC in LB media containing 100 g/ml ampicillin.

NGM plates containing 1mM IPTG were seeded with 250 l of RNAi culture and allowed to dry overnight. Six L4-stage -syn worms and ten same-age -syn; mir-239 mutant worms were transferred onto corresponding RNAi plates to lay eggs overnight at 20°C. The

76 offspring worms were continuously fed with bacteria containing dsRNA of EV or individual clones of the 26 validated targets. Transfer worms to corresponding fresh RNAi plates was conducted as needed to avoid starvation.

C. elegans neurodegeneration analysis

DA neurodegeneration in worms was scored as previously described 43. Briefly, synchronized worms were cultured at 20°C to day 7 or 8 after hatching. DA neuronal health was examined by GFP and scored under a fluorescent microscopy (Nikon Eclipse E800) at analysis dates. Worms were considered normal when all six anterior DA neurons (four CEPs and two ADEs) were present without any visible signs of degeneration such as process blebbing, cell body rounding or missing neurons. If a worm displayed a neuron with such an altered neuronal morphology, the entire worm was considered non-normal. In total, at least

90 (30 animals in triplicate) adult worms were analyzed with the investigator blinded to sample identity. A Cool Snap CCD camera (Photometrics) driven by MetaMorph software

(Molecular Devices) was utilized to acquire representative images of anterior DA neurons.

Statistics were analyzed using an Student’s t-test, one-way ANOVA with a Dunnett post hoc test, or two-way ANOVA with a Sidak’s post-hoc test (GraphPad Prism software).

Oxidative stress assays

Sodium azide treatment: We followed a previously published method to test oxidative stress resistance 32. Briefly, worms were transferred onto 35mm worm plates containing

100M sodium azide for 12 h prior to analysis, which occurred at days 4, 9,14 and 19. Five worms are placed into the well of a 96-well plate containing 50 l M9 buffer and then digital videos of animal movement (n= 10 worms/replicate, 3 replicates in total/condition) are

77 recorded for 1 min. using a Worm Tracker (MBF Biosciences) 32. The thrashing frequency for each animal was quantified by counting body bends, defined as a direction change of bending at the middle of worm body 44,45. A two-way ANOVA with a Sidak post hoc test

(GraphPad Prism) was used for following statistical analysis. Paraquat treatment: Paraquat experiments were performed as previously described 6,20. 45 synchronized worms were treated with 200 mM paraquat for 6 hours on day 7 post-hatching. The occurrence of death was monitored every hour. The dead worms were scored as “1” and censored worms were marked as “0”. Statistics were analyzed using Coherent survival analysis followed by

Mantel-Cox test (GraphPad Prism software).

References

1. He, L. & Hannon, G. J. MicroRNAs: small RNAs with a big role in gene regulation. Nat Rev Genet 5, 522-531, doi:10.1038/nrg1379 (2004).

2. Treusch, S. et al. Functional links between Abeta toxicity, endocytic trafficking, and Alzheimer's disease risk factors in yeast. Science 334, 1241-1245, doi:10.1126/science.1213210 (2011).

3. Tardiff, D. F. et al. Yeast reveal a "druggable" Rsp5/Nedd4 network that ameliorates alpha-synuclein toxicity in neurons. Science 342, 979-983, doi:10.1126/science.1245321 (2013).

4. Knight, A. L. et al. The glycolytic enzyme, GPI, is a functionally conserved modifier of dopaminergic neurodegeneration in Parkinson's models. Cell Metab 20, 145-157, doi:10.1016/j.cmet.2014.04.017 (2014).

5. Miska, E. A. et al. Most Caenorhabditis elegans microRNAs are individually not essential for development or viability. PLoS Genetics 3, e215, doi:10.1371/journal.pgen.0030215 (2007).

6. de Lencastre, A. et al. MicroRNAs Both Promote and Antagonize Longevity in C. elegans. Current Biology 20, 2159-2168, doi:10.1016/j.cub.2010.11.015 (2010).

7. Ruan, Q., Harrington, A. J., Caldwell, K. A., Caldwell, G. A. & Standaert, D. G. VPS41, a protein involved in lysosomal trafficking, is protective in Caenorhabditis elegans and mammalian cellular models of Parkinson's disease. Neurobiology of disease 37, 330-338, doi:10.1016/j.nbd.2009.10.011 (2010).

8. Harrington, A. J., Yacoubian, T. A., Slone, S. R., Caldwell, K. A. & Caldwell, G. A. Functional analysis of VPS41-mediated neuroprotection in Caenorhabditis elegans and mammalian models of Parkinson's disease. The Journal of neuroscience : the official journal of the Society for Neuroscience 32, 2142-2153, doi:10.1523/jneurosci.2606-11.2012 (2012).

9. Gitler, A. D. et al. The Parkinson's disease protein alpha-synuclein disrupts cellular Rab homeostasis. Proc Natl Acad Sci U S A 105, 145-150, doi:10.1073/pnas.0710685105 (2008).

10. Kim, H. et al. The Small GTPase RAC1/CED-10 Is Essential in Maintaining Dopaminergic Neuron Function and Survival Against alpha-Synuclein-Induced Toxicity. Molecular neurobiology, doi:10.1007/s12035-018-0881-7 (2018).

11. Berkowitz, L. A. et al. Application of a C. elegans Dopamine Neuron Degeneration Assay for the Validation of Potential Parkinson's Disease Genes. Journal of Visualized Experiments, doi:10.3791/835 (2008).

12. Harrington, A. J., Knight, A. L., Caldwell, G. A. & Caldwell, K. A. Caenorhabditis elegans as a model system for identifying effectors of alpha-synuclein misfolding and dopaminergic cell death associated with Parkinson's disease. Methods (San Diego, Calif.) 53, 220-225, doi:10.1016/j.ymeth.2010.12.036 (2011).

13. Vlachos, I. S. et al. DIANA-TarBase v7.0: indexing more than half a million experimentally supported miRNA:mRNA interactions. Nucleic acids research 43, D153-D159, doi:10.1093/nar/gku1215 (2014).

14. Fernandez, J. R., Grant, M. D., Tulli, N. M., Karkowski, L. M. & McClearn, G. E. Differences in locomotor activity across the lifespan of Drosophila melanogaster. Experimental gerontology 34, 621-631 (1999).

15. Le Bourg, E. The rate of living theory. Spontaneous locomotor activity, aging and longevity in Drosophila melanogaster. Experimental gerontology 22, 359-369 (1987).

16. Siwak, C. T., Murphey, H. L., Muggenburg, B. A. & Milgram, N. W. Age-dependent decline in locomotor activity in dogs is environment specific. Physiology & behavior 75, 65-70 (2002).

17. Scimonelli, T., Marucco, M. & Celis, M. E. Age-related changes in grooming behavior and motor activity in female rats. Physiology & behavior 66, 481-484 (1999).

18. Lin, Y. J., Seroude, L. & Benzer, S. Extended life-span and stress resistance in the Drosophila mutant methuselah. Science 282, 943-946 (1998).

19. Honda, Y. & Honda, S. The daf-2 gene network for longevity regulates oxidative stress resistance and Mn-superoxide dismutase gene expression in Caenorhabditis elegans. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 13, 1385-1393 (1999).

20. Li, J. et al. Caenorhabditis elegans HCF-1 functions in longevity maintenance as a DAF-16 regulator. PLoS biology 6, e233, doi:10.1371/journal.pbio.0060233 (2008).

21. Martinez, B. & Peplow, P. V. MicroRNAs in Parkinson's disease and emerging therapeutic targets. Neural Regeneration Research 12, 1945-1959, doi:10.4103/1673- 5374.221147 (2017).

22. Tatura, R. et al. Parkinson's disease: SNCA-, PARK2-, and LRRK2- targeting microRNAs elevated in cingulate gyrus. Parkinsonism & related disorders 33, 115- 121, doi:10.1016/j.parkreldis.2016.09.028 (2016).

23. Hoss, A. G., Labadorf, A., Beach, T. G., Latourelle, J. C. & Myers, R. H. microRNA Profiles in Parkinson's Disease Prefrontal Cortex. Frontiers in Aging Neuroscience 8, 36, doi:10.3389/fnagi.2016.00036 (2016)

24. Martins, M. et al. Convergence of miRNA expression profiling, alpha-synuclein interacton and GWAS in Parkinson's disease. PloS one 6, e25443, doi:10.1371/journal.pone.0025443 (2011).

25. Botta-Orfila, T. et al. Identification of blood serum micro-RNAs associated with idiopathic and LRRK2 Parkinson's disease. Journal of neuroscience research 92, 1071-1077, doi:10.1002/jnr.23377 (2014).

26. Burgos, K. et al. Profiles of extracellular miRNA in cerebrospinal fluid and serum from patients with Alzheimer's and Parkinson's diseases correlate with disease status and features of pathology. PloS one 9, e94839, doi:10.1371/journal.pone.0094839 (2014).

27. Gui, Y., Liu, H., Zhang, L., Lv, W. & Hu, X. Altered microRNA profiles in cerebrospinal fluid exosome in Parkinson disease and Alzheimer disease. Oncotarget 6, 37043-37053, doi:10.18632/oncotarget.6158 (2015).

28. Thome, A. D., Harms, A. S., Volpicelli-Daley, L. A. & Standaert, D. G. microRNA- 155 Regulates Alpha-Synuclein-Induced Inflammatory Responses in Models of Parkinson Disease. The Journal of neuroscience : the official journal of the Society for Neuroscience 36, 2383-2390, doi:10.1523/jneurosci.3900-15.2016 (2016).

29. Daugherty, M. et al. Complete reconstitution of the human coenzyme A biosynthetic pathway via comparative genomics. J Biol Chem 277, 21431-21439, doi:10.1074/jbc.M201708200 (2002).

30. Dusi, S. et al. Exome Sequence Reveals Mutations in CoA Synthase as a Cause of Neurodegeneration with Brain Iron Accumulation. The American Journal of Human Genetics 94, 11-22, doi:10.1016/j.ajhg.2013.11.008.

31. Zhou, B. et al. A novel pantothenate kinase gene (PANK2) is defective in Hallervorden-Spatz syndrome. Nat Genet 28, 345-349, doi:10.1038/ng572 (2001).

32. Zhang, S., Glukhova, S. A., Caldwell, K. A. & Caldwell, G. A. NCEH-1 modulates cholesterol metabolism and protects against alpha-synuclein toxicity in a C. elegans model of Parkinson's disease. Hum Mol Genet 26, 3823-3836, doi:10.1093/hmg/ddx269 (2017).

33. Chaudhuri, A. D., Kabaria, S., Choi, D. C., Mouradian, M. M. & Junn, E. MicroRNA- 7 Promotes Glycolysis to Protect against 1-Methyl-4-phenylpyridinium-induced Cell Death. J Biol Chem 290, 12425-12434, doi:10.1074/jbc.M114.625962 (2015).

34. Harman, D. Aging: overview. Annals of the New York Academy of Sciences 928, 1-21 (2001).

35. Maritim, A. C., Sanders, R. A. & Watkins, J. B., 3rd. Diabetes, oxidative stress, and antioxidants: a review. Journal of biochemical and molecular toxicology 17, 24-38, doi:10.1002/jbt.10058 (2003).

36. Lushchak, V. I. Environmentally induced oxidative stress in aquatic animals. Aquatic toxicology (Amsterdam, Netherlands) 101, 13-30, doi:10.1016/j.aquatox.2010.10.006 (2011).

37. Glenn, C. F. et al. Behavioral deficits during early stages of aging in Caenorhabditis elegans result from locomotory deficits possibly linked to muscle frailty. The journals

of gerontology. Series A, Biological sciences and medical sciences 59, 1251-1260 (2004).

38. Calixto, A., Chelur, D., Topalidou, I., Chen, X. & Chalfie, M. Enhanced neuronal RNAi in C. elegans using SID-1. Nature methods 7, 554-559, doi:10.1038/nmeth.1463 (2010).

39. Brenner, S. The genetics of Caenorhabditis elegans. Genetics 77, 71-94 (1974).

40. Cooper, A. A. et al. -Synuclein Blocks ER-Golgi Traffic and Rab1 Rescues Neuron Loss in Parkinson’s Models. Science (New York, N.Y.) 313, 324-328, doi:10.1126/science.1129462 (2006).

41. Lionaki, E. & Tavernarakis, N. in Methods in Molecular Biology 473-484 (Humana Press, 2012).

42. Kamath, R. Genome-wide RNAi screening in Caenorhabditis elegans. Methods 30, 313-321, doi:10.1016/s1046-2023(03)00050-1 (2003).

43. Cao, S. Torsin-Mediated Protection from Cellular Stress in the Dopaminergic Neurons of Caenorhabditis elegans. Journal of Neuroscience 25, 3801-3812, doi:10.1523/jneurosci.5157-04.2005 (2005).

44. Michaelson, L. C. elegans - A Practical Approach. Oxford University Press (2000).

45. Locke, C. J. et al. Pharmacogenetic analysis reveals a post-developmental role for Rac GTPases in Caenorhabditis elegans GABAergic neurotransmission. Genetics 183, 1357-1372, doi:10.1534/genetics.109.106880 (2009).

CHAPTER FOUR

INVESTIGATING A ROLE OF TOR-2 IN CONTROLLING ER EXPORT OF AMPA RECEPTORS IN THE REGULATION OF EXCITABILITY IN C. ELEGANS

This work was performed by Chuan Xu, G. Blake Parker, Guy A. Caldwell, Kim A.

Caldwell. Chuan Xu, G. Blake Parker collected all the data. Guy A. Caldwell and Kim A.

Caldwell contributed to the editing.

Abstract

Human torsinA, encoded by the DYT1 gene, is an ER resident chaperone protein that has been identified to be responsible for a human movement disorder called early-onset torsion dystonia. Tor-2 is a C. elegans homologue of human torsinA and it is expressed in a few neurons in the adult hermaphrodite, including the AVE interneurons. AVE neurons are responsible for driving backwards movement, which is regulated by AMPA type glutamate receptors. Changes in synaptic AMPA receptor levels have been proposed to be a key regulatory event in synaptic plasticity. Our data have revealed that tor-2 mutation, or pan- neuronal RNAi knockdown of tor-2, results in behavioral defects indicative of glutamatergic involvement. Specifically, defective backwards movement and nose touch response. We also discovered that pan-neuronal tor-2 RNAi caused a significant decrease in abundance of the

AMPA type glutamate receptor, GLR-1, in the ventral nerve cord. We propose that TOR-2 activity modulates the trafficking of AMPA-type glutamate receptors and further regulates synaptic excitability in C. elegans. This discovery provides insight into the cellular pathogenesis associated with early-onset torsion dystonia.

Introduction

Mutation in the DYT1 gene is responsible for an autosomal dominant inherited human movement disorder called early-onset torsion dystonia. DYT1 dystonia is characterized by painful, involuntary sustained muscle contractions and abnormal postures. Human torsinA is an ER-resident protein and is a member of the large and structurally diverse of family of

AAA+ proteins that hydrolyze ATP to release energy for remodeling substrates. While clinical manifestations are well-documented, the underlying molecular mechanisms by which torsinA modulates neuronal activity remain to be fully elucidated. Biochemical and in vivo evidence suggests that torsinA, like other AAA+ proteins, has molecular chaperone activity

1,2. For instance, a quantitative readout for ER stress response in C. elegans has demonstrated that torsinA functions as a homeostatic regulator of ER stress 3. Likewise, overexpression of human wild-type (WT) torsinA and related invertebrate orthologs can suppress the accumulation of misfolded proteins 1,4.

C. elegans tor-2 is a homologue of human torsinA; it is only expressed in a few neurons in the adult hermaphrodite, including the AVE interneurons, which are responsible for driving backwards movement 5. AVE interneurons have glutamatergic receptors, which are a major excitatory neurotransmitter in the nervous system 6. There are two distinct types of glutamate receptors in C. elegans, N-methyl-d-aspartate (NMDA) and amino-3-hydroxy-5- methylisoxazoleproprionic acid (AMPA) receptors. For convenience, glutamate receptors also can be divided into two major types: ionotropic and metabotropic receptors. Furthermore, ionotropic receptors also are classified into NMDA and non-NMDA receptors based on whether binding to NMDA agonist or not 7. AMPA-type ionotropic glutamate receptors

(iGluRs) is the major excitatory neurotransmitter receptor in the developing and adult vertebrate CNS 8. The cerebellar overactivity observed in neuroimaging studies of patients may indirectly reflect abnormal glutamate signaling 9. Glutamate receptors activities are often related with synaptic plasticity. The understanding about how these receptors are rapidly moved into and out of synaptic membranes can provide us profound implications about the mechanisms behind a lot of related CNS disorders 10. Several trafficking pathways for glutamate receptors have been gradually identified, including PDZ-domain contain proteins

11, membrane-associated guanylate kinases (MAGUKs) 12, transmembrane AMPAR regulatory proteins (TARPs) 13, and pentraxins, which cluster iGluRs through extracellular protein interactions 14.

AMPA receptors (AMPARs) are an ionotropic transmembrane receptor for glutamate that mediate the majority of fast excitatory neurotransmission in the central nervous system

(CNS). These ligand-gated ion channels are composed of combinations of four separate subunits (GluA1-4) and have been indicated critically important for nearly all aspects of brain function including learning, memory, and cognition. The abnormality in the processes that control AMPAR assembly, trafficking and synaptic expression of AMPARs can be linked with many psychiatric and neurological and neurodegenerative disorders 15. For example, decreased level of AMPARs has been showed as one of the earliest cell biological manifestations of dementia in Alzheimer disease (AD). Defective AMPAR trafficking caused by soluble amyloid-β (Aβ) is a major causative agent of synaptic dysfunction in AD. It has been reported that Aβ oligomers bind in close proximity to GluA2‑containing complexes.

The inhibition of Aβ oligomer binding and synaptic loss by AMPAR antagonist further indicated that Aβ may affect AMPAR trafficking by binding directly to the GluA2 protein complex 16,17. Furthermore, it has been implied that glutamate receptor activation, specifically

AMPA receptor activation, can induce dystonia in mice model. Moreover, the dystonia caused by AMPA receptor agonists can be reduced by AMPA receptor antagonists 18.

Changes in synaptic AMPA receptor levels have been proposed to be key regulatory events in the dynamic changes in neuronal synaptic efficacy, also termed synaptic plasticity

19. These processes have been well-studied in C. elegans using one of the AMPA glutamate receptors subunits as a model. Mutations in glr-1, the first characterized AMPA subtype in C. elegans, shows that when this gene is eliminated that there is a mutant behavioral response whereby the animals will no longer respond to nose touch stimulation and they are also disrupted in the “spontaneous” switch from forward to backward movement in C. elegan 20-22.

These simple C. elegans behavioral phenotypes can be used to examine other proteins and/or chemicals that influence the cellular processing of AMPA receptors, including biosynthesis, membrane trafficking and synaptic targeting, and ultimately degradation of these AMPA receptors. These processes are controlled by numerous intracellular regulatory proteins 23. For

86 instance, two classes of transmembrane proteins (TARP and SOL) play critical roles as auxiliary proteins in AMPAR trafficking and function 24,25. In C. elegans, the activity of the neuronal-specific cyclin-dependent kinase, CDK-5, has been reported to play a role in the abundance of glutamate receptor subunit, GLR-1, in the ventral nerve cord. Loss of CDK-5 activity led to decreased abundance of GLR-1 in the ventral nerve cord and consequently changed GLR-1- dependent behaviors including nose touch response and reversal frequency

26. The ubiquitin-conjugating enzyme variant UEV-1 also has been recognized as a regulator of AMPAR trafficking in C. elegans. Loss of UEV-1 activity resulted in the accumulation of

GLR-1 in neuron cell bodies and there were corresponding defects of GLR-1-dependent behaviors in C. elegans 27.

The members of the newly characterized AMPAR auxiliary proteins, the Cornichon homologs (CNIHs), also have been identified playing a role in the AMPAR biogenesis and

ER export process 28. Mutation of cni-1, the sole cornichon homologue in C. elegans, indicated a hyper-reverse phenotype associated with increased glutamatergic synaptic transmission. A corresponding increase in the number of synaptic GLR-1 receptors, together with larger glutamate-gated current, was revealed in cni-1 mutant worms 29.

We took note of these data within the context of the Human Liver Proteome Project where these scientists determined that human cornichon 4 physically interacted with human torsinA.

Cornichons are a family of ER-localized transmembrane proteins that directly interact with

AMPARs immediately after translation in the ER to promote export from the ER and Golgi complex 29,30. Based on torsinA chaperone activity in the ER, we asked that whether there was a role for tor-2, the homologue of human torsinA, in GLR-1 neurotransmitter receptor trafficking from the ER to the plasma membrane in C. elegans. Here, we revealed that tor-2 mutation or pan-neuronal RNAi knockdown of tor-2 in C. elegans disrupted backwards movement and nose touch response, behavioral phenotypes indicative of glutamatergic

87 involvement. Furthermore, we found that pan-neuronal tor-2 RNAi in C. elegans caused a significant decrease in abundance of the AMPA type glutamate receptor, GLR-1, in the ventral nerve cord. The identification of tor-2 in the AMPAR trafficking suggests a connection between glutamatergic signaling and the etiology of early-onset torsion dystonia.

Results tor-2 RNAi causes cell non-autonomous ER stress

Several papers have reported that human torsinA, an ER resident protein, has chaperone activity 3. Here, we wanted to determine if tor-2, the C. elegans homologue of human torsinA, is associated with an ER stress response. Therefore, we knocked down tor-2 in

3 worms expressing an ER stress reporter, Phsp-4::GFP, using previously established methods .

Worms fed tor-2 dsRNA resulted in significant increased ER stress, as measured using GFP intensity, compared with control (Fig. 4.1 A and 4.1 B). Since tor-2 is known to be expressed in few neurons and some muscles but not in the intestine 5, this suggests that knocking down tor-2 causes a cell non-autonomous effect on the intestine in C. elegans.

Figure 4.1. Systematic tor-2 RNAi increases ER stress in worm strain expressing Phsp-4::GFP. (A) Representative fluorescent micrographs of animals expressing the ER stress reporter hsp- 4::GFP in a worm strain fed with RNAi negative control EV (upper panel) and with dsRNA tor-2 (lower panel). The image depicts the anterior region of C. elegans and the boxed region indicates the 100×100 μm region just below the pharynx where measurements of fluorescence were taken in all animals. (B) Bar chart representing normalized GFP pixel intensity values measured consistently in the boxed region shown in A. Values are the mean ± s.d. of three experiments where 30 animals were analyzed per replicate (*P <0.01; Student’s t- test).

Pan-neuronal tor-2 RNAi causes nose touch response and backwards movement defects

Two behaviors, nose touch response and reversal frequency, are associated with the GLR-

1 subunit of AMPARs in C. elegans. Gentle touch to the nose of the worm results in a mechanosensory reflex that induces backward locomotion. Worms expressing loss-of- function glr-1 are less responsive to nose touch and decreased reverse frequency compared with control 6,20,22,31-33. Conversely, overexpression gain of function glr-1 causes increased reversal frequencies compared with controls 22. TOR-2 is known to be expressed in a few neurons, including AVE interneurons 5, which drives backwards movement in C. elegans

(Fig. 4.2). To explore the effect of TOR-2 activity on GLR-1 function, we knocked down tor-

2 in a worm strain that allows for pan-neuronal RNAi (Tu3401[Pmyo-2::mCherry, Punc-119::sid-1

V]; sid-1(pk3321)]) 34 and assayed backwards movement behavior at day 3 post-hatching. We found that worms fed tor-2 dsRNA showed significantly decreased percentage of successful

89 nose touch responses (Fig. 4.3 A) and backwards frequency (Fig. 4.3 B) compared with EV

RNAi. Compared with EV control worms, who had a rate of ~6 backwards turns in 3 minutes, the worms fed tor-2 dsRNA and the positive control glr-1 RNAi worms, both had ~3 backwards turns in 3 minutes (Fig. 4.3 A). Thus, it indicated that tor-2 activity can regulate glr-1 mediated behaviors. For the nose touch response, tor-2 RNAi worms had ~50% response of successful avoidance in 10 trials compared with control EV RNAi worms which had ~80% (Fig. 4.3 B).

Figure 4.2. A schematic of locomotion circuitry that controls the initiation of backward movement in C. elegans. When sensory neurons (ASH, FLP and QLQ) receive the sensory stimulus, they release the neurotransmitter glutamate into glutamate receptors in interneurons (AVE, AVA and AVD) and the command interneurons directly synapse onto downstream motor neurons to drive backwards locomotion in C. elegans. Modified from Source: Zheng et al., 1999, Neuron.

Figure 4.3. Treatment of pan-neuronal RNAi worm strain TU3401 with dsRNA tor-2 causes significant defects on backwards movements and nose touch response (A, B) Graphical representation of C. elegans strain TU3401[Pmyo-2::mCherry, Punc-119::sid-1; sid-1(pk3321)] following tor-2 or glr-1 knockdown by RNAi. For both RNAi experimental conditions, synchronized C. elegans were analyzed at day 3 post-hatching. RNAi bacteria, which do not express an RNAi clone (EV), were used as a negative control. (A) Backwards number in 3 minutes were recorded in Tu3401 C. elegans fed with EV or dsRNA tor-2 using a worm tracker. glr-1 RNAi was used as a positive control. Data are reported as the mean ± SD, where at least 30 worms were analyzed per condition. *P < 0.05, one-way ANOVA (B) Nose touch assay was executed in C. elegans after EV or dsRNA tor-2 treatment. An eyebrow hair was required for gentle touch on the C. elegans nose. Successful avoidance corresponded to the initiation of backwards locomotion after gentle nose touch. At least 15 animals were tested, with 10 trials for each animal. Data are expressed as % response of successful avoidance and reported as the mean ± SD, *P < 0.05, Student’s t-test.

We also wanted to determine whether loss-of-function via a tor-2 mutant would show similar effects on these two behaviors in C. elegans. Here we examined these phenotypes in the worm strain tor-2(tm697) at day 3 post-hatching. For this mutant, most of the coding region containing the ATP binding and hydrolysis sites, are removed, suggesting that this mutation is likely a loss-of-function mutation. Compared with wild-type N2, mutant tor-2 worms also showed a significant decrease in the percentage of worms displaying successful avoidance for the nose touch response (Fig. 4.4 A) and defects on the reversal frequency (Fig.

4.4 B). This further confirmed that the decreased activity of tor-2 caused the defects on the behaviors mediated by glr-1 in C. elegans.

Figure 4.4. Loss of function of tor-2 mutation causes significant defects on backwards movement and nose touch response in C. elegans (A, B) Graphical representation of C. elegans strain wild-type N2, tor-2(tm697) or loss of function glr-1(n2461). Synchronized C. elegans were analyzed at day 3 post-hatching. (A) Backwards movements that occurred for 3 minutes intervals were scored after video recording using a worm tracker. Data are reported as the mean ± SD, with at least 30 worms per condition. *P < 0.05, one-way ANOVA. (B) Nose touch response assay was executed in N2 wild-type or tor-2 mutation worms. An eyebrow hair was used in a gentle touch assay. Successful avoidance corresponded to backward locomotion initiation after gentle nose touch. At least 15 animals were tested in 10 trials. Response was expressed as % response of successful avoidance. Data are reported as the mean ± SD, *P < 0.05, Student’s t-test.

Tor-2 mutation or Pan-neuronal tor-2 RNAi decreases omega turns frequency in C. elegans

Omega turns are identified when worms make turns in which a single head swing leads to a reorientation of more than 135 degree resembling an omega (Ω) symbol. Glutamate has been indicated as a key regulator of Ω turns in C. elegans. Null glr-1 mutants display a decrease in the number of Ω turns per minute 35,36. Since we have revealed that an activity change of TOR-2 can disrupt GLR-1 mediated behaviors, as described for the nose touch response and reversal frequency, we decided to examine a third behavioral response, the Ω turn frequency. We examined both tor-2 mutants and pan-neuronal tor-2 RNAi worms. We found that compared with corresponding controls, worms expressing tor-2(tm697) or tor-2 pan-neuronal knockdown caused decreased Ω turn frequencies (Fig. 4.5 A and 4.5 B). For

92 instance, wild-type N2 worms have 6 Ω turns in 3 min, while tor-2(tm697) mutant worms have 4 Ω turns (Fig. 4.5 A).

Figure 4.5. Loss of function of tor-2 mutation and pan-neuronal tor-2 RNAi decreases omega turn frequency in C. elegans (A, B) Graphical representation of C. elegans strain wild-type N2 and tor-2(tm697) or pan neuronal RNAi worm strain TU3401[Pmyo-2::mCherry, Punc- 119::sid-1; sid-1(pk3321)] following EV, glr-1 or tor-2 RNAi knockdown. Synchronized C. elegans were analyzed at day 3 post-hatching. (A, B) Ω shape number over the course of 3 minutes was recorded using a worm tracker. Data are reported as the mean ± SD, using at least 30 worms per condition. Data were analyzed using Student’s t-test; *P < 0.05 (A) or one-way ANOVA, *P < 0.05 (B).

Pan-neuronal tor-2 RNAi decreases GLR-1 synaptic number in the ventral nerve cord

It was demonstrated that transgenic worms expressing a GFP-tagged GLR-1 construct

(Pglr-1::GLR-1::GFP) that GLR-1 is localized in the ventral nerve cord (VNC) interneurons and that it has a punctate pattern 37. By using two presynaptic markers, synaptobrevin (SNB-

1) and the vesicular glutamate transporter (VGLUT) which is encoded by the eat-4 gene, it was revealed that GLR-1::GFP puncta are usually closely apposed to both SNB-1 and EAT-4 puncta, but nonoverlapping. Further analysis suggested more than 80% GLR-1::GFP puncta correspond to postsynaptic elements, indicating that GLR-1 is postsynaptic glutamate receptors 31.

We crossed the Pglr-1::DsRed2 worms to a transgenic line that expresses Ptor-2::GFP. We found that TOR-2 was also expressed in the ventral nerve cord. To confirm the colocalization,

93 we first outcrossed the worm strain NC1750 (hdIs32 [glr-1::DsRed2]. gvEx173 [opt-3::GFP

+ rol-6 (su1006)]) to generate worms expressing Pglr-1::DsRed2, then further crossed it with worms expressing Ptor-2::GFP. Since TOR-2 was fused with GFP and GLR-1 was fused with dsRed, we were able to visualize colocalization of these proteins. Areas of overlap appear yellow. In this experiment, we determined that TOR-2 is expressed in the nervous system and there were areas of colocalization with GLR-1 in the ventral nerve cord in C. elegans. (Fig.

4.6 B)

Glutamate receptors numbers are associated with synaptic plasticity. Because TOR-2 is an

ER resident protein that could be involved in the cellular trafficking of GLR-1, a polytopic membrane-bound receptor protein, we generated a worm strain which can knockdown tor-2 pan-neuronally in a worm strain (KP1148) expressing Pglr-1::GLR-1::GFP. We fed this worm strain with tor-2, glr-1 dsRNA or EV and analyzed the distribution of GLR-1::GFP at day 3 post-hatching using the fluorescent microscope according to the previously published method

38 ,39 (Fig. 4.6 C). We discovered that compared with negative control EV, the worms fed tor-2 dsRNA or positive control glr-1 RNAi, both have significantly decreased GFP intensity, indicating that the glutamate receptor GLR-1 significantly decreased in the VNC in these two conditions. For glr-1 RNAi worms, there was ~40% decrease of the relative GFP intensity compared with EV, while for tor-2 RNAi worms, the decreased value was ~ 20%.

Furthermore, we asked whether this decreased synaptic number of GLR-1 occurs because of decreased transcription levels of glr-1. To test this, we did a qRT-PCR experiment and compared glr-1 mRNA levels between wild-type N2 and tor-2(tm697) mutation worms. We found there was no significant difference between them (Fig. 4.6 D), indicating that the change of GLR-1 synaptic number is not occurring at the transcription level, but is likely occurring during translation or post-translational processing.

Figure 4.6. Synaptic levels of Pglr-1::GLR-1::GFP are decreased in pan neuronal TU3401 worm strain followed by tor-2 RNAi. Synchronized C. elegans were analyzed at day 3 post- hatching. (A) Pglr-1::GLR-1::GFP is expressed in a punctate pattern along process of (ventral nerve cord) VNC. A schematic of GLR-1 localization in the VNC in C. elegans showing synaptic densities. (B) TOR-2 is expressed in the nervous system where it colocalizes with GLR-1 in the ventral nerve cord in C. elegans. Transgenic worm strain UA319 that expressed Ptor-2::GFP, rol; Pglr-1::DsRed2 were analyzed at day 3 post-hatching. (C) Quantification of relative GFP pixel intensity in the worm strain UA318 that expressed (Pglr-1::GLR-1::GFP; Pmyo-2::mCherry, Punc-119::sid-1; sid-1(pk3321)) fed with EV or dsRNA tor-2 or glr-1 RNAi bacteria. Data were analyzed using one-way ANOVA; *P < 0.05 (D) qRT-PCR experiment showed that the mRNA level of glr-1 in tor-2(tm697) mutant worms is nonsignificant different with wide-type N2 worms. Data was analyzed using Student’s t-test.

Discussion

Previous studies have revealed that regulation of AMPA receptor abundance at postsynaptic elements plays an important role for regulating synaptic strength. Our data demonstrate that TOR-2, the homologue of human torsinA, can modulate the abundance of

GLR-1, a subunit of AMPARs, in the ventral nerve cord in C. elegans. Furthermore, pan- neuronally tor-2 RNAi or a loss of function tor-2 mutation, can result in glr-1-mediated behaviors (nose touch response, reversal frequency). We speculate that decreased abundance of glr-1 in the ventral nerve cord contributes to these disrupted behaviors.

The ER lumen is the predominant site for torsinA localization, which is consistent with its role in protein folding and trafficking 40-42. As a characterized ER-resident chaperone protein,

96 torsinA has been shown involving in several cell activities such as an ER stress regulator 3, and interfering with the trafficking of proteins through the ER secretory pathway 43.

Interestingly, torsinA has been reported to interact with cornichons, the evolutionarily conserved ER protein, in the human liver project. Cornichon protein has been shown to modify the properties of vertebrate AMPARs in heterologous cells 44,45. Further study regarding CNI-1, the sole cornichon homolog in C. elegans, has indicated that Cornichons control ER export of AMPA receptors to regulate synaptic excitability. The export of

AMPARs is unregulated in cni-1 mutants and it resulted in increased transport of receptors, larger synaptic currents and neuronal hyperexcitability 29. Thus, based on the chaperone role of human torsinA and the interaction with cornichons, future studies for this project should involve examining TOR-2 mediated GLR-1 trafficking together with CNI-1 in C. elegans. An additional study would be to investigate whether there is a physical interaction between TOR-

2 and CNI-1, as the human liver project identified a direct interaction between torsinA and cornichon. Future directions also include exploring the role of tor-2 in the secretory pathway of glr-1 from the ER in detail and how the activity of TOR-2 impacts the glutamate-gated current.

On a final note, AMPARs are enriched in the central nervous system (CNS). Regulated trafficking of AMPA receptors (AMPARs) are an important mechanism that underlies the activity-dependent modification of synaptic strength. Our study is beginning to reveal that the trafficking process of glutamate receptors might be a contributor to early-onset torsion dystonia.

Methods and materials

Nematode strains

Nematodes were maintained using standard procedures 46. The strain N2 Bristol (N2),

NC1750 (hdIs32 [glr-1::DsRed2]; gvEx173[opt-3::GFP + rol-6(su1006)]) and

KP1148(nuIs25[Pglr-1::glr-1::GFP + lin-15(+)]) were provided by Caenorhabditis Genetics

Center (CGC), which is funded by NIH Office of Research Infrastructure Programs (P40

OD010440). Pan-neuronal specific RNAi worm strain TU3401 [sid-1(pk3321); uIS69(Punc-

119::sid-1, pCFJ90 (Pmyo-2::mCherry))] was generously provided by Martin Chalfie (Columbia

University, New York, NY) 47. The BIOLOGICAL RESOURCE strain tor-2(tm697) was provided by the MITANI Lab through the National Bio-Resource Project of the MEXT,

Japan. Worm strain UA317(baIn37[Ptor-2::GFP, rol-6 (su1006)]) was generated by previous student Songsong Cao in our lab. KP1148 was crossed with TU3401 to generate the double strain UA318. NC1750 was outcrossed with N2 worms 3 times to create a strain carrying only the hdIs32[Pglr-1::DsRed2] transgene, which was then further crossed with worm strain

Ptor-2::GFP, rol to generate strain UA319.

RNAi treatments

EV, tor-2 and glr-1 RNAi feeding clones were purchased from Geneservice. Bacteria containing these plasmids were isolated and grown overnight in LB media with 100μg/ml ampicillin. Nematode growth media plates containing 1μM IPTG were seeded with RNAi feeding clones and allowed to dry. L4 staged hermaphrodites were transferred to corresponding RNAi plates and allowed to lay eggs overnight to synchronize the F1 progeny.

Behavior assay

The nose touch response assay was performed as described previously 48. Briefly, worms were placed on a thin lawn of Escherichia coli and hair pick was required for gentle touch on the nose of worms. At least 15 worms per condition were tested. Each animal was scored 10 trials. Data were analyzed using Student’s t-test (GraphPad Prism software). For locomotion assays, worms were first transferred to an unseeded NGM plate. After 2 min of equilibration, the number of reversals/min was recorded for a total of 3 min/animal by using worm tracker.

At least 30 worms per condition were tested. Statistics were analyzed using unpaired

Student’s t-test or one-way ANOVA with a Dunnett post hoc test (GraphPad Prism software).

Fluorescent microscopy

Fluorescent microscopy was performed using a Nikon Eclipse E800 epifluorescence microscope. A Cool Snap CCD camera (Photometrics) driven by MetaMorph software

(Molecular Devices) was used to acquire fluorescent images. For colocalization, fluorescent images were obtained using 40x objective with cooled CCD camera. Transgenic worms that expressed Ptor-2::GFP, rol; Pglr-1:: DsRed2 were analyzed at day 3 post-hatching. To quantitate the fluorescence levels of GFP-tagged proteins, following the previous published paper 38,39.

Briefly, Z-stacks of fluorescent images were captured and maximum intensity projections were obtained using MetaMorph software under 100 X 1.4 NA PlanApo objective. Exposure times were chosen to fill the 12-bit dynamic range without saturation. Pixel intensity was measured for each animal by using a line scan measurement in MetaMorph by quantifying ventral nerve cord for each sample. An integrated optical density (IOD) score was obtained by summing the pixel intensity values for total fluorescent intensity for each nematode sample. At least 30 worms per condition were tested. Data were analyzed using one-way

ANOVA with a Dunnett post hoc test (GraphPad Prism software).

Real-time qRT-PCR

qPCR reactions were performed using IQ SYBR Green Supermix (Bio-Rad) with the

CFX96 Real-Time System (Bio-Rad) as described previously 49. ama-1, tba-1, and gpd-2 were used as internal controls. Relative glr-1 mRNA expression levels were normalized using these reference control genes. The following previously described primer sequences were used: tba-1, ama-1, and gpd-2 50. For glr-1 primers, full-length gene sequences were obtained from WormBase, and primers were designed by the Primer 3 software and evaluated for potential secondary structures of the amplicon by MFOLD software. MFOLD analysis was performed by adjusting the values to 50 mM Na+, 3 mM Mg2+, and 60°C annealing temperature. Data were analyzed using Student’s t-test (GraphPad Prism software).

The following C. elegans primers (5’-3’) were used: glr-1 forward: GAA GTT TCA GTG AGA CCT CC; glr-1 reverse: CGC TTT GAC AGA AGT ATT GGT T; tba-1 forward: GTACACTCCACTGATCTCTGCTGACAAG; tba-1 reverse: CTCTGTACAAGAGGCAAACAGCCATG; ama-1 forward: CCTACGATGTATCGAGGCAAA; ama-1 reverse: CCTCCCTCCGGTGTAATAATG; gpd-2 forward: CTCCATCGACTACATGGTCTACTTG; gpd-2 reverse: AGCTGGGTCTCTTGAGTTGTAGAC.

100

References

1. Caldwell, G. A. et al. Suppression of polyglutamine-induced protein aggregation in Caenorhabditis elegans by torsin proteins. Hum Mol Genet 12, 307-319 (2003).

2. Burdette, A. J., Churchill, P. F., Caldwell, G. A. & Caldwell, K. A. The early-onset torsion dystonia-associated protein, torsinA, displays molecular chaperone activity in vitro. Cell Stress & Chaperones 15, 605-617, doi:10.1007/s12192-010-0173-2 (2010).

3. Chen, P. et al. The early-onset torsion dystonia-associated protein, torsinA, is a homeostatic regulator of endoplasmic reticulum stress response. Hum Mol Genet 19, 3502-3515, doi:10.1093/hmg/ddq266 (2010).

4. McLean, P. J. et al. TorsinA and heat shock proteins act as molecular chaperones: suppression of alpha-synuclein aggregation. J Neurochem 83, 846-854 (2002).

5. Cao, S., Gelwix, C. C., Caldwell, K. A. & Caldwell, G. A. Torsin-Mediated Protection from Cellular Stress in the Dopaminergic Neurons of Caenorhabditis elegans. The Journal of Neuroscience 25, 3801-3812, doi:10.1523/jneurosci.5157-04.2005 (2005).

6. Brockie, P. J., Madsen, D. M., Zheng, Y., Mellem, J. & Maricq, A. V. Differential Expression of Glutamate Receptor Subunits in the Nervous System of Caenorhabditis elegans and Their Regulation by the Homeodomain Protein UNC-42. The Journal of Neuroscience 21, 1510-1522 (2001).

7. Kannampalli, P. & Sengupta, J. N. Role of Principal Ionotropic and Metabotropic Receptors in Visceral Pain. Journal of Neurogastroenterology and Motility 21, 147- 158, doi:10.5056/jnm15026 (2015).

8. Bowie, D. Redefining the classification of AMPA-selective ionotropic glutamate receptors. J Physiol 590, 49-61, doi:10.1113/jphysiol.2011.221689 (2012).

9. Wilson, B. K. & Hess, E. J. Symptomatic animal models for dystonia. Movement disorders : official journal of the Movement Disorder Society 28, 982-989, doi:10.1002/mds.25526 (2013).

10. Collingridge, G. L., Isaac, J. T. & Wang, Y. T. Receptor trafficking and synaptic plasticity. Nat Rev Neurosci 5, 952-962, doi:10.1038/nrn1556 (2004).

11. Willard, S. S. & Koochekpour, S. Glutamate, Glutamate Receptors, and Downstream Signaling Pathways. International Journal of Biological Sciences 9, 948-959, doi:10.7150/ijbs.6426 (2013).

101

12. Elias, G. M. & Nicoll, R. A. Synaptic trafficking of glutamate receptors by MAGUK scaffolding proteins. Trends in cell biology 17, 343-352, doi:10.1016/j.tcb.2007.07.005 (2007).

13. Payne, H. L. The role of transmembrane AMPA receptor regulatory proteins (TARPs) in neurotransmission and receptor trafficking (Review). Molecular membrane biology 25, 353-362, doi:10.1080/09687680801986480 (2008).

14. Garcia-Nafria, J., Herguedas, B., Watson, J. F. & Greger, I. H. The dynamic AMPA receptor extracellular region: a platform for synaptic protein interactions. J Physiol 594, 5449-5458, doi:10.1113/jp271844 (2016).

15. Henley, J. M. & Wilkinson, K. A. Synaptic AMPA receptor composition in development, plasticity and disease. Nat Rev Neurosci 17, 337-350, doi:10.1038/nrn.2016.37 (2016).

16. Walsh, D. M. & Selkoe, D. J. A beta oligomers - a decade of discovery. J Neurochem 101, 1172-1184, doi:10.1111/j.1471-4159.2006.04426.x (2007).

17. Zhao, W. Q. et al. Inhibition of calcineurin-mediated endocytosis and alpha-amino-3- hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors prevents amyloid beta oligomer-induced synaptic disruption. J Biol Chem 285, 7619-7632, doi:10.1074/jbc.M109.057182 (2010).

18. Fan, X., Hughes, K. E., Jinnah, H. A. & Hess, E. J. Selective and sustained alpha- amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor activation in cerebellum induces dystonia in mice. J Pharmacol Exp Ther 340, 733-741, doi:10.1124/jpet.111.190082 (2012).

19. Wang, G. J. et al. GRLD-1 regulates cell-wide abundance of glutamate receptor through post-transcriptional regulation. Nature neuroscience 13, 1489-1495, doi:10.1038/nn.2667 (2010).

20. Hart, A. C., Sims, S. & Kaplan, J. M. Synaptic code for sensory modalities revealed by C. elegans GLR-1 glutamate receptor. Nature 378, 82-85, doi:10.1038/378082a0 (1995).

21. Maricq, A. V., Peckol, E., Driscoll, M. & Bargmann, C. I. Mechanosensory signalling in C. elegans mediated by the GLR-1 glutamate receptor. Nature 378, 78-81, doi:10.1038/378078a0 (1995).

22. Zheng, Y., Brockie, P. J., Mellem, J. E., Madsen, D. M. & Maricq, A. V. Neuronal Control of Locomotion in C. elegans Is Modified by a Dominant Mutation in the GLR-1 Ionotropic Glutamate Receptor. Neuron 24, 347-361, doi:https://doi.org/10.1016/S0896-6273(00)80849-1 (1999).

23. Anggono, V. & Huganir, R. L. Regulation of AMPA Receptor Trafficking and Synaptic Plasticity. Current opinion in neurobiology 22, 461-469, doi:10.1016/j.conb.2011.12.006 (2012).

102

24. Jackson, Alexander C. & Nicoll, Roger A. The Expanding Social Network of Ionotropic Glutamate Receptors: TARPs and Other Transmembrane Auxiliary Subunits. Neuron 70, 178-199, doi:10.1016/j.neuron.2011.04.007.

25. Zheng, Y. et al. SOL-1 is an auxiliary subunit that modulates the gating of GLR-1 glutamate receptors in Caenorhabditis elegans. Proc Natl Acad Sci U S A 103, 1100- 1105, doi:10.1073/pnas.0504612103 (2006).

26. Juo, P., Harbaugh, T., Garriga, G. & Kaplan, J. M. CDK-5 Regulates the Abundance of GLR-1 Glutamate Receptors in the Ventral Cord of Caenorhabditis elegans. Molecular biology of the cell 18, 3883-3893, doi:10.1091/mbc.E06-09-0818 (2007).

27. Kramer, L. B. et al. UEV-1 Is an Ubiquitin-Conjugating Enzyme Variant That Regulates Glutamate Receptor Trafficking in C. elegans Neurons. PloS one 5, e14291, doi:10.1371/journal.pone.0014291 (2010).

28. Shi, Y. et al. Functional comparison of the effects of TARPs and cornichons on AMPA receptor trafficking and gating. Proceedings of the National Academy of Sciences 107, 16315-16319, doi:10.1073/pnas.1011706107 (2010).

29. Brockie, P. J. et al. Cornichons control ER export of AMPA receptors to regulate synaptic excitability. Neuron 80, 129-142, doi:10.1016/j.neuron.2013.07.028 (2013).

30. Haering, S. C., Tapken, D., Pahl, S. & Hollmann, M. Auxiliary Subunits: Shepherding AMPA Receptors to the Plasma Membrane. Membranes 4, 469-490, doi:10.3390/membranes4030469 (2014).

31. Burbea, M., Dreier, L., Dittman, J. S., Grunwald, M. E. & Kaplan, J. M. Ubiquitin and AP180 regulate the abundance of GLR-1 glutamate receptors at postsynaptic elements in C. elegans. Neuron 35, 107-120 (2002).

32. Juo, P. & Kaplan, J. M. The anaphase-promoting complex regulates the abundance of GLR-1 glutamate receptors in the ventral nerve cord of C. elegans. Current biology : CB 14, 2057-2062, doi:10.1016/j.cub.2004.11.010 (2004).

33. Schaefer, H. & Rongo, C. KEL-8 Is a Substrate Receptor for CUL3-dependent Ubiquitin Ligase That Regulates Synaptic Glutamate Receptor Turnover. Molecular biology of the cell 17, 1250-1260, doi:10.1091/mbc.E05-08-0794 (2006).

34. Harrington, A. J., Hamamichi, S., Caldwell, G. A. & Caldwell, K. A. C. elegans as a model organism to investigate molecular pathways involved with Parkinson's disease.

Developmental dynamics : an official publication of the American Association of Anatomists 239, 1282-1295, doi:10.1002/dvdy.22231 (2010).

35. Chalasani, S. H. et al. Dissecting a circuit for olfactory behaviour in Caenorhabditis elegans. Nature 450, 63-70, doi:10.1038/nature06292 (2007).

36. Campbell, J. C., Polan-Couillard, L. F., Chin-Sang, I. D. & Bendena, W. G. NPR-9, a Galanin-Like G-Protein Coupled Receptor, and GLR-1 Regulate Interneuronal Circuitry Underlying Multisensory Integration of Environmental Cues in Caenorhabditis elegans. PLoS Genetics 12, doi:10.1371/journal.pgen.1006050 (2016). 103

37. Rongo, C., Whitfield, C. W., Rodal, A., Kim, S. K. & Kaplan, J. M. LIN-10 is a shared component of the polarized protein localization pathways in neurons and epithelia. Cell 94, 751-759 (1998).

38. Shim, J., Umemura, T., Nothstein, E. & Rongo, C. The unfolded protein response regulates glutamate receptor export from the endoplasmic reticulum. Molecular biology of the cell 15, 4818-4828, doi:10.1091/mbc.E04-02-0108 (2004).

39. Chun, D. K., McEwen, J. M., Burbea, M. & Kaplan, J. M. UNC-108/Rab2 regulates postendocytic trafficking in Caenorhabditis elegans. Molecular biology of the cell 19, 2682-2695, doi:10.1091/mbc.E07-11-1120 (2008).

40. Goodchild, R. E., Kim, C. E. & Dauer, W. T. Loss of the dystonia-associated protein torsinA selectively disrupts the neuronal nuclear envelope. Neuron 48, 923-932, doi:10.1016/j.neuron.2005.11.010 (2005).

41. Hewett, J. et al. Mutant torsinA, responsible for early-onset torsion dystonia, forms membrane inclusions in cultured neural cells. Hum Mol Genet 9, 1403-1413 (2000).

42. Callan, A. C., Bunning, S., Jones, O. T., High, S. & Swanton, E. Biosynthesis of the dystonia-associated AAA+ ATPase torsinA at the endoplasmic reticulum. The Biochemical journal 401, 607-612, doi:10.1042/bj20061313 (2007).

43. Hewett, J. W. et al. Mutant torsinA interferes with protein processing through the secretory pathway in DYT1 dystonia cells. Proc Natl Acad Sci U S A 104, 7271-7276, doi:10.1073/pnas.0701185104 (2007).

44. Coombs, I. D. et al. Cornichons modify channel properties of recombinant and glial AMPA receptors. The Journal of neuroscience : the official journal of the Society for Neuroscience 32, 9796-9804, doi:10.1523/jneurosci.0345-12.2012 (2012).

45. Kato, A. S. et al. Hippocampal AMPA receptor gating controlled by both TARP and cornichon proteins. Neuron 68, 1082-1096, doi:10.1016/j.neuron.2010.11.026 (2010).

46 Brenner, S. The genetics of Caenorhabditis elegans. Genetics 77, 71-94 (1974).

47. Calixto, A., Chelur, D., Topalidou, I., Chen, X. & Chalfie, M. Enhanced neuronal RNAi in C. elegans using SID-1. Nature methods 7, 554-559, doi:10.1038/nmeth.1463 (2010).

48. Kaplan, J. M. & Horvitz, H. R. A dual mechanosensory and chemosensory neuron in Caenorhabditis elegans. Proc Natl Acad Sci U S A 90, 2227-2231 (1993).

49. Thompson, M. L. et al. TorsinA rescues ER-associated stress and locomotive defects in C. elegans models of ALS. Dis Model Mech 7, 233-243, doi:10.1242/dmm.013615 (2014).

50. Hoogewijs, D., Houthoofd, K., Matthijssens, F., Vandesompele, J. & Vanfleteren, J. R. Selection and validation of a set of reliable reference genes for quantitative sod gene expression analysis in C. elegans. BMC molecular biology 9, 9, doi:10.1186/1471- 2199-9-9 (2008)

104

CHAPTER FIVE

CONCLUSIONS AND FUTURE DIRECTIONS

Introduction

C. elegans is a versatile model organism, whereby genetic alterations to the genome can be used to determine cellular effects of proteins of interest and/or chemicals in response to various environmental or cellular stressors. We applied this to the analysis of Parkinson’s disease (PD) and early-onset torsion dystonia. Powerful gene editing tools such as RNAi, gene overexpression and genetic mutations assist with the analysis of movement disorders when examining the cellular pathogenesis underlying PD and Dystonia in C. elegans. For C. elegans PD models, the generation of transgenic C. elegans overexpressing human - synuclein (-syn) and GFP specifically in the dopaminergic (DA) neurons provides a platform for the identification of genes, chemicals, and cellular mechanisms that suppress or enhance DA neurotoxicity in C. elegans. For the analysis of early-onset torsion dystonia in C. elegans, modeling the molecular chaperone activity of torsinA in C. elegans has opened an avenue for studies of ER lumen activities of this protein 1,2.

In my study, in Chapter 2, chemical compensation identified four chemicals including meclofenoxate hydrochloride, cyclosporin A, sulfaphenazole and choline as putative protective compounds for mitochondrial phospholipid depletion induced neurodegeneration

105 in C. elegans with -syn expression in DA neurons. In Chapter 3, mir-239 mutants displayed a robust neuroprotective role against -syn-induced neurotoxicity. Several downstream targets of miR-239 were identified by a targeted RNAi screen in the C. elegans -syn- induced neurotoxicity model. In chapter 4, tor-2 was indicated as possible regulator in the trafficking process of an AMPA receptor subunit, GLR-1; these data suggest a possible connection between glutamatergic transmission and the etiology of dystonia. Through all of these studies of movement disorders, our findings strongly support the use of C. elegans as a leading model organism for the analysis of cell biology and genetic features associated with

PD and dystonia.

Conclusion

Chemical compensation rescues mitochondrial phospholipid depletion induced neurodegeneration in a C. elegans PD model

As described in Chapter 2, the conserved enzyme phosphatidylserine decarboxylase

(Psd1/yeast; PSD-1/worms) involved in the synthesis process of phosphatidylethanolamine

(PE). Previous study found that in S. cerevisiae, the co-occurrence of low PE and -syn in psd1Δ cells triggers mitochondrial defects, stress in the endoplasmic reticulum, misprocessing of glycosylphosphatidylinositol-anchored proteins, and a 3-fold increase in the level of -syn. To identify drugs to rescue this phenotype, a screening from the Prestwick library of 1121 Food and Drug Administration-approved drugs using psd1Δ + -syn cells revealed meclofenoxate hydrochloride, cyclosporin A, sulfaphenazole and choline as putative protective compounds. By using our C. elegans neurodegenerative model in which -syn is expressed specifically in the DA neurons with psd-1 depleted by RNAi, the protective activity of these drugs was confirmed. Moreover, all the four drugs rescued worms

106 expressing -syn in DA neurons that were deficient in the phospholipid cardiolipin (CL) following cardiolipin synthase (crls-1) depletion by RNAi. To answer how these drugs might function to compensate for mitochondrial phospholipid depletion in our PD models, there are different explanations based on the different drugs identified. When considering the drug meclofenoxate hydrochloride, it could scavenge radicals or increase the level of phosphatidyl choline (PC). The drug cyclosporin A might inhibit mitochondrial permeability transition pore (mPTP), which is a non-specific pore that forms in the inner mitochondrial membranes in response to high matrix Ca++, arachidonic acid 3, ceramide 4, inorganic phosphate, and many other factors. Here, we suspect that -syn or low PE and/or low CL are possible inciters of mPTP. Sulfaphenazole is known to inhibit CYP2J2, which is involved in the metabolism of lipids. When considering the drug choline, it could increase PC levels which compensates for low PE. We know that supplemental ethanolamine (ETA), which is converted to PE via the Kennedy pathway enzymes in the ER, increased the level of CL in psd1Δ yeast cells. We suggest supplemental choline compensating for low CL by increasing

PE is the same or similar homeostatic mechanism by which ethanolamine (ETA) did 5.

The mutation of mir-239 produces a neuroprotective role and prolonged lifespan in C. elegans of PD model

MicroRNAs (miRNAs) represent a class of small noncoding RNAs of ∼ 20 nt in length involving in the regulation of gene expression at the post-transcriptional level by degrading their target mRNAs and/or inhibiting their translation through sequence specific binding of the 3’untranslated regions of target messenger RNAs 6. While miRNAs have been shown to important for developmental processes, such as cell proliferation and differentiation, their disruption has also been linked with neurodegenerative disease pathogenesis, including

Parkinson’s disease 7.

107

As described in Chapter 3, we demonstrated a specific miRNA, mir-239, when mutant, induced a robust resistance to-syn-induced neurodegeneration in C. elegans. We functionally investigated 26 expression-validated targets of mir-239 regulation via conditional knockdown using a dopaminergic neuron-specific RNAi-sensitive strain. We demonstrated that depletion of nine gene products significantly reduced neuron protection of mir-239, indicating these genes act as downstream targets. These candidates include the gene products that function in fat metabolism (C46C11.1(hosl-1)), are mediated by steroid hormones (T27D12.1, F57G12.1, F46G10.2), are involved in reproduction and larval development (F25H9.6), muscle functions (F45D3.2, R09A8.5); embryo development

(T10G3.1, F27D4.2(lsy22)). Three genes from these targets have human orthologs:

C46C11.1(hosl-1), an ortholog of human hormone-sensitive lipase (LIPE); T2712.1, an ortholog of human SLC17A5 (solute carrier family 17 member 5); F25H9.6, an ortholog of human Ppcdc (Phosphopantothenoylcysteine decarboxylase).

Furthermore, we found mir-239 mutant prolonged life span in C. elegans with or without

-syn expression. Moreover, mir-239 mutant better resisted oxidative stress from sodium azide and paraquat treatment compared with control worms. The most popular theory about aging is reactive oxygen species (ROS) accumulation leading to functional alterations, pathological conditions and finally cell death 8,9. We suggest the ability of mir-239 mutant to resist the oxidative stress makes contribution for extending lifespan in C. elegans. tor-2 might function as a regulator in the trafficking process of an AMPA receptors subunit, GLR-1

AMPA type glutamate receptors (AMPARs) are a class of ionotropic glutamate receptors which are a major class of heteromeric ligand-gated ion channels and regulate rapid neurotransmission and synaptic plasticity in the vertebrate central nervous system (CNS) 10.

GLR-1 is the first characterized one subunit of AMPA type glutamate receptors in the C.

108 elegans. The activity of GLR-1 has been indicated to be necessary for mediating the behavioral response to light nose touch and the frequency with which animals change locomotory direction in response to sensory cues such as food. Mutations in glr-1 disrupted the backing response to nose touch stimulation 11,12. This defect was demonstrated from faulty glutamatergic neurotransmission between sensory neurons (primarily the ASH polymodal sensory neurons) and the command interneurons that express GLR-1 13. After the initial characterization of glr-1, additional behavioral paradigms linked with glr-1 were developed.

For example, glutamate has been indicated as a key regulator of omega turns in C. elegans.

Null glr-1 mutants showed a decrease in the number of omega turns per minute 14,15. Tor-2 is a homolog of human torsinA which are associated with the early-onset torsinA dystonia, an autosomal dominant movement disorder. Tor-2 expressed in few neurons including the command interneuron AVE, in which AMPA type glutamate receptors expressed 16. We found pan-neuronal tor-2 RNAi or loss of function mutation disrupted the glr-1 mediated behavior

(nose touch response, reversal frequency and omega turns frequency) which indicated the activity of tor-2 are linked with glr-1 mediated behaviors. These disrupted behaviors by tor-2 might possibly result from the decreased level of GLR-1. GLR-1 expressed in the ventral nerve cord (VNC) in a puncta pattern in C. elegans. Interestingly, we found TOR-2 are colocalized with GLR-1 in the VNC. To investigate how TOR-2 regulate the activity of GLR-

1, we measured the synaptic number level of GLR-1 in VNC under pan-neuronal tor-2 RNAi.

Our data indicated that compared with control EV, worms fed with tor-2 dsRNA have significant decreased level of GLR-1. We further confirmed that glr-1 mRNA level doesn’t change when tor-2 is depleted, implying the synaptic number change of GLR-1 is not in the transcriptional process, but likely in the translational or post-translational process.

109

Discussion and future direction

In chapter 2, we demonstrated exogenous complement with meclofenoxate hydrochloride, cyclosporin A or sulfaphenazole rescued mitochondrial phospholipid depletion including phosphatidylserine decarboxylase (psd-1) or cardiolipin synthase (crls-1) induced neurodegeneration in C. elegans with -syn expression in DA neuron. Future directions might focus on the proposed molecular mechanisms regarding the neuron protective role of these compounds. We know that the drug meclofenoxate hydrochloride scavenges radicals or increases the level of PC. Therefore, we could examine the impact of meclofenoxate hydrochloride on reactive oxygen species (ROS) level in C. elegans in future studies. The drug cyclosporin A is thought to inhibit mPTP, which is a nonspecific, voltage-dependent mitochondria permeability transition pore, permeant to any molecules less than 1.5 kDa in mass 17 that is formed under certain pathological conditions such as traumatic brain injury

(TBI) and spinal cord injury (SCI) 18. mPTP opening leads to the ROS accumulation and disrupt mitochondrial function 19,20. Since suspect that -syn or low PE and/or low CL are possible inciters of mPTP, we can try to test ROS level and mitochondria functions such as mitochondrial membrane potential (MMP) and ATP level with or without cyclosporin A in future studies to examine this. Sulfaphenazole is known to inhibit CYP2J2, which is involved in the metabolism of lipids. Therefore, in future studies we will knockdown this CYP2J2 and examine the impact on -syn induced neurotoxicity in C. elegans. We suspect supplemental choline increased PC which compensates for low PE and CL 5.To examine the neuroprotective mechanism of choline, we could supplement PC and analyze the impact on neuroprotection in C. elegans. It would be also interesting to test these compounds in higher eukaryotic PD model organisms, such as mice.

110

In chapter 3, we determined that an intersection of longevity and epigenetic modification for -syn neurotoxicity was linked to mir-239 and its targets. A mir-239 mutant showed significant resistance to -syn-induced neurotoxicity while overexpression of mir-239 in C. elegans displayed enhanced neurodegeneration in the presence of -syn. We identified nine gene targets of mir-239 with a neuroprotective role. These nine genes products function in different cellular pathways, including fat metabolism, are mediated by steroid hormones, are involved in reproduction and larval development, muscle functions and embryo development.

This study demonstrated that mechanisms underlying these pathways is associated with the pathogenesis of PD-related DA neurodegeneration. Three of the gene targets have human homologs [C46C11.1(hosl-1), T27D12.1, F25H9.6], indicating a shared pathway with gene candidates that should be further studied for neuroprotective mechanisms.

For the human ortholog of C46C11.1 (hosl-1), HSL (hormone-sensitive lipase), is also called LIPE, and it is an 84 kDa phosphoprotein that is the rate-limiting enzyme in the catalytic breakdown of triglycerides. The function of HSL is to predominantly hydrolyze diglycerides. It also hydrolyzes monoglycerides (MGs), retinylesters, and cholesterol esters

21. One possible mechanism for the reason that HOSL-1 is neuroprotective is because it is associated with the cholesterol metabolism pathway. HSL can hydrolyze cholesterol ester into free cholesterol. Our lab previously identified NCEH-1 (neutral cholesterol ester hydrolase 1) as a functional modifier of -syn-induced neurotoxicity in C. elegans 22. The NCEH-1 gene product functions to liberate metabolically active free cholesterol from intracellularly stored cholesterol esters in a manner similar to HSL. To further explore the role of hosl-1, we can determine if it is neuroprotective using overexpression studies in C. elegans DA neurons with

-syn. If C. elegans HOSL-1 is neuroprotective when overexpressed, we will also determine if human HSL is neuroprotective when overexpressed. There are also two deleterious single

111 nucleotide polymorphisms (SNPs) in human HSL, as determined by mining the SNPs3D database. This database analyzes non-synonymous missense SNPs in gene coding regions that are present in at least 1% of the human populations from the NCBI dbSNP databases. We will then use site-directed mutagenesis to make the two SNPs in HSL which have not been previously analyzed (R611C and G742R). We will then create transgenic animals that express these changes and determine if HSL(R611C) or HSL(G742R) abrogates the DA neuroprotection provided by wild-type HSL against -syn. Additional studies using HOSL-1 or HSL will involve titrating doses of cholesterol in the nematode growth media because

NCEH-1 neuroprotection is dependent on the presence of cholesterol, especially in aged worms 22. We would like to know if HOSL or HSL would also display neuroprotective role in a manner similar to NCEH-1.

Another positive candidate with a human ortholog was T27D12.1; in humans it is referred to as SLC17A5 and it encodes sialin, a lysosomal membrane transporter for sialic acid.

Different mutations in SLC17A5 cause either an autosomal recessive lysosomal storage disorder (LSD) or an autosomal recessive neurodegenerative disorder, which affects infants.

Notably, in a recent genetic association study designed to identify new genes implicated in lysosomal causes of PD, three new candidates were identified; among these was SLC17A5 23.

This study is important because previous mutants linked to lysosomal causes of PD have been shown to be potent determinants of this movement disorder. As examples, mutations in glucocerebrosidase (GBA) and lysosomal type 5 P-type ATPase (ATP13A2) both greatly increase the risk of developing PD 24. Lysosomes are an organelle responsible for clearing up long-lived proteins, such as aggregate-prone -syn, and for removing old or damaged organelles, such as mitochondria. Thus, consistent with what Robak and coworkers have identified, SLC17A5 loss-of-function may increase vulnerability to -syn-mediated mechanisms in PD. Future studies with this protein will involve overexpressing the wild-type 112

SLC17A5 protein (both C. elegans and human) to determine if they rescue -syn-induced neurodegeneration. Additionally, as there are well-characterized autosomal recessive mutations associated with LSD, these mutations can be introduced into the human protein and then transgenic C. elegans created with the changes to examine the impact on -syn-induced neuroprotection. There are also two non-synonymous missense mutations in SLC17A5 predicted by the SNPs3D database. These changes might not cause a devastating disease, such as PD on their own, but, perhaps, in combination with a low-level environmental stressor (paraquat or rotenone), they might push neurons over a threshold into a dysfunctional state. Considering that they are found in a small percentage of the population, these changes are worth modeling in the nematode (A43T and M185V).

The human ortholog of the third positive candidate, F25H9.6, is phosphopantothenoylcysteine decarboxylase (Ppcdc). This enzyme is utilized during cellular respiration and is necessary for the biosynthesis of coenzyme A (CoA). Eukaryotic cells obtain CoA exclusively via the uptake of extracellular precursors, especially pantothenic acid

(vitamin B5), which is intracellularly converted through five conserved enzymatic reactions into CoA. One of intermediate enzyme is Ppcdc (Phosphopantothenoylcysteine decarboxylase) which converts 4’ -Phosphopantothenoylcysteine into 4’- Phosphopantetheine

25. While no one has examined Ppcdc as a neurodegenerative risk factor to date, mutations in de novo CoA biosynthesis has been implicated in hereditary forms of neurodegenerative diseases. For instance, two human enzymes involved in CoA biosynthesis, pantothenate kinase (PANK), which is the first enzyme in the CoA biosynthetic pathway and bifunctional coenzyme A synthase (phosphopantetheine adenylyltransferase (PPAT) and dephospho-CoA kinase (DPCK)), which catalyzes the fourth and fifth sequential steps of CoA biosynthetic pathway 26, are associated with a hereditary disease whereby there is brain iron accumulation

(NBIA) and neurodegeneration 27,28. Further studies on determining the extent of Ppcdc 113 involvement in neuroprotection could involve determining if overexpression of the C. elegans or human homologs are neuroprotective against -syn-induced neurodegeneration. There are also three different isoleucine to methionine missense mutations in the coding region of this gene that could be examined, as well (positions 21, 46, and 78).

In chapter 4, we turned our attention to dystonia, a different movement disorder. Here, we demonstrated pan-neuronal tor-2 RNAi or tor-2 loss of function mutation, disrupted glr-1 mediated behaviors including nose touch response, reversal frequency and omega turn frequency. We deduced that these disrupted behaviors by tor-2 possibly result from the decreased synaptic number of GLR-1. glr-1 mRNA level didn’t change when tor-2 was depleted, therefore the change of GLR-1 synaptic number is not occurring at the transcription level, but is likely occurring during translation or post-translational process. To examine this, we can measure GLR-1 protein levels in the future study. Moreover, it will be meaningful to evaluate the contribution of TOR-2 protein function to AMPAR-mediated currents by patch- clamp electrophysiological analysis. Moreover, it would be interesting to see whether adding exogenous glutamate to worms would compensate for the disrupted behaviors caused by downregulated GLR-1 glutamate receptors.

Additionally, the Human Liver Proteome Project determined that human cornichon 4 physically interacts with torsinA. Cornichons are a family of ER-localized transmembrane proteins that directly interact with AMPARs immediately after translation in the ER to promote export from the ER and Golgi complex. Mutation of cni-1, the sole cornichon homologue in C. elegans, shows a hyper reversal phenotype associated with increased glutamatergic synaptic transmission 29. Thus, we hypothesize that CNI-1 and TOR-2 genetically interact and moreover, suspect that they also physically interact in C. elegans. In the future, it will be important to show a physical interaction between CNI-1 and TOR-2 in C.

114 elegans. TOR-2 is only expressed in a few neuron cells in C. elegans indicating a limitation to obtain enough protein samples if overexpression it in neurons. Although bacterial expression system can compensate this drawback and provide a large quantity of protein sample, it is still not suitable because of lacking protein post-translational modifications.

Therefore, to examine physical interaction between CNI-1 and TOR-2, I suggest that we overexpress both of them in the relatively large size of body wall muscle cells in C. elegans followed by a pull- down assay. If we can show that they do interact, it will suggest that

TOR-2 regulates the trafficking of GLR-1 from the ER to the plasma membrane. Additional experiments include examining glycosylation of GLR-1::GFP by analyzing the sensitivity to the endoglycosidase Endo H which removes N-glycans containing immature carbohydrate chains, that are typically found in the ER. In the Golgi, carbohydrate chains undergo further modifications, rendering them resistant to Endo H 30,31. To do this, we can determine the anterograde delivery process of GLR-1 from ER to Golgi when TOR-2 is depleted in C. elegans.

In summary, C. elegans can provide immeasurable opportunities for studying PD and dystonia. These experimental conclusions and future directions collectively offer a tangible platform on which further advances pertaining to therapeutic strategies and cellular mechanisms associated with PD and Dystonia can be built.

115

References

1. Chen, P. et al. The early-onset torsion dystonia-associated protein, torsinA, is a homeostatic regulator of endoplasmic reticulum stress response. Hum Mol Genet 19, 3502-3515, doi:10.1093/hmg/ddq266 (2010).

2. Cao, S. et al. Chemical enhancement of torsinA function in cell and animal models of torsion dystonia. Disease Models & Mechanisms 3, 386-396, doi:10.1242/dmm.003715 (2010).

3. Penzo, D. et al. Arachidonic acid released by phospholipase A(2) activation triggers Ca(2+)-dependent apoptosis through the mitochondrial pathway. J Biol Chem 279, 25219-25225, doi:10.1074/jbc.M310381200 (2004).

4. Arora, A. S., Jones, B. J., Patel, T. C., Bronk, S. F. & Gores, G. J. Ceramide induces hepatocyte cell death through disruption of mitochondrial function in the rat. Hepatology (Baltimore, Md.) 25, 958-963, doi:10.1002/hep.510250428 (1997).

5. Wang, S. et al. Phosphatidylethanolamine deficiency disrupts -synuclein homeostasis in yeast and worm models of Parkinson disease. Proceedings of the National Academy of Sciences 111, E3976-E3985, doi:10.1073/pnas.1411694111 (2014).

6. Li, Z. & Rana, T. M. Molecular Mechanisms of RNA-Triggered Gene Silencing Machineries. Accounts of Chemical Research 45, 1122-1131, doi:10.1021/ar200253u (2012).

7. Chan, A. W. S. & Kocerha, J. The Path to microRNA Therapeutics in Psychiatric and Neurodegenerative Disorders. Front Genet 3, doi:10.3389/fgene.2012.00082 (2012).

8. Nocchi, L., Daly, D. M., Chapple, C. & Grundy, D. Induction of oxidative stress causes functional alterations in mouse urothelium via a TRPM8-mediated mechanism: implications for aging. Aging cell 13, 540-550, doi:10.1111/acel.12208 (2014).

9. Harman, D. Aging: a theory based on free radical and radiation chemistry. Journal of gerontology 11, 298-300 (1956).

10. Hollmann, M., O'Shea-Greenfield, A., Rogers, S. W. & Heinemann, S. Cloning by functional expression of a member of the glutamate receptor family. Nature 342, 643- 648, doi:10.1038/342643a0 (1989).

11. Hart, A. C., Sims, S. & Kaplan, J. M. Synaptic code for sensory modalities revealed by C. elegans GLR-1 glutamate receptor. Nature 378, 82-85, doi:10.1038/378082a0 (1995).

116

12. Maricq, A. V., Peckol, E., Driscoll, M. & Bargmann, C. I. Mechanosensory signalling in C. elegans mediated by the GLR-1 glutamate receptor. Nature 378, 78-81, doi:10.1038/378078a0 (1995).

13. Mellem, J. E., Brockie, P. J., Zheng, Y., Madsen, D. M. & Maricq, A. V. Decoding of polymodal sensory stimuli by postsynaptic glutamate receptors in C. elegans. Neuron 36, 933-944 (2002).

14. Chalasani, S. H. et al. Dissecting a circuit for olfactory behaviour in Caenorhabditis elegans. Nature 450, 63-70, doi:10.1038/nature06292 (2007).

15. Campbell, J. C., Polan-Couillard, L. F., Chin-Sang, I. D. & Bendena, W. G. NPR-9, a Galanin-Like G-Protein Coupled Receptor, and GLR-1 Regulate Interneuronal Circuitry Underlying Multisensory Integration of Environmental Cues in Caenorhabditis elegans. PLoS Genetics 12, doi:10.1371/journal.pgen.1006050 (2016).

16. Caldwell, G. A. et al. Suppression of polyglutamine-induced protein aggregation in Caenorhabditis elegans by torsin proteins. Hum Mol Genet 12, 307-319 (2003).

17. Bernardi, P. Mitochondrial Transport of Cations: Channels, Exchangers, and Permeability Transition. Physiological Reviews 79, 1127-1155, doi:10.1152/physrev.1999.79.4.1127 (1999).

18. Veech, R. L., Valeri, C. R. & VanItallie, T. B. The mitochondrial permeability transition pore provides a key to the diagnosis and treatment of traumatic brain injury. IUBMB life 64, 203-207, doi:10.1002/iub.590 (2012).

19. Liu, Y. et al. Shengmai Formula suppressed over-activated Ras/MAPK pathway in C. elegans by opening mitochondrial permeability transition pore via regulating cyclophilin D. Sci Rep 6, 38934, doi:10.1038/srep38934 (2016).

20. Zhao, Y. et al. Vanadium compounds induced mitochondria permeability transition pore (PTP) opening related to oxidative stress. Journal of inorganic biochemistry 104, 371-378, doi:10.1016/j.jinorgbio.2009.11.007 (2010).

21. Zechner, R. & Langin, D. Hormone-sensitive lipase deficiency in humans. Cell Metab 20, 199-201, doi:10.1016/j.cmet.2014.07.018 (2014).

22. Zhang, S., Glukhova, S. A., Caldwell, K. A. & Caldwell, G. A. NCEH-1 modulates cholesterol metabolism and protects against -synuclein toxicity in a C. elegans model of Parkinson’s disease. Human Molecular Genetics 26, 3823-3836, doi:10.1093/hmg/ddx269 (2017).

23. Robak, L. A. et al. Excessive burden of lysosomal storage disorder gene variants in Parkinson's disease. Brain 140, 3191-3203, doi:10.1093/brain/awx285 (2017).

24. Sidransky, E. et al. Multicenter analysis of glucocerebrosidase mutations in Parkinson's disease. The New England journal of medicine 361, 1651-1661, doi:10.1056/NEJMoa0901281 (2009).

117

25. Srinivasan, B. et al. Extracellular 4'-phosphopantetheine is a source for intracellular coenzyme A synthesis. Nature chemical biology 11, 784-792, doi:10.1038/nchembio.1906 (2015).

26. Daugherty, M. et al. Complete reconstitution of the human coenzyme A biosynthetic pathway via comparative genomics. J Biol Chem 277, 21431-21439, doi:10.1074/jbc.M201708200 (2002).

27. Dusi, S. et al. Exome Sequence Reveals Mutations in CoA Synthase as a Cause of Neurodegeneration with Brain Iron Accumulation. The American Journal of Human Genetics 94, 11-22, doi:10.1016/j.ajhg.2013.11.008.

28. Zhou, B. et al. A novel pantothenate kinase gene (PANK2) is defective in Hallervorden-Spatz syndrome. Nat Genet 28, 345-349, doi:10.1038/ng572 (2001).

29. Brockie, P. J. et al. Cornichons control ER export of AMPA receptors to regulate synaptic excitability. Neuron 80, 129-142, doi:10.1016/j.neuron.2013.07.028 (2013).

30. Burbea, M., Dreier, L., Dittman, J. S., Grunwald, M. E. & Kaplan, J. M. Ubiquitin and AP180 regulate the abundance of GLR-1 glutamate receptors at postsynaptic elements in C. elegans. Neuron 35, 107-120 (2002).

31. Chun, D. K., McEwen, J. M., Burbea, M. & Kaplan, J. M. UNC-108/Rab2 regulates postendocytic trafficking in Caenorhabditis elegans. Molecular biology of the cell 19, 2682-2695, doi:10.1091/mbc.E07-11-1120 (2008).

118